Subjective Stress & Criminal Intent (Brauer, Day, & Kotlaja)
Jon Brauer
2024-08-02
1 Preamble
This document contains all code associated with the “Does subjective stress cause criminal intent” paper (Brauer, Day, & Kotlaja), including data wrangling, modeling, and visualizations. Warning: the document is long - excessively so - because it is intended to serve as a digital lab notebook of sorts that transparently documents decisions made as we navigated through the garden of forking paths on the way to publication. Some readers may appreciate this format, while others will not; c’est la vie.
In order to successfully knit into a single document, the full code was separated into several “child” RMD files that were compiled and sequentially knitted within this “mother” document. Further, within those separate child files, many additional files (e.g., modified datasets; model fits; posterior contrasts) were cached in various subfolders to reduce memory and computational demands and drastically speed up processing. All cached files may be deleted then, upon deletion, may be subsequently cached again by, e.g., running each individual “child” file. However, the full document will take a substantial amount of computational resources and time to run with the cached files, so delete these at your own risk. Also, you will need a computer with sufficient memory and processing power to render the final merged document (for comparison, I used a Dell XPS with Intel i9 2.4GHz processor & 32GB of RAM).
Also, final manuscript writing and pre-publication review inevitably result in changes. I plan to include any major additions to code and results; however, such changes could result in labeling misalignment between this document and the text. For example, initially I had generated DAGs outside this lab notebook, but ultimately decided to include them in the supplement and add them as a formal “Figure 1” in the manuscript. As a result, this “bumped” each figure label by one in the manuscript: Figure 1 in this supplement is now labeled as Figure 2 in the current manuscript; Figure 2 in the supplement is Figure 3 in the manuscript, and so on. After pre-publication review and (hopefully) acceptance, I hope to find time to relabel everything to match.
With all that said, buckle up for a bumpy ride.
2 Figure 1: Stress distributions (RQ1A)
(RMD FILE: BDK_2023_Stress_1_Fig1)
2.1 Reading Data, Packages, & Options
Load libraries & settings
#Load necessary libraries and set options.
# Start by installing "here" and "groundhog" packages if not installed
# initial_packages <- c("here", "groundhog")
# # Install packages not yet installed
# installed_packages <- initial_packages %in% rownames(installed.packages())
# if (any(installed_packages == FALSE)) {
# install.packages(initial_packages[!installed_packages])
# }
#If using groundhog, then load necessary libraries with "here" & "groundhog":
library(here)
#start by loading here package for self-referential file directory structure
# library(groundhog)
# load groundhog package for reproducible date-specified library loading
# here()
# check here() working directory
# get.groundhog.folder()
# check default groundhog working directory - may be diff than here()
# set.groundhog.folder(here())
# set groundhog working directory to same as here()
# groundhog.day="2024-07-15"
# assign date as object (at least two days b4 today)
# NOTE: must set date at least two *months* b4 today w/github install
# groundhog inconsistent after v3 update; removed from workflow Feb 2024
# commented out but kept script to help future reproductions
# groundhog.library("
library('tidyverse')
library('haven')
library('rstanarm')
library('brms')
library('bayesplot')
library('tidybayes')
library('modelr')
library('scico')
library('ggdist')
library('ggridges')
library('ggpubr')
library('ggnewscale')
library('viridis')
library('rstan')
library('StanHeaders')
library('parallel')
library('parallelly')
library('future')
library('emmeans')
library('patchwork')
library('mirt')
library('bmlm')
library('mediation')
library('bayestestR')
library('qgraph')
library('summarytools')
library('ggstance')
library('gridGraphics')
library('grid')
library('gridExtra')
library('xfun')
library('ggh4x')
library('cmdstanr')
library('polycor')
library('latticeExtra')
library('gt')
library(modelsummary)
library(naniar)
# ", groundhog.day)
#library(MASS) #conflicts w/select(), if attached, call dplyr::select()
#library(psych) #conflicts with alpha (figure transparency)
#NOTE: brm() models specified with "cmdstanr" backend instead of "rstan" for compiling
# Need to install RTools(4.3) outside R first to get cmdstanr to work. See:
# https://mc-stan.org/cmdstanr/articles/cmdstanr.html
# Once loaded, additional steps may be necessary to install cmdstan.
# E.g., run following to install cmdstanr, load library, & check cmdstan:
# install.packages("cmdstanr",repos = c("https://mc-stan.org/r-packages/",
# getOption("repos")))
# library(cmdstanr)
# cmdstanr::check_cmdstan_toolchain(fix = TRUE)
# check_cmdstan_toolchain()
#set default knit & scientific notation options
knitr::opts_chunk$set(echo=TRUE, warning=FALSE, message=FALSE, fig.align="center")
options(scipen = 999, digits = 2) #set global option to two decimals
# identify & set # of available cores
# replaced parallel::detectCores()
# https://www.jottr.org/2022/12/05/avoid-detectcores/
nCoresphys <- parallelly::availableCores(logical = FALSE)
theme_set(bayesplot::theme_default())
options(mc.cores = parallelly::availableCores(logical=FALSE))
#options(mc.cores = 4, logical=FALSE)
#alt: can specify exact number of cores desired here &/or in models/ppchecks
# graphics.off() # This closes all of R's graphics windows.
# rm(list=ls()) # Careful! This clears all of R's memory!
print("T/F: Root 'here()' folder contains subfolder 'Output'")## [1] "T/F: Root 'here()' folder contains subfolder 'Output'"#check for Output folder to save figures, create if one does not exist
ifelse(dir.exists(here("Output")), TRUE, dir.create(here("Output")))## [1] TRUERead data & create mean/median-split indicators for community SES
#NOTE: COMMENTED OUT SCRIPT BELOW SAVES ANALYSIS SUBSET FROM ORIGINAL FULL DATASET FOR PUBLIC SHARING
# ANALYSIS DATASET SAVED AND SHARED (/1_Data_Files/stress_dat.Rdata)
# (FULL ORIGINAL SPSS DATAFILE IS INCLUDED HERE)
# stress_dat <- read_spss(here("1_Data_Files/Datasets",'Merged_Bangladesh_Data_W12.sav'))
# names(stress_dat) <- tolower(names(stress_dat))
# stress_dat <- stress_dat %>%
# dplyr::select(
# slno, centerw1, centerw2, w_now1, n_villw1, #id vars
# q10_1w1, q10_2w1, q10_3w1, q10_4w1,
# q10_5w1, q10_6w1, q10_7w1, #stress w1
# q10_1w2, q10_2w2, q10_3w2, q10_4w2,
# q10_5w2, q10_6w2, q10_7w2, #stress w2
# q25_1w1, q25_2w1, q25_4w1,
# q25_5w1, q25_6w1, q25_8w1, #past crime w1
# q25_1w2, q25_2w2, q25_4w2,
# q25_5w2, q25_6w2, q25_7w2, #past crime w2
# q26_1w1, q26_2w1, q26_4w1,
# q26_5w1, q26_6w1, q26_8w1, #criminal intent w1
# q26_1w2, q26_2w2, q26_4w2,
# q26_5w2, q26_6w2, q26_7w2, #criminal intent w2
# q12_1w1, q12_2w1, q12_3w1, q12_4w1,
# q12_5w1, q12_6w1, q12_7w1, #negative emotions w1
# q12_1w2, q12_2w2, q12_3w2, q12_4w2,
# q12_5w2, q12_6w2, q12_7w2, #negative emotions w2
# q11_1w1, q11_2w1, q11_3w1,
# q11_6w1, q11_7w1, q11_8w1, #financial hardship/SES w1
# q11_1w2, q11_2w2, q11_3w2,
# q11_6w2, q11_7w2, q11_8w2, #financial hardship/SES w2
# q1w1, areaw1, areaw2, q2w1, q7w1, q8khaw1, q5w1 #basic demographics
# ) %>%
# mutate( #Community-level area ID (urban ward/rural village)
# wardvill = w_now1 + n_villw1
# ) %>%
# group_by(wardvill) %>% #generate anonymous area ID
# mutate(area_id = cur_group_id()) %>%
# ungroup() %>%
# rename(
# id = slno,
# agew1 = q2w1, #rename demographics
# educw1 = q7w1,
# kidsw1 = q8khaw1,
# stmonyw1 = q10_4w1, #rename stress items
# stmonyw2 = q10_4w2,
# sttranw1 = q10_5w1,
# sttranw2 = q10_5w2,
# strespw1 = q10_2w1,
# strespw2 = q10_2w2,
# stfairw1 = q10_3w1,
# stfairw2 = q10_3w2,
# stjobw1 = q10_1w1,
# stjobw2 = q10_1w2,
# stthftw1 = q10_6w1,
# stthftw2 = q10_6w2,
# stmugw1 = q10_7w1,
# stmugw2 = q10_7w2,
# pstthflt5w1 = q25_1w1, #rename past crime w1 items
# pstthfgt5w1 = q25_2w1,
# pstthreatw1 = q25_4w1,
# pstharmw1 = q25_5w1,
# pstusedrgw1 = q25_6w1,
# psthackw1 = q25_8w1,
# pstthflt5w2 = q25_1w2, #rename past crime w2 items
# pstthfgt5w2 = q25_2w2,
# pstthreatw2 = q25_4w2,
# pstharmw2 = q25_5w2,
# pstusedrgw2 = q25_6w2,
# psthackw2 = q25_7w2,
# prjthflt5w1 = q26_1w1, #rename criminal intent w1 items
# prjthfgt5w1 = q26_2w1,
# prjthreatw1 = q26_4w1,
# prjharmw1 = q26_5w1,
# prjusedrgw1 = q26_6w1,
# prjhackw1 = q26_8w1,
# prjthflt5w2 = q26_1w2, #rename criminal intent w2 items
# prjthfgt5w2 = q26_2w2,
# prjthreatw2 = q26_4w2,
# prjharmw2 = q26_5w2,
# prjusedrgw2 = q26_6w2,
# prjhackw2 = q26_7w2,
# depcantgow1 = q12_1w1, #rename negative emotions w1 items
# depeffortw1 = q12_2w1,
# deplonelyw1 = q12_3w1,
# depbluesw1 = q12_4w1,
# depunfairw1 = q12_5w1,
# depmistrtw1 = q12_6w1,
# depbetrayw1 = q12_7w1,
# depcantgow2 = q12_1w2, #rename negative emotions w2 items
# depeffortw2 = q12_2w2,
# deplonelyw2 = q12_3w2,
# depbluesw2 = q12_4w2,
# depunfairw2 = q12_5w2,
# depmistrtw2 = q12_6w2,
# depbetrayw2 = q12_7w2
# ) %>%
# mutate(
# female = if_else(q1w1 == 1, 0, 1),
# rural = if_else(areaw1 == 2, 1, 0), #(no rural/urban chg from areaw1-areaw2)
# marriedw1 = if_else(q5w1 == 2, 1, 0),
# sesaw1 = q11_1w1 + q11_2w1 + q11_3w1 + #used create area (L2) SES
# q11_6w1 + q11_7w1 + q11_8w1, #(reverse of indiv hardship below)
# depunfairw1 = na_if(depunfairw1, 9) #recode missing obs from 9 to NA
# ) %>%
# group_by(area_id) %>% #create area (L2) SES
# mutate(
# L2sesw1 = mean(sesaw1)
# ) %>%
# ungroup() %>%
# mutate_at(vars(c(q11_1w1, q11_2w1, q11_3w1, q11_6w1, q11_7w1, q11_8w1,
# q11_1w2, q11_2w2, q11_3w2, q11_6w2, q11_7w2, q11_8w2)),
# function(x) 6-x) %>% #reverse-code SES items for indiv financial hardship
# mutate( #create financial hardship sum scale
# osfinaw1 = q11_1w1 + q11_2w1 + q11_3w1 + q11_6w1 + q11_7w1 + q11_8w1,
# osfinaw2 = q11_1w2 + q11_2w2 + q11_3w2 + q11_6w2 + q11_7w2 + q11_8w2
# ) %>%
# dplyr::select( #drop ward/village identifiers & redundant cols
# -c(w_now1, n_villw1, wardvill, q1w1, areaw1, areaw2, q5w1, sesaw1,
# q11_1w1, q11_2w1, q11_3w1, q11_6w1, q11_7w1, q11_8w1,
# q11_1w2, q11_2w2, q11_3w2, q11_6w2, q11_7w2, q11_8w2)
# )
#
# # freq(stress_dat$depunfairw1)
#
# save(stress_dat, file = here("1_Data_Files/Datasets/stress_dat.Rdata"))2.1.1 Read data subset into R
#load Rdata file containing Dhaka W1/W2 stress analysis subset
load(here("1_Data_Files/Datasets/stress_dat.Rdata"))
options(scipen = 999, digits = 3) #change global option to three decimals Now that the data subset are read into R, let’s skim the data. A Pearson’s R correlation matrix is also included for future meta-analytic purposes. However, note that a key focus here will be on moving beyond metric normality assumptions underlying traditional linear model summaries (e.g., Pearson’s r; linear regression) and, instead, providing more valid and precise summaries of “bivariate” correlations (at T1 and change corrs that decompose within-person & between-person variance) using ordinal modeling approaches.
2.1.2 Skim data & correlations
2.1.2.1 Preview data
stress_dat %>% zap_label() %>% head() %>% gt()| id | centerw1 | centerw2 | stjobw1 | strespw1 | stfairw1 | stmonyw1 | sttranw1 | stthftw1 | stmugw1 | stjobw2 | strespw2 | stfairw2 | stmonyw2 | sttranw2 | stthftw2 | stmugw2 | pstthflt5w1 | pstthfgt5w1 | pstthreatw1 | pstharmw1 | pstusedrgw1 | psthackw1 | pstthflt5w2 | pstthfgt5w2 | pstthreatw2 | pstharmw2 | pstusedrgw2 | psthackw2 | prjthflt5w1 | prjthfgt5w1 | prjthreatw1 | prjharmw1 | prjusedrgw1 | prjhackw1 | prjthflt5w2 | prjthfgt5w2 | prjthreatw2 | prjharmw2 | prjusedrgw2 | prjhackw2 | depcantgow1 | depeffortw1 | deplonelyw1 | depbluesw1 | depunfairw1 | depmistrtw1 | depbetrayw1 | depcantgow2 | depeffortw2 | deplonelyw2 | depbluesw2 | depunfairw2 | depmistrtw2 | depbetrayw2 | agew1 | educw1 | kidsw1 | area_id | female | rural | marriedw1 | L2sesw1 | osfinaw1 | osfinaw2 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 1 | 1 | 5 | 2 | 1 | 4 | 5 | 1 | 1 | 5 | 2 | 1 | 4 | 5 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 3 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 3 | 1 | 1 | 1 | 1 | 1 | 1 | 3 | 4 | 1 | 2 | 1 | 1 | 1 | 1 | 2 | 4 | 2 | 3 | 2 | 2 | 26 | 14 | 0 | 2 | 0 | 0 | 0 | 16.9 | 24 | 24 |
| 2 | 1 | 1 | 5 | 5 | 4 | 5 | 2 | 1 | 1 | 5 | 5 | 5 | 5 | 2 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 4 | 2 | 1 | 3 | 4 | 1 | 1 | 4 | 4 | 1 | 3 | 3 | 3 | 3 | 24 | 7 | 0 | 2 | 0 | 0 | 0 | 16.9 | 21 | 20 |
| 3 | 1 | 1 | 3 | 1 | 1 | 5 | 5 | 1 | 3 | 3 | 1 | 1 | 5 | 4 | 1 | 3 | 1 | 1 | 1 | 2 | 1 | 1 | 1 | 1 | 1 | 2 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 4 | 5 | 2 | 5 | 4 | 2 | 4 | 4 | 5 | 5 | 5 | 2 | 1 | 60 | 8 | 3 | 2 | 0 | 0 | 0 | 16.9 | 27 | 27 |
| 4 | 1 | 1 | 4 | 5 | 5 | 1 | 2 | 5 | 1 | 5 | 5 | 5 | 1 | 2 | 4 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 3 | 1 | 1 | 3 | 1 | 1 | 3 | 1 | 1 | 1 | 2 | 1 | 1 | 2 | 3 | 1 | 1 | 1 | 2 | 3 | 3 | 3 | 3 | 3 | 23 | 14 | 0 | 2 | 0 | 0 | 0 | 16.9 | 13 | 13 |
| 5 | 1 | NA | 4 | 5 | 5 | 5 | 5 | 5 | 4 | NA | NA | NA | NA | NA | NA | NA | 1 | 1 | 1 | 1 | 1 | 1 | NA | NA | NA | NA | NA | NA | 1 | 1 | 1 | 1 | 1 | 1 | NA | NA | NA | NA | NA | NA | 3 | 1 | 1 | 2 | 1 | 2 | 1 | NA | NA | NA | NA | NA | NA | NA | 46 | 12 | 2 | 2 | 0 | 0 | 1 | 16.9 | 17 | NA |
| 6 | 1 | 1 | 1 | 2 | 3 | 1 | 3 | 1 | 1 | 1 | 2 | 3 | 1 | 3 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 4 | 4 | 1 | 1 | 3 | 1 | 1 | 5 | 4 | 1 | 2 | 3 | 3 | 1 | 37 | 15 | 4 | 2 | 1 | 0 | 1 | 16.9 | 18 | 18 |
2.1.2.2 Column names & labels
setNames(stack(lapply(stress_dat, label))[2:1], c("Varcode", "Variables")) %>% gt()| Varcode | Variables |
|---|---|
| id | Serial Number |
| centerw1 | Center |
| centerw2 | Center |
| stjobw1 | Getting a job that you really enjoy? |
| strespw1 | Earning respect from others around you? |
| stfairw1 | Being treated fairly by others around you? |
| stmonyw1 | Having enough money to buy the things you need? |
| sttranw1 | Having reliable daily transportation? |
| stthftw1 | Other people stealing from you? |
| stmugw1 | Other people mugging or assaulting you? |
| stjobw2 | Getting a job that you really enjoy? |
| strespw2 | Earning respect from others around you? |
| stfairw2 | Being treated fairly by others around you? |
| stmonyw2 | Having enough money to buy the things you need? |
| sttranw2 | Having reliable daily transportation? |
| stthftw2 | Having your money or property stolen by other people? |
| stmugw2 | Being physically attacked or assaulted by other people |
| pstthflt5w1 | Took money or property that didn't belong to you worth less than Tk 150. |
| pstthfgt5w1 | Took money or property that didn't belong to you worth Tk 150 or more. |
| pstthreatw1 | Threatened to use violence on someone else. |
| pstharmw1 | Physically harmed someone else on purpose. |
| pstusedrgw1 | Used ganja or other illegal drug. |
| psthackw1 | Attempt to access another person's private information, such as a bank account or computer files, without his or her knowledge or permission |
| pstthflt5w2 | Took money or property that didn't belong to you worth less than Tk 150. |
| pstthfgt5w2 | Took money or property that didn't belong to you worth Tk 150 or more. |
| pstthreatw2 | Threatened to use violence on someone else. |
| pstharmw2 | Physically harmed someone else on purpose. |
| pstusedrgw2 | Used ganja or other illegal drug. |
| psthackw2 | Attempt to access another person’s private information, such as a bank account or computer files, without his or her knowledge or permission. |
| prjthflt5w1 | Take money or property that didn't belong to you worth less than Tk 150. |
| prjthfgt5w1 | Take money or property that didn't belong to you worth Tk 150 or more. |
| prjthreatw1 | Threaten to use violence on someone else. |
| prjharmw1 | Physically harm someone else on purpose. |
| prjusedrgw1 | Use ganja or other illegal drug. |
| prjhackw1 | Attempt to access another person's private information, such as a bank account or computer files, without his or her knowledge or permission |
| prjthflt5w2 | Take money or property that didn't belong to you worth less than Tk 150. |
| prjthfgt5w2 | Take money or property that didn't belong to you worth Tk 150 or more. |
| prjthreatw2 | Threaten to use violence on someone else. |
| prjharmw2 | Physically harm someone else on purpose. |
| prjusedrgw2 | Use ganja or other illegal drug. |
| prjhackw2 | Attempt to access another person’s private information, such as a bank account or computer files, without his or her knowledge or permission. |
| depcantgow1 | Felt you just could not get going |
| depeffortw1 | Felt everything was an effort |
| deplonelyw1 | Felt lonely |
| depbluesw1 | Felt you could not shake the blues |
| depunfairw1 | Felt like your life circumstances are unfair |
| depmistrtw1 | Felt mistreated by others |
| depbetrayw1 | Felt betrayed by people you care about |
| depcantgow2 | Felt you just could not get going |
| depeffortw2 | Felt everything was an effort |
| deplonelyw2 | Felt lonely |
| depbluesw2 | Felt you could not shake the blues |
| depunfairw2 | Felt like your life circumstances are unfair |
| depmistrtw2 | Felt mistreated by others |
| depbetrayw2 | Felt betrayed by people you care about |
| agew1 | What is the respondent's age? |
| educw1 | How many years of formal education do you have? |
| kidsw1 | IF YES TO 8a. How many children do you have? |
| area_id | NA |
| female | NA |
| rural | NA |
| marriedw1 | NA |
| L2sesw1 | NA |
| osfinaw1 | NA |
| osfinaw2 | NA |
2.1.2.3 Skim data
datasummary_skim(stress_dat)| Unique | Missing Pct. | Mean | SD | Min | Median | Max | Histogram | |
|---|---|---|---|---|---|---|---|---|
| Serial Number | 600 | 0 | 300.5 | 173.3 | 1.0 | 300.5 | 600.0 |  |
| Center | 1 | 0 | 1.0 | 0.0 | 1.0 | 1.0 | 1.0 | |
| 2 | 18 | 1.0 | 0.0 | 1.0 | 1.0 | 1.0 | |
|
| Getting a job that you really enjoy? | 5 | 0 | 3.3 | 1.4 | 1.0 | 3.0 | 5.0 |  |
| Earning respect from others around you? | 5 | 0 | 3.3 | 1.5 | 1.0 | 4.0 | 5.0 |  |
| Being treated fairly by others around you? | 5 | 0 | 3.4 | 1.5 | 1.0 | 4.0 | 5.0 |  |
| Having enough money to buy the things you need? | 5 | 0 | 3.2 | 1.1 | 1.0 | 3.0 | 5.0 |  |
| Having reliable daily transportation? | 5 | 0 | 3.2 | 1.1 | 1.0 | 3.0 | 5.0 |  |
| Other people stealing from you? | 5 | 0 | 2.3 | 1.2 | 1.0 | 2.0 | 5.0 |  |
| Other people mugging or assaulting you? | 5 | 0 | 2.0 | 1.1 | 1.0 | 2.0 | 5.0 |  |
| Getting a job that you really enjoy? | 6 | 18 | 3.3 | 1.4 | 1.0 | 4.0 | 5.0 |  |
| Earning respect from others around you? | 6 | 18 | 3.5 | 1.3 | 1.0 | 4.0 | 5.0 |  |
| Being treated fairly by others around you? | 6 | 18 | 3.5 | 1.3 | 1.0 | 4.0 | 5.0 |  |
| Having enough money to buy the things you need? | 6 | 18 | 3.3 | 1.1 | 1.0 | 3.0 | 5.0 |  |
| Having reliable daily transportation? | 6 | 18 | 3.3 | 1.0 | 1.0 | 3.0 | 5.0 |  |
| Having your money or property stolen by other people? | 6 | 18 | 2.3 | 1.2 | 1.0 | 2.0 | 5.0 |  |
| Being physically attacked or assaulted by other people | 6 | 18 | 2.0 | 1.1 | 1.0 | 2.0 | 5.0 |  |
| Took money or property that didn't belong to you worth less than Tk 150. | 5 | 0 | 1.2 | 0.6 | 1.0 | 1.0 | 5.0 |  |
| Took money or property that didn't belong to you worth Tk 150 or more. | 4 | 0 | 1.2 | 0.5 | 1.0 | 1.0 | 4.0 |  |
| Threatened to use violence on someone else. | 4 | 0 | 1.1 | 0.3 | 1.0 | 1.0 | 4.0 |  |
| Physically harmed someone else on purpose. | 3 | 0 | 1.1 | 0.4 | 1.0 | 1.0 | 3.0 |  |
| Used ganja or other illegal drug. | 4 | 0 | 1.1 | 0.4 | 1.0 | 1.0 | 4.0 |  |
| Attempt to access another person's private information, such as a bank account or computer files, without his or her knowledge or permission | 4 | 0 | 1.1 | 0.3 | 1.0 | 1.0 | 4.0 |  |
| Took money or property that didn't belong to you worth less than Tk 150. | 5 | 18 | 1.2 | 0.5 | 1.0 | 1.0 | 4.0 |  |
| Took money or property that didn't belong to you worth Tk 150 or more. | 4 | 18 | 1.2 | 0.5 | 1.0 | 1.0 | 3.0 |  |
| Threatened to use violence on someone else. | 4 | 18 | 1.0 | 0.3 | 1.0 | 1.0 | 3.0 |  |
| Physically harmed someone else on purpose. | 4 | 18 | 1.1 | 0.3 | 1.0 | 1.0 | 3.0 |  |
| Used ganja or other illegal drug. | 5 | 18 | 1.1 | 0.4 | 1.0 | 1.0 | 4.0 |  |
| Attempt to access another person’s private information, such as a bank account or computer files, without his or her knowledge or permission. | 4 | 18 | 1.0 | 0.2 | 1.0 | 1.0 | 3.0 |  |
| Take money or property that didn't belong to you worth less than Tk 150. | 5 | 0 | 1.1 | 0.5 | 1.0 | 1.0 | 5.0 |  |
| Take money or property that didn't belong to you worth Tk 150 or more. | 3 | 0 | 1.1 | 0.4 | 1.0 | 1.0 | 3.0 |  |
| Threaten to use violence on someone else. | 3 | 0 | 1.1 | 0.3 | 1.0 | 1.0 | 3.0 |  |
| Physically harm someone else on purpose. | 4 | 0 | 1.1 | 0.3 | 1.0 | 1.0 | 4.0 |  |
| Use ganja or other illegal drug. | 3 | 0 | 1.1 | 0.3 | 1.0 | 1.0 | 3.0 |  |
| Attempt to access another person's private information, such as a bank account or computer files, without his or her knowledge or permission | 4 | 0 | 1.0 | 0.3 | 1.0 | 1.0 | 4.0 |  |
| Take money or property that didn't belong to you worth less than Tk 150. | 5 | 18 | 1.2 | 0.5 | 1.0 | 1.0 | 4.0 |  |
| Take money or property that didn't belong to you worth Tk 150 or more. | 5 | 18 | 1.1 | 0.4 | 1.0 | 1.0 | 4.0 |  |
| Threaten to use violence on someone else. | 4 | 18 | 1.1 | 0.3 | 1.0 | 1.0 | 3.0 |  |
| Physically harm someone else on purpose. | 4 | 18 | 1.0 | 0.3 | 1.0 | 1.0 | 3.0 |  |
| Use ganja or other illegal drug. | 5 | 18 | 1.1 | 0.3 | 1.0 | 1.0 | 5.0 |  |
| Attempt to access another person’s private information, such as a bank account or computer files, without his or her knowledge or permission. | 3 | 18 | 1.0 | 0.1 | 1.0 | 1.0 | 2.0 |  |
| Felt you just could not get going | 5 | 0 | 3.0 | 1.2 | 1.0 | 3.0 | 5.0 |  |
| Felt everything was an effort | 5 | 0 | 2.5 | 1.0 | 1.0 | 2.0 | 5.0 |  |
| Felt lonely | 5 | 0 | 2.5 | 1.2 | 1.0 | 2.0 | 5.0 |  |
| Felt you could not shake the blues | 5 | 0 | 2.5 | 1.1 | 1.0 | 2.0 | 5.0 |  |
| Felt like your life circumstances are unfair | 6 | 0 | 2.4 | 1.2 | 1.0 | 2.0 | 5.0 |  |
| Felt mistreated by others | 5 | 0 | 2.1 | 1.2 | 1.0 | 2.0 | 5.0 |  |
| Felt betrayed by people you care about | 5 | 0 | 2.1 | 1.1 | 1.0 | 2.0 | 5.0 |  |
| Felt you just could not get going | 6 | 18 | 3.0 | 1.3 | 1.0 | 3.0 | 5.0 |  |
| Felt everything was an effort | 6 | 18 | 2.6 | 1.0 | 1.0 | 3.0 | 5.0 |  |
| Felt lonely | 6 | 18 | 2.7 | 1.3 | 1.0 | 3.0 | 5.0 |  |
| Felt you could not shake the blues | 6 | 18 | 2.2 | 1.0 | 1.0 | 2.0 | 5.0 |  |
| Felt like your life circumstances are unfair | 6 | 18 | 2.9 | 1.2 | 1.0 | 3.0 | 5.0 |  |
| Felt mistreated by others | 6 | 18 | 2.4 | 1.1 | 1.0 | 2.0 | 5.0 |  |
| Felt betrayed by people you care about | 6 | 18 | 2.3 | 1.1 | 1.0 | 2.0 | 5.0 |  |
| What is the respondent's age? | 47 | 0 | 32.8 | 11.7 | 19.0 | 30.0 | 75.0 |  |
| How many years of formal education do you have? | 19 | 0 | 10.4 | 3.9 | 1.0 | 10.0 | 20.0 |  |
| IF YES TO 8a. How many children do you have? | 8 | 0 | 1.4 | 1.3 | 0.0 | 1.0 | 7.0 |  |
| area_id | 31 | 0 | 15.7 | 8.9 | 1.0 | 15.5 | 31.0 |  |
| female | 2 | 0 | 0.5 | 0.5 | 0.0 | 0.5 | 1.0 |  |
| rural | 2 | 0 | 0.4 | 0.5 | 0.0 | 0.0 | 1.0 |  |
| marriedw1 | 2 | 0 | 0.8 | 0.4 | 0.0 | 1.0 | 1.0 |  |
| L2sesw1 | 27 | 0 | 19.5 | 1.5 | 16.4 | 19.1 | 22.6 |  |
| osfinaw1 | 23 | 0 | 16.5 | 5.0 | 6.0 | 17.0 | 28.0 |  |
| osfinaw2 | 23 | 18 | 16.4 | 4.4 | 6.0 | 16.0 | 27.0 |  |
2.1.2.4 Pearson’s r matrix
stress_dat %>%
zap_label() %>%
dplyr::select(stmonyw1, sttranw1, strespw1, stfairw1, stjobw1, stthftw1, stmugw1,
stmonyw2, sttranw2, strespw2, stfairw2, stjobw2, stthftw2, stmugw2,
prjthflt5w1, prjthfgt5w1, prjthreatw1, prjharmw1, prjusedrgw1, prjhackw1,
prjthflt5w2, prjthfgt5w2, prjthreatw2, prjharmw2, prjusedrgw2, prjhackw2,
depcantgow1, depeffortw1, deplonelyw1, depbluesw1, depunfairw1, depmistrtw1, depbetrayw1,
depcantgow2, depeffortw2, deplonelyw2, depbluesw2, depunfairw2, depmistrtw2, depbetrayw2) %>%
datasummary_correlation(method="pearson")| stmonyw1 | sttranw1 | strespw1 | stfairw1 | stjobw1 | stthftw1 | stmugw1 | stmonyw2 | sttranw2 | strespw2 | stfairw2 | stjobw2 | stthftw2 | stmugw2 | prjthflt5w1 | prjthfgt5w1 | prjthreatw1 | prjharmw1 | prjusedrgw1 | prjhackw1 | prjthflt5w2 | prjthfgt5w2 | prjthreatw2 | prjharmw2 | prjusedrgw2 | prjhackw2 | depcantgow1 | depeffortw1 | deplonelyw1 | depbluesw1 | depunfairw1 | depmistrtw1 | depbetrayw1 | depcantgow2 | depeffortw2 | deplonelyw2 | depbluesw2 | depunfairw2 | depmistrtw2 | depbetrayw2 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| stmonyw1 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| sttranw1 | .64 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| strespw1 | .09 | .06 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stfairw1 | .13 | .09 | .87 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stjobw1 | .34 | .23 | .39 | .36 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stthftw1 | .10 | .14 | .24 | .29 | .17 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stmugw1 | .01 | .01 | .30 | .34 | .19 | .57 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stmonyw2 | .82 | .55 | .10 | .13 | .32 | .06 | .03 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| sttranw2 | .51 | .77 | .03 | .04 | .19 | .08 | .03 | .51 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| strespw2 | .11 | .04 | .86 | .79 | .38 | .18 | .28 | .13 | .05 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stfairw2 | .13 | .07 | .80 | .89 | .37 | .22 | .30 | .15 | .07 | .82 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stjobw2 | .30 | .20 | .40 | .35 | .88 | .10 | .16 | .29 | .20 | .40 | .38 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stthftw2 | .08 | .07 | .18 | .21 | .14 | .84 | .46 | .04 | .09 | .15 | .18 | .13 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| stmugw2 | .05 | .03 | .24 | .30 | .16 | .48 | .83 | .04 | .03 | .22 | .25 | .16 | .45 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjthflt5w1 | -.03 | -.05 | .09 | .12 | .10 | .04 | .03 | .00 | .04 | .14 | .14 | .16 | .08 | .02 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjthfgt5w1 | -.03 | .00 | .05 | .07 | .13 | .12 | .06 | -.03 | .04 | .11 | .09 | .15 | .17 | .03 | .61 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjthreatw1 | -.03 | .00 | .09 | .11 | .13 | .05 | .01 | -.05 | -.01 | .16 | .13 | .15 | .11 | .00 | .35 | .43 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjharmw1 | -.03 | -.02 | .03 | .04 | .06 | .00 | .04 | -.04 | -.03 | .08 | .04 | .08 | .05 | .04 | .21 | .31 | .51 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjusedrgw1 | -.03 | .00 | .10 | .07 | .11 | -.01 | .06 | .00 | -.02 | .13 | .11 | .12 | -.02 | .03 | .14 | .27 | .37 | .26 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjhackw1 | .05 | .02 | -.03 | -.01 | .09 | .00 | -.04 | .00 | .03 | .09 | .11 | .18 | .02 | -.01 | .14 | .23 | .30 | .12 | .14 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjthflt5w2 | -.01 | -.06 | .11 | .13 | .11 | .08 | .01 | .03 | -.02 | .11 | .15 | .10 | .07 | .00 | .71 | .52 | .28 | .23 | .16 | .16 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjthfgt5w2 | .01 | .01 | .11 | .10 | .13 | .09 | .00 | .05 | .01 | .10 | .11 | .13 | .11 | .01 | .57 | .64 | .33 | .26 | .15 | .14 | .82 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjthreatw2 | .02 | .04 | .13 | .12 | .12 | .13 | -.01 | .02 | .00 | .16 | .10 | .13 | .15 | -.01 | .25 | .33 | .67 | .36 | .20 | .25 | .27 | .33 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjharmw2 | -.02 | .04 | .09 | .09 | .08 | .09 | .03 | .00 | -.03 | .10 | .06 | .11 | .07 | .05 | .19 | .29 | .38 | .61 | .18 | .07 | .27 | .32 | .51 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjusedrgw2 | .02 | .07 | .14 | .09 | .14 | .02 | .03 | .04 | -.02 | .13 | .12 | .14 | .00 | .05 | .15 | .19 | .14 | .10 | .56 | .08 | .27 | .29 | .37 | .36 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| prjhackw2 | .00 | .01 | .06 | .04 | .08 | .06 | -.01 | .04 | -.02 | .07 | .04 | .12 | .02 | .04 | .03 | .05 | .01 | .03 | -.02 | .04 | .13 | .15 | .20 | .42 | .34 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . | . |
| depcantgow1 | .10 | .03 | .08 | .08 | .05 | .18 | .15 | .11 | .08 | .07 | .08 | -.01 | .11 | .13 | .00 | .01 | -.03 | -.03 | .00 | -.04 | -.01 | -.01 | -.05 | .00 | -.08 | -.01 | 1 | . | . | . | . | . | . | . | . | . | . | . | . | . |
| depeffortw1 | .11 | .07 | .19 | .16 | .15 | .17 | .22 | .05 | .05 | .14 | .13 | .15 | .12 | .21 | .00 | .03 | .00 | -.03 | -.01 | -.03 | -.01 | .00 | -.01 | .02 | .00 | .05 | .35 | 1 | . | . | . | . | . | . | . | . | . | . | . | . |
| deplonelyw1 | .07 | .03 | .17 | .17 | .03 | .14 | .17 | .05 | .04 | .19 | .20 | .04 | .12 | .16 | .04 | .08 | .02 | -.03 | .10 | -.02 | .07 | .04 | .04 | .01 | .11 | .05 | .19 | .22 | 1 | . | . | . | . | . | . | . | . | . | . | . |
| depbluesw1 | .10 | .07 | .11 | .13 | .08 | .15 | .12 | .09 | .07 | .02 | .05 | .08 | .07 | .10 | -.01 | -.02 | -.02 | .03 | .04 | -.03 | -.06 | -.10 | -.05 | .00 | -.03 | .05 | .13 | .16 | .26 | 1 | . | . | . | . | . | . | . | . | . | . |
| depunfairw1 | .24 | .17 | .06 | .07 | .12 | .11 | .12 | .14 | .17 | .04 | .07 | .11 | .06 | .11 | .05 | -.01 | .09 | .02 | .04 | -.01 | .05 | .00 | .06 | .09 | .03 | .01 | .15 | .28 | .22 | .26 | 1 | . | . | . | . | . | . | . | . | . |
| depmistrtw1 | .07 | .00 | .26 | .29 | .07 | .26 | .37 | .05 | .00 | .25 | .27 | .07 | .23 | .39 | .01 | -.01 | .00 | -.02 | .03 | -.07 | -.02 | -.01 | -.01 | -.02 | .02 | .01 | .13 | .17 | .28 | .25 | .25 | 1 | . | . | . | . | . | . | . | . |
| depbetrayw1 | .09 | .04 | .30 | .33 | .15 | .31 | .43 | .11 | .04 | .30 | .33 | .17 | .28 | .41 | .07 | .08 | .12 | .07 | .15 | .03 | .12 | .12 | .09 | .08 | .14 | .05 | .18 | .20 | .26 | .12 | .16 | .64 | 1 | . | . | . | . | . | . | . |
| depcantgow2 | -.05 | -.06 | -.13 | -.10 | -.11 | .06 | .06 | .00 | .00 | -.11 | -.08 | -.08 | .08 | .08 | .00 | .01 | -.09 | -.03 | -.09 | -.11 | -.01 | .01 | -.05 | .02 | -.05 | .02 | .15 | .07 | .07 | .02 | -.02 | -.01 | -.03 | 1 | . | . | . | . | . | . |
| depeffortw2 | .00 | -.03 | -.01 | -.02 | .01 | .02 | .02 | .00 | .01 | -.01 | -.01 | .05 | .05 | .04 | .02 | .00 | -.09 | -.04 | -.04 | -.09 | .01 | .02 | -.06 | .01 | -.01 | .05 | .03 | .12 | .08 | .06 | .08 | .09 | .01 | .45 | 1 | . | . | . | . | . |
| deplonelyw2 | -.08 | -.04 | -.04 | -.07 | .03 | -.06 | .01 | -.08 | -.05 | -.04 | -.07 | .03 | -.04 | .05 | -.03 | .05 | .03 | .00 | .10 | .07 | -.04 | .00 | .03 | .01 | .12 | .06 | -.04 | .05 | .15 | -.03 | .01 | .01 | .00 | .26 | .29 | 1 | . | . | . | . |
| depbluesw2 | .00 | -.05 | .01 | .01 | .00 | -.04 | .02 | -.03 | -.06 | -.05 | -.07 | .00 | -.01 | .05 | .00 | -.01 | .05 | .09 | .01 | .00 | .03 | .04 | .01 | .01 | .00 | .05 | -.02 | .08 | .08 | .06 | .07 | .06 | .06 | .13 | .24 | .36 | 1 | . | . | . |
| depunfairw2 | .04 | .01 | .05 | .02 | .13 | .04 | .07 | .04 | .08 | .11 | .10 | .19 | .10 | .05 | .12 | .12 | .10 | .04 | .13 | .05 | .13 | .14 | .03 | .03 | .09 | .05 | .01 | .13 | .07 | .01 | .11 | .10 | .15 | .06 | .30 | .22 | .27 | 1 | . | . |
| depmistrtw2 | .05 | .02 | .06 | .09 | .05 | .04 | -.04 | .06 | .08 | .04 | .07 | .08 | .09 | .02 | .09 | .08 | .05 | .03 | -.04 | -.02 | .07 | .07 | .09 | .11 | .05 | .14 | .04 | .09 | -.03 | .02 | .01 | .02 | .10 | .10 | .22 | .13 | .23 | .38 | 1 | . |
| depbetrayw2 | .10 | .07 | .11 | .13 | .13 | .10 | .04 | .12 | .08 | .12 | .11 | .16 | .14 | .07 | .12 | .12 | .15 | .10 | .06 | .13 | .14 | .16 | .17 | .15 | .11 | .14 | -.01 | .11 | .07 | .09 | .07 | .08 | .17 | .05 | .20 | .13 | .28 | .34 | .54 | 1 |
Let’s also take a look at some missingness patterns due to attrition from T1 to T2.
2.1.3 Examine Missingness due to Attrition (T1 to T2)
missdf <- stress_dat %>%
dplyr::select(centerw2, stmonyw1, sttranw1, strespw1, stfairw1, stjobw1, stthftw1,
stmugw1, prjthflt5w1, prjthfgt5w1, prjthreatw1,
prjharmw1, prjusedrgw1, prjhackw1, depcantgow1, depeffortw1,
deplonelyw1, depbluesw1, depunfairw1, depmistrtw1, depbetrayw1)
shadowdat <- bind_shadow(missdf)
#missing plot function
missplot <- function(df, x_var, y_var) {
ggplot(df, aes(x = .data[[x_var]], colour = .data[[y_var]])) +
geom_density()
}
# list approach did not work
# plot_list <- colnames(shadowdat)[-1] %>%
# map( ~ missplot(shadowdat, colnames(shadowdat)[1], .x))
mp1 <- missplot(shadowdat, "stmonyw1", "centerw2_NA")
mp2 <- missplot(shadowdat, "sttranw1", "centerw2_NA")
mp3 <- missplot(shadowdat, "strespw1", "centerw2_NA")
mp4 <- missplot(shadowdat, "stfairw1", "centerw2_NA")
mp5 <- missplot(shadowdat, "stjobw1", "centerw2_NA")
mp6 <- missplot(shadowdat, "stthftw1", "centerw2_NA")
mp7 <- missplot(shadowdat, "stmugw1", "centerw2_NA")
mp8 <- missplot(shadowdat, "prjthflt5w1", "centerw2_NA")
mp9 <- missplot(shadowdat, "prjthfgt5w1", "centerw2_NA")
mp10 <- missplot(shadowdat, "prjthreatw1", "centerw2_NA")
mp11 <- missplot(shadowdat, "prjharmw1", "centerw2_NA")
mp12 <- missplot(shadowdat, "prjusedrgw1", "centerw2_NA")
mp13 <- missplot(shadowdat, "prjhackw1", "centerw2_NA")
mp14 <- missplot(shadowdat, "depcantgow1", "centerw2_NA")
mp15 <- missplot(shadowdat, "depeffortw1", "centerw2_NA")
mp16 <- missplot(shadowdat, "deplonelyw1", "centerw2_NA")
mp17 <- missplot(shadowdat, "depbluesw1", "centerw2_NA")
mp18 <- missplot(shadowdat, "depunfairw1", "centerw2_NA")
mp19 <- missplot(shadowdat, "depmistrtw1", "centerw2_NA")
mp20 <- missplot(shadowdat, "depbetrayw1", "centerw2_NA")2.1.3.1 Stress x attrition
mpstress <- (mp1 + mp2) / (mp3 + mp4) / (mp5 + guide_area()) / (mp6 + mp7) +
plot_layout(guides = 'collect', widths = c(4,.6)) +
plot_annotation(
title = 'Stress by Attrition')
mpstress
2.1.3.2 Crim intent x attrition
mpprj <- (mp8 + mp9) / (mp10 + mp11) / (mp12 + mp13) / guide_area() +
plot_layout(guides = 'collect', widths = c(4,.6)) +
plot_annotation(
title = 'Crim intent by Attrition')
mpprj
2.1.3.3 Neg emotions x attrition
mpdep <- (mp14 + mp15) / (mp16 + mp17) / (mp18 + mp19) / (mp20 + guide_area()) +
plot_layout(guides = 'collect', widths = c(4,.6)) +
plot_annotation(
title = 'Crim intent by Attrition')
mpdep
2.1.3.4 Item means x attrition
varsw1 <- c("stmonyw1", "sttranw1", "strespw1", "stfairw1", "stjobw1", "stthftw1",
"stmugw1", "pstthflt5w1", "pstthfgt5w1", "pstthreatw1", "pstharmw1",
"pstusedrgw1", "psthackw1", "prjthflt5w1", "prjthfgt5w1", "prjthreatw1",
"prjharmw1", "prjusedrgw1", "prjhackw1", "depcantgow1", "depeffortw1",
"deplonelyw1", "depbluesw1", "depunfairw1", "depmistrtw1", "depbetrayw1")
stress_dat %>%
bind_shadow() %>%
group_by(centerw2_NA) %>%
summarise_at(.vars = varsw1,
.funs = c("mean"),
na.rm = TRUE) %>%
gt()| centerw2_NA | stmonyw1 | sttranw1 | strespw1 | stfairw1 | stjobw1 | stthftw1 | stmugw1 | pstthflt5w1 | pstthfgt5w1 | pstthreatw1 | pstharmw1 | pstusedrgw1 | psthackw1 | prjthflt5w1 | prjthfgt5w1 | prjthreatw1 | prjharmw1 | prjusedrgw1 | prjhackw1 | depcantgow1 | depeffortw1 | deplonelyw1 | depbluesw1 | depunfairw1 | depmistrtw1 | depbetrayw1 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| !NA | 3.15 | 3.16 | 3.31 | 3.4 | 3.27 | 2.22 | 1.96 | 1.20 | 1.16 | 1.07 | 1.10 | 1.11 | 1.06 | 1.14 | 1.11 | 1.08 | 1.06 | 1.06 | 1.04 | 3.0 | 2.48 | 2.52 | 2.50 | 2.41 | 2.15 | 2.17 |
| NA | 3.24 | 3.36 | 3.41 | 3.5 | 3.23 | 2.50 | 1.92 | 1.24 | 1.14 | 1.01 | 1.05 | 1.05 | 1.06 | 1.14 | 1.09 | 1.02 | 1.04 | 1.04 | 1.06 | 2.9 | 2.49 | 2.49 | 2.41 | 2.24 | 1.91 | 1.77 |
With those basic data exploration tasks out of the way, we can begin answering our research questions. First, we need to do a bit of data wrangling.
2.1.4 Data wrangling
stress.wide <- stress_dat %>%
dplyr::filter(centerw2==1) %>% #retain n=489/600 Rs in both waves
dplyr::select(-c(centerw1, centerw2)) %>%
mutate( #standardize community (L2) SES
L2sesw1z = (L2sesw1 - mean(L2sesw1))/sd(L2sesw1)
)
# freq(stress.wide)
#summarize median community (L2) SES value (-.268),
#save as vector (value) to create community groups
L2SESw1zmed <- as_vector(stress.wide %>%
summarise_at(c("L2sesw1z"), median))
# L2SESw1zmed
#create community-level rural/urban + median-split L2 SES factor variable,
#dichotomize crime outcomes
stress.wide <- stress.wide %>%
mutate(
#Create rural/ses catetories:
#split communities based on mean ses
rural.ses.avg = 0,
rural.ses.avg = ifelse(rural == 1 & L2sesw1z <= 0, 1,
rural.ses.avg), #Rural/Low SES
rural.ses.avg = ifelse(rural == 1 & L2sesw1z > 0, 2,
rural.ses.avg), #Rural/High SES
rural.ses.avg = ifelse(rural == 0 & L2sesw1z <= 0, 3,
rural.ses.avg), #Urban/Low SES
rural.ses.avg = ifelse(rural == 0 & L2sesw1z > 0, 4,
rural.ses.avg), #Urban/High SES
rural.ses.avg = factor(rural.ses.avg, levels = c(1, 2, 3, 4)),
#split communities based on median ses
rural.ses.med = 0,
rural.ses.med = ifelse(rural == 1 & L2sesw1z <= L2SESw1zmed, 1,
rural.ses.med), #Rural/Low SES
rural.ses.med = ifelse(rural == 1 & L2sesw1z > L2SESw1zmed, 2,
rural.ses.med), #Rural/High SES
rural.ses.med = ifelse(rural == 0 & L2sesw1z <= L2SESw1zmed, 3,
rural.ses.med), #Urban/Low SES
rural.ses.med = ifelse(rural == 0 & L2sesw1z > L2SESw1zmed, 4,
rural.ses.med), #Urban/High SES
rural.ses.med = factor(rural.ses.med, levels = c(1, 2, 3, 4)),
stmonych12 = stmonyw2 - stmonyw1, #stress change vars (for plotting descriptives)
sttranch12 = sttranw2 - sttranw1,
strespch12 = strespw2 - strespw1,
stfairch12 = stfairw2 - stfairw1,
stjobch12 = stjobw2 - stjobw1,
stthftch12 = stthftw2 - stthftw1,
stmugch12 = stmugw2 - stmugw1
) %>%
rename(
pstthflt5w1di = pstthflt5w1, #rename past crime w1 items (binary)
pstthfgt5w1di = pstthfgt5w1,
pstthreatw1di = pstthreatw1,
pstharmw1di = pstharmw1,
pstusedrgw1di = pstusedrgw1,
psthackw1di = psthackw1,
pstthflt5w2di = pstthflt5w2, #rename past crime w2 items (binary)
pstthfgt5w2di = pstthfgt5w2,
pstthreatw2di = pstthreatw2,
pstharmw2di = pstharmw2,
pstusedrgw2di = pstusedrgw2,
psthackw2di = psthackw2,
prjthflt5w1di = prjthflt5w1, #rename criminal intent w1 items (binary)
prjthfgt5w1di = prjthfgt5w1,
prjthreatw1di = prjthreatw1,
prjharmw1di = prjharmw1,
prjusedrgw1di = prjusedrgw1,
prjhackw1di = prjhackw1,
prjthflt5w2di = prjthflt5w2, #rename criminal intent w2 items (binary)
prjthfgt5w2di = prjthfgt5w2,
prjthreatw2di = prjthreatw2,
prjharmw2di = prjharmw2,
prjusedrgw2di = prjusedrgw2,
prjhackw2di = prjhackw2
) %>%
mutate_at(vars(c( #dichotomize crime outcomes
pstthflt5w1di, pstthfgt5w1di, pstthreatw1di,
pstharmw1di, pstusedrgw1di, psthackw1di,
pstthflt5w2di, pstthfgt5w2di, pstthreatw2di,
pstharmw2di, pstusedrgw2di, psthackw2di,
prjthflt5w1di, prjthfgt5w1di, prjthreatw1di,
prjharmw1di, prjusedrgw1di, prjhackw1di,
prjthflt5w2di, prjthfgt5w2di, prjthreatw2di,
prjharmw2di, prjusedrgw2di, prjhackw2di)),
list(~ if_else(. > 1, 1, 0))
)
# freq(stress.wide)
options(scipen = 999, digits = 2) #change global option back to two decimals
#head(stress.wide)
#view(stress.wide)# table(stress.wide$rural.ses.med, stress.wide$rural)Now that we have read in, explored, and begun wrangling the data, we can start answering our research questions.
2.2 RQ1a: Stress Prevalance by Wave (Full Sample)
How often did participants report stressing or worrying in T1 and T2 - overall and specifically about financial, relational, occupational, or victimization issues?
As in our “Stress in Bangladesh” paper, we start by displaying the distributions of all seven subjective stress items at T1 and T2, as well as within-person change distributions, in Figure 1 below.
#Begin by transforming all 14 stress items (7 per wave, two waves) into new factor variables
#Create vector for variables I'm turning into factors:
stress_vars <- c("stmonyw1", "stmonyw2", "sttranw1", "sttranw2", "strespw1", "strespw2", "stfairw1","stfairw2",
"stjobw1", "stjobw2", "stthftw1", "stthftw2", "stmugw1", "stmugw2")
stress_vars_fct <- c("stmonyw1_fct", "stmonyw2_fct", "sttranw1_fct", "sttranw2_fct", "strespw1_fct", "strespw2_fct", "stfairw1_fct","stfairw2_fct",
"stjobw1_fct", "stjobw2_fct", "stthftw1_fct", "stthftw2_fct", "stmugw1_fct", "stmugw2_fct")
stress_vars_levels = c( "Never (1)" = "1",
"Rarely (2)" = "2",
"Sometimes (3)" = "3",
"Often (4)" = "4",
"Very often (5)" = "5")
stress.wide <- stress.wide %>%
mutate(stmonyw1_fct = fct_recode(as.factor(stmonyw1), !!!stress_vars_levels),
stmonyw2_fct = fct_recode(as.factor(stmonyw2), !!!stress_vars_levels),
sttranw1_fct = fct_recode(as.factor(sttranw1), !!!stress_vars_levels),
sttranw2_fct = fct_recode(as.factor(sttranw2), !!!stress_vars_levels),
strespw1_fct = fct_recode(as.factor(strespw1), !!!stress_vars_levels),
strespw2_fct = fct_recode(as.factor(strespw2), !!!stress_vars_levels),
stfairw1_fct = fct_recode(as.factor(stfairw1), !!!stress_vars_levels),
stfairw2_fct = fct_recode(as.factor(stfairw2), !!!stress_vars_levels),
stjobw1_fct = fct_recode(as.factor(stjobw1), !!!stress_vars_levels),
stjobw2_fct = fct_recode(as.factor(stjobw2), !!!stress_vars_levels),
stthftw1_fct = fct_recode(as.factor(stthftw1), !!!stress_vars_levels),
stthftw2_fct = fct_recode(as.factor(stthftw2), !!!stress_vars_levels),
stmugw1_fct = fct_recode(as.factor(stmugw1), !!!stress_vars_levels),
stmugw2_fct = fct_recode(as.factor(stmugw2), !!!stress_vars_levels))
#Wrangle data to get both T1 & T2 onto same figure
tab1dataw1 <- stress.wide %>%
dplyr::select(id, stmonyw1, sttranw1, strespw1, stfairw1, stjobw1, stthftw1, stmugw1) %>%
dplyr::mutate(wave = "Time 1")
# tab1dataw1
tab1dataw2 <- stress.wide %>%
dplyr::select(id, stmonyw2, sttranw2, strespw2, stfairw2, stjobw2, stthftw2, stmugw2) %>%
dplyr::mutate(wave = "Time 2")
# tab1dataw2
items_vars_levels = c("stmonyw1", "sttranw1", "strespw1", "stfairw1", "stjobw1", "stthftw1", "stmugw1")
items_vars_labels = c("Money for\nnecessities" = "stmonyw1",
"Reliable\ntransportation" = "sttranw1",
"Being treated\nwith respect" = "strespw1",
"Being treated\nfairly" = "stfairw1",
"Finding job\nyou enjoy" = "stjobw1",
"Others stealing\nfrom you" = "stthftw1",
"Others\nassaulting you" = "stmugw1")
items_vars_wave1 = c("Money for necessities, T1" = "stmonyw1",
"Reliable transportation, T1" = "sttranw1",
"Being treated w/respect, T1" = "strespw1",
"Being treated fairly, T1" = "stfairw1",
"Finding job you enjoy, T1" = "stjobw1",
"Others stealing from you, T1" = "stthftw1",
"Others assaulting you, T1" = "stmugw1")
items_vars_levelsw2 = c("stmonyw2", "sttranw2", "strespw2", "stfairw2", "stjobw2", "stthftw2", "stmugw2")
items_vars_labelsw2 = c("Money for\nnecessities" = "stmonyw2",
"Reliable\ntransportation" = "sttranw2",
"Being treated\nwith respect" = "strespw2",
"Being treated\nfairly" = "stfairw2",
"Finding job\nyou enjoy" = "stjobw2",
"Others stealing\nfrom you" = "stthftw2",
"Others\nassaulting you" = "stmugw2")
items_vars_wave2 = c("Money for necessities, T2" = "stmonyw2",
"Reliable transportation, T2" = "sttranw2",
"Being treated w/respect, T2" = "strespw2",
"Being treated fairly, T2" = "stfairw2",
"Finding job you enjoy, T2" = "stjobw2",
"Others stealing from you, T2" = "stthftw2",
"Others assaulting you, T2" = "stmugw2")
#pivot data and combine
#Wave 1
tab1datalgw1 <- tab1dataw1 %>%
pivot_longer(cols = c("stmonyw1", "sttranw1", "strespw1", "stfairw1", "stjobw1", "stthftw1", "stmugw1"),
names_to = "items", values_to = "response") %>%
mutate(items = factor(items, levels = items_vars_levels),
items_fct = fct_recode(items, !!!items_vars_labels),
items_fct_wave = fct_recode(items, !!!items_vars_wave1),
response_fct = fct_recode(as.factor(response), !!!stress_vars_levels))
# levels(tab1datalgw1$items)
# levels(tab1datalgw1$items_fct)
# levels(tab1datalgw1$items_fct_wave)
# levels(tab1datalgw1$response_fct)
#calculate mean and standard error and sd for each variable
tab1datalgw1_sum <- tab1datalgw1 %>%
dplyr::group_by(items) %>%
dplyr::summarize(mean_se(response), sd = sd(response)) %>%
rename(mean = y,
min95 = ymin,
max95 = ymax)
#merge summary data with individual data
tab1datalgw1 <- full_join(tab1datalgw1, tab1datalgw1_sum)
#Wave 2
tab1datalgw2 <- tab1dataw2 %>%
pivot_longer(cols = c("stmonyw2", "sttranw2", "strespw2", "stfairw2", "stjobw2", "stthftw2", "stmugw2"),
names_to = "items", values_to = "response") %>%
mutate(items = factor(items, levels = items_vars_levelsw2),
items_fct = fct_recode(items, !!!items_vars_labelsw2),
items_fct_wave = fct_recode(items, !!!items_vars_wave2),
response_fct = fct_recode(as.factor(response), !!!stress_vars_levels))
# levels(tab1datalgw2$items)
# levels(tab1datalgw2$items_fct)
# levels(tab1datalgw2$items_fct_wave)
# levels(tab1datalgw2$response_fct)
#calculate mean and standard error and sd for each variable
tab1datalgw2_sum <- tab1datalgw2 %>%
dplyr::group_by(items) %>%
dplyr::summarize(mean_se(response), sd = sd(response)) %>%
rename(mean = y,
min95 = ymin,
max95 = ymax)
#merge summary data with individual data
tab1datalgw2 <- full_join(tab1datalgw2, tab1datalgw2_sum)
tab1datalg <- bind_rows(tab1datalgw1, tab1datalgw2) %>%
mutate(wave = as.factor(wave)) %>%
arrange(id, wave)
# head(tab1datalg)
# levels(tab1datalg$items)
# levels(tab1datalg$items_fct)
# levels(tab1datalg$items_fct_wave)
# levels(tab1datalg$response_fct)2.2.1 FIGURE 1: Subj Stress by Wave
#NOTE: Cannot properly align italicized expression as below - consider ggtext package w/html tags instead
# expression(italic("Note: N") ~ "=489 respondents participating at both survey waves. \nInterval plots display item mean (horizontal dash), 50% (thick vertical bar) and 95% (thin vertical bar) intervals. \nBar charts display proportion of full sample reporting each item response category.")
# E99D53 "#883E3A"
Fig1 <- tab1datalg %>%
ggplot() +
geom_bar(aes(y = response_fct, fill = wave), width = .7, stat="count") +
scale_fill_manual(values = c("#E99D53", "#E0754F")) +
facet_grid(rows = vars(items_fct), cols = vars(wave), switch = "both") + #switch y-axis strip titles to left side of plot
scale_color_manual(values = c("#E99D53", "#E0754F")) +
stat_pointinterval(data = tab1datalg, aes(y = response, color = wave), .width = c(0.5, 0.95),
point_interval = mean_qi, show_point = TRUE, shape=3,
position = position_nudge(x = -15)) +
# labs(title = "FIGURE 1: Subjective Stress Item Distributions\n ",
# subtitle = "How often do you stress or worry about...",
# caption = "Note: N=489 respondents participating at both survey waves. Interval plots show item mean (horizontal \ntick) and 50% (thick vertical bar) and 95% (thin vertical bar) intervals. Bars display proportion \nof full sample reporting each response category, from 'never' (bottom bar) to 'very often' (top bar)."
# ) +
theme(axis.title = element_blank(), #removes axis titles
panel.grid=element_blank(), #removes grid lines
# plot.title.position = "plot", #changes default title location from graph to whole plot
# plot.title = element_text(face = "bold", size = 12),
# plot.subtitle = element_text(face = "italic", size = 11),
# plot.caption.position = "plot",
# plot.caption = element_text(face = "plain", size = 10, hjust=0, vjust=0),
legend.position = "none",
legend.title = element_blank(),
axis.line = element_blank(),
axis.ticks.x = element_line(),
axis.text.x = element_text(size = 7, angle = 0),
#strip.text.x = element_blank(), #remove x-axis strip titles (Time 1 and Time 2)
axis.text.y = element_text(size = 7, angle = 0, hjust = 1),
strip.text.y.left = element_text(angle = 0, hjust=0),
strip.placement = "outside",
strip.background = element_blank(),
# strip.background = element_rect(fill="white"),
axis.title.x = element_blank(),
axis.title.y = element_blank()) +
scale_x_continuous(limit=c(-16,220), breaks=c("0"=0,".1"=0.1*489,".2"=0.2*489,".3"=0.3*489,".4"=0.4*489))
# Fig1
# Add change distributions to Fig1
stchgitems <- c("stmonych12", "sttranch12", "strespch12", "stfairch12",
"stjobch12", "stthftch12", "stmugch12")
stress_chg_levels = c("Decrease (-2+)"= "-2",
"Decrease (-1)" = "-1",
"No change (0)" = "0",
"Increase (+1)" = "1",
"Increase (+2+)" = "2")
tab1chgdata <- stress.wide %>%
dplyr::select(id, stmonych12, sttranch12, strespch12, stfairch12, stjobch12, stthftch12, stmugch12) %>%
mutate(wave = "T2-T1") %>%
pivot_longer(cols = c("stmonych12", "sttranch12", "strespch12", "stfairch12", "stjobch12", "stthftch12", "stmugch12"),
names_to = "items", values_to = "response") %>%
mutate(items = factor(items, levels = stchgitems),
items_fct = factor(items),
response = if_else(response >2, 2, response),
response = if_else(response < -2, -2, response),
response_fct = fct_recode(as.factor(response), !!!stress_chg_levels))
#calculate mean and standard error and sd for each variable
tab1chgdata_sum <- tab1chgdata %>%
dplyr::group_by(items) %>%
dplyr::summarize(mean_se(response), sd = sd(response)) %>%
rename(mean = y,
min95 = ymin,
max95 = ymax)
#merge summary data with individual data
tab1chgdatalg <- full_join(tab1chgdata, tab1chgdata_sum)
Fig1chg <- tab1chgdatalg %>%
ggplot() +
# facet_wrap(~items_fct, nrow = 7, strip.position = "left") +
geom_bar(aes(y = response, fill = wave), width = .7, stat="count") +
facet_grid(rows = vars(items_fct), cols = vars(wave), switch = "both") +
scale_fill_manual(values = "#883E3A") +
scale_color_manual(values = "#883E3A") +
scale_y_continuous(breaks=c(-2,-1,0,1,2),
labels=c("Decrease (-2+)", "Decrease (-1)", "No Change (0)",
"Increase (+1)", "Increase (+2+)")) +
stat_pointinterval(data = tab1chgdatalg, aes(y = response, color = wave), .width = c(0.5, 0.95),
point_interval = mean_qi, show_point = TRUE, shape=3,
position = position_nudge(x = -15)) +
theme(axis.title.y = element_blank(), #removes axis titles
strip.text.y = element_blank(),
axis.title.x = element_blank(),
panel.grid=element_blank(), #removes grid lines
legend.position = "none",
legend.title = element_blank(),
axis.line = element_blank(),
axis.text.y = element_text(size = 7, angle = 0),
axis.ticks.x = element_line(),
axis.text.x = element_text(size = 7, angle = 0)) +
scale_x_continuous(limit=c(-16,350), breaks=c("0"=0,".1"=0.1*489,".2"=0.2*489,
".3"=0.3*489,".4"=0.4*489,
".5"=0.5*489, ".6"=0.6*489,
".7"=0.7*489))
# Fig1chg
design <- "
112
112
"
# labs(title = "FIGURE 1: Subjective Stress Item Distributions\n ",
# subtitle = "How often do you stress or worry about...",
# caption = "Note: N=489 respondents participating at both survey waves. Interval plots show item mean (horizontal \ntick) and 50% (thick vertical bar) and 95% (thin vertical bar) intervals. Bars display proportion \nof full sample reporting each response category, from 'never' (bottom bar) to 'very often' (top bar)."
# ) +
Figure1 <- Fig1 + Fig1chg +
plot_layout(design=design) +
plot_annotation(
title = 'FIGURE 1 Subjective Stress Item Distributions',
subtitle = 'How often do you stress or worry about...',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Interval plots show item mean (horizontal tick), 50% (thick vertical bar), and 95% (thin vertical bar) intervals. Bars display proportion of full sample reporting each item response category (Time 1; Time 2) or degrees of change in item response categories (T2-T1).', width=140)) &
theme(plot.title = element_text(size=12, face="bold"),
plot.subtitle = element_text(face="italic"),
plot.caption = element_text(size=8, hjust = 0)) #move caption to left of plot
#Export to image
ggsave("Figure1.jpeg", Figure1, width=6.5, height=9, path=here("Output"))
Figure1
#Grab Figure 1 plot data for specific counts/props/percents to use in text
plt_b <- ggplot_build(Fig1)
fig1data <- as.data.frame(plt_b$data[[1]]) %>%
mutate(ynew=as.integer(y))
# fig1data
# glimpse(fig1data)
#Write function to calculate specific response category percentages
panel_catpct <- function(figpanel, ycatvalue) {
paneldata <- as_tibble(fig1data %>% filter(PANEL==figpanel))
catcounts <- as.numeric(paneldata %>% filter(y %in% ycatvalue) %>% .$x)
catpct <- ((sum(catcounts))/489)*100
return(catpct)
}
#First value input is Fig 1 panel of desired stress item (top left = 1, top right = 2, second left = 3, etc.)
#Second value/vector is desired response category(ies) (1 = never, 2 = rarely, 3 = sometimes, etc.)
#Example: Percent reporting *sometimes* (cat=3) or *often* (cat=4) stress abt money
#panel_catpct(1,c(3,4))As Figure 1 shows, the observed response patterns for the first two financial stress items are approximately normally distributed in both waves, with the majority of respondents reporting somewhat or often stressing about money (T1: 61.55%; T2: 62.78%) or transportation (T1: 58.9%; T2: 65.64%) and relatively few reportedly never stressing about these things (range: 3.07% to 7.77%). In contrast, the observed item distributions for relational and job-related stress items do not exhibit symmetric decay about the midpoint characteristic of normal distributions; rather, approximately half of our Bangladeshi respondents reportedly often or very often stress about being treated with respect (T1: 52.15%; T2: 52.76%), being treated fairly (T1: 55.83%; T2: 55.42%), and finding a job they enjoy (T1: 49.28%; T2: 51.12%), whereas a sizeable proportion of respondents report never experiencing these types of stress (range: 8.79% to 16.36%). Finally, response distributions for subjective stress about criminal victimization exhibit high positive skew, with the majority of respondents reporting never or rarely stressing about others stealing from them (T1: 63.39%; T2: 62.17%) or assaulting them (T1: 71.78%; T2: 69.94%). In contrast, very few respondents report very often stressing about such victimization (range: 2.04% to 5.11%), which might be expected given the relative rarity of crime and criminal victimization. Of course, frequency and potency of stress are distinct characteristics, and general strain theory suggests that victimization-related stress may be especially potent and criminogenic despite being comparatively rare.
save(stress.wide, file = here("1_Data_Files/Datasets/stress_wide.Rdata"))3 Figure 2: Stress by community (RQ1B)
(RMD FILE: BDK_2023_Stress_2_Fig2)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE3.1 RQ1B: Social Distributions of Stress by Community (T1)
RQ1B (Social distributions): Do these levels of subjective stress vary systematically across rural/urban communities with high/low aggregate SES? Specifically, do residents of low-SES urban communities report more frequently stressing or worrying about these various potential sources of stress?
3.1.1 From observations to models of subjective stress ordinal responses
Instead of plotting observed response variable distributions to visualize differences in stress prevalence across communities, we will recover these response distributions from posterior estimates (medians) generated by Bayesian sequential ordinal probit models for each stress item. While we could simply plot the observed frequencies for each community group as we did above by wave for the full sample, recovering these estimates from Bayesian posterior prediction distributions will serve a couple useful purposes. First, recall this is a probability sample of residents in the Dhaka metropolitan region of Bangladesh. A (Bayesian, in this case) statistical model permits us to visualize and communicate uncertainty surrounding these observed proportions as estimates of population parameters, such as by plotting 50% and 95% uncertainty (e.g., posterior HDI or credible) intervals around the estimated proportions for every response category for a given item.
Second, ultimately we wish to estimate the association between stress and its posited outcomes (criminal intent; depressive symptoms) by including the stress items as predictors in outcome models. Typically, researchers simply include ordinal items or ordered composite scales as predictor variables in regression models and treat them as metric or continuous normal variables without much thought given to their distributional properties. Yet, such practices can have disastrous and often under-appreciated consequences for model inferences (see Liddell & Kruschke 2017). Instead, by building ordinal models predicting response category probabilities for each stress item, we can generate ordinal probit threshold estimates that can subsequently be used in outcome regression models to more accurately estimate, for instance, the average change in an outcome like criminal behavior or depressive symptoms that is associated with a one-unit increase in a stress item’s ordered response categories (e.g., from never to rarely; see … brms paper). Hence, building ordinal descriptive models of our stress item distributions that accurately recover observed data is an essential first step towards building outcome models that accurately estimate associations between those outcomes and their ordinal predictors.
We estimate both cumulative and category-specific threshold models for all seven stress items, then use the threshold estimates (and uncertainty from posterior distributions) from each of the seven category-specific models to plot the stress response distributions (and uncertainty bounds) by community in Figure 2. Most of this section is devoted to exploring the appropriateness of these modeling decisions, including comparisons to standard linear models, posterior predictive checks, and conditional effects plots. If you wish, you can skip this and jump directly to the stress distribtions in Figure 2.
3.1.2 Comparing Metric & Ordinal Models of Subjective Stress
Now, we turn to visualizations of stress patterns across rural/urban and low/high SES communities. To do this, we estimate both cumulative and category-specific models for all seven stress items, then used the threshold estimates (and uncertainty from posterior distributions) from each of the seven category-specific models to plot the stress response distributions (and uncertainty bounds) by community group in Figure 2. In our “Stress in Bangladesh” paper, we noted that the supplemenary materials contain more information about the appropriateness of these modeling decisions, including comparisons to standard linear models, posterior predictive checks, and conditional effects plots. This section contains that supplementary information for the models used to analyze RQ1 (descriptive distributions for stress items).
Before generating models that recovered stress item response distributions for each community group, we want to briefly illustrate how simple descriptive models that impose metric normality assumptions on ordinal stress items - as is the case in a typical regression model with ordinal items entered as continuous predictors - is a poor fit for the observed data.
We then show that the quite parsimonious cumulative probit model (which is used in regression models for RQ2 and beyond) does a much better job of recovering the observed stress item response distributions in the entire sample and in each community group. From there, we generate plots of stress item response distributions across community groups from estimates generated in more complex sequential ordinal models that specify category-specific effects of each rural/SES community subgroup on stress item response thresholds. These more complex models generate threshold estimates that accurately recover observed data distributions in the entire sample as well as in each community group, thereby ensuring that our model-based summaries of the stress item distributions across groups is consistent with observed item distributions, which is essential to accurately answer RQ1.
After answering questions about community variations in levels and sources of stress, we will later return to simpler cumulative probit models of stress items in assessing stress-outcome associations. Cumulative probit models are reasonably accurate and quite parsimonious, whereas the large number of threshold parameters estimated in more complex sequential category-specific models can pose problems when using relatively small samples to predict rare event outcomes like crime.
3.1.3 (Not-so-)Normal models: Stress abt money & assault
We begin with the Bayesian equivalent of simple means or intercept-only model. As implied by the name, the intercept-only model for an individual stress in the full sample estimates only one parameter - the intercept or mean value of overall stress. In a classical (frequentist) model, this estimate of the mean stress level would be computed directly from the data or likelihood. In a Bayesian model, this estimate represents the center of a posterior sample of estimates that each are generated by combining the likelihood (e.g., observed mean in the data) with a prior distribution of plausible values for the parameter in question. In constructing a Bayesian model, one might specify a highly diffuse or “flat” prior distribution(s), which would result in posterior estimates that are generated primarily by the likelihood; with diffuse priors, Bayesian model estimates generally will converge with estimates generated from comparable classical statistical models. Alternatively, one might specify stronger or “more informative” prior distributions that, for instance, specify a certain range of values as highly implausible or even as impossible; doing so will result in posterior estimates that deviate from likelihood-only estimates in predictable, prior-specified ways. For a simple example, a classical model of crime counts might sometimes generate negative predicted counts for some subgroups and improbably extreme positive counts for other subgroups; in a Bayesian model, one could specify an a priori distribution of plausible values for which negative counts are impossible and extreme positive counts are improbable. In this study, we specify “weakly informative” or “regularizing” priors (McElreath) that are sufficiently diffuse as to allow for a wide range of likelihood-driven estimates yet are informative or “tight” enough to down-weight and pull in extreme or highly implausible likelihood-based estimates; such priors are designed to generate better out-of-sample predictive properties than likelihood-based estimates, particularly in small (sub)samples.
We start by illustrating the poor fit between Gaussian normal
distributional assumptions and these ordinal stress items by focusing on
two individual stress items at wave one: stress about having enough
money to buy necessities (stmonyw1) and stress
about being assaulted (stmugw1). As we noted above,
the financial stress item item exhibits an approximately normal
distribution with nearly symmetric decay about the midpoint, whereas the
victimization stress item has a positively skewed distribution. These
full-sample item distributions can be seen in Figure 1 above; they are
also replicated for convenience in the frequency bar graphs below.
#load stress.wide (from Fig1 Rmd)
load(here("1_Data_Files/Datasets/stress_wide.Rdata"))
stress.wide <- zap_labels(stress.wide)
stress.wide <- zap_label(stress.wide)3.1.3.1 Example stress item response distributions
3.1.3.1.1 stmonyw1 (stress about money, time 1)
ggplot(stress.wide, aes(stmonyw1)) +
geom_bar()
3.1.3.1.2 stmugw1 (stress about assault, time 1)
ggplot(stress.wide, aes(stmugw1)) +
geom_bar()
To estimate means in an intercept-only Bayesian model, one might
specify a normally distributed prior distribution centered at the
measurement scale midpoint (mu=3), and perhaps specify sd=2 for a
relatively flat prior that allows for a wide range of plausible values -
as well as implausible values given measured range is from 1 to 5. In
fact, this is the default scaled prior distribution for the intercept
estimate that will be used in the rstanarm model below.
This prior distribution would be very weakly informative for this
distribution (i.e., a one standard deviation unit decrease or increase
is specified equivalent to the variable’s min and max values,
respectively), meaning that it should result in the likelihood (data)
driving the posterior estimates. Moreover, the posterior estimates
generated by this model will converge with results generated using
classical frequentist approaches (many of which also assume
distributional normality). The figure below helps with visualizing this
prior. The black normal curve displays prior probability values for a
range of a priori theoretically plausible response categories.
The bulk of this prior distribution falls within the red lines
indicating the limits of the measured response values of 1 and 5 for our
stress items. However, a substantial amount of the probability
distribution also covers values for “implausible” category responses
laying outside these bounds. Hence, non-zero probabilities for stress
categories such as “0” or “6” are technically permitted by the prior
distribution though, when combined with the likelihood, estimated
response probabilities for these categories should converge to zero
since there are no observed data beyond these bounds and ample data
within them.
#simulate intercept prior for linear models predicting specific stress items
#Center at midpoint x=3 and rescale to sd=2
a <- rnorm(1e4,3,2)
# mean(a)
minx <- 1
maxx <- 5
curve(dnorm(x,3,2),from=-5,to=10,n=200)
curve((x-1), add = TRUE, col = "red", lwd = 2)
curve((x-5), add = TRUE, col = "red", lwd = 2)
Now let’s estimate overall and group-specific means with
rstanarm models, which automatically scale the variable
priors, using the same procedures as before. We can generate a prior
summary after estimating the model to see that the scaled prior
specified automatically by the model is same as the one graphically
depicted in the example above.
3.1.3.2 stmonyw1: Normal model, full sample
We begin with the stmonyw1 intercept-only model in the
full sample, …
#Bayesian models estimating *full sample mean* and specific rural/ses subgroup means for specific stress at T1.
#Items: *stmonyw1* sttranw1 strespw1 stfairw1 stjobw1 stthftw1 stmugw1
#Stress about money at T1: Intercept only, Normal prior
#Midpoint of plausible range [1,5] = 3
#Manual caching for stan_glm - file & file_refit built into brms
if (file.exists(here("Models","stmony.rds"))) {
stmony <- readRDS("Models/stmony.rds")
} else {
stmony <- stan_glm(stmonyw1 ~ 1,
family = gaussian(),
data = stress.wide,
prior_intercept = normal(3, 2),
chains = 4,
cores = nCoresphys,
seed = 8675309
)
saveRDS(stmony, "Models/stmony.rds")
}
out.stmony <- ppchecks(stmony) #runs user-written "ppchecks" function that request common checks3.1.3.2.1 PPcheck(density)
out.stmony[[3]]
3.1.3.2.2 PPcheck(histogram)
out.stmony[[5]]
3.1.3.2.3 PPcheck(mean)
out.stmony[[6]]
3.1.3.2.4 PPcheck(median)
out.stmony[[7]]
3.1.3.2.5 PPcheck(boxplots)
out.stmony[[4]]
3.1.3.2.6 Fit summary
out.stmony[[1]]##
## Model Info:
## function: stan_glm
## family: gaussian [identity]
## formula: stmonyw1 ~ 1
## algorithm: sampling
## sample: 4000 (posterior sample size)
## priors: see help('prior_summary')
## observations: 489
## predictors: 1
##
## Estimates:
## mean sd 2.5% 97.5%
## (Intercept) 3.15 0.05 3.05 3.24
## sigma 1.10 0.03 1.04 1.17
##
## Fit Diagnostics:
## mean sd 2.5% 97.5%
## mean_PPD 3.15 0.07 3.01 3.29
##
## The mean_ppd is the sample average posterior predictive distribution of the outcome variable (for details see help('summary.stanreg')).
##
## MCMC diagnostics
## mcse Rhat n_eff
## (Intercept) 0.00 1.00 2794
## sigma 0.00 1.00 2455
## mean_PPD 0.00 1.00 3355
## log-posterior 0.02 1.00 1610
##
## For each parameter, mcse is Monte Carlo standard error, n_eff is a crude measure of effective sample size, and Rhat is the potential scale reduction factor on split chains (at convergence Rhat=1).3.1.3.2.7 Prior summary
out.stmony[[2]]## Priors for model 'modelfit'
## ------
## Intercept (after predictors centered)
## ~ normal(location = 3, scale = 2)
##
## Auxiliary (sigma)
## Specified prior:
## ~ exponential(rate = 1)
## Adjusted prior:
## ~ exponential(rate = 0.91)
## ------
## See help('prior_summary.stanreg') for more details… followed by the stmonyw1 group-specific means
model.
3.1.3.3 stmonyw1: Normal model, grouped
#Bayesian models estimating full sample mean and *specific rural/ses subgroup means* for specific stress at T1.
#Items: *stmonyw1* sttranw1 strespw1 stfairw1 stjobw1 stthftw1 stmugw1
#Stress about money at T1: Group intercepts, Normal prior
if (file.exists(here("Models","stmony.rurses"))) {
stmony.rurses <- readRDS("Models/stmony.rurses")
} else {
stmony.rurses <- update(stmony, formula. = ~ 0 + rural.ses.med)
saveRDS(stmony.rurses, "Models/stmony_rurses.rds")
}
out.stmony.rurses <- ppchecks(stmony.rurses)3.1.3.3.1 PPcheck(density)
out.stmony.rurses[[3]]
3.1.3.3.2 PPcheck(histogram)
out.stmony.rurses[[5]]
3.1.3.3.3 PPcheck(mean)
out.stmony.rurses[[6]]
3.1.3.3.4 PPcheck(median)
out.stmony.rurses[[7]]
3.1.3.3.5 PPcheck(boxplots)
out.stmony.rurses[[4]]
3.1.3.3.6 PPcheck(grouped-mean)
pp_check(stmony.rurses, plotfun = "stat_grouped", stat = "mean", group = "rural.ses.med")
3.1.3.3.7 PPcheck(grouped-median)
pp_check(stmony.rurses, plotfun = "stat_grouped", stat = "median", group = "rural.ses.med")
3.1.3.3.8 Fit summary
out.stmony.rurses[[1]]##
## Model Info:
## function: stan_glm
## family: gaussian [identity]
## formula: stmonyw1 ~ rural.ses.med - 1
## algorithm: sampling
## sample: 4000 (posterior sample size)
## priors: see help('prior_summary')
## observations: 489
## predictors: 4
##
## Estimates:
## mean sd 2.5% 97.5%
## rural.ses.med1 2.93 0.10 2.73 3.13
## rural.ses.med2 3.10 0.10 2.90 3.29
## rural.ses.med3 3.41 0.09 3.23 3.59
## rural.ses.med4 3.10 0.11 2.90 3.31
## sigma 1.09 0.04 1.02 1.16
##
## Fit Diagnostics:
## mean sd 2.5% 97.5%
## mean_PPD 3.15 0.07 3.01 3.29
##
## The mean_ppd is the sample average posterior predictive distribution of the outcome variable (for details see help('summary.stanreg')).
##
## MCMC diagnostics
## mcse Rhat n_eff
## rural.ses.med1 0.00 1.00 7020
## rural.ses.med2 0.00 1.00 6261
## rural.ses.med3 0.00 1.00 6531
## rural.ses.med4 0.00 1.00 5985
## sigma 0.00 1.00 5326
## mean_PPD 0.00 1.00 5051
## log-posterior 0.04 1.00 1962
##
## For each parameter, mcse is Monte Carlo standard error, n_eff is a crude measure of effective sample size, and Rhat is the potential scale reduction factor on split chains (at convergence Rhat=1).3.1.3.3.9 Prior summary
out.stmony.rurses[[2]]## Priors for model 'modelfit'
## ------
##
## Coefficients
## Specified prior:
## ~ normal(location = [0,0,0,...], scale = [2.5,2.5,2.5,...])
## Adjusted prior:
## ~ normal(location = [0,0,0,...], scale = [6.52,6.32,6.03,...])
##
## Auxiliary (sigma)
## Specified prior:
## ~ exponential(rate = 1)
## Adjusted prior:
## ~ exponential(rate = 0.91)
## ------
## See help('prior_summary.stanreg') for more details3.1.3.4 stmugw1: Normal model, full sample
Next, the stmugw1 full sample model…
#Bayesian models estimating *full sample mean* and specific rural/ses subgroup means for specific stress at T1.
#Items: stmonyw1 sttranw1 strespw1 stfairw1 stjobw1 stthftw1 *stmugw1*
#Stress about assault at T1: Intercept only, Normal prior
#Midpoint of plausible range [1,5] = 3
if (file.exists(here("Models","stmug"))) {
stmug <- readRDS("Models/stmug")
} else {
stmug <- stan_glm(stmugw1 ~ 1,
family = gaussian(),
data = stress.wide,
prior_intercept = normal(3, 2),
#chains = 4, cores = 8,
seed = 8675309)
saveRDS(stmug, "Models/stmug.rds")
}
out.stmug <- ppchecks(stmug)3.1.3.4.1 PPcheck(density)
out.stmug[[3]]
3.1.3.4.2 PPcheck(histogram)
out.stmug[[5]]
3.1.3.4.3 PPcheck(mean)
out.stmug[[6]]
3.1.3.4.4 PPcheck(median)
out.stmug[[7]]
3.1.3.4.5 PPcheck(boxplots)
out.stmug[[4]]
3.1.3.4.6 Fit summary
out.stmug[[1]]##
## Model Info:
## function: stan_glm
## family: gaussian [identity]
## formula: stmugw1 ~ 1
## algorithm: sampling
## sample: 4000 (posterior sample size)
## priors: see help('prior_summary')
## observations: 489
## predictors: 1
##
## Estimates:
## mean sd 2.5% 97.5%
## (Intercept) 1.96 0.05 1.86 2.05
## sigma 1.06 0.03 0.99 1.12
##
## Fit Diagnostics:
## mean sd 2.5% 97.5%
## mean_PPD 1.96 0.07 1.82 2.09
##
## The mean_ppd is the sample average posterior predictive distribution of the outcome variable (for details see help('summary.stanreg')).
##
## MCMC diagnostics
## mcse Rhat n_eff
## (Intercept) 0.00 1.00 2905
## sigma 0.00 1.00 2677
## mean_PPD 0.00 1.00 3325
## log-posterior 0.03 1.00 1228
##
## For each parameter, mcse is Monte Carlo standard error, n_eff is a crude measure of effective sample size, and Rhat is the potential scale reduction factor on split chains (at convergence Rhat=1).3.1.3.4.7 Prior summary
out.stmug[[2]]## Priors for model 'modelfit'
## ------
## Intercept (after predictors centered)
## ~ normal(location = 3, scale = 2)
##
## Auxiliary (sigma)
## Specified prior:
## ~ exponential(rate = 1)
## Adjusted prior:
## ~ exponential(rate = 0.95)
## ------
## See help('prior_summary.stanreg') for more details…and the stmugw1 group-specific means model.
3.1.3.5 stmugw1: Normal model, grouped
#Bayesian models estimating full sample mean and *specific rural/ses subgroup means* for specific stress at T1.
#Items: stmonyw1 sttranw1 strespw1 stfairw1 stjobw1 stthftw1 *stmugw1*
#Stress about assault at T1: Group intercepts, Normal prior
if (file.exists(here("Models","stmug.rurses"))) {
stmug.rurses <- readRDS("Models/stmug.rurses")
} else {
stmug.rurses <- update(stmug, formula. = ~ 0 + rural.ses.med)
saveRDS(stmug.rurses, "Models/stmug_rurses.rds")
}
out.stmug.rurses <- ppchecks(stmug.rurses)3.1.3.5.1 PPcheck(density)
out.stmug.rurses[[3]]
3.1.3.5.2 PPcheck(histogram)
out.stmug.rurses[[5]]
3.1.3.5.3 PPcheck(mean)
out.stmug.rurses[[6]]
3.1.3.5.4 PPcheck(median)
out.stmug.rurses[[7]]
3.1.3.5.5 PPcheck(boxplots)
out.stmug.rurses[[4]]
3.1.3.5.6 PPcheck(grouped-mean)
pp_check(stmug.rurses, plotfun = "stat_grouped", stat = "mean", group = "rural.ses.med")
3.1.3.5.7 PPcheck(grouped-median)
pp_check(stmug.rurses, plotfun = "stat_grouped", stat = "median", group = "rural.ses.med")
3.1.3.5.8 Fit summary
out.stmug.rurses[[1]]##
## Model Info:
## function: stan_glm
## family: gaussian [identity]
## formula: stmugw1 ~ rural.ses.med - 1
## algorithm: sampling
## sample: 4000 (posterior sample size)
## priors: see help('prior_summary')
## observations: 489
## predictors: 4
##
## Estimates:
## mean sd 2.5% 97.5%
## rural.ses.med1 1.65 0.10 1.46 1.83
## rural.ses.med2 1.90 0.09 1.72 2.10
## rural.ses.med3 2.19 0.09 2.02 2.36
## rural.ses.med4 2.04 0.10 1.84 2.23
## sigma 1.04 0.03 0.98 1.11
##
## Fit Diagnostics:
## mean sd 2.5% 97.5%
## mean_PPD 1.96 0.07 1.83 2.09
##
## The mean_ppd is the sample average posterior predictive distribution of the outcome variable (for details see help('summary.stanreg')).
##
## MCMC diagnostics
## mcse Rhat n_eff
## rural.ses.med1 0.00 1.00 5720
## rural.ses.med2 0.00 1.00 4876
## rural.ses.med3 0.00 1.00 4560
## rural.ses.med4 0.00 1.00 5039
## sigma 0.00 1.00 5324
## mean_PPD 0.00 1.00 4868
## log-posterior 0.04 1.00 1936
##
## For each parameter, mcse is Monte Carlo standard error, n_eff is a crude measure of effective sample size, and Rhat is the potential scale reduction factor on split chains (at convergence Rhat=1).3.1.3.5.9 Prior summary
out.stmug.rurses[[2]]## Priors for model 'modelfit'
## ------
##
## Coefficients
## Specified prior:
## ~ normal(location = [0,0,0,...], scale = [2.5,2.5,2.5,...])
## Adjusted prior:
## ~ normal(location = [0,0,0,...], scale = [6.25,6.06,5.78,...])
##
## Auxiliary (sigma)
## Specified prior:
## ~ exponential(rate = 1)
## Adjusted prior:
## ~ exponential(rate = 0.95)
## ------
## See help('prior_summary.stanreg') for more detailsA couple things are noteworthy here. First, the overall and group-specific means were recovered from the predicted distributions generated by these “normal” models. This is unsurprising, since this is essentially what these models are designed to do. However, the medians were not recovered by predicted distributions. Other posterior predicted checks, such as observed versus predicted (Y vs. Yrep) density overlays or histograms, also show poor fit between ordinal variable distributions and metric normality assumptions.
Descriptive and inferential problems stemming from poor fit in metric linear models will be minimized when analyzing ordinal items or multi-item ordinal scales with a substantial number of categories and with distributions characterized by symmetric decay about a centrally located mean (e.g., “stmony” above). Problems will be exacerbated in metric linear models for ordinal items (or multi-item ordinal scales) with few categories and/or with distributions that do not resembles a Gaussian normal distribution (e.g., “stmug” above).
Yet, alternative ordinal models can better handle each of these situations. So, while treating a general stress scale comprised of multiple ordinal items as metric might be defensible, particularly if the scale exhibits symmetric decay about the mean, analyzing individual stress-specific ordinal items as metric is potentially more problematic. Since a key goal is to accurately describe the item-level relationships between stress and theorized outcomes, traditional linear (i.e., metric normal) models are not a wise choice.
3.1.4 Ordinal, not normal, models should be ordinary
Alternatively, let’s consider an ordinal model for the specific stress items. First, we need to recode specific stress variables as factors. We will also create a new community group factor that is more descriptive (i.e., “Rural/Low SES” instead of “1”), since we will want to start paying attention to specific group distributions soon.
#First, need to recode specific stress variables as factors.
#Items: stmonyw1 sttranw1 strespw1 stfairw1 stjobw1 stthftw1 stmugw1
#Also create new community "group" variable.
stress.wide2 <- stress.wide %>%
mutate(
stmonyw1f = factor(stmonyw1, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
sttranw1f = factor(sttranw1, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
strespw1f = factor(strespw1, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stfairw1f = factor(stfairw1, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stjobw1f = factor(stjobw1, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stthftw1f = factor(stthftw1, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stmugw1f = factor(stmugw1, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
rural.ses.med2 = 0,
rural.ses.med2 = ifelse(rural.ses.med == 1, "Rural/Low SES", rural.ses.med2), #Rural/Low SES
rural.ses.med2 = ifelse(rural.ses.med == 2, "Rural/High SES", rural.ses.med2), #Rural/High SES
rural.ses.med2 = ifelse(rural.ses.med == 3, "Urban/Low SES", rural.ses.med2), #Urban/Low SES
rural.ses.med2 = ifelse(rural.ses.med == 4, "Urban/High SES", rural.ses.med2), #Urban/High SES
rural.ses.med2 = factor(rural.ses.med2)
)
# head(stress.wide2)Next, let’s consider a cumulative probit brms model to
see if we can recover overall and group-specific means and medians for
the ordinal stress items as well as their specific ordinal category
values or “thresholds.” Recall, the metric normal models above recovered
means well, which is what they are designed to do, but the posterior
predictive check plots (e.g., density and histogram) show that the
underlying normal distribution assumption results in model predictions
that fail to recover the observed ordinal item values. We will start
with the stmonyw1 item again.
3.1.4.1 stmonyw1: Cumul. probit, full sample
#Revise function to call basic summaries & ppchecks
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit, probs = c(0.025, 0.975), digits = 2) #summarize results
priorsum <- prior_summary(modelfit) #check priors
ppcheck <- pp_check(modelfit, cores=1) #posterior predictive checks
allchecks <- list(fitsummary, priorsum, ppcheck)
return(allchecks)
}#Ordinal cumulative probit model predicting stress about money at T1
#Stress about money at T1 - Thresholds only, Cumulative probit model:
stmony.ord <-
brm(data = stress.wide2,
family = cumulative("probit"),
stmonyw1f ~ 1,
prior(normal(0, 4), class = Intercept), #see kruschke/kurtz re: priors for this model
iter = 2000, warmup = 1000,
chains = 4,
cores = nCoresphys,
backend = "cmdstanr",
#threads = threading(2), #consider within-chain threading if machine has at least 8 physical cores
seed = 8675309,
file = "Models/stmony.ord",
file_refit = "on_change"
)
out.stmony.ord <- ppchecks(stmony.ord) #posterior predictive checks
# Plot fit
# plot(stmony.ord, ask=FALSE) 3.1.4.1.1 PPcheck(bars, full sample)
pp_check(stmony.ord, type="bars", ndraws=1000, cores=1) +
scale_x_continuous("Stress abt money, T1", breaks = 1:5) +
scale_y_continuous(NULL, breaks = NULL, expand = expansion(mult = c(0, 0.05))) +
ggtitle("Data with posterior predictions (cumul-ord, fullsamp)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.9, .8))
3.1.4.1.2 PPcheck(density, full sample)
out.stmony.ord[[3]]
3.1.4.1.3 Fit summary (full sample)
out.stmony.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stmonyw1f ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -1.43 0.08 -1.60 -1.27 1.00 2391 2566
## Intercept[2] -0.60 0.06 -0.72 -0.49 1.00 3634 3273
## Intercept[3] 0.27 0.06 0.16 0.38 1.00 3904 3417
## Intercept[4] 1.23 0.07 1.09 1.37 1.00 3636 3339
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.1.4 Prior summary (full sample)
out.stmony.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)3.1.4.2 stmonyw1: Cumul. probit, grouped
#Ordinal cumulative probit model predicting stress about money at T1
#Stress about money at T1: Thresholds + group differences, Cumul probit
stmony.ord.rurses <- update(stmony.ord,
formula. = ~ 1 + rural.ses.med2,
newdata = stress.wide2,
file = "Models/stmony.ord.rurses",
file_refit = "on_change"
)
out.stmony.ord.rurses <- ppchecks(stmony.ord.rurses) #posterior predictive checks
# Plot fit
# plot(stmony.ord.rurses, ask=FALSE) 3.1.4.2.1 PPcheck(bars, grouped)
pp_check(stmony.ord.rurses, type="bars_grouped",
group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt money, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cumul-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.2.2 PPcheck(density, grouped)
out.stmony.ord.rurses[[3]]
3.1.4.2.3 Fit summary (grouped)
out.stmony.ord.rurses[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stmonyw1f ~ rural.ses.med2
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -1.39 0.12 -1.62 -1.17 1.00 2460
## Intercept[2] -0.56 0.10 -0.74 -0.37 1.00 2649
## Intercept[3] 0.33 0.10 0.13 0.52 1.00 2534
## Intercept[4] 1.30 0.11 1.09 1.52 1.00 3007
## rural.ses.med2RuralDLowSES -0.16 0.13 -0.43 0.11 1.00 3223
## rural.ses.med2UrbanDHighSES 0.01 0.13 -0.26 0.27 1.00 3014
## rural.ses.med2UrbanDLowSES 0.31 0.13 0.06 0.56 1.00 3003
## Tail_ESS
## Intercept[1] 2606
## Intercept[2] 2934
## Intercept[3] 3015
## Intercept[4] 3465
## rural.ses.med2RuralDLowSES 2705
## rural.ses.med2UrbanDHighSES 2803
## rural.ses.med2UrbanDLowSES 2966
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.2.4 Prior summary (grouped)
out.stmony.ord.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## (flat) b rural.ses.med2RuralDLowSES
## (flat) b rural.ses.med2UrbanDHighSES
## (flat) b rural.ses.med2UrbanDLowSES
## normal(0, 4) Intercept
## normal(0, 4) Intercept 1
## normal(0, 4) Intercept 2
## normal(0, 4) Intercept 3
## normal(0, 4) Intercept 4
## source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)As expected, this model is a much better fit to the data than the metric normal model. The overall unconditional model recovers the observed data very well; however, the group means model does not quite recover the observed data for all five categories within each of the four groups.
Rather than specifying a cumulative ordinal model, a sequential
ordinal model with category-specific effects of community group on
stress might do a better job of recovering the observed data in the
group means models. The cumulative model is more parsimonious, but it is
over- and under-predicting values for some categories of
stmonyw1, and it cannot be modified to specify
category-specific effects.
The sequential ordinal model has an intuitive theoretical derivation - it relies on the logic, built into truly ordinal variables, that in order to achieve a category k, one must first achieve all lower categories 1 through k – 1. The sequential model, as well as the mathematically similar adjacent category model, allows for specification of category-specific effects of a predictor (for details, see Burkner & Vuorre, Appendix A).
3.1.4.3 stmonyw1: Seq. probit, cat-specific, grouped
#See if sequential model also improves recovery of observed data compared to cumulative model
prior_seq <- prior(normal(0, 5), class = b) +
prior(normal(0, 5), class = Intercept)
#see burkner & vuorre re: priors for this model
# https://journals.sagepub.com/doi/full/10.1177/2515245918823199
stmony.seq.rurses <- update(stmony.ord,
formula. = ~ 1 + cs(rural.ses.med2),
prior = prior_seq,
newdata = stress.wide2,
family = sratio("cloglog"),
file = "Models/stmony.seq.rurses",
file_refit = "on_change"
)
out.stmony.seq.rurses <- ppchecks(stmony.seq.rurses) #posterior predictive checks
# Plot fit
# plot(stmony.seq.rurses, ask=FALSE)3.1.4.3.1 PPcheck(bars, cat-spec)
pp_check(stmony.seq.rurses, type="bars_grouped",
group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt money, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cat-spec-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.3.2 PPcheck(density, cat-spec)
out.stmony.seq.rurses[[3]]
3.1.4.3.3 Fit summary (cat-spec)
out.stmony.seq.rurses[[1]]## Family: sratio
## Links: mu = cloglog; disc = identity
## Formula: stmonyw1f ~ cs(rural.ses.med2)
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept[1] -2.89 0.39 -3.70 -2.20 1.00
## Intercept[2] -1.49 0.20 -1.91 -1.12 1.00
## Intercept[3] -0.18 0.14 -0.47 0.10 1.00
## Intercept[4] 0.31 0.20 -0.10 0.69 1.00
## rural.ses.med2RuralDLowSES[1] -0.82 0.47 -1.78 0.08 1.00
## rural.ses.med2RuralDLowSES[2] -0.03 0.29 -0.60 0.55 1.00
## rural.ses.med2RuralDLowSES[3] -0.02 0.21 -0.42 0.38 1.00
## rural.ses.med2RuralDLowSES[4] -0.02 0.30 -0.59 0.57 1.00
## rural.ses.med2UrbanDHighSES[1] -0.61 0.50 -1.58 0.37 1.00
## rural.ses.med2UrbanDHighSES[2] -0.26 0.28 -0.82 0.28 1.00
## rural.ses.med2UrbanDHighSES[3] 0.59 0.25 0.12 1.07 1.00
## rural.ses.med2UrbanDHighSES[4] 0.11 0.28 -0.43 0.65 1.00
## rural.ses.med2UrbanDLowSES[1] 0.35 0.58 -0.78 1.54 1.00
## rural.ses.med2UrbanDLowSES[2] 0.10 0.28 -0.46 0.63 1.00
## rural.ses.med2UrbanDLowSES[3] 0.71 0.22 0.29 1.14 1.00
## rural.ses.med2UrbanDLowSES[4] 0.16 0.25 -0.33 0.66 1.00
## Bulk_ESS Tail_ESS
## Intercept[1] 2666 2149
## Intercept[2] 2197 2246
## Intercept[3] 2319 2740
## Intercept[4] 2394 2192
## rural.ses.med2RuralDLowSES[1] 2770 2547
## rural.ses.med2RuralDLowSES[2] 2629 2921
## rural.ses.med2RuralDLowSES[3] 2768 3130
## rural.ses.med2RuralDLowSES[4] 2998 2954
## rural.ses.med2UrbanDHighSES[1] 2793 2918
## rural.ses.med2UrbanDHighSES[2] 2584 2591
## rural.ses.med2UrbanDHighSES[3] 3311 2975
## rural.ses.med2UrbanDHighSES[4] 2728 2793
## rural.ses.med2UrbanDLowSES[1] 2965 2869
## rural.ses.med2UrbanDLowSES[2] 2504 2675
## rural.ses.med2UrbanDLowSES[3] 3168 2859
## rural.ses.med2UrbanDLowSES[4] 2726 2627
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.3.4 Prior summary (cat-spec)
out.stmony.seq.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## normal(0, 5) b
## normal(0, 5) b rural.ses.med2RuralDLowSES
## normal(0, 5) b rural.ses.med2UrbanDHighSES
## normal(0, 5) b rural.ses.med2UrbanDLowSES
## normal(0, 5) Intercept
## normal(0, 5) Intercept 1
## normal(0, 5) Intercept 2
## normal(0, 5) Intercept 3
## normal(0, 5) Intercept 4
## source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)As suspected, this model does a much better job of recovering the observed data. Also, it is the specification of category-specific effects (not the specification of a sequential model) that improves recovery of the observed data distribution (not shown). Of course, the improved (over)fit to the data was purchased at the steep cost of estimating nine additional parameters, which is not too unwieldy for this simple descriptive model but would become problematic in more complex multivariate regression models with our sample.
For this type of model, it can be helpful to plot the estimated
relationship between community group and stmonyw1 for each
threshold or response category of the stress outcome variable.
conditional_effects(stmony.seq.rurses, "rural.ses.med2", categorical = TRUE)
The figure shows that respondents in rural communities are much more likely to report “sometimes” (=3) stressing about money compared to urban respondents - this is the modal response category for rural residents. In contrast, the modal category is “often” (=4) for urban residents - and this pattern is especially pronounced among respondents in low SES urban areas. Overall, then, it appears that urban residents are more likely than rural residents to stress about having money to pay for necessities.
We will continue examining the fit of both the more parsimonious
cumulative model and the more flexible sequential model from here. The
sequential category-specific model will ensure that we accurately
recover the observed data when plotting descriptive response
distributions for stress items by community (Figure 2). Later, when
building models to predict our key outcomes (criminal behavior; criminal
intent; negative emotions), we will switch back to cumulative probit
models of stress and use the built-in cumulative ordinal predictor
specification with mo(). As these `stmonyw1’ results show,
a cumulative ordinal model specification might not result in a perfect
fit relative to the observed data, but it is a substantial improvement
on assuming a normal distribution, and this relatively small loss of fit
(at least compared to metric normal models) is a worthwhile trade-off
for the cumulative model’s parsimony as our models get more complex. If
we had a much larger sample and/or our key outcome variables were not so
rare, we might also be able to include the stress item predictors as
sequential category-specific ordinal thresholds, such as by recoding
categorical versions of our stress items with “sequential difference
coding” (see p.23: https://bpspsychub.onlinelibrary.wiley.com/doi/10.1111/bmsp.12195).
With our data, though, we will need to accept the loss of precision for
the sake of parsimony in our ordinal modeling.
# CHECK LOO POSTERIOR PREDICTIVE COMPARISON? 3.1.4.4 stmugw1: Cumul. probit, full sample
Let’s continue the example by following the same procedure with our
stmugw1 variable.
#Ordinal cumulative probit model predicting stress about assault at T1
#Stress about assault at T1 - Thresholds only, Cumulative probit model:
stmug.ord <-
brm(data = stress.wide2,
family = cumulative("probit"),
stmugw1f ~ 1,
prior(normal(0, 4), class = Intercept), #see kruschke/kurz re: priors for this model
#file = here("2_Models", "stmug.ord"),
#Save fit - CAUTION: despite model changes, will *NOT* rerun the model if `file` exists,
iter = 2000, warmup = 1000,
chains = 4,
cores = nCoresphys,
backend = "cmdstanr",
#threads = threading(2), #consider within-chain threading if machine has at least 8 physical cores
seed = 8675309,
file = "Models/stmug.ord",
file_refit = "on_change"
)
out.stmug.ord <- ppchecks(stmug.ord) #posterior predictive checks3.1.4.4.1 PPcheck(bars, full sample)
pp_check(stmug.ord, type="bars", ndraws=1000, cores=1) +
scale_x_continuous("Stress abt assault, T1", breaks = 1:5) +
scale_y_continuous(NULL, breaks = NULL, expand = expansion(mult = c(0, 0.05))) +
ggtitle("Data with posterior predictions (cumul-ord, fullsamp)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.9, .8))
3.1.4.4.2 PPcheck(density, full sample)
out.stmug.ord[[3]]
3.1.4.4.3 Fit summary (full sample)
out.stmug.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stmugw1f ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -0.16 0.06 -0.27 -0.05 1.00 3599 2936
## Intercept[2] 0.58 0.06 0.46 0.70 1.00 4875 3399
## Intercept[3] 1.31 0.08 1.16 1.46 1.00 4858 3219
## Intercept[4] 2.06 0.13 1.82 2.33 1.00 4276 3150
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.4.4 Prior summary (full sample)
out.stmug.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)3.1.4.5 stmugw1: Cumul. probit, grouped
#Ordinal cumulative probit model predicting stress about assault at T1
#Stress about assault at T1: Thresholds + group differences, Cumul probit
stmug.ord.rurses <- update(stmug.ord,
formula. = ~ 1 + rural.ses.med2,
newdata = stress.wide2,
file = "Models/stmug.ord.rurses",
file_refit = "on_change"
)
out.stmug.ord.rurses <- ppchecks(stmug.ord.rurses) #posterior predictive checks3.1.4.5.1 PPcheck(bars, grouped)
pp_check(stmug.ord.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt assault, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cumul-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.5.2 PPcheck(density, grouped)
stmug.ord.rurses[[3]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## (flat) b rural.ses.med2RuralDLowSES
## (flat) b rural.ses.med2UrbanDHighSES
## (flat) b rural.ses.med2UrbanDLowSES
## normal(0, 4) Intercept
## normal(0, 4) Intercept 1
## normal(0, 4) Intercept 2
## normal(0, 4) Intercept 3
## normal(0, 4) Intercept 4
## source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.5.3 Conditional effects (grouped)
conditional_effects(stmug.ord.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.5.4 Fit summary (grouped)
out.stmug.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stmugw1f ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -0.16 0.06 -0.27 -0.05 1.00 3599 2936
## Intercept[2] 0.58 0.06 0.46 0.70 1.00 4875 3399
## Intercept[3] 1.31 0.08 1.16 1.46 1.00 4858 3219
## Intercept[4] 2.06 0.13 1.82 2.33 1.00 4276 3150
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.5.5 Prior summary (grouped)
out.stmug.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)The cumulative model does a much better job recovering the observed
data for the stmugw1 item, relative to the earlier
stmonyw1 model. Also, as expected, it is a much better fit
than the metric normal model. Let’s compare the sequential model with
category-specific effects for the group means model.
3.1.4.6 stmugw1: Seq. probit, cat-specific, grouped
#See if sequential model w/category specific effects improves recovery of observed data
stmug.seq.rurses <- update(stmug.ord,
formula. = ~ 1 + cs(rural.ses.med2),
prior = prior_seq,
newdata = stress.wide2,
family = sratio("cloglog"),
file = "Models/stmug.seq.rurses",
file_refit = "on_change"
)
out.stmug.seq.rurses <- ppchecks(stmug.seq.rurses) #posterior predictive checks3.1.4.6.1 PPcheck(bars, cat-spec)
pp_check(stmug.seq.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt assault, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cat-spec-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.6.2 PPcheck(density, cat-spec)
out.stmug.seq.rurses[[3]]
3.1.4.6.3 Conditional effects (cat-spec)
conditional_effects(stmug.seq.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.6.4 Fit summary (cat-spec)
out.stmug.seq.rurses[[1]]## Family: sratio
## Links: mu = cloglog; disc = identity
## Formula: stmugw1f ~ cs(rural.ses.med2)
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept[1] -0.47 0.14 -0.75 -0.21 1.00
## Intercept[2] -0.61 0.19 -0.98 -0.25 1.00
## Intercept[3] 0.42 0.20 0.01 0.80 1.00
## Intercept[4] 3.13 1.72 0.73 7.26 1.00
## rural.ses.med2RuralDLowSES[1] -0.06 0.19 -0.43 0.31 1.00
## rural.ses.med2RuralDLowSES[2] -0.95 0.25 -1.43 -0.47 1.00
## rural.ses.med2RuralDLowSES[3] -0.55 0.42 -1.41 0.28 1.00
## rural.ses.med2RuralDLowSES[4] -1.90 3.84 -10.34 4.64 1.00
## rural.ses.med2UrbanDHighSES[1] 0.10 0.20 -0.30 0.50 1.00
## rural.ses.med2UrbanDHighSES[2] -0.15 0.26 -0.65 0.36 1.00
## rural.ses.med2UrbanDHighSES[3] 0.63 0.32 0.02 1.30 1.00
## rural.ses.med2UrbanDHighSES[4] 3.15 1.74 0.61 7.20 1.00
## rural.ses.med2UrbanDLowSES[1] 0.27 0.19 -0.11 0.64 1.00
## rural.ses.med2UrbanDLowSES[2] 0.10 0.25 -0.40 0.58 1.00
## rural.ses.med2UrbanDLowSES[3] 0.64 0.28 0.10 1.19 1.00
## rural.ses.med2UrbanDLowSES[4] 2.71 1.74 0.23 6.90 1.00
## Bulk_ESS Tail_ESS
## Intercept[1] 2009 1822
## Intercept[2] 1837 2546
## Intercept[3] 1782 2204
## Intercept[4] 1330 1292
## rural.ses.med2RuralDLowSES[1] 2458 2802
## rural.ses.med2RuralDLowSES[2] 2148 2583
## rural.ses.med2RuralDLowSES[3] 2958 3058
## rural.ses.med2RuralDLowSES[4] 3096 2281
## rural.ses.med2UrbanDHighSES[1] 2381 2436
## rural.ses.med2UrbanDHighSES[2] 2262 3066
## rural.ses.med2UrbanDHighSES[3] 2276 2843
## rural.ses.med2UrbanDHighSES[4] 1357 1418
## rural.ses.med2UrbanDLowSES[1] 2441 2122
## rural.ses.med2UrbanDLowSES[2] 2146 2712
## rural.ses.med2UrbanDLowSES[3] 2249 2298
## rural.ses.med2UrbanDLowSES[4] 1339 1479
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.6.5 Prior summary (cat-spec)
out.stmug.seq.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## normal(0, 5) b
## normal(0, 5) b rural.ses.med2RuralDLowSES
## normal(0, 5) b rural.ses.med2UrbanDHighSES
## normal(0, 5) b rural.ses.med2UrbanDLowSES
## normal(0, 5) Intercept
## normal(0, 5) Intercept 1
## normal(0, 5) Intercept 2
## normal(0, 5) Intercept 3
## normal(0, 5) Intercept 4
## source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)Again, the parsimonious cumulative ordinal model recovers the observed group means data well and, as expected, the sequential category-specific model recovers the observed data even better - but at the cost of estimating substantially more threshold parameters.
As far as descriptive patterns, urban residents are less likely to report “never” stressing about being victims of assault and more likely to “often” or “very often” report stressing about assault.
Below are the ordinal models for the remaining stress items that we
used to generate Figure 2.
#### sttranw1: Cumul. probit, full & grouped {.tabset}
#Ordinal cumulative probit models predicting stress about transport at T1
#Items: stmonyw1 *sttranw1* strespw1 stfairw1 stjobw1 stthftw1 stmugw1
#Stress about transportation at T1 - Thresholds only, Cumulative probit model:
sttran.ord <-
brm(data = stress.wide2,
family = cumulative("probit"),
sttranw1f ~ 1,
prior(normal(0, 4), class = Intercept), #see kruschke/kurz re: priors for this model
iter = 2000, warmup = 1000,
chains = 4,
cores = nCoresphys,
backend = "cmdstanr",
#threads = threading(2), #consider within-chain threading if machine has at least 8 physical cores
seed = 8675309,
file = "Models/sttran.ord",
file_refit = "on_change"
)
out.sttran.ord <- ppchecks(sttran.ord) #posterior predictive checks
#Stress about transportation at T1: Thresholds + group differences, Cumul probit
sttran.ord.rurses <- update(sttran.ord,
formula. = ~ 1 + rural.ses.med2,
newdata = stress.wide2,
file = "Models/sttran.ord.rurses",
file_refit = "on_change"
)
out.sttran.ord.rurses <- ppchecks(sttran.ord.rurses) #posterior predictive checks3.1.4.6.6 PPcheck(bars, full sample)
pp_check(sttran.ord, type="bars", ndraws=1000, cores=1) +
scale_x_continuous("Stress abt transport, T1", breaks = 1:5) +
scale_y_continuous(NULL, breaks = NULL, expand = expansion(mult = c(0, 0.05))) +
ggtitle("Data with posterior predictions (cumul-ord, fullsamp)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.9, .8))
3.1.4.6.7 PPcheck(bars, grouped)
pp_check(sttran.ord.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt transport, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cumul-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.6.8 PPcheck(density, full sample)
out.sttran.ord[[3]]
3.1.4.6.9 PPcheck(density, grouped)
out.sttran.ord.rurses[[3]]
3.1.4.6.10 Conditional effects (grouped)
conditional_effects(sttran.ord.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.6.11 Fit summary (full sample)
out.sttran.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: sttranw1f ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -1.70 0.10 -1.90 -1.51 1.00 2179 1863
## Intercept[2] -0.55 0.06 -0.66 -0.43 1.00 3989 3382
## Intercept[3] 0.31 0.06 0.20 0.42 1.00 4192 2732
## Intercept[4] 1.18 0.07 1.05 1.33 1.00 3687 3378
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.6.12 Prior summary (full sample)
out.sttran.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)3.1.4.6.13 Fit summary (grouped)
out.sttran.ord.rurses[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: sttranw1f ~ rural.ses.med2
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -1.67 0.13 -1.92 -1.41 1.00 2570
## Intercept[2] -0.49 0.10 -0.69 -0.30 1.00 2955
## Intercept[3] 0.38 0.10 0.19 0.57 1.00 3161
## Intercept[4] 1.28 0.11 1.07 1.50 1.00 3144
## rural.ses.med2RuralDLowSES -0.24 0.13 -0.50 0.01 1.00 3458
## rural.ses.med2UrbanDHighSES 0.17 0.13 -0.10 0.43 1.00 3390
## rural.ses.med2UrbanDLowSES 0.30 0.13 0.05 0.55 1.00 3269
## Tail_ESS
## Intercept[1] 2764
## Intercept[2] 2969
## Intercept[3] 2958
## Intercept[4] 2962
## rural.ses.med2RuralDLowSES 2915
## rural.ses.med2UrbanDHighSES 2888
## rural.ses.med2UrbanDLowSES 3077
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.6.14 Prior summary (grouped)
out.sttran.ord.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## (flat) b rural.ses.med2RuralDLowSES
## (flat) b rural.ses.med2UrbanDHighSES
## (flat) b rural.ses.med2UrbanDLowSES
## normal(0, 4) Intercept
## normal(0, 4) Intercept 1
## normal(0, 4) Intercept 2
## normal(0, 4) Intercept 3
## normal(0, 4) Intercept 4
## source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.7 sttranw1: Seq. probit, cat-specific, grouped
#See if sequential model w/category specific effects improves recovery of observed data
sttran.seq.rurses <- update(sttran.ord,
formula. = ~ 1 + cs(rural.ses.med2),
prior = prior_seq,
newdata = stress.wide2,
family = sratio("cloglog"),
file = "Models/sttran.seq.rurses",
file_refit = "on_change"
)
out.sttran.seq.rurses <- ppchecks(sttran.seq.rurses) #posterior predictive checks3.1.4.7.1 PPcheck(bars, cat-spec)
pp_check(sttran.seq.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt transport, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cat-spec-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.7.2 PPcheck(density, cat-spec)
out.sttran.seq.rurses[[3]]
3.1.4.7.3 Conditional effects (cat-spec)
conditional_effects(sttran.seq.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.7.4 Fit summary (cat-spec)
out.sttran.seq.rurses[[1]]## Family: sratio
## Links: mu = cloglog; disc = identity
## Formula: sttranw1f ~ cs(rural.ses.med2)
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept[1] -3.50 0.50 -4.58 -2.63 1.00
## Intercept[2] -1.26 0.18 -1.66 -0.92 1.00
## Intercept[3] -0.22 0.14 -0.52 0.05 1.00
## Intercept[4] 0.39 0.20 -0.01 0.76 1.00
## rural.ses.med2RuralDLowSES[1] -0.83 0.61 -2.06 0.33 1.00
## rural.ses.med2RuralDLowSES[2] -0.55 0.24 -1.00 -0.07 1.00
## rural.ses.med2RuralDLowSES[3] 0.30 0.24 -0.16 0.78 1.00
## rural.ses.med2RuralDLowSES[4] -0.27 0.29 -0.81 0.29 1.00
## rural.ses.med2UrbanDHighSES[1] -0.87 0.61 -2.11 0.27 1.00
## rural.ses.med2UrbanDHighSES[2] 0.51 0.32 -0.08 1.17 1.00
## rural.ses.med2UrbanDHighSES[3] 0.06 0.21 -0.36 0.47 1.00
## rural.ses.med2UrbanDHighSES[4] 0.89 0.31 0.31 1.52 1.00
## rural.ses.med2UrbanDLowSES[1] 1.01 0.93 -0.69 2.98 1.00
## rural.ses.med2UrbanDLowSES[2] 0.04 0.26 -0.47 0.53 1.00
## rural.ses.med2UrbanDLowSES[3] 0.66 0.22 0.25 1.09 1.00
## rural.ses.med2UrbanDLowSES[4] 0.27 0.25 -0.22 0.76 1.00
## Bulk_ESS Tail_ESS
## Intercept[1] 2398 2296
## Intercept[2] 2437 2207
## Intercept[3] 2899 2995
## Intercept[4] 2545 2981
## rural.ses.med2RuralDLowSES[1] 2595 2439
## rural.ses.med2RuralDLowSES[2] 2703 2655
## rural.ses.med2RuralDLowSES[3] 3604 3082
## rural.ses.med2RuralDLowSES[4] 2835 2806
## rural.ses.med2UrbanDHighSES[1] 2564 2617
## rural.ses.med2UrbanDHighSES[2] 3027 2573
## rural.ses.med2UrbanDHighSES[3] 3326 3129
## rural.ses.med2UrbanDHighSES[4] 2904 2892
## rural.ses.med2UrbanDLowSES[1] 2845 2497
## rural.ses.med2UrbanDLowSES[2] 2914 2939
## rural.ses.med2UrbanDLowSES[3] 3320 3502
## rural.ses.med2UrbanDLowSES[4] 2812 2896
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.7.5 Prior summary (cat-spec)
out.sttran.seq.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## normal(0, 5) b
## normal(0, 5) b rural.ses.med2RuralDLowSES
## normal(0, 5) b rural.ses.med2UrbanDHighSES
## normal(0, 5) b rural.ses.med2UrbanDLowSES
## normal(0, 5) Intercept
## normal(0, 5) Intercept 1
## normal(0, 5) Intercept 2
## normal(0, 5) Intercept 3
## normal(0, 5) Intercept 4
## source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.8 strespw1: Cumul. probit, full & grouped
#Ordinal cumulative probit models predicting stress about respect at T1
#Items: stmonyw1 sttranw1 *strespw1* stfairw1 stjobw1 stthftw1 stmugw1
#Stress about respect at T1 - Thresholds only, Cumulative probit model:
stresp.ord <-
brm(data = stress.wide2,
family = cumulative("probit"),
strespw1f ~ 1,
prior(normal(0, 4), class = Intercept), #see kruschke/kurz re: priors for this model
iter = 2000, warmup = 1000,
chains = 4,
cores = nCoresphys,
backend = "cmdstanr",
#threads = threading(2), #consider within-chain threading if machine has at least 8 physical cores
seed = 8675309,
file = "Models/stresp.ord",
file_refit = "on_change"
)
out.stresp.ord <- ppchecks(stresp.ord) #posterior predictive checks
#Stress about respect at T1: Thresholds + group differences, Cumul probit
stresp.ord.rurses <- update(stresp.ord,
formula. = ~ 1 + rural.ses.med2,
newdata = stress.wide2,
file = "Models/stresp.ord.rurses",
file_refit = "on_change"
)
out.stresp.ord.rurses <- ppchecks(stresp.ord.rurses) #posterior predictive checks3.1.4.8.1 PPcheck(bars, full sample)
pp_check(stresp.ord, type="bars", ndraws=1000, cores=1) +
scale_x_continuous("Stress abt respect, T1", breaks = 1:5) +
scale_y_continuous(NULL, breaks = NULL, expand = expansion(mult = c(0, 0.05))) +
ggtitle("Data with posterior predictions (cumul-ord, fullsamp)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.9, .8))
3.1.4.8.2 PPcheck(bars, grouped)
pp_check(stresp.ord.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt respect, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cumul-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.8.3 PPcheck(density, full sample)
out.stresp.ord[[3]]
3.1.4.8.4 PPcheck(density, grouped)
out.stresp.ord.rurses[[3]]
3.1.4.8.5 Conditional effects (grouped)
conditional_effects(stresp.ord.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.8.6 Fit summary (full sample)
out.stresp.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: strespw1f ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -1.02 0.07 -1.15 -0.89 1.00 2767 2619
## Intercept[2] -0.43 0.06 -0.55 -0.32 1.00 4400 3214
## Intercept[3] -0.05 0.06 -0.16 0.06 1.00 3625 3152
## Intercept[4] 0.59 0.06 0.47 0.71 1.00 4459 3396
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.8.7 Prior summary (full sample)
out.stresp.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)3.1.4.8.8 Fit summary (grouped)
out.stresp.ord.rurses[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: strespw1f ~ rural.ses.med2
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -0.82 0.11 -1.03 -0.62 1.00 3033
## Intercept[2] -0.23 0.10 -0.43 -0.03 1.00 3506
## Intercept[3] 0.15 0.10 -0.04 0.34 1.00 3436
## Intercept[4] 0.80 0.10 0.60 0.99 1.00 3215
## rural.ses.med2RuralDLowSES 0.20 0.14 -0.07 0.47 1.00 3656
## rural.ses.med2UrbanDHighSES 0.50 0.14 0.23 0.78 1.00 3436
## rural.ses.med2UrbanDLowSES 0.17 0.13 -0.08 0.43 1.00 3582
## Tail_ESS
## Intercept[1] 2916
## Intercept[2] 2974
## Intercept[3] 3178
## Intercept[4] 3061
## rural.ses.med2RuralDLowSES 2788
## rural.ses.med2UrbanDHighSES 3039
## rural.ses.med2UrbanDLowSES 3074
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.8.9 Prior summary (grouped)
out.stresp.ord.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## (flat) b rural.ses.med2RuralDLowSES
## (flat) b rural.ses.med2UrbanDHighSES
## (flat) b rural.ses.med2UrbanDLowSES
## normal(0, 4) Intercept
## normal(0, 4) Intercept 1
## normal(0, 4) Intercept 2
## normal(0, 4) Intercept 3
## normal(0, 4) Intercept 4
## source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.9 strespw1: Seq. probit, cat-specific, grouped
#See if sequential model w/category specific effects improves recovery of observed data
stresp.seq.rurses <- update(stresp.ord,
formula. = ~ 1 + cs(rural.ses.med2),
prior = prior_seq,
newdata = stress.wide2,
family = sratio("cloglog"),
file = "Models/stresp.seq.rurses",
file_refit = "on_change"
)
out.stresp.seq.rurses <- ppchecks(stresp.seq.rurses) #posterior predictive checks3.1.4.9.1 PPcheck(bars, cat-spec)
pp_check(stresp.seq.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt respect, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cat-spec-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.9.2 PPcheck(density, cat-spec)
out.stresp.seq.rurses[[3]]
3.1.4.9.3 Conditional effects (cat-spec)
conditional_effects(stresp.seq.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.9.4 Fit summary (cat-spec)
out.stresp.seq.rurses[[1]]## Family: sratio
## Links: mu = cloglog; disc = identity
## Formula: strespw1f ~ cs(rural.ses.med2)
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept[1] -1.47 0.19 -1.86 -1.11 1.00
## Intercept[2] -1.62 0.24 -2.12 -1.17 1.00
## Intercept[3] -1.33 0.23 -1.81 -0.91 1.00
## Intercept[4] 0.09 0.17 -0.24 0.41 1.00
## rural.ses.med2RuralDLowSES[1] 1.08 0.39 0.35 1.87 1.00
## rural.ses.med2RuralDLowSES[2] -0.49 0.31 -1.09 0.11 1.00
## rural.ses.med2RuralDLowSES[3] -0.25 0.32 -0.88 0.38 1.00
## rural.ses.med2RuralDLowSES[4] 0.50 0.26 -0.01 1.00 1.00
## rural.ses.med2UrbanDHighSES[1] 1.03 0.39 0.31 1.85 1.00
## rural.ses.med2UrbanDHighSES[2] -0.17 0.33 -0.81 0.49 1.00
## rural.ses.med2UrbanDHighSES[3] 0.59 0.39 -0.14 1.38 1.00
## rural.ses.med2UrbanDHighSES[4] 0.98 0.26 0.48 1.50 1.00
## rural.ses.med2UrbanDLowSES[1] -0.07 0.26 -0.59 0.44 1.00
## rural.ses.med2UrbanDLowSES[2] 0.08 0.34 -0.58 0.76 1.00
## rural.ses.med2UrbanDLowSES[3] 0.17 0.33 -0.48 0.80 1.00
## rural.ses.med2UrbanDLowSES[4] 0.85 0.26 0.34 1.35 1.00
## Bulk_ESS Tail_ESS
## Intercept[1] 3119 3054
## Intercept[2] 2224 2330
## Intercept[3] 2371 2269
## Intercept[4] 2509 2964
## rural.ses.med2RuralDLowSES[1] 3933 3305
## rural.ses.med2RuralDLowSES[2] 2501 2798
## rural.ses.med2RuralDLowSES[3] 2810 2709
## rural.ses.med2RuralDLowSES[4] 2997 3130
## rural.ses.med2UrbanDHighSES[1] 3577 2599
## rural.ses.med2UrbanDHighSES[2] 2597 2633
## rural.ses.med2UrbanDHighSES[3] 3094 3152
## rural.ses.med2UrbanDHighSES[4] 3176 2197
## rural.ses.med2UrbanDLowSES[1] 3272 2796
## rural.ses.med2UrbanDLowSES[2] 2501 2693
## rural.ses.med2UrbanDLowSES[3] 2877 2261
## rural.ses.med2UrbanDLowSES[4] 2765 3140
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.9.5 Prior summary (cat-spec)
out.stresp.seq.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## normal(0, 5) b
## normal(0, 5) b rural.ses.med2RuralDLowSES
## normal(0, 5) b rural.ses.med2UrbanDHighSES
## normal(0, 5) b rural.ses.med2UrbanDLowSES
## normal(0, 5) Intercept
## normal(0, 5) Intercept 1
## normal(0, 5) Intercept 2
## normal(0, 5) Intercept 3
## normal(0, 5) Intercept 4
## source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.10 stfairw1: Cumul. probit, full & grouped
#Ordinal cumulative probit models predicting stress about fair treatment at T1
#Items: stmonyw1 sttranw1 strespw1 *stfairw1* stjobw1 stthftw1 stmugw1
#Stress about fair treatment at T1 - Thresholds only, Cumulative probit model:
stfair.ord <-
brm(data = stress.wide2,
family = cumulative("probit"),
stfairw1f ~ 1,
prior(normal(0, 4), class = Intercept), #see kruschke/kurz re: priors for this model
iter = 2000, warmup = 1000,
chains = 4,
cores = nCoresphys,
backend = "cmdstanr",
#threads = threading(2), #consider within-chain threading if machine has at least 8 physical cores
seed = 8675309,
file = "Models/stfair.ord",
file_refit = "on_change"
)
out.stfair.ord <- ppchecks(stfair.ord) #posterior predictive checks
#Stress about fair treatment at T1: Thresholds + group differences, Cumul probit
stfair.ord.rurses <- update(stfair.ord,
formula. = ~ 1 + rural.ses.med2,
newdata = stress.wide2,
file = "Models/stfair.ord.rurses",
file_refit = "on_change"
)
out.stfair.ord.rurses <- ppchecks(stfair.ord.rurses) #posterior predictive checks3.1.4.10.1 PPcheck(bars, full sample)
pp_check(stfair.ord, type="bars", ndraws=1000, cores=1) +
scale_x_continuous("Stress abt fair trtmt, T1", breaks = 1:5) +
scale_y_continuous(NULL, breaks = NULL, expand = expansion(mult = c(0, 0.05))) +
ggtitle("Data with posterior predictions (cumul-ord, fullsamp)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.9, .8))
3.1.4.10.2 PPcheck(bars, grouped)
pp_check(stfair.ord.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt fair trtmt, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cumul-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.10.3 PPcheck(density, full sample)
out.stfair.ord[[3]]
3.1.4.10.4 PPcheck(density, grouped)
out.stfair.ord.rurses[[3]]
3.1.4.10.5 Conditional effects (grouped)
conditional_effects(stfair.ord.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.10.6 Fit summary (full sample)
out.stfair.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stfairw1f ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -1.03 0.07 -1.16 -0.90 1.00 2990 2819
## Intercept[2] -0.43 0.06 -0.54 -0.31 1.00 4615 3790
## Intercept[3] -0.14 0.06 -0.25 -0.03 1.00 4787 3504
## Intercept[4] 0.43 0.06 0.32 0.55 1.00 5003 3324
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.10.7 Prior summary (full sample)
out.stfair.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)3.1.4.10.8 Fit summary (grouped)
out.stfair.ord.rurses[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stfairw1f ~ rural.ses.med2
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -0.94 0.11 -1.15 -0.72 1.00 2522
## Intercept[2] -0.33 0.10 -0.54 -0.13 1.00 2847
## Intercept[3] -0.04 0.10 -0.24 0.16 1.00 2877
## Intercept[4] 0.54 0.10 0.35 0.75 1.00 2968
## rural.ses.med2RuralDLowSES -0.06 0.14 -0.32 0.21 1.00 2869
## rural.ses.med2UrbanDHighSES 0.38 0.15 0.10 0.67 1.00 3420
## rural.ses.med2UrbanDLowSES 0.12 0.13 -0.16 0.38 1.00 3122
## Tail_ESS
## Intercept[1] 2887
## Intercept[2] 3003
## Intercept[3] 3092
## Intercept[4] 3030
## rural.ses.med2RuralDLowSES 2878
## rural.ses.med2UrbanDHighSES 3055
## rural.ses.med2UrbanDLowSES 2780
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.10.9 Prior summary (grouped)
out.stfair.ord.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## (flat) b rural.ses.med2RuralDLowSES
## (flat) b rural.ses.med2UrbanDHighSES
## (flat) b rural.ses.med2UrbanDLowSES
## normal(0, 4) Intercept
## normal(0, 4) Intercept 1
## normal(0, 4) Intercept 2
## normal(0, 4) Intercept 3
## normal(0, 4) Intercept 4
## source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.11 stfairw1: Seq. probit, cat-specific, grouped
#See if sequential model w/category specific effects improves recovery of observed data
stfair.seq.rurses <- update(stfair.ord,
formula. = ~ 1 + cs(rural.ses.med2),
prior = prior_seq,
newdata = stress.wide2,
family = sratio("cloglog"),
file = "Models/stfair.seq.rurses",
file_refit = "on_change"
)
out.stfair.seq.rurses <- ppchecks(stfair.seq.rurses) #posterior predictive checks3.1.4.11.1 PPcheck(bars, cat-spec)
pp_check(stfair.seq.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt fair trtmt, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cat-spec-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.11.2 PPcheck(density, cat-spec)
out.stfair.seq.rurses[[3]]
3.1.4.11.3 Conditional effects (cat-spec)
conditional_effects(stfair.seq.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.11.4 Fit summary (cat-spec)
out.stfair.seq.rurses[[1]]## Family: sratio
## Links: mu = cloglog; disc = identity
## Formula: stfairw1f ~ cs(rural.ses.med2)
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept[1] -1.65 0.22 -2.10 -1.25 1.00
## Intercept[2] -1.29 0.20 -1.69 -0.92 1.00
## Intercept[3] -2.25 0.36 -3.01 -1.60 1.00
## Intercept[4] -0.40 0.17 -0.75 -0.07 1.00
## rural.ses.med2RuralDLowSES[1] 0.33 0.34 -0.31 1.02 1.00
## rural.ses.med2RuralDLowSES[2] -0.18 0.28 -0.73 0.37 1.00
## rural.ses.med2RuralDLowSES[3] -1.14 0.42 -1.98 -0.34 1.00
## rural.ses.med2RuralDLowSES[4] 0.11 0.27 -0.41 0.65 1.00
## rural.ses.med2UrbanDHighSES[1] 0.85 0.40 0.11 1.66 1.00
## rural.ses.med2UrbanDHighSES[2] 0.40 0.31 -0.21 1.02 1.00
## rural.ses.med2UrbanDHighSES[3] 0.08 0.52 -0.94 1.11 1.00
## rural.ses.med2UrbanDHighSES[4] 0.32 0.25 -0.18 0.83 1.00
## rural.ses.med2UrbanDLowSES[1] -0.14 0.29 -0.71 0.41 1.00
## rural.ses.med2UrbanDLowSES[2] 0.45 0.31 -0.15 1.07 1.00
## rural.ses.med2UrbanDLowSES[3] -0.51 0.44 -1.39 0.33 1.00
## rural.ses.med2UrbanDLowSES[4] 0.65 0.27 0.14 1.19 1.00
## Bulk_ESS Tail_ESS
## Intercept[1] 2678 2655
## Intercept[2] 2599 2473
## Intercept[3] 2172 2272
## Intercept[4] 2595 2873
## rural.ses.med2RuralDLowSES[1] 3114 3057
## rural.ses.med2RuralDLowSES[2] 3010 2971
## rural.ses.med2RuralDLowSES[3] 2479 2407
## rural.ses.med2RuralDLowSES[4] 3335 3305
## rural.ses.med2UrbanDHighSES[1] 3459 2701
## rural.ses.med2UrbanDHighSES[2] 3324 3294
## rural.ses.med2UrbanDHighSES[3] 2687 2556
## rural.ses.med2UrbanDHighSES[4] 3186 3034
## rural.ses.med2UrbanDLowSES[1] 3067 3034
## rural.ses.med2UrbanDLowSES[2] 3158 2744
## rural.ses.med2UrbanDLowSES[3] 2428 2479
## rural.ses.med2UrbanDLowSES[4] 3330 3203
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.11.5 Prior summary (cat-spec)
out.stfair.seq.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## normal(0, 5) b
## normal(0, 5) b rural.ses.med2RuralDLowSES
## normal(0, 5) b rural.ses.med2UrbanDHighSES
## normal(0, 5) b rural.ses.med2UrbanDLowSES
## normal(0, 5) Intercept
## normal(0, 5) Intercept 1
## normal(0, 5) Intercept 2
## normal(0, 5) Intercept 3
## normal(0, 5) Intercept 4
## source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.12 stjobw1: Cumul. probit, full & grouped
#Ordinal cumulative probit models predicting stress about job at T1
#Items: stmonyw1 sttranw1 strespw1 stfairw1 *stjobw1* stthftw1 stmugw1
#Stress about job at T1 - Thresholds only, Cumulative probit model:
stjob.ord <-
brm(data = stress.wide2,
family = cumulative("probit"),
stjobw1f ~ 1,
prior(normal(0, 4), class = Intercept), #see kruschke/kurz re: priors for this model
iter = 2000, warmup = 1000,
chains = 4,
cores = nCoresphys,
backend = "cmdstanr",
#threads = threading(2), #consider within-chain threading if machine has at least 8 physical cores
seed = 8675309,
file = "Models/stjob.ord",
file_refit = "on_change"
)
out.stjob.ord <- ppchecks(stjob.ord) #posterior predictive checks
#Stress about job at T1: Thresholds + group differences, Cumul probit
stjob.ord.rurses <- update(stjob.ord,
formula. = ~ 1 + rural.ses.med2,
newdata = stress.wide2,
file = "Models/stjob.ord.rurses",
file_refit = "on_change"
)
out.stjob.ord.rurses <- ppchecks(stjob.ord.rurses) #posterior predictive checks3.1.4.12.1 PPcheck(bars, full sample)
pp_check(stjob.ord, type="bars", ndraws=1000, cores=1) +
scale_x_continuous("Stress abt job, T1", breaks = 1:5) +
scale_y_continuous(NULL, breaks = NULL, expand = expansion(mult = c(0, 0.05))) +
ggtitle("Data with posterior predictions (cumul-ord, fullsamp)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.9, .8))
3.1.4.12.2 PPcheck(bars, grouped)
pp_check(stjob.ord.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt job, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cumul-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.12.3 PPcheck(density, full sample)
out.stjob.ord[[3]]
3.1.4.12.4 PPcheck(density, grouped)
out.stjob.ord.rurses[[3]]
3.1.4.12.5 Conditional effects (grouped)
conditional_effects(stjob.ord.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.12.6 Fit summary (full sample)
out.stjob.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stjobw1f ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -0.98 0.07 -1.12 -0.85 1.00 3114 2588
## Intercept[2] -0.48 0.06 -0.60 -0.37 1.00 4493 3359
## Intercept[3] 0.02 0.06 -0.09 0.13 1.00 4673 3418
## Intercept[4] 0.65 0.06 0.53 0.77 1.00 5329 3236
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.12.7 Prior summary (full sample)
out.stjob.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)3.1.4.12.8 Fit summary (grouped)
out.stjob.ord.rurses[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stjobw1f ~ rural.ses.med2
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -0.81 0.11 -1.02 -0.61 1.00 2957
## Intercept[2] -0.32 0.10 -0.51 -0.12 1.00 3321
## Intercept[3] 0.19 0.10 -0.00 0.39 1.00 3172
## Intercept[4] 0.84 0.10 0.64 1.05 1.00 3080
## rural.ses.med2RuralDLowSES 0.03 0.13 -0.24 0.28 1.00 3443
## rural.ses.med2UrbanDHighSES 0.39 0.14 0.12 0.66 1.00 3224
## rural.ses.med2UrbanDLowSES 0.31 0.13 0.06 0.56 1.00 3170
## Tail_ESS
## Intercept[1] 2806
## Intercept[2] 3421
## Intercept[3] 3497
## Intercept[4] 3037
## rural.ses.med2RuralDLowSES 3078
## rural.ses.med2UrbanDHighSES 2918
## rural.ses.med2UrbanDLowSES 2995
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.12.9 Prior summary (grouped)
out.stjob.ord.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## (flat) b rural.ses.med2RuralDLowSES
## (flat) b rural.ses.med2UrbanDHighSES
## (flat) b rural.ses.med2UrbanDLowSES
## normal(0, 4) Intercept
## normal(0, 4) Intercept 1
## normal(0, 4) Intercept 2
## normal(0, 4) Intercept 3
## normal(0, 4) Intercept 4
## source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.13 stjobw1: Seq. probit, cat-specific, grouped
#See if sequential model w/category specific effects improves recovery of observed data
stjob.seq.rurses <- update(stjob.ord,
formula. = ~ 1 + cs(rural.ses.med2),
prior = prior_seq,
newdata = stress.wide2,
family = sratio("cloglog"),
file = "Models/stjob.seq.rurses",
file_refit = "on_change"
)
out.stjob.seq.rurses <- ppchecks(stjob.seq.rurses) #posterior predictive checks3.1.4.13.1 PPcheck(bars, cat-spec)
pp_check(stjob.seq.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt job, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cat-spec-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.13.2 PPcheck(density, cat-spec)
out.stjob.seq.rurses[[3]]
3.1.4.13.3 Conditional effects (cat-spec)
conditional_effects(stjob.seq.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.13.4 Fit summary (cat-spec)
out.stjob.seq.rurses[[1]]## Family: sratio
## Links: mu = cloglog; disc = identity
## Formula: stjobw1f ~ cs(rural.ses.med2)
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept[1] -1.94 0.24 -2.45 -1.50 1.00
## Intercept[2] -1.30 0.20 -1.72 -0.93 1.00
## Intercept[3] -0.58 0.17 -0.92 -0.26 1.00
## Intercept[4] -0.14 0.20 -0.55 0.23 1.00
## rural.ses.med2RuralDLowSES[1] -0.34 0.33 -0.98 0.29 1.00
## rural.ses.med2RuralDLowSES[2] 0.71 0.35 0.04 1.40 1.00
## rural.ses.med2RuralDLowSES[3] 0.10 0.24 -0.35 0.59 1.00
## rural.ses.med2RuralDLowSES[4] -0.11 0.28 -0.64 0.45 1.00
## rural.ses.med2UrbanDHighSES[1] -0.44 0.32 -1.08 0.18 1.00
## rural.ses.med2UrbanDHighSES[2] 0.95 0.40 0.18 1.74 1.00
## rural.ses.med2UrbanDHighSES[3] 1.33 0.36 0.66 2.06 1.00
## rural.ses.med2UrbanDHighSES[4] 0.72 0.29 0.15 1.29 1.00
## rural.ses.med2UrbanDLowSES[1] -0.01 0.33 -0.66 0.61 1.00
## rural.ses.med2UrbanDLowSES[2] 0.12 0.28 -0.42 0.67 1.00
## rural.ses.med2UrbanDLowSES[3] 1.16 0.31 0.57 1.76 1.00
## rural.ses.med2UrbanDLowSES[4] 0.48 0.26 -0.03 1.00 1.00
## Bulk_ESS Tail_ESS
## Intercept[1] 2548 2579
## Intercept[2] 2967 2276
## Intercept[3] 3201 3164
## Intercept[4] 2229 2560
## rural.ses.med2RuralDLowSES[1] 2857 2796
## rural.ses.med2RuralDLowSES[2] 3268 3323
## rural.ses.med2RuralDLowSES[3] 3472 3233
## rural.ses.med2RuralDLowSES[4] 2505 3126
## rural.ses.med2UrbanDHighSES[1] 2813 2645
## rural.ses.med2UrbanDHighSES[2] 3635 2756
## rural.ses.med2UrbanDHighSES[3] 3615 2948
## rural.ses.med2UrbanDHighSES[4] 2672 2590
## rural.ses.med2UrbanDLowSES[1] 2988 2736
## rural.ses.med2UrbanDLowSES[2] 3454 2420
## rural.ses.med2UrbanDLowSES[3] 3649 3009
## rural.ses.med2UrbanDLowSES[4] 2657 2754
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.13.5 Prior summary (cat-spec)
out.stjob.seq.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## normal(0, 5) b
## normal(0, 5) b rural.ses.med2RuralDLowSES
## normal(0, 5) b rural.ses.med2UrbanDHighSES
## normal(0, 5) b rural.ses.med2UrbanDLowSES
## normal(0, 5) Intercept
## normal(0, 5) Intercept 1
## normal(0, 5) Intercept 2
## normal(0, 5) Intercept 3
## normal(0, 5) Intercept 4
## source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.14 stthftw1: Cumul. probit, full & grouped
#Ordinal cumulative probit models predicting stress about job at T1
#Items: stmonyw1 sttranw1 strespw1 stfairw1 stjobw1 *stthftw1* stmugw1
#Stress about theft at T1 - Thresholds only, Cumulative probit model:
stthft.ord <-
brm(data = stress.wide2,
family = cumulative("probit"),
stthftw1 ~ 1,
prior(normal(0, 4), class = Intercept), #see kruschke/kurz re: priors for this model
iter = 2000, warmup = 1000,
chains = 4,
cores = nCoresphys,
backend = "cmdstanr",
#threads = threading(2), #consider within-chain threading if machine has at least 8 physical cores
seed = 8675309,
file = "Models/stthft.ord",
file_refit = "on_change"
)
out.stthft.ord <- ppchecks(stthft.ord) #posterior predictive checks
#Stress about theft at T1: Thresholds + group differences, Cumul probit
stthft.ord.rurses <- update(stthft.ord,
formula. = ~ 1 + rural.ses.med2,
newdata = stress.wide2,
file = "Models/stthft.ord.rurses",
file_refit = "on_change"
)
out.stthft.ord.rurses <- ppchecks(stthft.ord.rurses) #posterior predictive checks3.1.4.14.1 PPcheck(bars, full sample)
pp_check(stthft.ord, type="bars", ndraws=1000, cores=1) +
scale_x_continuous("Stress abt theft, T1", breaks = 1:5) +
scale_y_continuous(NULL, breaks = NULL, expand = expansion(mult = c(0, 0.05))) +
ggtitle("Data with posterior predictions (cumul-ord, fullsamp)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.9, .8))
3.1.4.14.2 PPcheck(bars, grouped)
pp_check(stthft.ord.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt theft, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cumul-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.14.3 PPcheck(density, full sample)
out.stthft.ord[[3]]
3.1.4.14.4 PPcheck(density, grouped)
out.stthft.ord.rurses[[3]]
3.1.4.14.5 Conditional effects (grouped)
conditional_effects(stthft.ord.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.14.6 Fit summary (full sample)
out.stthft.ord[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stthftw1 ~ 1
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept[1] -0.42 0.06 -0.53 -0.30 1.00 3576 3138
## Intercept[2] 0.34 0.06 0.23 0.45 1.00 4996 3305
## Intercept[3] 1.02 0.07 0.89 1.16 1.00 4174 2856
## Intercept[4] 1.71 0.10 1.52 1.91 1.00 4521 3148
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.14.7 Prior summary (full sample)
out.stthft.ord[[2]]## prior class coef group resp dpar nlpar lb ub source
## normal(0, 4) Intercept user
## normal(0, 4) Intercept 1 (vectorized)
## normal(0, 4) Intercept 2 (vectorized)
## normal(0, 4) Intercept 3 (vectorized)
## normal(0, 4) Intercept 4 (vectorized)3.1.4.14.8 Fit summary (grouped)
out.stthft.ord.rurses[[1]]## Family: cumulative
## Links: mu = probit; disc = identity
## Formula: stthftw1 ~ rural.ses.med2
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -0.35 0.10 -0.55 -0.16 1.00 3479
## Intercept[2] 0.42 0.10 0.23 0.61 1.00 3840
## Intercept[3] 1.14 0.11 0.93 1.35 1.00 3739
## Intercept[4] 1.86 0.14 1.60 2.14 1.00 4251
## rural.ses.med2RuralDLowSES -0.31 0.14 -0.59 -0.04 1.00 3674
## rural.ses.med2UrbanDHighSES 0.27 0.14 -0.01 0.55 1.00 3799
## rural.ses.med2UrbanDLowSES 0.32 0.13 0.06 0.57 1.00 3539
## Tail_ESS
## Intercept[1] 2963
## Intercept[2] 3244
## Intercept[3] 3772
## Intercept[4] 3419
## rural.ses.med2RuralDLowSES 2865
## rural.ses.med2UrbanDHighSES 3179
## rural.ses.med2UrbanDLowSES 3114
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.14.9 Prior summary (grouped)
out.stthft.ord.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## (flat) b rural.ses.med2RuralDLowSES
## (flat) b rural.ses.med2UrbanDHighSES
## (flat) b rural.ses.med2UrbanDLowSES
## normal(0, 4) Intercept
## normal(0, 4) Intercept 1
## normal(0, 4) Intercept 2
## normal(0, 4) Intercept 3
## normal(0, 4) Intercept 4
## source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)3.1.4.15 stthftw1: Seq. probit, cat-specific, grouped
#See if sequential model w/category specific effects improves recovery of observed data
stthft.seq.rurses <- update(stthft.ord,
formula. = ~ 1 + cs(rural.ses.med2),
prior = prior_seq,
newdata = stress.wide2,
family = sratio("cloglog"),
file = "Models/stthft.seq.rurses",
file_refit = "on_change"
)
out.stthft.seq.rurses <- ppchecks(stthft.seq.rurses) #posterior predictive checks3.1.4.15.1 PPcheck(bars, cat-spec)
pp_check(stthft.seq.rurses, type="bars_grouped", group = "rural.ses.med2", ndraws=1000, cores = 1) +
scale_x_continuous("Stress abt theft, T1", breaks = 1:5) +
ggtitle("Data with posterior predictions (cat-spec-ord)",
subtitle = "N = 489") +
theme(legend.background = element_blank(),
legend.position = c(.5, 1))
3.1.4.15.2 PPcheck(density, cat-spec)
out.stjob.seq.rurses[[3]]
3.1.4.15.3 Conditional effects (cat-spec)
conditional_effects(stthft.seq.rurses, "rural.ses.med2", categorical = TRUE)
3.1.4.15.4 Fit summary (cat-spec)
out.stthft.seq.rurses[[1]]## Family: sratio
## Links: mu = cloglog; disc = identity
## Formula: stthftw1 ~ cs(rural.ses.med2)
## Data: stress.wide2 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept[1] -1.04 0.16 -1.38 -0.73 1.00
## Intercept[2] -0.29 0.15 -0.61 -0.00 1.00
## Intercept[3] 0.47 0.20 0.06 0.85 1.00
## Intercept[4] 0.71 0.44 -0.19 1.54 1.00
## rural.ses.med2RuralDLowSES[1] -0.41 0.21 -0.82 0.01 1.00
## rural.ses.med2RuralDLowSES[2] -0.38 0.22 -0.79 0.07 1.00
## rural.ses.med2RuralDLowSES[3] -0.39 0.35 -1.08 0.29 1.00
## rural.ses.med2RuralDLowSES[4] -3.58 3.21 -11.15 1.02 1.00
## rural.ses.med2UrbanDHighSES[1] -0.09 0.24 -0.55 0.38 1.00
## rural.ses.med2UrbanDHighSES[2] 0.70 0.26 0.20 1.23 1.00
## rural.ses.med2UrbanDHighSES[3] 0.98 0.29 0.39 1.55 1.00
## rural.ses.med2UrbanDHighSES[4] 0.32 0.50 -0.68 1.28 1.00
## rural.ses.med2UrbanDLowSES[1] -0.11 0.22 -0.55 0.32 1.00
## rural.ses.med2UrbanDLowSES[2] 0.68 0.24 0.21 1.16 1.00
## rural.ses.med2UrbanDLowSES[3] 1.08 0.28 0.52 1.64 1.00
## rural.ses.med2UrbanDLowSES[4] 0.81 0.49 -0.16 1.74 1.00
## Bulk_ESS Tail_ESS
## Intercept[1] 2479 2199
## Intercept[2] 2684 2474
## Intercept[3] 2517 2517
## Intercept[4] 2265 2550
## rural.ses.med2RuralDLowSES[1] 2958 2814
## rural.ses.med2RuralDLowSES[2] 3150 2871
## rural.ses.med2RuralDLowSES[3] 3402 3030
## rural.ses.med2RuralDLowSES[4] 2945 2037
## rural.ses.med2UrbanDHighSES[1] 2881 2888
## rural.ses.med2UrbanDHighSES[2] 3195 2876
## rural.ses.med2UrbanDHighSES[3] 3037 3000
## rural.ses.med2UrbanDHighSES[4] 2544 2542
## rural.ses.med2UrbanDLowSES[1] 2777 2902
## rural.ses.med2UrbanDLowSES[2] 3173 3140
## rural.ses.med2UrbanDLowSES[3] 2917 2879
## rural.ses.med2UrbanDLowSES[4] 2396 2712
##
## Further Distributional Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).3.1.4.15.5 Prior summary (cat-spec)
out.stthft.seq.rurses[[2]]## prior class coef group resp dpar nlpar lb ub
## normal(0, 5) b
## normal(0, 5) b rural.ses.med2RuralDLowSES
## normal(0, 5) b rural.ses.med2UrbanDHighSES
## normal(0, 5) b rural.ses.med2UrbanDLowSES
## normal(0, 5) Intercept
## normal(0, 5) Intercept 1
## normal(0, 5) Intercept 2
## normal(0, 5) Intercept 3
## normal(0, 5) Intercept 4
## source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)As with the money and assault stress variables, the cumulative probit model recovered stress item response distributions in the full sample and generally did an adequate but not perfect job of recovering grouped item distributions. In contrast, the sequential probit category-specific model recovered observed response in all four groups and for all five categories across all seven stress response items as expected. Thus, the threshold estimates and uncertainty from posterior distributions from each of the category-specific models above were used to plot Figure 2.
3.1.5 Visualizing ordinal thresholds
Before we created Figure 2 to help visualizing posterior predicted distributions for the stress response scales generated by the thresholds from our category-specific ordinal models, we also generated some plots to help readers try visualizing the thresholds themselves. Below, we plot the posterior distributions that generated the threshold estimates in the cumulative model applied to the full sample for each of the stress items.
Before doing so, we set a colorblind-friendly color palette.
#Change color palette
sl <- scico(palette = "lajolla", n = 9)
scales::show_col(sl)
We begin by plotting thresholds from the first stress item
(stress about money, or stmonyw1f). First, though,
it is well past time that I gave a huge shout-out to John Kruschke’s Doing
Bayesian Data Analysis and Solomon
Kurz’s bookdown translation to brms for explaining why/how to do all
this and generously sharing all their code publicly so I could copy and
apply it in my own work. None of this would be possible without the
tireless efforts and steadfast commitment to open science practices of
these scholars and of several others (e.g., Richard McElreath; Paul
Buerkner; Matthew Kay; Andrew Heiss).
#wrangle posterior
# posterior_samples(stmony.ord) # posterior_samples function is deprecated
postmony <-
as_draws_df(stmony.ord) %>%
dplyr::select(`b_Intercept[1]`:`b_Intercept[4]`, .iteration)
glimpse(postmony) ## Rows: 4,000
## Columns: 5
## $ `b_Intercept[1]` <dbl> -1.3, -1.3, -1.4, -1.3, -1.5, -1.5, -1.5, -1.4, -1.4,…
## $ `b_Intercept[2]` <dbl> -0.57, -0.57, -0.61, -0.53, -0.48, -0.66, -0.62, -0.5…
## $ `b_Intercept[3]` <dbl> 0.237, 0.235, 0.199, 0.280, 0.352, 0.296, 0.202, 0.25…
## $ `b_Intercept[4]` <dbl> 1.2, 1.3, 1.2, 1.3, 1.3, 1.2, 1.4, 1.2, 1.4, 1.3, 1.3…
## $ .iteration <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16…#Plot thresholds ala Kruschke/Kurz: https://bookdown.org/content/3686/ordinal-predicted-variable.html#the-case-of-metric-predictors
#Stress about money
means <-
postmony %>%
summarise_at(vars(`b_Intercept[1]`:`b_Intercept[4]`), mean) %>%
pivot_longer(everything(),
values_to = "mean")
postmony %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold)) %>%
ggplot(aes(x = threshold, y = theta_bar, color = name)) +
geom_vline(data = means,
aes(xintercept = mean, color = name),
linetype = 2) +
geom_point(alpha = 1/10) +
scale_color_scico_d(palette = "lajolla", begin = .25) +
ylab("mean threshold") +
theme(legend.position = "none")
For comparison, the plot below displays the ordinal thresholds for all seven stress items; recall, all seven are measured on the same response scale (1=Never to 5=Very often).
#Plot thresholds ala Kruschke/Kurz: https://bookdown.org/content/3686/ordinal-predicted-variable.html#the-case-of-metric-predictors
#Wrangle posteriors
#Items: stmonyw1 sttranw1 strespw1 stfairw1 stjobw1 *stthftw1* stmugw1
#Stress abt money posterior wrangled above (postmony)
#Stress abt transport
posttran <-
as_draws_df(sttran.ord) %>%
dplyr::select(`b_Intercept[1]`:`b_Intercept[4]`, .iteration)
#Stress abt respect
postresp <-
as_draws_df(stresp.ord) %>%
dplyr::select(`b_Intercept[1]`:`b_Intercept[4]`, .iteration)
#Stress abt fair trtmt
postfair <-
as_draws_df(stfair.ord) %>%
dplyr::select(`b_Intercept[1]`:`b_Intercept[4]`, .iteration)
#Stress abt job
postjob <-
as_draws_df(stjob.ord) %>%
dplyr::select(`b_Intercept[1]`:`b_Intercept[4]`, .iteration)
#Stress abt theft victimization
postthft <-
as_draws_df(stthft.ord) %>%
dplyr::select(`b_Intercept[1]`:`b_Intercept[4]`, .iteration)
#Stress abt assault victimization
postmug <-
as_draws_df(stmug.ord) %>%
dplyr::select(`b_Intercept[1]`:`b_Intercept[4]`, .iteration)
p <-
bind_rows(
# money
postmony %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold)),
# transport
posttran %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold)),
# respect
postresp %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold)),
# fair trtmt
postfair %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold)),
# job
postjob %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold)),
# theft victimization
postthft %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold)),
# assault victimization
postmug %>%
gather(name, threshold, -.iteration) %>%
group_by(.iteration) %>%
mutate(theta_bar = mean(threshold))
) %>%
# add an index
mutate(model = rep(c("Stress: money", "Stress: transport", "Stress: Respect", "Stress: Fairness", "Stress: Job", "Stress: Theft", "Stress: Assault"), each = n() / 7))
# compute the means by model and threshold for the vertical lines
means <-
p %>%
ungroup() %>%
group_by(model, name) %>%
summarise(mean = mean(threshold))
# plot!
p %>%
ggplot(aes(x = threshold, y = theta_bar)) +
geom_vline(data = means,
aes(xintercept = mean, color = name),
linetype = 2) +
geom_point(aes(color = name),
alpha = 1/10, size = 1/2) +
scale_color_scico_d(palette = "lajolla", begin = .25) +
ylab("mean threshold") +
theme(legend.position = "none") +
facet_wrap(~model, ncol = 3, scales = "free")
As these plots show, the thresholds capture the approximate normality
of the stress about money item, with the midpoint of the
threshold values - where x-axis equals 0 - being located
near the midpoint of the plots and relative balance around zero.
Meanwhile, these thresholds also capture the high positive skew in the
victimization items (theft; assault), with the
midpoint (x-axis = 0) loaded on the far left end of the
plot and the remaining values spread out to the right of zero. For more
information, see (Kruschke; Kurz).
Now let’s turn to summarizing and visualizing group differences in stress using the posterior distributions and threshold estimates from the category specific models. We can begin by generating posterior predicted response probabilities for each group.
3.1.6 Generating predicted response probabilities for stress by community
#Posterior predicted category probabilities for all 489 observations
#pp <- predict(stmony.rurses.ord)
#head(pp)
#Fitted values - posterior predicted response probabilities by group
#"Rural/Low SES"
pred.rurlow <- fitted(stmony.seq.rurses,
newdata = tibble(rural.ses.med2 = "Rural/Low SES"))
head(pred.rurlow)## , , P(Y = 1)
##
## Estimate Est.Error Q2.5 Q97.5
## [1,] 0.12 0.031 0.069 0.19
##
## , , P(Y = 2)
##
## Estimate Est.Error Q2.5 Q97.5
## [1,] 0.18 0.035 0.12 0.26
##
## , , P(Y = 3)
##
## Estimate Est.Error Q2.5 Q97.5
## [1,] 0.4 0.045 0.31 0.49
##
## , , P(Y = 4)
##
## Estimate Est.Error Q2.5 Q97.5
## [1,] 0.22 0.038 0.15 0.3
##
## , , P(Y = 5)
##
## Estimate Est.Error Q2.5 Q97.5
## [1,] 0.074 0.025 0.034 0.13#"Rural/High SES"
pred.rurhi <- fitted(stmony.seq.rurses,
newdata = tibble(rural.ses.med2 = "Rural/High SES"))
#head(pred.rurhi)
#"Urban/Low SES"
pred.urblow <- fitted(stmony.seq.rurses,
newdata = tibble(rural.ses.med2 = "Urban/Low SES"))
#head(pred.urblow)
#"Urban/High SES"
pred.urbhi <- fitted(stmony.seq.rurses,
newdata = tibble(rural.ses.med2 = "Urban/High SES"))
#head(pred.urbhi)
#plot help
#https://solomonkurz.netlify.app/post/would-you-like-all-your-posteriors-in-one-plot/We could wrangle and plot these posterior predicted values, Instead, we plotted predicted stress response distributions directly from the posterior samples in Figure 2. This is the same as the Figure 2 found in the “Stress in Bangladesh” paper.
3.1.7 FIGURE 2: Stress by Community
# Set a color scale to colorblind-safe lajolla palette
# uses library(scico)
#Change color palette
sl <- scico::scico(palette = "lajolla", n = 9)
# scales::show_col(sl)#Instead of manually calculating conditional means & plotting, use "add_fitted_draws":
#http://mjskay.github.io/tidybayes/articles/tidy-brms.html
#See "esoph_plot" example...
#function to grab & save fig2 plot data as object for each stress item
get_pp_data <- function(fitdraw, varname, varlabel) {
stress.wide2 %>%
data_grid(rural.ses.med2) %>%
# add_fitted_draws(fitdraw) %>% #depricated - use add_epred_draws()
add_epred_draws(fitdraw) %>%
mutate(stress_fct = .category,
stress_var = varname,
stress_label = varlabel
)
}
#Get posterior prediction draws for Figure 2 plots for each stress item
#Can add directly to ggplot, though separate files may be useful for advanced faceting
stmony_pp_data <- get_pp_data(stmony.seq.rurses, "stmonyw1f","Money")
sttran_pp_data <- get_pp_data(sttran.seq.rurses, "sttranw1f","Transportation")
stresp_pp_data <- get_pp_data(stresp.seq.rurses, "strespw1f","Respect")
stfair_pp_data <- get_pp_data(stfair.seq.rurses, "stfairw1f","Fair treatment")
stjob_pp_data <- get_pp_data(stjob.seq.rurses, "stjobw1f","Job")
stthft_pp_data <- get_pp_data(stthft.seq.rurses, "stthftw1f","Theft")
stmug_pp_data <- get_pp_data(stmug.seq.rurses, "stmugw1f","Assault")
#Create label for community types
commlabs <- c("Rural/Low SES"="Rural/\nLowSES",
"Rural/High SES"="Rural/\nHighSES",
"Urban/Low SES"="Urban/\nLowSES",
"Urban/High SES"="Urban/\nHighSES")
#function to find & drop leading zeroes (used for x-axis label)
dropLeadingZero <- function(l){
str_replace(l, '0(?=.)', '')
}
#function to create same/similar plots for each stress item (to combine into Fig2)
#NOTE: To generalize to new/data & cols, see:
#https://www.infoworld.com/article/3410295/how-to-write-your-own-ggplot2-functions-in-r.html
#when treating colname as input, need to wrap in {{colname}}
#e.g., aes(x = {{myxcol}},y = {{myycol}}, fill = {{myycol}})
#then change to function(pp_data, myxcol, myycol, y_label)
#and function call like this: plot_item(stmony_pp_data, .value, stress_fct, "Money")
#NOTE: change .value (add_fitted_draws) to .epred (add_epred draws)
plot_item <- function(pp_data, y_label) {
ggplot(data = pp_data, aes(x = .epred, y = stress_fct, fill = stress_fct)) +
coord_cartesian(expand = FALSE, xlim=c(0,.58)) +
stat_summary(fun = median, geom = "bar", width = 1, color = "white") +
stat_pointinterval(color="dark grey", .width=.95, alpha=.7, size=.01, show.legend = FALSE) +
# facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom") +
scale_x_continuous(limits=c(0, .64), breaks=c(0,.2,.4),
labels = dropLeadingZero) +
scale_fill_scico_d(palette = "lajolla", begin=.2, end=.8,
guide = guide_legend(reverse = TRUE),
labels = c("Never", "Rarely", "Sometimes", "Often", "Very often")) +
ylab(y_label)
}
#Set theme for figure 2 plots
#NOTE: for global theme setting (e.g., blog), see `thematic` package
#after update, changed axis.title.x to axis.title.x.bottom to remove .epred title
fig2theme.base <- theme_set(
theme_classic() +
theme(strip.background = element_blank(),
strip.placement = "outside",
# axis.title.x.bottom = element_blank(),
axis.title.x = element_blank(),
axis.line.x = element_line(size=.25),
axis.ticks.x = element_line(size=.25),
axis.text.y = element_blank(),
axis.line.y = element_blank(),
axis.ticks.y = element_blank(),
legend.title = element_blank(),
legend.position = "right",
strip.text.x = element_text(size = 8))
)
fig2theme.top <- theme(
strip.text = element_blank(),
strip.text.x = element_blank(),
axis.text.x = element_blank(),
axis.line.x = element_line(color="grey90", size=.25),
axis.ticks.x = element_line(color="grey90", size=.25)
)
stmony_plot <- plot_item(stmony_pp_data, "Money") +
facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom") +
fig2theme.base +
fig2theme.top
sttran_plot <- plot_item(sttran_pp_data, "Transportation") +
facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom") +
fig2theme.base +
fig2theme.top
stresp_plot <- plot_item(stresp_pp_data, "Respect") +
facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom") +
fig2theme.base +
fig2theme.top
stfair_plot <- plot_item(stfair_pp_data, "Fair treatment") +
facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom") +
fig2theme.base +
fig2theme.top
stjob_plot <- plot_item(stjob_pp_data, "Job") +
facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom",
labeller = labeller(rural.ses.med2= as_labeller(commlabs))) +
fig2theme.base +
fig2theme.top
stthft_plot <- plot_item(stthft_pp_data, "Theft") +
facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom",
labeller = labeller(rural.ses.med2= as_labeller(commlabs))) +
fig2theme.base
stmug_plot <- plot_item(stmug_pp_data, "Assault") +
facet_wrap(~rural.ses.med2, nrow = 1, strip.position = "bottom",
labeller = labeller(rural.ses.med2= as_labeller(commlabs))) +
fig2theme.base
Fig2 <- stmony_plot + sttran_plot +
stresp_plot + stfair_plot +
stjob_plot + guide_area() +
stthft_plot + stmug_plot +
plot_layout(ncol = 2, guides = 'collect') +
plot_annotation(
title = 'FIGURE 2 Subjective Stress Distributions (T1), by Community',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. Bar charts display T1 stress item response distributions by \ncommunity (rural/urban; high/low SES), calculated as the median of the posterior predictive distribution (PPD) for each \nresponse category within a community from sequential ordinal models specifying response category-specific effects of \ncommunity. Grey interval plots display PPD median (dot) and 95% uncertainty interval (line). Median estimates accurately \nrecover observed proportions for each response category within each group.') &
theme(legend.justification = "center", #align legend in center of guide_area()
legend.text = element_text(size=8, face="italic"),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.title = element_blank(),
legend.key.width = unit(1.5, "cm"), #adjust width of key (color box) in legend
legend.key.height = unit(.85, "cm"), #adjust width of key (color box) in legend
#legend.key.size = unit(2, "cm"), #adjusts overall size of square of key (color box) in legend
axis.title.x.bottom = element_blank() #removes axis titles
# strip.background = element_rect(color="white", fill="white")
)
Fig2
#Export to image
ggsave("Figure2.jpeg", Fig2, width=6.5, height=9, path=here("Output"))
#Grab Figure 2 plot data to view & set limits for coord_cartesian xlim
# plt_b <- ggplot_build(fig2)
# fig2data1 <- as.data.frame(plt_b$data[[1]])
# fig2data2 <- as.data.frame(plt_b$data[[2]])
#stmug, rural/low ses, "never" cat: 95%ci(xmax) < 0.58 (=.576)
#set xlim in data to .64 (captures all posterior values & sets equal scale for all plots)
#max value for 100% posterior = .637
#zoom plot to view coord_cartesian to xlim=.58 (will not cut off data)Our primary aim with this figure was to visualize stress response distributions and community differences in these distributions to answer RQ1b, which asked whether levels and sources of stress vary at T1 across rural/urban communities with high/low aggregate SES. In general, respondents residing in urban areas report more frequently experiencing stress from all sources - financial, interpersonal, job, and victimization.
Now we move from describing stress distributions to examining stress as a theoretically posited predictor of criminal intent and depressive symptoms. Recall, when we include the stress items as predictors, we will want to properly account for their ordinal nature in our models. We will do so by relying on thresholds from monotonic cumulative ordinal probit models rather than treating the stress items as continuous metric variables in regression models (the standard approach).
save(stress.wide2, file = here("1_Data_Files/Datasets/stress_wide2.Rdata"))4 Modeling T1 correlations (RQ2A)
(RMD FILE: BDK_2023_Stress_3_T1corr_mods)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE4.1 RQ2: Stress Deficit
RQ2 (Stress deficit): Is subjective stress positively correlated with self-reported criminal intent and negative emotions?
We will use two different estimation methods to answer this question - a cross-sectional “between-person” and a longitudinal “within-person” change approach, respectively. To do so, we will rephrase the question twice to better correspond to each estimation strategy:
RQ2A: (Stress deficit; Between-person): Do individuals who report higher levels of subjective stress at Time 1 (T1) also have a higher probability of reporting criminal intentions or negative emotions compared to those reporting less stress at T1?
RQ2B: (Stress deficit; Within-person): Are within-person increases in subjective stress from T1 to T2 (i.e., T2-T1) correlated with within-person increases (T2-T1) in the probability of reporting criminal intent or negative emotions?
4.1.1 Visualizing prevalence of past crime, criminal intent, & negative emotions
Before begin to answer these questions with models of crime and negative emotions, let’s examine prevalence for these outcomes.
Answering this question requires a switch from modeling stress (which, as noted, will become useful soon in getting the predictor side of our model correct) to modeling three constructs measuring theorized causal outcomes of stress: criminal behavior, criminal intent, and depressive symptoms. We will start with the crime outcomes.
These data contain items indicating how often participants reportedly engaged in past criminal behavior in the past two years (originally from 1=Never to 5=Very often) and comparable items that ask the participants estimate the future likelihood that they will engage in these same criminal acts (originally from 1=No chance to 5=High probability), referred to hereafter for brevity as criminal intent. We focus on items measuring six different acts: theft less than 5BAM (approximately equivalent purchasing power in 2013 as $5USD); theft greater than 5BAM; threats to use violence on someone; physically harming someone else on purpose; using marijuana or other illegal drugs; and attempting to access another person’s private information (e.g., bank account; computer files) without permission. Despite having five measured response categories, as is typical with crime items, the vast majority of responses are “1” (Never) and responses exceeding “2” (Rarely or Little chance) are extremely uncommon for these items. Due to the lack of variability and in an attempt to minimize problems caused by empty cell frequencies, each of these items is dichotomized as “0” (=Never or No chance) or “1” (=At least rarely or Little chance or greater). As such, these items are modeled using distributions that are appropriate for binary responses (e.g., Bernoulli or binomial with logit or probit link).
In analyses of stress-outcome, we limit our focus to criminal intent. Examining associations between current reports of stress in the past week and criminal behavior in the past two years would fail to establish proper causal ordering among variables implied by strain or stress process theories, whereas current reports of stress in the past week and future behavioral intentions in theory provides more appropriate causal ordering. Yet, there is high stability in past crime and criminal intent responses, with an average polychoric correlation of rho=0.84 and range from rho=0.8 to rho=0.9 between the original T1 ordinal item pairs. Additionally, supplementary figures will illustrate similar item response distributions and highly comparable stress-crime associations at T1.
load(here("1_Data_Files/Datasets/stress_dat.Rdata"))
#calculate individual & avg tetrachoric corrs
# original ordinal past crime & criminal intent items at T1
rho1 <- polychor(stress_dat$pstthflt5w1, stress_dat$prjthflt5w1)
# rho1 #.82
rho2 <- polychor(stress_dat$pstthfgt5w1, stress_dat$prjthfgt5w1)
# rho2 #.8
rho3 <- polychor(stress_dat$pstthreatw1, stress_dat$prjthreatw1)
# rho3 #.84
rho4 <- polychor(stress_dat$pstharmw1, stress_dat$prjharmw1)
# rho4 #.84
rho5 <- polychor(stress_dat$pstusedrgw1, stress_dat$prjusedrgw1)
# rho5 #.9
rho6 <- polychor(stress_dat$psthackw1, stress_dat$prjhackw1)
# rho6 #.83
rhoavg <- (rho1 + rho2 + rho3 + rho4 + rho5 + rho6)/6
# rhoavg #.84In addition, we will assess seven binary items related to experiences with negative emotions in which respondents were asked how often in the past week they: felt you could not get going; felt everything was an effort; felt lonely; felt you could not shake the blues; felt like your life circumstances were unfair; felt mistreated by others; and felt betrayed by people you care about. The first four are classic items from the popular CESD scale’s “depressive affect” factor (CITE). The remaining three items were modified (see also 2008 survey & JQ Coercion piece) to capture an affective sense of unfairness or mistreatment, which are theorized to be subjective consequences of coercion or criminogenic stress. For all items, the five original response categories were recoded so that “0” indicates infrequently (Never; Rarely; Sometimes) and “1” indicates frequently (Often; Very often) experiencing theses depressive symptoms or feelings in the past week.
Supplemental Figure 1 below visualizes the prevalence rates for each of the crime and negative emotion items.
First, we need to create a factor version of each crime and affect item for use in binary models. We also need to transform our stress items into integers to model their monotonic ordinal effects using cumulative probit thresholds.
#load stress.wide (from Fig1 Rmd)
load(here("1_Data_Files/Datasets/stress_wide.Rdata"))
stress.wide <- zap_labels(stress.wide)
stress.wide <- zap_label(stress.wide)
load(here("1_Data_Files/Datasets/stress_wide2.Rdata"))# cesdw1: q12_1w1 q12_2w1 q12_3w1 q12_4w1 q12_5mw1 q12_6w1 q12_7w1 ;
# cesdw2: q12_1w2 q12_2w2 q12_3w2 q12_4w2 q12_5w2 q12_6w2 q12_7w2 ;
#First, need to recode outcome variables as factors & stress items as integers for mo()
stress.wide3 <- stress.wide2 %>%
mutate(
pstthflt5w1f = factor(pstthflt5w1di, ordered=TRUE, levels = c(0,1)),
pstthfgt5w1f = factor(pstthfgt5w1di, ordered=TRUE, levels = c(0,1)),
pstthreatw1f = factor(pstthreatw1di, ordered=TRUE, levels = c(0,1)),
pstharmw1f = factor(pstharmw1di, ordered=TRUE, levels = c(0,1)),
pstusedrgw1f = factor(pstusedrgw1di, ordered=TRUE, levels = c(0,1)),
psthackw1f = factor(psthackw1di, ordered=TRUE, levels = c(0,1)),
pstanyw1f = factor(if_else(pstthflt5w1di == 1 | pstthfgt5w1di == 1 | pstthreatw1di == 1 |
pstharmw1di == 1 | pstusedrgw1di == 1 | psthackw1di == 1, 1, 0),
ordered=TRUE, levels = c(0,1)),
prjthflt5w1f = factor(prjthflt5w1di, ordered=TRUE, levels = c(0,1)),
prjthfgt5w1f = factor(prjthfgt5w1di, ordered=TRUE, levels = c(0,1)),
prjthreatw1f = factor(prjthreatw1di, ordered=TRUE, levels = c(0,1)),
prjharmw1f = factor(prjharmw1di, ordered=TRUE, levels = c(0,1)),
prjusedrgw1f = factor(prjusedrgw1di, ordered=TRUE, levels = c(0,1)),
prjhackw1f = factor(prjhackw1di, ordered=TRUE, levels = c(0,1)),
prjanyw1f = factor(if_else(prjthflt5w1di == 1 | prjthfgt5w1di == 1 | prjthreatw1di == 1 |
prjharmw1di == 1 | prjusedrgw1di == 1 | prjhackw1di == 1, 1, 0),
ordered=TRUE, levels = c(0,1)),
# pstcrmvarw1 = pstthflt5w1f + pstthfgt5w1f + pstthreatw1f +
# pstharmw1f + pstusedrgw1f + psthackw1f,
# pstcrmvarw1f = factor(pstcrmvarw1, ordered=TRUE, levels = c(0,1,2,3,4,5,6)),
# prjcrmvarw1 = prjthflt5w1f + prjthfgt5w1f + prjthreatw1f +
# prjharmw1f + prjusedrgw1f + prjhackw1f,
# prjcrmvarw1f = factor(prjcrmvarw1, ordered=TRUE, levels = c(0,1,2,3,4,5,6)),
depcantgow1di = if_else(depcantgow1 %in% c(4,5), 1, 0),
depcantgow1f = factor(depcantgow1di, ordered=TRUE, levels = c(0,1)),
depeffortw1di = if_else(depeffortw1 %in% c(4,5), 1, 0),
depeffortw1f = factor(depeffortw1di, ordered=TRUE, levels = c(0,1)),
deplonelyw1di = if_else(deplonelyw1 %in% c(4,5), 1, 0),
deplonelyw1f = factor(deplonelyw1di, ordered=TRUE, levels = c(0,1)),
depbluesw1di = if_else(depbluesw1 %in% c(4,5), 1, 0),
depbluesw1f = factor(depbluesw1di, ordered=TRUE, levels = c(0,1)),
depunfairw1di = if_else(depunfairw1 %in% c(4,5), 1, 0),
depunfairw1f = factor(depunfairw1di, ordered=TRUE, levels = c(0,1)),
depmistrtw1di = if_else(depmistrtw1 %in% c(4,5), 1, 0),
depmistrtw1f = factor(depmistrtw1di, ordered=TRUE, levels = c(0,1)),
depbetrayw1di = if_else(depbetrayw1 %in% c(4,5), 1, 0),
depbetrayw1f = factor(depbetrayw1di, ordered=TRUE, levels = c(0,1)),
stmonyw1i = recode(stmonyw1f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
sttranw1i = recode(sttranw1f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
strespw1i = recode(strespw1f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stfairw1i = recode(stfairw1f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stjobw1i = recode(stjobw1f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stthftw1i = recode(stthftw1f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stmugw1i = recode(stmugw1f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer()
)
#ordered cat predictor - remaining stress subscales & general stress
# #stress subscales - levels of ordered factors
# levels(stress.wide3$ssperw1f)
# levels(stress.wide3$ssjobw1f)
# levels(stress.wide3$ssvicw1f)
# summary(stress.wide3$ssgenw1)
# table(stress.wide3$ssgenw1)
# head(stress.wide3)
# make sure it worked
# stress.wide3 %>% janitor::tabyl("depcantgow1f")
# stress.wide3 %>% janitor::tabyl("depeffortw1f")
# stress.wide3 %>% janitor::tabyl("deplonelyw1f")
# stress.wide3 %>% janitor::tabyl("depbluesw1f")
# stress.wide3 %>% janitor::tabyl("depunfairw1f")
# stress.wide3 %>% janitor::tabyl("depmistrtw1f")
# stress.wide3 %>% janitor::tabyl("depbetrayw1f")
# stress.wide3 %>%
# distinct(stmonyw1f, stmonyw1i) %>%
# arrange(stmonyw1i)
# stress.wide3 %>%
# distinct(sttranw1f, sttranw1i) %>%
# arrange(sttranw1i)
# stress.wide3 %>%
# distinct(strespw1f, strespw1i) %>%
# arrange(strespw1i)
# stress.wide3 %>%
# distinct(stfairw1f, stfairw1i) %>%
# arrange(stfairw1i)
# stress.wide3 %>%
# distinct(stjobw1f, stjobw1i) %>%
# arrange(stjobw1i)
# stress.wide3 %>%
# distinct(stthftw1f, stthftw1i) %>%
# arrange(stthftw1i)
# stress.wide3 %>%
# distinct(stmugw1f, stmugw1i) %>%
# arrange(stmugw1i) Next, we visualize frequency distributions for the dichotomous items measuring past participation in six crimes, future intent to participate in six crimes, and seven indicators of negative emotions (four classic CESD depressive symptoms and three “criminogenic” negative emotions). Given the extreme rarity of self-reported crime in these data, we also add a binary indicator of any past participation or future intent to engage in any of the six measured criminal behaviors. Later, and in our “Stress in Bangladesh” paper, we limit focus to criminal intent and negative emotions and drop the past crime behaviors. Though models of past crime and criminal intent tend to generate comparable patterns, past crime items fail to properly establish causal order and, thus, estimates from models predicting past crime are less plausibly interpreted as estimates of the underlying causal processes examined here.
4.2 Supp. Figure 1 - Outcome Distributions
The distributions for each of the six past crime items, six projected crime items, and seven depressive symptom items are shown in the supplemental figure below.
#NOTE: Consider "stacked" bar charts with 21 crime/dep outcome on y axix & a single bar displaying prop 1 or 0
#Exemplar plot
# ggplot(data = stress.wide3, aes(y = pstthflt5w1f, fill=pstthflt5w1f)) +
# geom_bar(aes(y = pstthflt5w1f, fill=factor(pstthflt5w1f), stat="count")) +
# coord_cartesian(expand = FALSE, xlim=c(0,450)) +
# scale_fill_scico_d(palette = "lajolla", begin=.8, end=.4,
# guide = guide_legend(reverse = TRUE),
# labels = c("0", "1+"))
# myplot <- ggplot(tips, aes(day)) +
# geom_bar(aes(y = (..count..)/sum(..count..))) +
# scale_y_continuous(labels=scales::percent) +
# ylab("relative frequencies")
#Try this with xlab("Relative frequencies")
psty = c("pstthflt5w1f", "pstthfgt5w1f", "pstthreatw1f", "pstharmw1f",
"pstusedrgw1f", "psthackw1f")
#Create function to plot crime items
plot_outcome <- function(outcome_data, outcome_item, y_label) {
ggplot(data = outcome_data, aes(y = {{outcome_item}}, fill={{outcome_item}})) +
geom_bar(aes(y = {{outcome_item}}), stat="count") +
coord_cartesian(expand = FALSE, xlim=c(0,500)) +
scale_fill_scico_d(palette = "lajolla", begin=.8, end=.4,
guide = guide_legend(reverse = TRUE),
labels = c("0", "1"))+
ylab(y_label) +
theme(axis.title.y=element_text(size=8),
axis.title.x = element_blank(),
legend.position="none") +
scale_x_continuous(limit=c(-16,500), breaks=c("0"=0,".1"=0.1*489,".2"=0.2*489,
".3"=0.3*489,".4"=0.4*489,
".5"=0.5*489, ".6"=0.6*489,
".7"=0.7*489, ".8"=0.8*489,
".9"=0.9*489))
}
#Plot outcome items using function
pstthflt5w1fplot <- plot_outcome(stress.wide3, pstthflt5w1f, "Past Theft <5BAM")
pstthfgt5w1fplot <- plot_outcome(stress.wide3, pstthfgt5w1f, "Past Theft >5BAM")
pstthreatw1fplot <- plot_outcome(stress.wide3, pstthreatw1f, "Past Threat")
pstharmw1fplot <- plot_outcome(stress.wide3, pstharmw1f, "Past Harm")
pstusedrgw1fplot <- plot_outcome(stress.wide3, pstusedrgw1f, "Past Use Drugs")
psthackw1fplot <- plot_outcome(stress.wide3, psthackw1f, "Past Hack")
pstanyw1fplot <- plot_outcome(stress.wide3, pstanyw1f, "Any Past Crime")
prjthflt5w1fplot <- plot_outcome(stress.wide3, prjthflt5w1f, "Theft <5BAM Intent")
prjthfgt5w1fplot <- plot_outcome(stress.wide3, prjthfgt5w1f, "Theft >5BAM Intent")
prjthreatw1fplot <- plot_outcome(stress.wide3, prjthreatw1f, "Threat Intent")
prjharmw1fplot <- plot_outcome(stress.wide3, prjharmw1f, "Harm Intent")
prjusedrgw1fplot <- plot_outcome(stress.wide3, prjusedrgw1f, "Use Drugs Intent")
prjhackw1fplot <- plot_outcome(stress.wide3, prjhackw1f, "Hack Intent")
prjanyw1fplot <- plot_outcome(stress.wide3, prjanyw1f, "Any Crime Intent")
depcantgow1fplot <- plot_outcome(stress.wide3, depcantgow1f, "Can't Get Going")
depeffortw1fplot <- plot_outcome(stress.wide3, depeffortw1f, "Everything Effort")
deplonelyw1fplot <- plot_outcome(stress.wide3, deplonelyw1f, "Lonely")
depbluesw1fplot <- plot_outcome(stress.wide3, depbluesw1f, "Can't Shake Blues")
depunfairw1fplot <- plot_outcome(stress.wide3, depunfairw1f, "Felt Life Unfair")
depmistrtw1fplot <- plot_outcome(stress.wide3, depmistrtw1f, "Felt Mistreated")
depbetrayw1fplot <- plot_outcome(stress.wide3, depbetrayw1f, "Felt Betrayed")
#Combine separate plots with patchwork
OutcomeFig <- pstthflt5w1fplot + prjthflt5w1fplot + depcantgow1fplot +
pstthfgt5w1fplot + prjthfgt5w1fplot + depeffortw1fplot +
pstthreatw1fplot + prjthreatw1fplot + deplonelyw1fplot +
pstharmw1fplot + prjharmw1fplot + depbluesw1fplot +
pstusedrgw1fplot + prjusedrgw1fplot + depunfairw1fplot +
psthackw1fplot + prjhackw1fplot + depmistrtw1fplot +
pstanyw1fplot + prjanyw1fplot + depbetrayw1fplot +
plot_layout(ncol = 3) +
plot_annotation(
title = 'SUPPLEMENTAL FIGURE 1 Prevalence of Past Crime, Criminal Intent, & Negative\nEmotions (T1)',
# subtitle = 'Subtitle',
caption = 'Note: N=489 respondents participating at both survey waves.') &
theme(plot.title = element_text(size=12, face="bold"),
# plot.subtitle = element_text(size=12),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
)
OutcomeFig 
#Export to image
ggsave("SuppFigure1.jpeg", OutcomeFig, width=6.5, height=9, path=here("Output"))Prevalence is quite low for all items, with reports of past crime or criminal intent being especially rare compared to experiences with negative emotions. However, our collapsed indicators of “any” past or intended crime have prevalence rates that are more comparable to negative emotions items.
4.3 RQ2A: Stress-Outcome Correlations at T1: Traditional Between-Person Method
RQ2A: (Stress deficit; Between-person): Do individuals who report higher levels of subjective stress at Time 1 (T1) also have a higher probability of reporting criminal intentions or negative emotions compared to those reporting less stress at T1?
To answer this, we must move from describing stress distributions to examining stress as a theoretically posited predictor of crime and depressive symptoms. Recall, when we include the stress items as predictors, we will want to properly account for their ordinal nature in our models. We will do so by relying on thresholds from monotonic cumulative ordinal probit models rather than treating the stress items as continuous metric variables in regression models (the standard approach).
4.3.1 Posterior Predictive Checks for Bivariate Stress/Outcome Models
In the “Stress in Bangladesh” paper, we present results from a series of simple bivariate stress/outcome models that regress each outcome item, or each indicator of criminal intent and negative emotions, separately on each indicator of stress at T1 (see Figure 3). We specify a Bernoulli distribution with a logit link for each outcome item and monotonic cumulative ordinal probit thresholds for each stress predictor.. Below, we present code for these models as well as all results and posterior checks for each of those models used to answer RQ2 (bivariate associations between stress and crime/depressive symptom outcomes). For transparency’s sake, we also begin by presenting results (and later supplementary figures) from T1 models predicting past crime, though, as noted above, these models are not presented in the final paper.
4.3.2 Past crime items: Bivarate models
4.3.2.1 Bivariate Corr: T1 stmonyw1/past crime
#Bivariate: past crime items ~ mo(stmonyw1i)
# prior <- get_prior(mvbind(pstthflt5w1f, pstthfgt5w1f) ~ 1 + mo(stmonyw1i),
# data = stress.wide3,
# family = "bernoulli")
# prior1 <- c(prior(normal(0, 2), class = Intercept, resp = pstthflt5w1f),
# prior(normal(0, 2), class = Intercept, resp = pstthfgt5w1f),
# prior(normal(0, 2), class = Intercept, resp = pstthreatw1f),
# prior(normal(0, 2), class = Intercept, resp = pstharmw1f),
# prior(normal(0, 2), class = Intercept, resp = pstusedrgw1f),
# prior(normal(0, 2), class = Intercept, resp = psthackw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = pstthflt5w1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = pstthfgt5w1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = pstthreatw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = pstharmw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = pstusedrgw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = psthackw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = pstthflt5w1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = pstthfgt5w1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = pstthreatw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = pstharmw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = pstusedrgw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = psthackw1f)
# )
#Vectorize priors:
#list of colnames for past crime DVs
pstdv_names <- noquote(c("pstthflt5w1f", "pstthfgt5w1f", "pstthreatw1f", "pstharmw1f",
"pstusedrgw1f", "psthackw1f"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = pstdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmonyw1i',
resp = pstdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmonyw1i1',
resp = pstdv_names))
# NOTE: consider using future::plan(multisession()) & rstan for parallelizing models
# plan(multiprocess(workers=nCoresphys)) #multiprocess depricated
# if (file.exists(here("Models","allpstcrime_stmony_fit.rds"))) {
# allpstcrime.stmony.fit <- readRDS("Models/allpstcrime_stmony_fit.rds")
# } else {}
#Manual caching unnecessary - use file & file_refit built into brms
allpstcrime.stmony.fit <-
brm(
mvbind(pstthflt5w1f, pstthfgt5w1f, pstthreatw1f, pstharmw1f, pstusedrgw1f, psthackw1f) ~ 1 + mo(stmonyw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
# threads = threading(2), #within-chain threading may speed up intensive models
backend = "cmdstanr", #may need to reinstall RStan
# future = TRUE, #requires rstan not cmdstanr
seed = 8675309,
file = "Models/allpstcrime_stmony_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar past crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="pstthflt5w1f")
ppcheckdv2 <-pp_check(modelfit, resp="pstthfgt5w1f")
ppcheckdv3 <-pp_check(modelfit, resp="pstthreatw1f")
ppcheckdv4 <-pp_check(modelfit, resp="pstharmw1f")
ppcheckdv5 <-pp_check(modelfit, resp="pstusedrgw1f")
ppcheckdv6 <-pp_check(modelfit, resp="psthackw1f")
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6,
plotcoefs, plotcoefs2)
return(allchecks)
}
out.allpstcrime.stmony.fit <- ppchecks(allpstcrime.stmony.fit)
# plot fit
# plot(allpstcrime.stmony.fit, variable = "^bsp_",
# regex = TRUE, nvariables=3, ask=FALSE)
# mcmc_plot(allpstcrime.stmony.fit, variable = "^bsp_", regex = TRUE)4.3.2.1.1 Coefficient plot (intervals)
out.allpstcrime.stmony.fit[[10]]
4.3.2.1.2 Coefficient plot (distributions)
out.allpstcrime.stmony.fit[[9]]
4.3.2.1.3 PPcheck (density)
p1 <- out.allpstcrime.stmony.fit[[3]] + labs(title = "Past Theft <5BAM (T1)")
p2 <- out.allpstcrime.stmony.fit[[4]] + labs(title = "Past Theft >5BAM (T1)")
p3 <- out.allpstcrime.stmony.fit[[5]] + labs(title = "Past Threat (T1)")
p4 <- out.allpstcrime.stmony.fit[[6]] + labs(title = "Past Harm (T1)")
p5 <- out.allpstcrime.stmony.fit[[7]] + labs(title = "Past Use Drugs (T1)")
p6 <- out.allpstcrime.stmony.fit[[8]] + labs(title = "Past Hack (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.2.1.4 Fit summary
out.allpstcrime.stmony.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: pstthflt5w1f ~ 1 + mo(stmonyw1i)
## pstthfgt5w1f ~ 1 + mo(stmonyw1i)
## pstthreatw1f ~ 1 + mo(stmonyw1i)
## pstharmw1f ~ 1 + mo(stmonyw1i)
## pstusedrgw1f ~ 1 + mo(stmonyw1i)
## psthackw1f ~ 1 + mo(stmonyw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_Intercept -1.93 0.29 -2.53 -1.39 1.00 2992
## pstthfgt5w1f_Intercept -2.03 0.30 -2.64 -1.45 1.00 3227
## pstthreatw1f_Intercept -3.15 0.40 -4.00 -2.41 1.00 3446
## pstharmw1f_Intercept -2.48 0.35 -3.19 -1.80 1.00 2843
## pstusedrgw1f_Intercept -2.71 0.36 -3.46 -2.04 1.00 3177
## psthackw1f_Intercept -3.31 0.43 -4.20 -2.50 1.00 3791
## pstthflt5w1f_mostmonyw1i 0.04 0.11 -0.17 0.27 1.00 3162
## pstthfgt5w1f_mostmonyw1i -0.04 0.12 -0.27 0.21 1.00 3372
## pstthreatw1f_mostmonyw1i 0.09 0.15 -0.20 0.40 1.00 3514
## pstharmw1f_mostmonyw1i -0.04 0.14 -0.30 0.26 1.00 2893
## pstusedrgw1f_mostmonyw1i 0.03 0.14 -0.25 0.32 1.00 3087
## psthackw1f_mostmonyw1i 0.02 0.17 -0.29 0.36 1.01 3766
## Tail_ESS
## pstthflt5w1f_Intercept 2893
## pstthfgt5w1f_Intercept 3189
## pstthreatw1f_Intercept 2542
## pstharmw1f_Intercept 2498
## pstusedrgw1f_Intercept 2940
## psthackw1f_Intercept 3248
## pstthflt5w1f_mostmonyw1i 3157
## pstthfgt5w1f_mostmonyw1i 3210
## pstthreatw1f_mostmonyw1i 2433
## pstharmw1f_mostmonyw1i 2624
## pstusedrgw1f_mostmonyw1i 2833
## psthackw1f_mostmonyw1i 3058
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_mostmonyw1i1[1] 0.26 0.15 0.04 0.60 1.00 5468
## pstthflt5w1f_mostmonyw1i1[2] 0.23 0.14 0.03 0.58 1.00 5433
## pstthflt5w1f_mostmonyw1i1[3] 0.24 0.14 0.03 0.57 1.00 6220
## pstthflt5w1f_mostmonyw1i1[4] 0.26 0.14 0.04 0.58 1.00 5685
## pstthfgt5w1f_mostmonyw1i1[1] 0.25 0.14 0.04 0.57 1.00 6237
## pstthfgt5w1f_mostmonyw1i1[2] 0.27 0.16 0.04 0.63 1.00 4600
## pstthfgt5w1f_mostmonyw1i1[3] 0.23 0.14 0.03 0.55 1.00 6631
## pstthfgt5w1f_mostmonyw1i1[4] 0.24 0.15 0.03 0.57 1.00 5360
## pstthreatw1f_mostmonyw1i1[1] 0.26 0.15 0.04 0.60 1.00 5223
## pstthreatw1f_mostmonyw1i1[2] 0.23 0.14 0.03 0.56 1.00 5901
## pstthreatw1f_mostmonyw1i1[3] 0.24 0.14 0.03 0.55 1.00 7032
## pstthreatw1f_mostmonyw1i1[4] 0.27 0.14 0.05 0.59 1.00 6459
## pstharmw1f_mostmonyw1i1[1] 0.25 0.15 0.04 0.58 1.00 6962
## pstharmw1f_mostmonyw1i1[2] 0.26 0.15 0.04 0.60 1.00 4600
## pstharmw1f_mostmonyw1i1[3] 0.24 0.14 0.03 0.56 1.00 6241
## pstharmw1f_mostmonyw1i1[4] 0.25 0.15 0.04 0.58 1.00 4469
## pstusedrgw1f_mostmonyw1i1[1] 0.26 0.15 0.04 0.58 1.00 6840
## pstusedrgw1f_mostmonyw1i1[2] 0.24 0.15 0.03 0.58 1.00 5708
## pstusedrgw1f_mostmonyw1i1[3] 0.24 0.14 0.03 0.57 1.00 6729
## pstusedrgw1f_mostmonyw1i1[4] 0.26 0.15 0.04 0.59 1.00 5798
## psthackw1f_mostmonyw1i1[1] 0.26 0.15 0.04 0.59 1.00 6647
## psthackw1f_mostmonyw1i1[2] 0.24 0.15 0.03 0.57 1.00 7530
## psthackw1f_mostmonyw1i1[3] 0.24 0.14 0.04 0.56 1.00 6962
## psthackw1f_mostmonyw1i1[4] 0.26 0.15 0.04 0.59 1.00 5872
## Tail_ESS
## pstthflt5w1f_mostmonyw1i1[1] 2808
## pstthflt5w1f_mostmonyw1i1[2] 3194
## pstthflt5w1f_mostmonyw1i1[3] 3000
## pstthflt5w1f_mostmonyw1i1[4] 3261
## pstthfgt5w1f_mostmonyw1i1[1] 2713
## pstthfgt5w1f_mostmonyw1i1[2] 3124
## pstthfgt5w1f_mostmonyw1i1[3] 2640
## pstthfgt5w1f_mostmonyw1i1[4] 3023
## pstthreatw1f_mostmonyw1i1[1] 2046
## pstthreatw1f_mostmonyw1i1[2] 2443
## pstthreatw1f_mostmonyw1i1[3] 3193
## pstthreatw1f_mostmonyw1i1[4] 3251
## pstharmw1f_mostmonyw1i1[1] 2531
## pstharmw1f_mostmonyw1i1[2] 2939
## pstharmw1f_mostmonyw1i1[3] 2879
## pstharmw1f_mostmonyw1i1[4] 3253
## pstusedrgw1f_mostmonyw1i1[1] 2807
## pstusedrgw1f_mostmonyw1i1[2] 2640
## pstusedrgw1f_mostmonyw1i1[3] 2459
## pstusedrgw1f_mostmonyw1i1[4] 2964
## psthackw1f_mostmonyw1i1[1] 2891
## psthackw1f_mostmonyw1i1[2] 2920
## psthackw1f_mostmonyw1i1[3] 3268
## psthackw1f_mostmonyw1i1[4] 2676
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.2.1.5 Prior summary
out.allpstcrime.stmony.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b psthackw1f
## normal(0, 0.25) b mostmonyw1i psthackw1f
## (flat) b pstharmw1f
## normal(0, 0.25) b mostmonyw1i pstharmw1f
## (flat) b pstthfgt5w1f
## normal(0, 0.25) b mostmonyw1i pstthfgt5w1f
## (flat) b pstthflt5w1f
## normal(0, 0.25) b mostmonyw1i pstthflt5w1f
## (flat) b pstthreatw1f
## normal(0, 0.25) b mostmonyw1i pstthreatw1f
## (flat) b pstusedrgw1f
## normal(0, 0.25) b mostmonyw1i pstusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept psthackw1f
## normal(0, 2) Intercept pstharmw1f
## normal(0, 2) Intercept pstthfgt5w1f
## normal(0, 2) Intercept pstthflt5w1f
## normal(0, 2) Intercept pstthreatw1f
## normal(0, 2) Intercept pstusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 psthackw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 pstharmw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 pstthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 pstthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 pstthreatw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 pstusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.2.2 Bivariate Corr: T1 sttranw1/past crime
#Bivariate: past crime items ~ mo(sttranw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = pstdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mosttranw1i',
resp = pstdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mosttranw1i1',
resp = pstdv_names))
allpstcrime.sttran.fit <- brm(
mvbind(pstthflt5w1f, pstthfgt5w1f, pstthreatw1f, pstharmw1f, pstusedrgw1f, psthackw1f) ~ 1 + mo(sttranw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allpstcrime_sttran_fit",
file_refit = "on_change"
)
out.allpstcrime.sttran.fit <- ppchecks(allpstcrime.sttran.fit)4.3.2.2.1 Coefficient plot (intervals)
out.allpstcrime.sttran.fit[[10]]
4.3.2.2.2 Coefficient plot (distributions)
out.allpstcrime.sttran.fit[[9]]
4.3.2.2.3 PPcheck (density)
p1 <- out.allpstcrime.sttran.fit[[3]] + labs(title = "Past Theft <5BAM (T1)")
p2 <- out.allpstcrime.sttran.fit[[4]] + labs(title = "Past Theft >5BAM (T1)")
p3 <- out.allpstcrime.sttran.fit[[5]] + labs(title = "Past Threat (T1)")
p4 <- out.allpstcrime.sttran.fit[[6]] + labs(title = "Past Harm (T1)")
p5 <- out.allpstcrime.sttran.fit[[7]] + labs(title = "Past Use Drugs (T1)")
p6 <- out.allpstcrime.sttran.fit[[8]] + labs(title = "Past Hack (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.2.2.4 Fit summary
out.allpstcrime.sttran.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: pstthflt5w1f ~ 1 + mo(sttranw1i)
## pstthfgt5w1f ~ 1 + mo(sttranw1i)
## pstthreatw1f ~ 1 + mo(sttranw1i)
## pstharmw1f ~ 1 + mo(sttranw1i)
## pstusedrgw1f ~ 1 + mo(sttranw1i)
## psthackw1f ~ 1 + mo(sttranw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_Intercept -1.88 0.29 -2.41 -1.24 1.00 3190
## pstthfgt5w1f_Intercept -2.19 0.32 -2.81 -1.56 1.00 3069
## pstthreatw1f_Intercept -3.17 0.40 -3.97 -2.39 1.00 3433
## pstharmw1f_Intercept -2.35 0.37 -3.08 -1.61 1.00 3161
## pstusedrgw1f_Intercept -2.81 0.39 -3.61 -2.06 1.00 3234
## psthackw1f_Intercept -3.15 0.43 -4.01 -2.35 1.00 3531
## pstthflt5w1f_mosttranw1i 0.03 0.12 -0.22 0.25 1.00 3094
## pstthfgt5w1f_mosttranw1i 0.04 0.13 -0.22 0.29 1.00 3349
## pstthreatw1f_mosttranw1i 0.11 0.15 -0.20 0.41 1.00 3513
## pstharmw1f_mosttranw1i -0.09 0.15 -0.37 0.20 1.00 3252
## pstusedrgw1f_mosttranw1i 0.08 0.16 -0.22 0.40 1.00 2864
## psthackw1f_mosttranw1i -0.05 0.16 -0.35 0.27 1.00 3281
## Tail_ESS
## pstthflt5w1f_Intercept 2794
## pstthfgt5w1f_Intercept 2458
## pstthreatw1f_Intercept 2682
## pstharmw1f_Intercept 2603
## pstusedrgw1f_Intercept 3021
## psthackw1f_Intercept 2555
## pstthflt5w1f_mosttranw1i 2835
## pstthfgt5w1f_mosttranw1i 2769
## pstthreatw1f_mosttranw1i 2585
## pstharmw1f_mosttranw1i 2893
## pstusedrgw1f_mosttranw1i 3059
## psthackw1f_mosttranw1i 2632
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_mosttranw1i1[1] 0.26 0.15 0.04 0.60 1.00 5617
## pstthflt5w1f_mosttranw1i1[2] 0.23 0.14 0.03 0.57 1.00 7525
## pstthflt5w1f_mosttranw1i1[3] 0.24 0.14 0.03 0.58 1.00 6841
## pstthflt5w1f_mosttranw1i1[4] 0.27 0.15 0.04 0.60 1.00 5623
## pstthfgt5w1f_mosttranw1i1[1] 0.26 0.15 0.04 0.59 1.00 6506
## pstthfgt5w1f_mosttranw1i1[2] 0.23 0.14 0.03 0.55 1.00 8372
## pstthfgt5w1f_mosttranw1i1[3] 0.24 0.14 0.04 0.56 1.00 7284
## pstthfgt5w1f_mosttranw1i1[4] 0.27 0.15 0.04 0.61 1.00 5683
## pstthreatw1f_mosttranw1i1[1] 0.24 0.14 0.04 0.56 1.00 5745
## pstthreatw1f_mosttranw1i1[2] 0.23 0.14 0.03 0.55 1.00 6804
## pstthreatw1f_mosttranw1i1[3] 0.25 0.15 0.03 0.58 1.00 6329
## pstthreatw1f_mosttranw1i1[4] 0.28 0.16 0.04 0.63 1.00 6932
## pstharmw1f_mosttranw1i1[1] 0.26 0.14 0.04 0.58 1.00 5677
## pstharmw1f_mosttranw1i1[2] 0.28 0.15 0.04 0.62 1.00 5089
## pstharmw1f_mosttranw1i1[3] 0.24 0.14 0.03 0.55 1.00 7652
## pstharmw1f_mosttranw1i1[4] 0.23 0.14 0.03 0.56 1.00 4586
## pstusedrgw1f_mosttranw1i1[1] 0.26 0.15 0.04 0.59 1.00 6222
## pstusedrgw1f_mosttranw1i1[2] 0.22 0.14 0.03 0.54 1.00 5365
## pstusedrgw1f_mosttranw1i1[3] 0.22 0.14 0.03 0.54 1.00 6311
## pstusedrgw1f_mosttranw1i1[4] 0.30 0.17 0.04 0.65 1.00 4260
## psthackw1f_mosttranw1i1[1] 0.26 0.14 0.04 0.58 1.00 5490
## psthackw1f_mosttranw1i1[2] 0.25 0.14 0.04 0.57 1.00 5690
## psthackw1f_mosttranw1i1[3] 0.24 0.14 0.04 0.57 1.00 8421
## psthackw1f_mosttranw1i1[4] 0.25 0.14 0.04 0.56 1.00 6678
## Tail_ESS
## pstthflt5w1f_mosttranw1i1[1] 3225
## pstthflt5w1f_mosttranw1i1[2] 2765
## pstthflt5w1f_mosttranw1i1[3] 2825
## pstthflt5w1f_mosttranw1i1[4] 2929
## pstthfgt5w1f_mosttranw1i1[1] 2502
## pstthfgt5w1f_mosttranw1i1[2] 2353
## pstthfgt5w1f_mosttranw1i1[3] 2970
## pstthfgt5w1f_mosttranw1i1[4] 3354
## pstthreatw1f_mosttranw1i1[1] 2905
## pstthreatw1f_mosttranw1i1[2] 2814
## pstthreatw1f_mosttranw1i1[3] 3129
## pstthreatw1f_mosttranw1i1[4] 2721
## pstharmw1f_mosttranw1i1[1] 2419
## pstharmw1f_mosttranw1i1[2] 2875
## pstharmw1f_mosttranw1i1[3] 3000
## pstharmw1f_mosttranw1i1[4] 2836
## pstusedrgw1f_mosttranw1i1[1] 2650
## pstusedrgw1f_mosttranw1i1[2] 2935
## pstusedrgw1f_mosttranw1i1[3] 2859
## pstusedrgw1f_mosttranw1i1[4] 2850
## psthackw1f_mosttranw1i1[1] 2810
## psthackw1f_mosttranw1i1[2] 2655
## psthackw1f_mosttranw1i1[3] 3170
## psthackw1f_mosttranw1i1[4] 3134
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.2.2.5 Prior summary
out.allpstcrime.sttran.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b psthackw1f
## normal(0, 0.25) b mosttranw1i psthackw1f
## (flat) b pstharmw1f
## normal(0, 0.25) b mosttranw1i pstharmw1f
## (flat) b pstthfgt5w1f
## normal(0, 0.25) b mosttranw1i pstthfgt5w1f
## (flat) b pstthflt5w1f
## normal(0, 0.25) b mosttranw1i pstthflt5w1f
## (flat) b pstthreatw1f
## normal(0, 0.25) b mosttranw1i pstthreatw1f
## (flat) b pstusedrgw1f
## normal(0, 0.25) b mosttranw1i pstusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept psthackw1f
## normal(0, 2) Intercept pstharmw1f
## normal(0, 2) Intercept pstthfgt5w1f
## normal(0, 2) Intercept pstthflt5w1f
## normal(0, 2) Intercept pstthreatw1f
## normal(0, 2) Intercept pstusedrgw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 psthackw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 pstharmw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 pstthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 pstthflt5w1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 pstthreatw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 pstusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.2.3 Bivariate Corr: T1 strespw1/past crime
#Bivariate: past crime items ~ mo(strespw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = pstdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostrespw1i',
resp = pstdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostrespw1i1',
resp = pstdv_names))
allpstcrime.stresp.fit <- brm(
mvbind(pstthflt5w1f, pstthfgt5w1f, pstthreatw1f, pstharmw1f, pstusedrgw1f, psthackw1f) ~ 1 + mo(strespw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allpstcrime_stresp_fit",
file_refit = "on_change")
out.allpstcrime.stresp.fit <- ppchecks(allpstcrime.stresp.fit)4.3.2.3.1 Coefficient plot (intervals)
out.allpstcrime.stresp.fit[[10]]
4.3.2.3.2 Coefficient plot (distributions)
out.allpstcrime.stresp.fit[[9]]
4.3.2.3.3 PPcheck (density)
p1 <- out.allpstcrime.stresp.fit[[3]] + labs(title = "Past Theft <5BAM (T1)")
p2 <- out.allpstcrime.stresp.fit[[4]] + labs(title = "Past Theft >5BAM (T1)")
p3 <- out.allpstcrime.stresp.fit[[5]] + labs(title = "Past Threat (T1)")
p4 <- out.allpstcrime.stresp.fit[[6]] + labs(title = "Past Harm (T1)")
p5 <- out.allpstcrime.stresp.fit[[7]] + labs(title = "Past Use Drugs (T1)")
p6 <- out.allpstcrime.stresp.fit[[8]] + labs(title = "Past Hack (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.2.3.4 Fit summary
out.allpstcrime.stresp.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: pstthflt5w1f ~ 1 + mo(strespw1i)
## pstthfgt5w1f ~ 1 + mo(strespw1i)
## pstthreatw1f ~ 1 + mo(strespw1i)
## pstharmw1f ~ 1 + mo(strespw1i)
## pstusedrgw1f ~ 1 + mo(strespw1i)
## psthackw1f ~ 1 + mo(strespw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_Intercept -2.42 0.28 -3.02 -1.92 1.00 3486
## pstthfgt5w1f_Intercept -2.50 0.31 -3.18 -1.93 1.00 4723
## pstthreatw1f_Intercept -3.36 0.38 -4.16 -2.64 1.00 4553
## pstharmw1f_Intercept -2.86 0.35 -3.63 -2.22 1.00 3905
## pstusedrgw1f_Intercept -3.29 0.37 -4.06 -2.60 1.00 4476
## psthackw1f_Intercept -3.40 0.41 -4.24 -2.66 1.00 5010
## pstthflt5w1f_mostrespw1i 0.24 0.09 0.07 0.42 1.00 3703
## pstthfgt5w1f_mostrespw1i 0.15 0.10 -0.04 0.36 1.00 4673
## pstthreatw1f_mostrespw1i 0.17 0.13 -0.08 0.42 1.00 4941
## pstharmw1f_mostrespw1i 0.12 0.11 -0.10 0.34 1.00 3981
## pstusedrgw1f_mostrespw1i 0.25 0.11 0.03 0.48 1.00 4756
## psthackw1f_mostrespw1i 0.06 0.14 -0.21 0.34 1.00 5004
## Tail_ESS
## pstthflt5w1f_Intercept 2597
## pstthfgt5w1f_Intercept 2857
## pstthreatw1f_Intercept 3134
## pstharmw1f_Intercept 2693
## pstusedrgw1f_Intercept 2955
## psthackw1f_Intercept 2671
## pstthflt5w1f_mostrespw1i 2731
## pstthfgt5w1f_mostrespw1i 2879
## pstthreatw1f_mostrespw1i 3124
## pstharmw1f_mostrespw1i 2907
## pstusedrgw1f_mostrespw1i 3127
## psthackw1f_mostrespw1i 3426
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_mostrespw1i1[1] 0.22 0.13 0.03 0.52 1.00 7913
## pstthflt5w1f_mostrespw1i1[2] 0.26 0.15 0.04 0.58 1.00 7742
## pstthflt5w1f_mostrespw1i1[3] 0.25 0.14 0.04 0.56 1.00 7768
## pstthflt5w1f_mostrespw1i1[4] 0.27 0.14 0.05 0.57 1.00 8553
## pstthfgt5w1f_mostrespw1i1[1] 0.29 0.16 0.05 0.63 1.00 6859
## pstthfgt5w1f_mostrespw1i1[2] 0.24 0.14 0.03 0.55 1.00 8499
## pstthfgt5w1f_mostrespw1i1[3] 0.23 0.14 0.03 0.54 1.00 7839
## pstthfgt5w1f_mostrespw1i1[4] 0.24 0.13 0.04 0.54 1.00 6889
## pstthreatw1f_mostrespw1i1[1] 0.24 0.14 0.03 0.57 1.00 8148
## pstthreatw1f_mostrespw1i1[2] 0.24 0.13 0.04 0.55 1.00 6830
## pstthreatw1f_mostrespw1i1[3] 0.26 0.15 0.04 0.59 1.00 7216
## pstthreatw1f_mostrespw1i1[4] 0.26 0.14 0.04 0.58 1.00 7220
## pstharmw1f_mostrespw1i1[1] 0.27 0.15 0.04 0.60 1.00 6900
## pstharmw1f_mostrespw1i1[2] 0.26 0.14 0.04 0.57 1.00 9042
## pstharmw1f_mostrespw1i1[3] 0.23 0.14 0.03 0.56 1.00 6819
## pstharmw1f_mostrespw1i1[4] 0.24 0.14 0.04 0.56 1.00 7185
## pstusedrgw1f_mostrespw1i1[1] 0.23 0.13 0.04 0.53 1.00 6997
## pstusedrgw1f_mostrespw1i1[2] 0.22 0.13 0.03 0.53 1.00 7594
## pstusedrgw1f_mostrespw1i1[3] 0.31 0.15 0.06 0.63 1.00 7061
## pstusedrgw1f_mostrespw1i1[4] 0.24 0.14 0.04 0.56 1.00 7387
## psthackw1f_mostrespw1i1[1] 0.26 0.15 0.03 0.61 1.00 6609
## psthackw1f_mostrespw1i1[2] 0.24 0.14 0.04 0.58 1.00 8017
## psthackw1f_mostrespw1i1[3] 0.25 0.14 0.04 0.57 1.00 6835
## psthackw1f_mostrespw1i1[4] 0.25 0.14 0.04 0.57 1.00 6695
## Tail_ESS
## pstthflt5w1f_mostrespw1i1[1] 2801
## pstthflt5w1f_mostrespw1i1[2] 2230
## pstthflt5w1f_mostrespw1i1[3] 2742
## pstthflt5w1f_mostrespw1i1[4] 3237
## pstthfgt5w1f_mostrespw1i1[1] 2254
## pstthfgt5w1f_mostrespw1i1[2] 2523
## pstthfgt5w1f_mostrespw1i1[3] 2897
## pstthfgt5w1f_mostrespw1i1[4] 2629
## pstthreatw1f_mostrespw1i1[1] 2746
## pstthreatw1f_mostrespw1i1[2] 2404
## pstthreatw1f_mostrespw1i1[3] 2795
## pstthreatw1f_mostrespw1i1[4] 2686
## pstharmw1f_mostrespw1i1[1] 2369
## pstharmw1f_mostrespw1i1[2] 2932
## pstharmw1f_mostrespw1i1[3] 2516
## pstharmw1f_mostrespw1i1[4] 3063
## pstusedrgw1f_mostrespw1i1[1] 2413
## pstusedrgw1f_mostrespw1i1[2] 2093
## pstusedrgw1f_mostrespw1i1[3] 2799
## pstusedrgw1f_mostrespw1i1[4] 2800
## psthackw1f_mostrespw1i1[1] 2556
## psthackw1f_mostrespw1i1[2] 2756
## psthackw1f_mostrespw1i1[3] 2432
## psthackw1f_mostrespw1i1[4] 2874
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.2.3.5 Prior summary
out.allpstcrime.stresp.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b psthackw1f
## normal(0, 0.25) b mostrespw1i psthackw1f
## (flat) b pstharmw1f
## normal(0, 0.25) b mostrespw1i pstharmw1f
## (flat) b pstthfgt5w1f
## normal(0, 0.25) b mostrespw1i pstthfgt5w1f
## (flat) b pstthflt5w1f
## normal(0, 0.25) b mostrespw1i pstthflt5w1f
## (flat) b pstthreatw1f
## normal(0, 0.25) b mostrespw1i pstthreatw1f
## (flat) b pstusedrgw1f
## normal(0, 0.25) b mostrespw1i pstusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept psthackw1f
## normal(0, 2) Intercept pstharmw1f
## normal(0, 2) Intercept pstthfgt5w1f
## normal(0, 2) Intercept pstthflt5w1f
## normal(0, 2) Intercept pstthreatw1f
## normal(0, 2) Intercept pstusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 psthackw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 pstharmw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 pstthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 pstthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 pstthreatw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 pstusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.2.4 Bivariate Corr: T1 stfairw1/past crime
#Bivariate: past crime items ~ mo(stfairw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = pstdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostfairw1i',
resp = pstdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostfairw1i1',
resp = pstdv_names))
allpstcrime.stfair.fit <- brm(
mvbind(pstthflt5w1f, pstthfgt5w1f, pstthreatw1f, pstharmw1f, pstusedrgw1f, psthackw1f) ~ 1 + mo(stfairw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allpstcrime_stfair_fit",
file_refit = "on_change"
)
out.allpstcrime.stfair.fit <- ppchecks(allpstcrime.stfair.fit)4.3.2.4.1 Coefficient plot (intervals)
out.allpstcrime.stfair.fit[[10]]
4.3.2.4.2 Coefficient plot (distributions)
out.allpstcrime.stfair.fit[[9]]
4.3.2.4.3 PPcheck (density)
p1 <- out.allpstcrime.stfair.fit[[3]] + labs(title = "Past Theft <5BAM (T1)")
p2 <- out.allpstcrime.stfair.fit[[4]] + labs(title = "Past Theft >5BAM (T1)")
p3 <- out.allpstcrime.stfair.fit[[5]] + labs(title = "Past Threat (T1)")
p4 <- out.allpstcrime.stfair.fit[[6]] + labs(title = "Past Harm (T1)")
p5 <- out.allpstcrime.stfair.fit[[7]] + labs(title = "Past Use Drugs (T1)")
p6 <- out.allpstcrime.stfair.fit[[8]] + labs(title = "Past Hack (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.2.4.4 Fit summary
out.allpstcrime.stfair.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: pstthflt5w1f ~ 1 + mo(stfairw1i)
## pstthfgt5w1f ~ 1 + mo(stfairw1i)
## pstthreatw1f ~ 1 + mo(stfairw1i)
## pstharmw1f ~ 1 + mo(stfairw1i)
## pstusedrgw1f ~ 1 + mo(stfairw1i)
## psthackw1f ~ 1 + mo(stfairw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_Intercept -2.41 0.26 -2.96 -1.92 1.00 4250
## pstthfgt5w1f_Intercept -2.47 0.31 -3.13 -1.91 1.00 2873
## pstthreatw1f_Intercept -3.49 0.41 -4.33 -2.72 1.00 3515
## pstharmw1f_Intercept -3.00 0.36 -3.76 -2.36 1.00 2955
## pstusedrgw1f_Intercept -3.28 0.39 -4.12 -2.57 1.00 3477
## psthackw1f_Intercept -3.50 0.42 -4.38 -2.72 1.00 3426
## pstthflt5w1f_mostfairw1i 0.23 0.08 0.07 0.40 1.00 4191
## pstthfgt5w1f_mostfairw1i 0.13 0.10 -0.06 0.33 1.00 2926
## pstthreatw1f_mostfairw1i 0.21 0.13 -0.04 0.46 1.00 3769
## pstharmw1f_mostfairw1i 0.17 0.11 -0.04 0.39 1.00 3009
## pstusedrgw1f_mostfairw1i 0.23 0.12 0.01 0.47 1.00 3437
## psthackw1f_mostfairw1i 0.09 0.14 -0.17 0.36 1.00 3517
## Tail_ESS
## pstthflt5w1f_Intercept 2836
## pstthfgt5w1f_Intercept 2588
## pstthreatw1f_Intercept 2840
## pstharmw1f_Intercept 2617
## pstusedrgw1f_Intercept 2667
## psthackw1f_Intercept 3092
## pstthflt5w1f_mostfairw1i 2769
## pstthfgt5w1f_mostfairw1i 2344
## pstthreatw1f_mostfairw1i 2847
## pstharmw1f_mostfairw1i 2753
## pstusedrgw1f_mostfairw1i 2876
## psthackw1f_mostfairw1i 3325
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_mostfairw1i1[1] 0.20 0.12 0.03 0.50 1.00 6257
## pstthflt5w1f_mostfairw1i1[2] 0.23 0.13 0.04 0.53 1.00 8749
## pstthflt5w1f_mostfairw1i1[3] 0.26 0.14 0.04 0.57 1.00 6598
## pstthflt5w1f_mostfairw1i1[4] 0.30 0.14 0.06 0.60 1.00 6341
## pstthfgt5w1f_mostfairw1i1[1] 0.27 0.15 0.04 0.60 1.00 5255
## pstthfgt5w1f_mostfairw1i1[2] 0.25 0.14 0.04 0.58 1.00 8072
## pstthfgt5w1f_mostfairw1i1[3] 0.26 0.14 0.04 0.57 1.00 6849
## pstthfgt5w1f_mostfairw1i1[4] 0.22 0.13 0.03 0.53 1.00 6735
## pstthreatw1f_mostfairw1i1[1] 0.23 0.13 0.04 0.55 1.00 7258
## pstthreatw1f_mostfairw1i1[2] 0.25 0.14 0.04 0.55 1.00 6725
## pstthreatw1f_mostfairw1i1[3] 0.27 0.15 0.04 0.61 1.00 5776
## pstthreatw1f_mostfairw1i1[4] 0.25 0.14 0.04 0.58 1.00 7025
## pstharmw1f_mostfairw1i1[1] 0.24 0.14 0.03 0.57 1.00 6856
## pstharmw1f_mostfairw1i1[2] 0.25 0.14 0.05 0.56 1.00 6755
## pstharmw1f_mostfairw1i1[3] 0.28 0.15 0.04 0.62 1.00 6325
## pstharmw1f_mostfairw1i1[4] 0.23 0.13 0.03 0.54 1.00 7288
## pstusedrgw1f_mostfairw1i1[1] 0.23 0.14 0.03 0.54 1.00 7141
## pstusedrgw1f_mostfairw1i1[2] 0.26 0.14 0.04 0.57 1.00 7207
## pstusedrgw1f_mostfairw1i1[3] 0.31 0.16 0.06 0.65 1.00 5474
## pstusedrgw1f_mostfairw1i1[4] 0.20 0.12 0.03 0.50 1.00 6296
## psthackw1f_mostfairw1i1[1] 0.24 0.14 0.03 0.57 1.00 6462
## psthackw1f_mostfairw1i1[2] 0.25 0.15 0.03 0.58 1.00 6839
## psthackw1f_mostfairw1i1[3] 0.27 0.16 0.04 0.63 1.00 6228
## psthackw1f_mostfairw1i1[4] 0.23 0.14 0.03 0.56 1.00 5553
## Tail_ESS
## pstthflt5w1f_mostfairw1i1[1] 2304
## pstthflt5w1f_mostfairw1i1[2] 2783
## pstthflt5w1f_mostfairw1i1[3] 2695
## pstthflt5w1f_mostfairw1i1[4] 2987
## pstthfgt5w1f_mostfairw1i1[1] 2534
## pstthfgt5w1f_mostfairw1i1[2] 2927
## pstthfgt5w1f_mostfairw1i1[3] 2419
## pstthfgt5w1f_mostfairw1i1[4] 2026
## pstthreatw1f_mostfairw1i1[1] 2361
## pstthreatw1f_mostfairw1i1[2] 2848
## pstthreatw1f_mostfairw1i1[3] 3032
## pstthreatw1f_mostfairw1i1[4] 2806
## pstharmw1f_mostfairw1i1[1] 2033
## pstharmw1f_mostfairw1i1[2] 2918
## pstharmw1f_mostfairw1i1[3] 3075
## pstharmw1f_mostfairw1i1[4] 2763
## pstusedrgw1f_mostfairw1i1[1] 2968
## pstusedrgw1f_mostfairw1i1[2] 2825
## pstusedrgw1f_mostfairw1i1[3] 2919
## pstusedrgw1f_mostfairw1i1[4] 2843
## psthackw1f_mostfairw1i1[1] 2423
## psthackw1f_mostfairw1i1[2] 2484
## psthackw1f_mostfairw1i1[3] 2957
## psthackw1f_mostfairw1i1[4] 2474
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.2.4.5 Prior summary
out.allpstcrime.stfair.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b psthackw1f
## normal(0, 0.25) b mostfairw1i psthackw1f
## (flat) b pstharmw1f
## normal(0, 0.25) b mostfairw1i pstharmw1f
## (flat) b pstthfgt5w1f
## normal(0, 0.25) b mostfairw1i pstthfgt5w1f
## (flat) b pstthflt5w1f
## normal(0, 0.25) b mostfairw1i pstthflt5w1f
## (flat) b pstthreatw1f
## normal(0, 0.25) b mostfairw1i pstthreatw1f
## (flat) b pstusedrgw1f
## normal(0, 0.25) b mostfairw1i pstusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept psthackw1f
## normal(0, 2) Intercept pstharmw1f
## normal(0, 2) Intercept pstthfgt5w1f
## normal(0, 2) Intercept pstthflt5w1f
## normal(0, 2) Intercept pstthreatw1f
## normal(0, 2) Intercept pstusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 psthackw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 pstharmw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 pstthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 pstthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 pstthreatw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 pstusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.2.5 Bivariate Corr: T1 stjobw1/past crime
#Bivariate: past crime items ~ mo(stjobw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = pstdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostjobw1i',
resp = pstdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostjobw1i1',
resp = pstdv_names))
allpstcrime.stjob.fit <- brm(
mvbind(pstthflt5w1f, pstthfgt5w1f, pstthreatw1f, pstharmw1f, pstusedrgw1f, psthackw1f) ~ 1 + mo(stjobw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allpstcrime_stjob_fit",
file_refit = "on_change"
)
out.allpstcrime.stjob.fit <- ppchecks(allpstcrime.stjob.fit)4.3.2.5.1 Coefficient plot (intervals)
out.allpstcrime.stjob.fit[[10]]
4.3.2.5.2 Coefficient plot (distributions)
out.allpstcrime.stjob.fit[[9]]
4.3.2.5.3 PPcheck (density)
p1 <- out.allpstcrime.stjob.fit[[3]] + labs(title = "Past Theft <5BAM (T1)")
p2 <- out.allpstcrime.stjob.fit[[4]] + labs(title = "Past Theft >5BAM (T1)")
p3 <- out.allpstcrime.stjob.fit[[5]] + labs(title = "Past Threat (T1)")
p4 <- out.allpstcrime.stjob.fit[[6]] + labs(title = "Past Harm (T1)")
p5 <- out.allpstcrime.stjob.fit[[7]] + labs(title = "Past Use Drugs (T1)")
p6 <- out.allpstcrime.stjob.fit[[8]] + labs(title = "Past Hack (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.2.5.4 Fit summary
out.allpstcrime.stjob.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: pstthflt5w1f ~ 1 + mo(stjobw1i)
## pstthfgt5w1f ~ 1 + mo(stjobw1i)
## pstthreatw1f ~ 1 + mo(stjobw1i)
## pstharmw1f ~ 1 + mo(stjobw1i)
## pstusedrgw1f ~ 1 + mo(stjobw1i)
## psthackw1f ~ 1 + mo(stjobw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_Intercept -2.24 0.25 -2.77 -1.78 1.00 3805
## pstthfgt5w1f_Intercept -2.16 0.27 -2.70 -1.67 1.00 3935
## pstthreatw1f_Intercept -3.71 0.41 -4.53 -2.97 1.00 3965
## pstharmw1f_Intercept -2.72 0.32 -3.40 -2.13 1.00 3585
## pstusedrgw1f_Intercept -3.44 0.39 -4.27 -2.72 1.00 3835
## psthackw1f_Intercept -4.11 0.48 -5.14 -3.24 1.00 3396
## pstthflt5w1f_mostjobw1i 0.17 0.09 0.00 0.34 1.00 4031
## pstthfgt5w1f_mostjobw1i 0.01 0.10 -0.18 0.20 1.00 3682
## pstthreatw1f_mostjobw1i 0.31 0.13 0.07 0.56 1.00 3885
## pstharmw1f_mostjobw1i 0.06 0.11 -0.15 0.29 1.00 3590
## pstusedrgw1f_mostjobw1i 0.31 0.12 0.07 0.56 1.00 3851
## psthackw1f_mostjobw1i 0.34 0.15 0.06 0.64 1.00 3324
## Tail_ESS
## pstthflt5w1f_Intercept 3063
## pstthfgt5w1f_Intercept 2828
## pstthreatw1f_Intercept 3144
## pstharmw1f_Intercept 2597
## pstusedrgw1f_Intercept 2989
## psthackw1f_Intercept 3115
## pstthflt5w1f_mostjobw1i 3247
## pstthfgt5w1f_mostjobw1i 3092
## pstthreatw1f_mostjobw1i 3028
## pstharmw1f_mostjobw1i 2880
## pstusedrgw1f_mostjobw1i 2883
## psthackw1f_mostjobw1i 2874
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_mostjobw1i1[1] 0.23 0.13 0.04 0.53 1.00 7169
## pstthflt5w1f_mostjobw1i1[2] 0.21 0.13 0.03 0.50 1.00 6519
## pstthflt5w1f_mostjobw1i1[3] 0.33 0.16 0.05 0.66 1.00 5986
## pstthflt5w1f_mostjobw1i1[4] 0.23 0.13 0.04 0.54 1.00 6576
## pstthfgt5w1f_mostjobw1i1[1] 0.25 0.15 0.04 0.59 1.00 5204
## pstthfgt5w1f_mostjobw1i1[2] 0.25 0.14 0.04 0.56 1.00 6651
## pstthfgt5w1f_mostjobw1i1[3] 0.25 0.15 0.03 0.61 1.00 6335
## pstthfgt5w1f_mostjobw1i1[4] 0.25 0.15 0.03 0.57 1.00 7548
## pstthreatw1f_mostjobw1i1[1] 0.22 0.13 0.03 0.51 1.00 7308
## pstthreatw1f_mostjobw1i1[2] 0.20 0.12 0.03 0.47 1.00 8685
## pstthreatw1f_mostjobw1i1[3] 0.28 0.15 0.05 0.60 1.00 6914
## pstthreatw1f_mostjobw1i1[4] 0.30 0.15 0.05 0.62 1.00 7305
## pstharmw1f_mostjobw1i1[1] 0.27 0.15 0.04 0.60 1.00 6281
## pstharmw1f_mostjobw1i1[2] 0.24 0.14 0.04 0.57 1.00 6940
## pstharmw1f_mostjobw1i1[3] 0.25 0.14 0.03 0.57 1.00 5702
## pstharmw1f_mostjobw1i1[4] 0.24 0.14 0.04 0.56 1.00 5546
## pstusedrgw1f_mostjobw1i1[1] 0.24 0.13 0.03 0.53 1.01 6796
## pstusedrgw1f_mostjobw1i1[2] 0.21 0.13 0.03 0.52 1.00 6596
## pstusedrgw1f_mostjobw1i1[3] 0.32 0.15 0.06 0.64 1.00 5934
## pstusedrgw1f_mostjobw1i1[4] 0.23 0.13 0.04 0.52 1.00 6568
## psthackw1f_mostjobw1i1[1] 0.22 0.13 0.03 0.53 1.00 7646
## psthackw1f_mostjobw1i1[2] 0.21 0.13 0.03 0.50 1.00 6905
## psthackw1f_mostjobw1i1[3] 0.31 0.16 0.05 0.64 1.00 6517
## psthackw1f_mostjobw1i1[4] 0.26 0.15 0.04 0.59 1.00 8080
## Tail_ESS
## pstthflt5w1f_mostjobw1i1[1] 2566
## pstthflt5w1f_mostjobw1i1[2] 2741
## pstthflt5w1f_mostjobw1i1[3] 2700
## pstthflt5w1f_mostjobw1i1[4] 2724
## pstthfgt5w1f_mostjobw1i1[1] 2478
## pstthfgt5w1f_mostjobw1i1[2] 2666
## pstthfgt5w1f_mostjobw1i1[3] 2858
## pstthfgt5w1f_mostjobw1i1[4] 2549
## pstthreatw1f_mostjobw1i1[1] 2809
## pstthreatw1f_mostjobw1i1[2] 2515
## pstthreatw1f_mostjobw1i1[3] 2810
## pstthreatw1f_mostjobw1i1[4] 2892
## pstharmw1f_mostjobw1i1[1] 2455
## pstharmw1f_mostjobw1i1[2] 2619
## pstharmw1f_mostjobw1i1[3] 2356
## pstharmw1f_mostjobw1i1[4] 2720
## pstusedrgw1f_mostjobw1i1[1] 2872
## pstusedrgw1f_mostjobw1i1[2] 2689
## pstusedrgw1f_mostjobw1i1[3] 2718
## pstusedrgw1f_mostjobw1i1[4] 3056
## psthackw1f_mostjobw1i1[1] 2654
## psthackw1f_mostjobw1i1[2] 2415
## psthackw1f_mostjobw1i1[3] 2620
## psthackw1f_mostjobw1i1[4] 2523
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.2.5.5 Prior summary
out.allpstcrime.stjob.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b psthackw1f
## normal(0, 0.25) b mostjobw1i psthackw1f
## (flat) b pstharmw1f
## normal(0, 0.25) b mostjobw1i pstharmw1f
## (flat) b pstthfgt5w1f
## normal(0, 0.25) b mostjobw1i pstthfgt5w1f
## (flat) b pstthflt5w1f
## normal(0, 0.25) b mostjobw1i pstthflt5w1f
## (flat) b pstthreatw1f
## normal(0, 0.25) b mostjobw1i pstthreatw1f
## (flat) b pstusedrgw1f
## normal(0, 0.25) b mostjobw1i pstusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept psthackw1f
## normal(0, 2) Intercept pstharmw1f
## normal(0, 2) Intercept pstthfgt5w1f
## normal(0, 2) Intercept pstthflt5w1f
## normal(0, 2) Intercept pstthreatw1f
## normal(0, 2) Intercept pstusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 psthackw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 pstharmw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 pstthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 pstthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 pstthreatw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 pstusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.2.6 Bivariate Corr: T1 stthftw1/past crime
#Bivariate: past crime items ~ mo(stthftw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = pstdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostthftw1i',
resp = pstdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostthftw1i1',
resp = pstdv_names))
allpstcrime.stthft.fit <- brm(
mvbind(pstthflt5w1f, pstthfgt5w1f, pstthreatw1f, pstharmw1f, pstusedrgw1f, psthackw1f) ~ 1 + mo(stthftw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allpstcrime_stthft_fit",
file_refit = "on_change"
)
out.allpstcrime.stthft.fit <- ppchecks(allpstcrime.stthft.fit)4.3.2.6.1 Coefficient plot (intervals)
out.allpstcrime.stthft.fit[[10]]
4.3.2.6.2 Coefficient plot (distributions)
out.allpstcrime.stthft.fit[[9]]
4.3.2.6.3 PPcheck (density)
p1 <- out.allpstcrime.stthft.fit[[3]] + labs(title = "Past Theft <5BAM (T1)")
p2 <- out.allpstcrime.stthft.fit[[4]] + labs(title = "Past Theft >5BAM (T1)")
p3 <- out.allpstcrime.stthft.fit[[5]] + labs(title = "Past Threat (T1)")
p4 <- out.allpstcrime.stthft.fit[[6]] + labs(title = "Past Harm (T1)")
p5 <- out.allpstcrime.stthft.fit[[7]] + labs(title = "Past Use Drugs (T1)")
p6 <- out.allpstcrime.stthft.fit[[8]] + labs(title = "Past Hack (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.2.6.4 Fit summary
out.allpstcrime.stthft.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: pstthflt5w1f ~ 1 + mo(stthftw1i)
## pstthfgt5w1f ~ 1 + mo(stthftw1i)
## pstthreatw1f ~ 1 + mo(stthftw1i)
## pstharmw1f ~ 1 + mo(stthftw1i)
## pstusedrgw1f ~ 1 + mo(stthftw1i)
## psthackw1f ~ 1 + mo(stthftw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_Intercept -1.93 0.18 -2.30 -1.59 1.00 5467
## pstthfgt5w1f_Intercept -2.27 0.20 -2.67 -1.87 1.00 5448
## pstthreatw1f_Intercept -3.17 0.30 -3.81 -2.62 1.00 5005
## pstharmw1f_Intercept -2.77 0.26 -3.30 -2.30 1.00 5631
## pstusedrgw1f_Intercept -2.87 0.25 -3.38 -2.40 1.00 4720
## psthackw1f_Intercept -3.22 0.30 -3.83 -2.63 1.00 5285
## pstthflt5w1f_mostthftw1i 0.07 0.10 -0.13 0.26 1.00 5146
## pstthfgt5w1f_mostthftw1i 0.11 0.11 -0.11 0.32 1.00 4889
## pstthreatw1f_mostthftw1i 0.15 0.14 -0.14 0.42 1.00 5367
## pstharmw1f_mostthftw1i 0.15 0.12 -0.08 0.39 1.00 5446
## pstusedrgw1f_mostthftw1i 0.16 0.13 -0.09 0.40 1.00 4864
## psthackw1f_mostthftw1i -0.04 0.16 -0.37 0.27 1.00 5444
## Tail_ESS
## pstthflt5w1f_Intercept 2711
## pstthfgt5w1f_Intercept 3345
## pstthreatw1f_Intercept 2959
## pstharmw1f_Intercept 3005
## pstusedrgw1f_Intercept 2939
## psthackw1f_Intercept 2830
## pstthflt5w1f_mostthftw1i 2632
## pstthfgt5w1f_mostthftw1i 3194
## pstthreatw1f_mostthftw1i 3307
## pstharmw1f_mostthftw1i 3184
## pstusedrgw1f_mostthftw1i 3080
## psthackw1f_mostthftw1i 2805
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_mostthftw1i1[1] 0.23 0.14 0.03 0.55 1.00 7468
## pstthflt5w1f_mostthftw1i1[2] 0.25 0.14 0.04 0.56 1.00 6172
## pstthflt5w1f_mostthftw1i1[3] 0.27 0.15 0.04 0.59 1.00 7078
## pstthflt5w1f_mostthftw1i1[4] 0.26 0.15 0.03 0.58 1.00 5868
## pstthfgt5w1f_mostthftw1i1[1] 0.24 0.14 0.03 0.56 1.00 6130
## pstthfgt5w1f_mostthftw1i1[2] 0.26 0.15 0.04 0.58 1.00 6539
## pstthfgt5w1f_mostthftw1i1[3] 0.24 0.14 0.04 0.57 1.00 6425
## pstthfgt5w1f_mostthftw1i1[4] 0.26 0.15 0.04 0.59 1.00 6116
## pstthreatw1f_mostthftw1i1[1] 0.27 0.15 0.04 0.60 1.00 5510
## pstthreatw1f_mostthftw1i1[2] 0.25 0.14 0.04 0.57 1.00 6393
## pstthreatw1f_mostthftw1i1[3] 0.24 0.14 0.03 0.56 1.00 6292
## pstthreatw1f_mostthftw1i1[4] 0.24 0.14 0.03 0.57 1.00 5649
## pstharmw1f_mostthftw1i1[1] 0.25 0.14 0.04 0.57 1.00 6319
## pstharmw1f_mostthftw1i1[2] 0.27 0.15 0.04 0.60 1.00 6740
## pstharmw1f_mostthftw1i1[3] 0.23 0.14 0.03 0.55 1.00 5451
## pstharmw1f_mostthftw1i1[4] 0.24 0.14 0.03 0.57 1.00 5830
## pstusedrgw1f_mostthftw1i1[1] 0.24 0.14 0.03 0.55 1.00 5262
## pstusedrgw1f_mostthftw1i1[2] 0.24 0.14 0.03 0.56 1.00 5280
## pstusedrgw1f_mostthftw1i1[3] 0.27 0.15 0.04 0.61 1.00 6071
## pstusedrgw1f_mostthftw1i1[4] 0.24 0.14 0.04 0.56 1.00 6727
## psthackw1f_mostthftw1i1[1] 0.25 0.14 0.03 0.58 1.00 6062
## psthackw1f_mostthftw1i1[2] 0.24 0.14 0.03 0.56 1.00 6783
## psthackw1f_mostthftw1i1[3] 0.25 0.14 0.04 0.59 1.00 7526
## psthackw1f_mostthftw1i1[4] 0.26 0.15 0.04 0.61 1.00 6533
## Tail_ESS
## pstthflt5w1f_mostthftw1i1[1] 2527
## pstthflt5w1f_mostthftw1i1[2] 2890
## pstthflt5w1f_mostthftw1i1[3] 2744
## pstthflt5w1f_mostthftw1i1[4] 2557
## pstthfgt5w1f_mostthftw1i1[1] 2442
## pstthfgt5w1f_mostthftw1i1[2] 2704
## pstthfgt5w1f_mostthftw1i1[3] 3231
## pstthfgt5w1f_mostthftw1i1[4] 2979
## pstthreatw1f_mostthftw1i1[1] 2477
## pstthreatw1f_mostthftw1i1[2] 2419
## pstthreatw1f_mostthftw1i1[3] 2494
## pstthreatw1f_mostthftw1i1[4] 2869
## pstharmw1f_mostthftw1i1[1] 2706
## pstharmw1f_mostthftw1i1[2] 3082
## pstharmw1f_mostthftw1i1[3] 2651
## pstharmw1f_mostthftw1i1[4] 2720
## pstusedrgw1f_mostthftw1i1[1] 2892
## pstusedrgw1f_mostthftw1i1[2] 2267
## pstusedrgw1f_mostthftw1i1[3] 2451
## pstusedrgw1f_mostthftw1i1[4] 2767
## psthackw1f_mostthftw1i1[1] 3020
## psthackw1f_mostthftw1i1[2] 2802
## psthackw1f_mostthftw1i1[3] 2742
## psthackw1f_mostthftw1i1[4] 3062
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.2.6.5 Prior summary
out.allpstcrime.stthft.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b psthackw1f
## normal(0, 0.25) b mostthftw1i psthackw1f
## (flat) b pstharmw1f
## normal(0, 0.25) b mostthftw1i pstharmw1f
## (flat) b pstthfgt5w1f
## normal(0, 0.25) b mostthftw1i pstthfgt5w1f
## (flat) b pstthflt5w1f
## normal(0, 0.25) b mostthftw1i pstthflt5w1f
## (flat) b pstthreatw1f
## normal(0, 0.25) b mostthftw1i pstthreatw1f
## (flat) b pstusedrgw1f
## normal(0, 0.25) b mostthftw1i pstusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept psthackw1f
## normal(0, 2) Intercept pstharmw1f
## normal(0, 2) Intercept pstthfgt5w1f
## normal(0, 2) Intercept pstthflt5w1f
## normal(0, 2) Intercept pstthreatw1f
## normal(0, 2) Intercept pstusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 psthackw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 pstharmw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 pstthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 pstthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 pstthreatw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 pstusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.2.7 Bivariate Corr: T1 stmugw1/past crime
#Bivariate: past crime items ~ mo(stmugw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = pstdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmugw1i',
resp = pstdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmugw1i1',
resp = pstdv_names))
allpstcrime.stmug.fit <- brm(
mvbind(pstthflt5w1f, pstthfgt5w1f, pstthreatw1f, pstharmw1f, pstusedrgw1f, psthackw1f) ~ 1 + mo(stmugw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allpstcrime_stmug_fit",
file_refit = "on_change"
)
out.allpstcrime.stmug.fit <- ppchecks(allpstcrime.stmug.fit)4.3.2.7.1 Coefficient plot (intervals)
out.allpstcrime.stmug.fit[[10]]
4.3.2.7.2 Coefficient plot (distributions)
out.allpstcrime.stmug.fit[[9]]
4.3.2.7.3 PPcheck (density)
p1 <- out.allpstcrime.stmug.fit[[3]] + labs(title = "Past Theft <5BAM (T1)")
p2 <- out.allpstcrime.stmug.fit[[4]] + labs(title = "Past Theft >5BAM (T1)")
p3 <- out.allpstcrime.stmug.fit[[5]] + labs(title = "Past Threat (T1)")
p4 <- out.allpstcrime.stmug.fit[[6]] + labs(title = "Past Harm (T1)")
p5 <- out.allpstcrime.stmug.fit[[7]] + labs(title = "Past Use Drugs (T1)")
p6 <- out.allpstcrime.stmug.fit[[8]] + labs(title = "Past Hack (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.2.7.4 Fit summary
out.allpstcrime.stmug.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: pstthflt5w1f ~ 1 + mo(stmugw1i)
## pstthfgt5w1f ~ 1 + mo(stmugw1i)
## pstthreatw1f ~ 1 + mo(stmugw1i)
## pstharmw1f ~ 1 + mo(stmugw1i)
## pstusedrgw1f ~ 1 + mo(stmugw1i)
## psthackw1f ~ 1 + mo(stmugw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_Intercept -1.93 0.18 -2.31 -1.60 1.00 4997
## pstthfgt5w1f_Intercept -2.32 0.21 -2.79 -1.93 1.00 4241
## pstthreatw1f_Intercept -3.11 0.29 -3.73 -2.60 1.00 3986
## pstharmw1f_Intercept -2.77 0.23 -3.26 -2.33 1.00 4833
## pstusedrgw1f_Intercept -2.84 0.26 -3.37 -2.37 1.00 5146
## psthackw1f_Intercept -3.40 0.31 -4.04 -2.85 1.00 4878
## pstthflt5w1f_mostmugw1i 0.08 0.11 -0.15 0.31 1.00 4860
## pstthfgt5w1f_mostmugw1i 0.16 0.12 -0.07 0.40 1.00 3981
## pstthreatw1f_mostmugw1i 0.12 0.16 -0.21 0.44 1.00 3791
## pstharmw1f_mostmugw1i 0.20 0.13 -0.06 0.46 1.00 5063
## pstusedrgw1f_mostmugw1i 0.18 0.15 -0.11 0.49 1.00 4672
## psthackw1f_mostmugw1i 0.15 0.17 -0.19 0.48 1.00 4309
## Tail_ESS
## pstthflt5w1f_Intercept 3185
## pstthfgt5w1f_Intercept 2789
## pstthreatw1f_Intercept 2692
## pstharmw1f_Intercept 3171
## pstusedrgw1f_Intercept 2830
## psthackw1f_Intercept 3292
## pstthflt5w1f_mostmugw1i 3184
## pstthfgt5w1f_mostmugw1i 3567
## pstthreatw1f_mostmugw1i 2766
## pstharmw1f_mostmugw1i 3064
## pstusedrgw1f_mostmugw1i 3008
## psthackw1f_mostmugw1i 3050
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## pstthflt5w1f_mostmugw1i1[1] 0.26 0.15 0.04 0.61 1.00 5458
## pstthflt5w1f_mostmugw1i1[2] 0.24 0.14 0.04 0.55 1.00 6252
## pstthflt5w1f_mostmugw1i1[3] 0.23 0.14 0.03 0.57 1.00 5679
## pstthflt5w1f_mostmugw1i1[4] 0.27 0.15 0.04 0.61 1.00 5843
## pstthfgt5w1f_mostmugw1i1[1] 0.31 0.16 0.05 0.63 1.00 5395
## pstthfgt5w1f_mostmugw1i1[2] 0.22 0.13 0.03 0.54 1.00 7049
## pstthfgt5w1f_mostmugw1i1[3] 0.21 0.13 0.03 0.52 1.00 5788
## pstthfgt5w1f_mostmugw1i1[4] 0.26 0.15 0.04 0.59 1.00 5418
## pstthreatw1f_mostmugw1i1[1] 0.30 0.16 0.05 0.63 1.00 4738
## pstthreatw1f_mostmugw1i1[2] 0.22 0.14 0.03 0.55 1.00 5399
## pstthreatw1f_mostmugw1i1[3] 0.23 0.14 0.03 0.55 1.00 5928
## pstthreatw1f_mostmugw1i1[4] 0.25 0.14 0.04 0.56 1.00 6365
## pstharmw1f_mostmugw1i1[1] 0.25 0.14 0.04 0.56 1.00 5731
## pstharmw1f_mostmugw1i1[2] 0.24 0.14 0.03 0.57 1.00 5368
## pstharmw1f_mostmugw1i1[3] 0.25 0.14 0.04 0.56 1.00 6598
## pstharmw1f_mostmugw1i1[4] 0.26 0.14 0.04 0.58 1.00 6466
## pstusedrgw1f_mostmugw1i1[1] 0.27 0.15 0.04 0.59 1.00 5819
## pstusedrgw1f_mostmugw1i1[2] 0.21 0.13 0.03 0.52 1.00 5816
## pstusedrgw1f_mostmugw1i1[3] 0.23 0.14 0.03 0.55 1.00 5943
## pstusedrgw1f_mostmugw1i1[4] 0.29 0.16 0.05 0.64 1.00 5484
## psthackw1f_mostmugw1i1[1] 0.27 0.15 0.04 0.60 1.00 5823
## psthackw1f_mostmugw1i1[2] 0.21 0.14 0.03 0.55 1.00 5672
## psthackw1f_mostmugw1i1[3] 0.24 0.14 0.04 0.58 1.00 5816
## psthackw1f_mostmugw1i1[4] 0.27 0.15 0.04 0.61 1.00 5853
## Tail_ESS
## pstthflt5w1f_mostmugw1i1[1] 2488
## pstthflt5w1f_mostmugw1i1[2] 2433
## pstthflt5w1f_mostmugw1i1[3] 2892
## pstthflt5w1f_mostmugw1i1[4] 2941
## pstthfgt5w1f_mostmugw1i1[1] 3269
## pstthfgt5w1f_mostmugw1i1[2] 3096
## pstthfgt5w1f_mostmugw1i1[3] 2566
## pstthfgt5w1f_mostmugw1i1[4] 2615
## pstthreatw1f_mostmugw1i1[1] 2939
## pstthreatw1f_mostmugw1i1[2] 2787
## pstthreatw1f_mostmugw1i1[3] 2746
## pstthreatw1f_mostmugw1i1[4] 3049
## pstharmw1f_mostmugw1i1[1] 2638
## pstharmw1f_mostmugw1i1[2] 2465
## pstharmw1f_mostmugw1i1[3] 2555
## pstharmw1f_mostmugw1i1[4] 2898
## pstusedrgw1f_mostmugw1i1[1] 3269
## pstusedrgw1f_mostmugw1i1[2] 3133
## pstusedrgw1f_mostmugw1i1[3] 2868
## pstusedrgw1f_mostmugw1i1[4] 3211
## psthackw1f_mostmugw1i1[1] 3000
## psthackw1f_mostmugw1i1[2] 2921
## psthackw1f_mostmugw1i1[3] 2781
## psthackw1f_mostmugw1i1[4] 2963
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.2.7.5 Prior summary
out.allpstcrime.stmug.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b psthackw1f
## normal(0, 0.25) b mostmugw1i psthackw1f
## (flat) b pstharmw1f
## normal(0, 0.25) b mostmugw1i pstharmw1f
## (flat) b pstthfgt5w1f
## normal(0, 0.25) b mostmugw1i pstthfgt5w1f
## (flat) b pstthflt5w1f
## normal(0, 0.25) b mostmugw1i pstthflt5w1f
## (flat) b pstthreatw1f
## normal(0, 0.25) b mostmugw1i pstthreatw1f
## (flat) b pstusedrgw1f
## normal(0, 0.25) b mostmugw1i pstusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept psthackw1f
## normal(0, 2) Intercept pstharmw1f
## normal(0, 2) Intercept pstthfgt5w1f
## normal(0, 2) Intercept pstthflt5w1f
## normal(0, 2) Intercept pstthreatw1f
## normal(0, 2) Intercept pstusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 psthackw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 pstharmw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 pstthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 pstthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 pstthreatw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 pstusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.3 Criminal intent items: Bivariate models
Now let’s repeat the process of building simple bivariate models for criminal intent by regressing each intent item on each stress indicator. As with past crime, we will specify a Bernoulli distribution with a logit link.
4.3.3.1 Bivariate Corr: T1 stmonyw1/crim intent
#Bivariate: criminal intent items ~ mo(stmonyw1i)
# prior1 <- c(prior(normal(0, 2), class = Intercept, resp = prjthflt5w1f),
# prior(normal(0, 2), class = Intercept, resp = prjthfgt5w1f),
# prior(normal(0, 2), class = Intercept, resp = prjthreatw1f),
# prior(normal(0, 2), class = Intercept, resp = prjharmw1f),
# prior(normal(0, 2), class = Intercept, resp = prjusedrgw1f),
# prior(normal(0, 2), class = Intercept, resp = prjhackw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = prjthflt5w1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = prjthfgt5w1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = prjthreatw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = prjharmw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = prjusedrgw1f),
# prior(normal(0, 0.25), class = b, coef = mostmonyw1i, resp = prjhackw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = prjthflt5w1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = prjthfgt5w1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = prjthreatw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = prjharmw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = prjusedrgw1f),
# prior(dirichlet(2, 2, 2, 2), class = simo, coef = mostmonyw1i1, resp = prjhackw1f)
# )
#Vectorize priors:
#list of colnames for projected crime DVs
prjdv_names <- noquote(c("prjthflt5w1f", "prjthfgt5w1f", "prjthreatw1f", "prjharmw1f",
"prjusedrgw1f", "prjhackw1f"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmonyw1i',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmonyw1i1',
resp = prjdv_names))
allprjcrime.stmony.fit <- brm(
mvbind(prjthflt5w1f, prjthfgt5w1f, prjthreatw1f, prjharmw1f, prjusedrgw1f, prjhackw1f) ~ 1 + mo(stmonyw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allprjcrime_stmony_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="prjthflt5w1f")
ppcheckdv2 <-pp_check(modelfit, resp="prjthfgt5w1f")
ppcheckdv3 <-pp_check(modelfit, resp="prjthreatw1f")
ppcheckdv4 <-pp_check(modelfit, resp="prjharmw1f")
ppcheckdv5 <-pp_check(modelfit, resp="prjusedrgw1f")
ppcheckdv6 <-pp_check(modelfit, resp="prjhackw1f")
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6,
plotcoefs, plotcoefs2)
return(allchecks)
}
out.allprjcrime.stmony.fit <- ppchecks(allprjcrime.stmony.fit)4.3.3.1.1 Coefficient plot (intervals)
out.allprjcrime.stmony.fit[[10]]
4.3.3.1.2 Coefficient plot (distributions)
out.allprjcrime.stmony.fit[[9]]
4.3.3.1.3 PPcheck (density)
p1 <- out.allprjcrime.stmony.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.allprjcrime.stmony.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.allprjcrime.stmony.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.allprjcrime.stmony.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.allprjcrime.stmony.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.allprjcrime.stmony.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.3.1.4 Fit summary
out.allprjcrime.stmony.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5w1f ~ 1 + mo(stmonyw1i)
## prjthfgt5w1f ~ 1 + mo(stmonyw1i)
## prjthreatw1f ~ 1 + mo(stmonyw1i)
## prjharmw1f ~ 1 + mo(stmonyw1i)
## prjusedrgw1f ~ 1 + mo(stmonyw1i)
## prjhackw1f ~ 1 + mo(stmonyw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_Intercept -2.15 0.30 -2.77 -1.59 1.00 3484
## prjthfgt5w1f_Intercept -2.37 0.31 -3.02 -1.79 1.00 2871
## prjthreatw1f_Intercept -2.53 0.35 -3.24 -1.85 1.00 3507
## prjharmw1f_Intercept -2.66 0.36 -3.38 -1.96 1.00 4205
## prjusedrgw1f_Intercept -2.85 0.40 -3.67 -2.07 1.00 3512
## prjhackw1f_Intercept -3.22 0.46 -4.12 -2.34 1.00 3721
## prjthflt5w1f_mostmonyw1i 0.01 0.12 -0.21 0.25 1.00 3333
## prjthfgt5w1f_mostmonyw1i 0.04 0.13 -0.19 0.29 1.00 3195
## prjthreatw1f_mostmonyw1i -0.13 0.14 -0.40 0.16 1.00 3336
## prjharmw1f_mostmonyw1i -0.20 0.15 -0.49 0.10 1.00 4322
## prjusedrgw1f_mostmonyw1i -0.17 0.16 -0.47 0.17 1.00 3506
## prjhackw1f_mostmonyw1i -0.10 0.18 -0.44 0.24 1.00 3616
## Tail_ESS
## prjthflt5w1f_Intercept 2690
## prjthfgt5w1f_Intercept 1986
## prjthreatw1f_Intercept 2651
## prjharmw1f_Intercept 3193
## prjusedrgw1f_Intercept 2641
## prjhackw1f_Intercept 2774
## prjthflt5w1f_mostmonyw1i 3036
## prjthfgt5w1f_mostmonyw1i 2871
## prjthreatw1f_mostmonyw1i 2796
## prjharmw1f_mostmonyw1i 3302
## prjusedrgw1f_mostmonyw1i 2598
## prjhackw1f_mostmonyw1i 2848
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_mostmonyw1i1[1] 0.26 0.15 0.03 0.60 1.00 6571
## prjthflt5w1f_mostmonyw1i1[2] 0.24 0.14 0.03 0.57 1.00 7815
## prjthflt5w1f_mostmonyw1i1[3] 0.24 0.14 0.03 0.56 1.00 7475
## prjthflt5w1f_mostmonyw1i1[4] 0.26 0.15 0.04 0.58 1.00 7101
## prjthfgt5w1f_mostmonyw1i1[1] 0.26 0.15 0.04 0.60 1.00 6105
## prjthfgt5w1f_mostmonyw1i1[2] 0.24 0.14 0.03 0.57 1.00 5347
## prjthfgt5w1f_mostmonyw1i1[3] 0.24 0.14 0.03 0.58 1.00 6666
## prjthfgt5w1f_mostmonyw1i1[4] 0.26 0.15 0.04 0.60 1.00 7077
## prjthreatw1f_mostmonyw1i1[1] 0.23 0.14 0.03 0.57 1.00 7117
## prjthreatw1f_mostmonyw1i1[2] 0.28 0.15 0.04 0.61 1.00 5452
## prjthreatw1f_mostmonyw1i1[3] 0.26 0.14 0.04 0.58 1.00 6609
## prjthreatw1f_mostmonyw1i1[4] 0.23 0.14 0.03 0.55 1.00 6066
## prjharmw1f_mostmonyw1i1[1] 0.23 0.14 0.04 0.55 1.00 7712
## prjharmw1f_mostmonyw1i1[2] 0.25 0.14 0.04 0.57 1.00 7507
## prjharmw1f_mostmonyw1i1[3] 0.30 0.16 0.05 0.63 1.00 5734
## prjharmw1f_mostmonyw1i1[4] 0.22 0.13 0.03 0.53 1.00 6381
## prjusedrgw1f_mostmonyw1i1[1] 0.23 0.13 0.03 0.53 1.00 7331
## prjusedrgw1f_mostmonyw1i1[2] 0.31 0.16 0.06 0.65 1.00 6542
## prjusedrgw1f_mostmonyw1i1[3] 0.25 0.14 0.04 0.57 1.00 6327
## prjusedrgw1f_mostmonyw1i1[4] 0.21 0.14 0.03 0.53 1.00 6066
## prjhackw1f_mostmonyw1i1[1] 0.26 0.15 0.04 0.59 1.00 5218
## prjhackw1f_mostmonyw1i1[2] 0.27 0.15 0.04 0.61 1.00 6124
## prjhackw1f_mostmonyw1i1[3] 0.23 0.14 0.03 0.54 1.00 6149
## prjhackw1f_mostmonyw1i1[4] 0.24 0.14 0.04 0.57 1.00 6132
## Tail_ESS
## prjthflt5w1f_mostmonyw1i1[1] 2362
## prjthflt5w1f_mostmonyw1i1[2] 2524
## prjthflt5w1f_mostmonyw1i1[3] 2627
## prjthflt5w1f_mostmonyw1i1[4] 2534
## prjthfgt5w1f_mostmonyw1i1[1] 2677
## prjthfgt5w1f_mostmonyw1i1[2] 1992
## prjthfgt5w1f_mostmonyw1i1[3] 2648
## prjthfgt5w1f_mostmonyw1i1[4] 2986
## prjthreatw1f_mostmonyw1i1[1] 2691
## prjthreatw1f_mostmonyw1i1[2] 2559
## prjthreatw1f_mostmonyw1i1[3] 2796
## prjthreatw1f_mostmonyw1i1[4] 2955
## prjharmw1f_mostmonyw1i1[1] 1940
## prjharmw1f_mostmonyw1i1[2] 2595
## prjharmw1f_mostmonyw1i1[3] 3122
## prjharmw1f_mostmonyw1i1[4] 3233
## prjusedrgw1f_mostmonyw1i1[1] 2637
## prjusedrgw1f_mostmonyw1i1[2] 3006
## prjusedrgw1f_mostmonyw1i1[3] 2785
## prjusedrgw1f_mostmonyw1i1[4] 2678
## prjhackw1f_mostmonyw1i1[1] 2900
## prjhackw1f_mostmonyw1i1[2] 3053
## prjhackw1f_mostmonyw1i1[3] 2900
## prjhackw1f_mostmonyw1i1[4] 2890
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.3.1.5 Prior summary
out.allprjcrime.stmony.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b prjhackw1f
## normal(0, 0.25) b mostmonyw1i prjhackw1f
## (flat) b prjharmw1f
## normal(0, 0.25) b mostmonyw1i prjharmw1f
## (flat) b prjthfgt5w1f
## normal(0, 0.25) b mostmonyw1i prjthfgt5w1f
## (flat) b prjthflt5w1f
## normal(0, 0.25) b mostmonyw1i prjthflt5w1f
## (flat) b prjthreatw1f
## normal(0, 0.25) b mostmonyw1i prjthreatw1f
## (flat) b prjusedrgw1f
## normal(0, 0.25) b mostmonyw1i prjusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept prjhackw1f
## normal(0, 2) Intercept prjharmw1f
## normal(0, 2) Intercept prjthfgt5w1f
## normal(0, 2) Intercept prjthflt5w1f
## normal(0, 2) Intercept prjthreatw1f
## normal(0, 2) Intercept prjusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 prjhackw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 prjharmw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 prjthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 prjthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 prjthreatw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 prjusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.3.2 Bivariate Corr: T1 sttranw1/crim intent
#Bivariate: criminal intent items ~ mo(sttranw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mosttranw1i',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mosttranw1i1',
resp = prjdv_names))
allprjcrime.sttran.fit <- brm(
mvbind(prjthflt5w1f, prjthfgt5w1f, prjthreatw1f, prjharmw1f, prjusedrgw1f, prjhackw1f) ~ 1 + mo(sttranw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allprjcrime_sttran_fit",
file_refit = "on_change"
)
out.allprjcrime.sttran.fit <- ppchecks(allprjcrime.sttran.fit)4.3.3.2.1 Coefficient plot (intervals)
out.allprjcrime.sttran.fit[[10]]
4.3.3.2.2 Coefficient plot (distributions)
out.allprjcrime.sttran.fit[[9]]
4.3.3.2.3 PPcheck (density)
p1 <- out.allprjcrime.sttran.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.allprjcrime.sttran.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.allprjcrime.sttran.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.allprjcrime.sttran.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.allprjcrime.sttran.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.allprjcrime.sttran.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.3.2.4 Fit summary
out.allprjcrime.sttran.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5w1f ~ 1 + mo(sttranw1i)
## prjthfgt5w1f ~ 1 + mo(sttranw1i)
## prjthreatw1f ~ 1 + mo(sttranw1i)
## prjharmw1f ~ 1 + mo(sttranw1i)
## prjusedrgw1f ~ 1 + mo(sttranw1i)
## prjhackw1f ~ 1 + mo(sttranw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_Intercept -1.99 0.31 -2.57 -1.36 1.00 3697
## prjthfgt5w1f_Intercept -2.40 0.32 -3.02 -1.78 1.00 3676
## prjthreatw1f_Intercept -2.70 0.37 -3.44 -1.96 1.00 3416
## prjharmw1f_Intercept -2.59 0.40 -3.37 -1.79 1.00 3362
## prjusedrgw1f_Intercept -3.06 0.41 -3.90 -2.28 1.00 3877
## prjhackw1f_Intercept -3.33 0.47 -4.27 -2.39 1.00 3120
## prjthflt5w1f_mosttranw1i -0.06 0.12 -0.30 0.17 1.00 3638
## prjthfgt5w1f_mosttranw1i 0.06 0.13 -0.19 0.30 1.00 3877
## prjthreatw1f_mosttranw1i -0.04 0.15 -0.34 0.24 1.00 3602
## prjharmw1f_mosttranw1i -0.22 0.16 -0.52 0.10 1.00 3664
## prjusedrgw1f_mosttranw1i -0.06 0.16 -0.37 0.26 1.00 3947
## prjhackw1f_mosttranw1i -0.04 0.18 -0.40 0.32 1.00 3432
## Tail_ESS
## prjthflt5w1f_Intercept 2704
## prjthfgt5w1f_Intercept 2889
## prjthreatw1f_Intercept 2657
## prjharmw1f_Intercept 2989
## prjusedrgw1f_Intercept 3130
## prjhackw1f_Intercept 2129
## prjthflt5w1f_mosttranw1i 2999
## prjthfgt5w1f_mosttranw1i 3209
## prjthreatw1f_mosttranw1i 2678
## prjharmw1f_mosttranw1i 3460
## prjusedrgw1f_mosttranw1i 3017
## prjhackw1f_mosttranw1i 3165
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_mosttranw1i1[1] 0.27 0.15 0.03 0.61 1.00 5118
## prjthflt5w1f_mosttranw1i1[2] 0.25 0.14 0.04 0.56 1.00 7251
## prjthflt5w1f_mosttranw1i1[3] 0.24 0.14 0.03 0.56 1.00 7215
## prjthflt5w1f_mosttranw1i1[4] 0.24 0.14 0.03 0.56 1.00 5177
## prjthfgt5w1f_mosttranw1i1[1] 0.25 0.15 0.03 0.59 1.00 5745
## prjthfgt5w1f_mosttranw1i1[2] 0.24 0.14 0.04 0.55 1.00 6604
## prjthfgt5w1f_mosttranw1i1[3] 0.24 0.14 0.04 0.56 1.00 6801
## prjthfgt5w1f_mosttranw1i1[4] 0.28 0.15 0.04 0.61 1.00 4829
## prjthreatw1f_mosttranw1i1[1] 0.26 0.15 0.04 0.59 1.00 6748
## prjthreatw1f_mosttranw1i1[2] 0.24 0.14 0.04 0.56 1.00 6576
## prjthreatw1f_mosttranw1i1[3] 0.25 0.14 0.03 0.56 1.00 6908
## prjthreatw1f_mosttranw1i1[4] 0.25 0.14 0.04 0.57 1.00 6029
## prjharmw1f_mosttranw1i1[1] 0.26 0.14 0.04 0.57 1.00 5831
## prjharmw1f_mosttranw1i1[2] 0.29 0.15 0.05 0.62 1.00 5557
## prjharmw1f_mosttranw1i1[3] 0.24 0.13 0.03 0.55 1.00 6579
## prjharmw1f_mosttranw1i1[4] 0.22 0.13 0.03 0.52 1.00 6348
## prjusedrgw1f_mosttranw1i1[1] 0.25 0.14 0.04 0.58 1.00 7972
## prjusedrgw1f_mosttranw1i1[2] 0.26 0.15 0.04 0.59 1.00 6531
## prjusedrgw1f_mosttranw1i1[3] 0.25 0.14 0.04 0.56 1.00 6098
## prjusedrgw1f_mosttranw1i1[4] 0.24 0.14 0.04 0.57 1.00 6881
## prjhackw1f_mosttranw1i1[1] 0.26 0.15 0.04 0.59 1.00 6531
## prjhackw1f_mosttranw1i1[2] 0.26 0.15 0.04 0.60 1.00 5651
## prjhackw1f_mosttranw1i1[3] 0.24 0.14 0.04 0.54 1.00 6715
## prjhackw1f_mosttranw1i1[4] 0.25 0.14 0.04 0.57 1.00 5016
## Tail_ESS
## prjthflt5w1f_mosttranw1i1[1] 2160
## prjthflt5w1f_mosttranw1i1[2] 2685
## prjthflt5w1f_mosttranw1i1[3] 3136
## prjthflt5w1f_mosttranw1i1[4] 3183
## prjthfgt5w1f_mosttranw1i1[1] 2595
## prjthfgt5w1f_mosttranw1i1[2] 2594
## prjthfgt5w1f_mosttranw1i1[3] 2595
## prjthfgt5w1f_mosttranw1i1[4] 2799
## prjthreatw1f_mosttranw1i1[1] 2534
## prjthreatw1f_mosttranw1i1[2] 2283
## prjthreatw1f_mosttranw1i1[3] 2858
## prjthreatw1f_mosttranw1i1[4] 3026
## prjharmw1f_mosttranw1i1[1] 2992
## prjharmw1f_mosttranw1i1[2] 2319
## prjharmw1f_mosttranw1i1[3] 2959
## prjharmw1f_mosttranw1i1[4] 2981
## prjusedrgw1f_mosttranw1i1[1] 3117
## prjusedrgw1f_mosttranw1i1[2] 2777
## prjusedrgw1f_mosttranw1i1[3] 3170
## prjusedrgw1f_mosttranw1i1[4] 2768
## prjhackw1f_mosttranw1i1[1] 2693
## prjhackw1f_mosttranw1i1[2] 2924
## prjhackw1f_mosttranw1i1[3] 3300
## prjhackw1f_mosttranw1i1[4] 2831
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.3.2.5 Prior summary
out.allprjcrime.sttran.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b prjhackw1f
## normal(0, 0.25) b mosttranw1i prjhackw1f
## (flat) b prjharmw1f
## normal(0, 0.25) b mosttranw1i prjharmw1f
## (flat) b prjthfgt5w1f
## normal(0, 0.25) b mosttranw1i prjthfgt5w1f
## (flat) b prjthflt5w1f
## normal(0, 0.25) b mosttranw1i prjthflt5w1f
## (flat) b prjthreatw1f
## normal(0, 0.25) b mosttranw1i prjthreatw1f
## (flat) b prjusedrgw1f
## normal(0, 0.25) b mosttranw1i prjusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept prjhackw1f
## normal(0, 2) Intercept prjharmw1f
## normal(0, 2) Intercept prjthfgt5w1f
## normal(0, 2) Intercept prjthflt5w1f
## normal(0, 2) Intercept prjthreatw1f
## normal(0, 2) Intercept prjusedrgw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 prjhackw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 prjharmw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 prjthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 prjthflt5w1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 prjthreatw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 prjusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.3.3 Bivariate Corr: T1 strespw1/crim intent
#Bivariate: criminal intent items ~ mo(strespw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostrespw1i',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostrespw1i1',
resp = prjdv_names))
allprjcrime.stresp.fit <- brm(
mvbind(prjthflt5w1f, prjthfgt5w1f, prjthreatw1f, prjharmw1f, prjusedrgw1f, prjhackw1f) ~ 1 + mo(strespw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allprjcrime_stresp_fit",
file_refit = "on_change"
)
out.allprjcrime.stresp.fit <- ppchecks(allprjcrime.stresp.fit)4.3.3.3.1 Coefficient plot (intervals)
out.allprjcrime.stresp.fit[[10]]
4.3.3.3.2 Coefficient plot (distributions)
out.allprjcrime.stresp.fit[[9]]
4.3.3.3.3 PPcheck (density)
p1 <- out.allprjcrime.stresp.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.allprjcrime.stresp.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.allprjcrime.stresp.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.allprjcrime.stresp.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.allprjcrime.stresp.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.allprjcrime.stresp.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.3.3.4 Fit summary
out.allprjcrime.stresp.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5w1f ~ 1 + mo(strespw1i)
## prjthfgt5w1f ~ 1 + mo(strespw1i)
## prjthreatw1f ~ 1 + mo(strespw1i)
## prjharmw1f ~ 1 + mo(strespw1i)
## prjusedrgw1f ~ 1 + mo(strespw1i)
## prjhackw1f ~ 1 + mo(strespw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_Intercept -2.55 0.30 -3.19 -2.00 1.00 4232
## prjthfgt5w1f_Intercept -2.57 0.31 -3.19 -2.00 1.00 3996
## prjthreatw1f_Intercept -3.31 0.39 -4.11 -2.60 1.00 4053
## prjharmw1f_Intercept -3.24 0.39 -4.05 -2.50 1.00 4704
## prjusedrgw1f_Intercept -3.56 0.41 -4.40 -2.83 1.00 3937
## prjhackw1f_Intercept -3.66 0.43 -4.55 -2.88 1.00 3794
## prjthflt5w1f_mostrespw1i 0.17 0.10 -0.02 0.37 1.00 4308
## prjthfgt5w1f_mostrespw1i 0.11 0.10 -0.09 0.31 1.00 4114
## prjthreatw1f_mostrespw1i 0.21 0.12 -0.03 0.45 1.00 3837
## prjharmw1f_mostrespw1i 0.07 0.13 -0.18 0.33 1.00 4699
## prjusedrgw1f_mostrespw1i 0.15 0.13 -0.11 0.42 1.00 4097
## prjhackw1f_mostrespw1i 0.09 0.14 -0.19 0.38 1.00 4080
## Tail_ESS
## prjthflt5w1f_Intercept 3142
## prjthfgt5w1f_Intercept 2753
## prjthreatw1f_Intercept 2456
## prjharmw1f_Intercept 3161
## prjusedrgw1f_Intercept 2707
## prjhackw1f_Intercept 2906
## prjthflt5w1f_mostrespw1i 3249
## prjthfgt5w1f_mostrespw1i 2950
## prjthreatw1f_mostrespw1i 2554
## prjharmw1f_mostrespw1i 2859
## prjusedrgw1f_mostrespw1i 2697
## prjhackw1f_mostrespw1i 2844
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_mostrespw1i1[1] 0.26 0.14 0.04 0.59 1.00 7241
## prjthflt5w1f_mostrespw1i1[2] 0.25 0.14 0.04 0.55 1.00 8092
## prjthflt5w1f_mostrespw1i1[3] 0.25 0.14 0.04 0.56 1.00 7926
## prjthflt5w1f_mostrespw1i1[4] 0.24 0.13 0.04 0.54 1.00 6706
## prjthfgt5w1f_mostrespw1i1[1] 0.26 0.15 0.04 0.59 1.00 6544
## prjthfgt5w1f_mostrespw1i1[2] 0.25 0.14 0.04 0.58 1.00 7592
## prjthfgt5w1f_mostrespw1i1[3] 0.24 0.14 0.04 0.55 1.00 6316
## prjthfgt5w1f_mostrespw1i1[4] 0.24 0.13 0.04 0.55 1.00 6785
## prjthreatw1f_mostrespw1i1[1] 0.24 0.14 0.03 0.54 1.00 7615
## prjthreatw1f_mostrespw1i1[2] 0.25 0.14 0.04 0.56 1.00 7579
## prjthreatw1f_mostrespw1i1[3] 0.26 0.14 0.04 0.58 1.00 6500
## prjthreatw1f_mostrespw1i1[4] 0.25 0.14 0.04 0.57 1.00 6677
## prjharmw1f_mostrespw1i1[1] 0.26 0.15 0.04 0.60 1.00 7034
## prjharmw1f_mostrespw1i1[2] 0.25 0.14 0.03 0.58 1.00 6753
## prjharmw1f_mostrespw1i1[3] 0.24 0.14 0.03 0.57 1.00 7317
## prjharmw1f_mostrespw1i1[4] 0.25 0.14 0.04 0.57 1.00 6599
## prjusedrgw1f_mostrespw1i1[1] 0.24 0.14 0.04 0.56 1.00 7771
## prjusedrgw1f_mostrespw1i1[2] 0.24 0.14 0.03 0.57 1.00 7238
## prjusedrgw1f_mostrespw1i1[3] 0.27 0.15 0.04 0.61 1.00 6398
## prjusedrgw1f_mostrespw1i1[4] 0.25 0.15 0.04 0.58 1.00 7796
## prjhackw1f_mostrespw1i1[1] 0.25 0.15 0.04 0.59 1.00 9074
## prjhackw1f_mostrespw1i1[2] 0.25 0.14 0.04 0.56 1.00 7174
## prjhackw1f_mostrespw1i1[3] 0.24 0.14 0.04 0.58 1.00 6448
## prjhackw1f_mostrespw1i1[4] 0.25 0.14 0.04 0.57 1.00 5699
## Tail_ESS
## prjthflt5w1f_mostrespw1i1[1] 2869
## prjthflt5w1f_mostrespw1i1[2] 3001
## prjthflt5w1f_mostrespw1i1[3] 3092
## prjthflt5w1f_mostrespw1i1[4] 3003
## prjthfgt5w1f_mostrespw1i1[1] 2252
## prjthfgt5w1f_mostrespw1i1[2] 2396
## prjthfgt5w1f_mostrespw1i1[3] 3410
## prjthfgt5w1f_mostrespw1i1[4] 2626
## prjthreatw1f_mostrespw1i1[1] 2540
## prjthreatw1f_mostrespw1i1[2] 2649
## prjthreatw1f_mostrespw1i1[3] 2849
## prjthreatw1f_mostrespw1i1[4] 2802
## prjharmw1f_mostrespw1i1[1] 2469
## prjharmw1f_mostrespw1i1[2] 2697
## prjharmw1f_mostrespw1i1[3] 2893
## prjharmw1f_mostrespw1i1[4] 2695
## prjusedrgw1f_mostrespw1i1[1] 2544
## prjusedrgw1f_mostrespw1i1[2] 2714
## prjusedrgw1f_mostrespw1i1[3] 2906
## prjusedrgw1f_mostrespw1i1[4] 2924
## prjhackw1f_mostrespw1i1[1] 2769
## prjhackw1f_mostrespw1i1[2] 2877
## prjhackw1f_mostrespw1i1[3] 3024
## prjhackw1f_mostrespw1i1[4] 3104
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.3.3.5 Prior summary
out.allprjcrime.stresp.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b prjhackw1f
## normal(0, 0.25) b mostrespw1i prjhackw1f
## (flat) b prjharmw1f
## normal(0, 0.25) b mostrespw1i prjharmw1f
## (flat) b prjthfgt5w1f
## normal(0, 0.25) b mostrespw1i prjthfgt5w1f
## (flat) b prjthflt5w1f
## normal(0, 0.25) b mostrespw1i prjthflt5w1f
## (flat) b prjthreatw1f
## normal(0, 0.25) b mostrespw1i prjthreatw1f
## (flat) b prjusedrgw1f
## normal(0, 0.25) b mostrespw1i prjusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept prjhackw1f
## normal(0, 2) Intercept prjharmw1f
## normal(0, 2) Intercept prjthfgt5w1f
## normal(0, 2) Intercept prjthflt5w1f
## normal(0, 2) Intercept prjthreatw1f
## normal(0, 2) Intercept prjusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 prjhackw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 prjharmw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 prjthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 prjthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 prjthreatw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 prjusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.3.4 Bivariate Corr: T1 stfairw1/crim intent
#Bivariate: criminal intent items ~ mo(stfairw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostfairw1i',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostfairw1i1',
resp = prjdv_names))
allprjcrime.stfair.fit <- brm(
mvbind(prjthflt5w1f, prjthfgt5w1f, prjthreatw1f, prjharmw1f, prjusedrgw1f, prjhackw1f) ~ 1 + mo(stfairw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allprjcrime_stfair_fit",
file_refit = "on_change"
)
out.allprjcrime.stfair.fit <- ppchecks(allprjcrime.stfair.fit)4.3.3.4.1 Coefficient plot (intervals)
out.allprjcrime.stfair.fit[[10]]
4.3.3.4.2 Coefficient plot (distributions)
out.allprjcrime.stfair.fit[[9]]
4.3.3.4.3 PPcheck (density)
p1 <- out.allprjcrime.stfair.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.allprjcrime.stfair.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.allprjcrime.stfair.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.allprjcrime.stfair.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.allprjcrime.stfair.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.allprjcrime.stfair.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.3.4.4 Fit summary
out.allprjcrime.stfair.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5w1f ~ 1 + mo(stfairw1i)
## prjthfgt5w1f ~ 1 + mo(stfairw1i)
## prjthreatw1f ~ 1 + mo(stfairw1i)
## prjharmw1f ~ 1 + mo(stfairw1i)
## prjusedrgw1f ~ 1 + mo(stfairw1i)
## prjhackw1f ~ 1 + mo(stfairw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_Intercept -2.75 0.32 -3.42 -2.18 1.00 3991
## prjthfgt5w1f_Intercept -2.76 0.31 -3.40 -2.17 1.00 4300
## prjthreatw1f_Intercept -3.53 0.39 -4.35 -2.81 1.00 4444
## prjharmw1f_Intercept -3.23 0.39 -4.06 -2.51 1.00 4002
## prjusedrgw1f_Intercept -3.44 0.42 -4.34 -2.66 1.00 3904
## prjhackw1f_Intercept -3.91 0.48 -4.90 -3.01 1.00 3949
## prjthflt5w1f_mostfairw1i 0.24 0.10 0.06 0.44 1.00 4532
## prjthfgt5w1f_mostfairw1i 0.19 0.10 -0.01 0.39 1.00 4425
## prjthreatw1f_mostfairw1i 0.28 0.12 0.05 0.52 1.00 3741
## prjharmw1f_mostfairw1i 0.07 0.13 -0.17 0.33 1.00 3956
## prjusedrgw1f_mostfairw1i 0.09 0.13 -0.16 0.36 1.00 3857
## prjhackw1f_mostfairw1i 0.19 0.15 -0.10 0.48 1.00 4159
## Tail_ESS
## prjthflt5w1f_Intercept 2198
## prjthfgt5w1f_Intercept 2683
## prjthreatw1f_Intercept 3073
## prjharmw1f_Intercept 2851
## prjusedrgw1f_Intercept 2810
## prjhackw1f_Intercept 2795
## prjthflt5w1f_mostfairw1i 2799
## prjthfgt5w1f_mostfairw1i 2899
## prjthreatw1f_mostfairw1i 3306
## prjharmw1f_mostfairw1i 3131
## prjusedrgw1f_mostfairw1i 2980
## prjhackw1f_mostfairw1i 2909
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_mostfairw1i1[1] 0.25 0.14 0.04 0.56 1.00 6001
## prjthflt5w1f_mostfairw1i1[2] 0.22 0.13 0.04 0.51 1.00 8112
## prjthflt5w1f_mostfairw1i1[3] 0.23 0.14 0.04 0.55 1.00 7096
## prjthflt5w1f_mostfairw1i1[4] 0.30 0.15 0.05 0.61 1.00 7778
## prjthfgt5w1f_mostfairw1i1[1] 0.24 0.13 0.04 0.54 1.00 8038
## prjthfgt5w1f_mostfairw1i1[2] 0.24 0.14 0.03 0.56 1.00 7433
## prjthfgt5w1f_mostfairw1i1[3] 0.24 0.14 0.04 0.56 1.00 8091
## prjthfgt5w1f_mostfairw1i1[4] 0.28 0.15 0.05 0.60 1.00 6527
## prjthreatw1f_mostfairw1i1[1] 0.21 0.12 0.03 0.51 1.00 8267
## prjthreatw1f_mostfairw1i1[2] 0.24 0.14 0.04 0.55 1.00 8298
## prjthreatw1f_mostfairw1i1[3] 0.25 0.14 0.04 0.56 1.00 6994
## prjthreatw1f_mostfairw1i1[4] 0.31 0.15 0.06 0.64 1.00 7787
## prjharmw1f_mostfairw1i1[1] 0.25 0.14 0.04 0.58 1.00 8067
## prjharmw1f_mostfairw1i1[2] 0.25 0.14 0.04 0.58 1.00 6918
## prjharmw1f_mostfairw1i1[3] 0.25 0.15 0.04 0.59 1.00 7931
## prjharmw1f_mostfairw1i1[4] 0.24 0.14 0.04 0.55 1.00 6872
## prjusedrgw1f_mostfairw1i1[1] 0.25 0.14 0.04 0.58 1.00 9948
## prjusedrgw1f_mostfairw1i1[2] 0.25 0.15 0.04 0.59 1.00 7962
## prjusedrgw1f_mostfairw1i1[3] 0.26 0.14 0.04 0.58 1.00 7592
## prjusedrgw1f_mostfairw1i1[4] 0.24 0.14 0.03 0.57 1.00 7723
## prjhackw1f_mostfairw1i1[1] 0.22 0.13 0.03 0.53 1.00 6906
## prjhackw1f_mostfairw1i1[2] 0.25 0.14 0.04 0.57 1.00 7537
## prjhackw1f_mostfairw1i1[3] 0.28 0.15 0.04 0.62 1.00 6839
## prjhackw1f_mostfairw1i1[4] 0.25 0.15 0.04 0.58 1.00 7244
## Tail_ESS
## prjthflt5w1f_mostfairw1i1[1] 2552
## prjthflt5w1f_mostfairw1i1[2] 2670
## prjthflt5w1f_mostfairw1i1[3] 2726
## prjthflt5w1f_mostfairw1i1[4] 2688
## prjthfgt5w1f_mostfairw1i1[1] 2738
## prjthfgt5w1f_mostfairw1i1[2] 2088
## prjthfgt5w1f_mostfairw1i1[3] 2824
## prjthfgt5w1f_mostfairw1i1[4] 2768
## prjthreatw1f_mostfairw1i1[1] 2814
## prjthreatw1f_mostfairw1i1[2] 2563
## prjthreatw1f_mostfairw1i1[3] 2858
## prjthreatw1f_mostfairw1i1[4] 2763
## prjharmw1f_mostfairw1i1[1] 2909
## prjharmw1f_mostfairw1i1[2] 2444
## prjharmw1f_mostfairw1i1[3] 2565
## prjharmw1f_mostfairw1i1[4] 3221
## prjusedrgw1f_mostfairw1i1[1] 2943
## prjusedrgw1f_mostfairw1i1[2] 2667
## prjusedrgw1f_mostfairw1i1[3] 2978
## prjusedrgw1f_mostfairw1i1[4] 2836
## prjhackw1f_mostfairw1i1[1] 2810
## prjhackw1f_mostfairw1i1[2] 2473
## prjhackw1f_mostfairw1i1[3] 2317
## prjhackw1f_mostfairw1i1[4] 2748
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.3.4.5 Prior summary
out.allprjcrime.stfair.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b prjhackw1f
## normal(0, 0.25) b mostfairw1i prjhackw1f
## (flat) b prjharmw1f
## normal(0, 0.25) b mostfairw1i prjharmw1f
## (flat) b prjthfgt5w1f
## normal(0, 0.25) b mostfairw1i prjthfgt5w1f
## (flat) b prjthflt5w1f
## normal(0, 0.25) b mostfairw1i prjthflt5w1f
## (flat) b prjthreatw1f
## normal(0, 0.25) b mostfairw1i prjthreatw1f
## (flat) b prjusedrgw1f
## normal(0, 0.25) b mostfairw1i prjusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept prjhackw1f
## normal(0, 2) Intercept prjharmw1f
## normal(0, 2) Intercept prjthfgt5w1f
## normal(0, 2) Intercept prjthflt5w1f
## normal(0, 2) Intercept prjthreatw1f
## normal(0, 2) Intercept prjusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 prjhackw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 prjharmw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 prjthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 prjthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 prjthreatw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 prjusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.3.5 Bivariate Corr: T1 stjobw1/crim intent
#Bivariate: criminal intent items ~ mo(stjobw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostjobw1i',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostjobw1i1',
resp = prjdv_names))
allprjcrime.stjob.fit <- brm(
mvbind(prjthflt5w1f, prjthfgt5w1f, prjthreatw1f, prjharmw1f, prjusedrgw1f, prjhackw1f) ~ 1 + mo(stjobw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allprjcrime_stjob_fit",
file_refit = "on_change"
)
out.allprjcrime.stjob.fit <- ppchecks(allprjcrime.stjob.fit)4.3.3.5.1 Coefficient plot (intervals)
out.allprjcrime.stjob.fit[[10]]
4.3.3.5.2 Coefficient plot (distributions)
out.allprjcrime.stjob.fit[[9]]
4.3.3.5.3 PPcheck (density)
p1 <- out.allprjcrime.stjob.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.allprjcrime.stjob.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.allprjcrime.stjob.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.allprjcrime.stjob.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.allprjcrime.stjob.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.allprjcrime.stjob.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.3.5.4 Fit summary
out.allprjcrime.stjob.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5w1f ~ 1 + mo(stjobw1i)
## prjthfgt5w1f ~ 1 + mo(stjobw1i)
## prjthreatw1f ~ 1 + mo(stjobw1i)
## prjharmw1f ~ 1 + mo(stjobw1i)
## prjusedrgw1f ~ 1 + mo(stjobw1i)
## prjhackw1f ~ 1 + mo(stjobw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_Intercept -2.87 0.32 -3.54 -2.28 1.00 3587
## prjthfgt5w1f_Intercept -3.10 0.34 -3.81 -2.49 1.00 3845
## prjthreatw1f_Intercept -3.71 0.42 -4.59 -2.95 1.00 3699
## prjharmw1f_Intercept -3.33 0.40 -4.15 -2.60 1.00 3128
## prjusedrgw1f_Intercept -3.81 0.44 -4.76 -3.00 1.00 3384
## prjhackw1f_Intercept -4.63 0.52 -5.74 -3.70 1.00 3660
## prjthflt5w1f_mostjobw1i 0.30 0.10 0.11 0.50 1.00 3856
## prjthfgt5w1f_mostjobw1i 0.33 0.10 0.13 0.54 1.00 3961
## prjthreatw1f_mostjobw1i 0.37 0.13 0.12 0.62 1.00 3779
## prjharmw1f_mostjobw1i 0.11 0.13 -0.14 0.37 1.00 3540
## prjusedrgw1f_mostjobw1i 0.24 0.14 -0.03 0.52 1.00 3041
## prjhackw1f_mostjobw1i 0.48 0.16 0.18 0.79 1.00 3555
## Tail_ESS
## prjthflt5w1f_Intercept 2729
## prjthfgt5w1f_Intercept 2478
## prjthreatw1f_Intercept 2753
## prjharmw1f_Intercept 2714
## prjusedrgw1f_Intercept 2962
## prjhackw1f_Intercept 2656
## prjthflt5w1f_mostjobw1i 3013
## prjthfgt5w1f_mostjobw1i 2849
## prjthreatw1f_mostjobw1i 2922
## prjharmw1f_mostjobw1i 2821
## prjusedrgw1f_mostjobw1i 3065
## prjhackw1f_mostjobw1i 2775
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_mostjobw1i1[1] 0.18 0.11 0.02 0.45 1.00 4640
## prjthflt5w1f_mostjobw1i1[2] 0.20 0.12 0.03 0.49 1.00 6269
## prjthflt5w1f_mostjobw1i1[3] 0.42 0.16 0.11 0.72 1.00 5628
## prjthflt5w1f_mostjobw1i1[4] 0.19 0.11 0.03 0.47 1.00 6821
## prjthfgt5w1f_mostjobw1i1[1] 0.18 0.11 0.02 0.45 1.00 7121
## prjthfgt5w1f_mostjobw1i1[2] 0.23 0.14 0.03 0.54 1.00 6704
## prjthfgt5w1f_mostjobw1i1[3] 0.35 0.16 0.08 0.68 1.00 5871
## prjthfgt5w1f_mostjobw1i1[4] 0.24 0.13 0.04 0.52 1.00 6596
## prjthreatw1f_mostjobw1i1[1] 0.20 0.12 0.03 0.47 1.00 6312
## prjthreatw1f_mostjobw1i1[2] 0.22 0.13 0.03 0.50 1.00 6621
## prjthreatw1f_mostjobw1i1[3] 0.30 0.15 0.06 0.61 1.00 6942
## prjthreatw1f_mostjobw1i1[4] 0.28 0.14 0.05 0.57 1.00 6596
## prjharmw1f_mostjobw1i1[1] 0.25 0.14 0.04 0.57 1.00 5805
## prjharmw1f_mostjobw1i1[2] 0.25 0.15 0.03 0.60 1.00 6321
## prjharmw1f_mostjobw1i1[3] 0.27 0.15 0.04 0.61 1.00 5408
## prjharmw1f_mostjobw1i1[4] 0.22 0.14 0.03 0.55 1.00 5669
## prjusedrgw1f_mostjobw1i1[1] 0.23 0.14 0.03 0.55 1.00 6843
## prjusedrgw1f_mostjobw1i1[2] 0.25 0.14 0.03 0.58 1.00 6942
## prjusedrgw1f_mostjobw1i1[3] 0.28 0.15 0.05 0.61 1.00 6209
## prjusedrgw1f_mostjobw1i1[4] 0.24 0.14 0.04 0.55 1.00 8515
## prjhackw1f_mostjobw1i1[1] 0.17 0.11 0.02 0.42 1.00 6447
## prjhackw1f_mostjobw1i1[2] 0.20 0.12 0.02 0.49 1.00 7177
## prjhackw1f_mostjobw1i1[3] 0.28 0.15 0.05 0.60 1.00 6916
## prjhackw1f_mostjobw1i1[4] 0.35 0.15 0.08 0.67 1.00 6385
## Tail_ESS
## prjthflt5w1f_mostjobw1i1[1] 2148
## prjthflt5w1f_mostjobw1i1[2] 2505
## prjthflt5w1f_mostjobw1i1[3] 3243
## prjthflt5w1f_mostjobw1i1[4] 2868
## prjthfgt5w1f_mostjobw1i1[1] 2441
## prjthfgt5w1f_mostjobw1i1[2] 2619
## prjthfgt5w1f_mostjobw1i1[3] 2605
## prjthfgt5w1f_mostjobw1i1[4] 2633
## prjthreatw1f_mostjobw1i1[1] 2639
## prjthreatw1f_mostjobw1i1[2] 2599
## prjthreatw1f_mostjobw1i1[3] 2950
## prjthreatw1f_mostjobw1i1[4] 2344
## prjharmw1f_mostjobw1i1[1] 2848
## prjharmw1f_mostjobw1i1[2] 2503
## prjharmw1f_mostjobw1i1[3] 2762
## prjharmw1f_mostjobw1i1[4] 2648
## prjusedrgw1f_mostjobw1i1[1] 2655
## prjusedrgw1f_mostjobw1i1[2] 2473
## prjusedrgw1f_mostjobw1i1[3] 2880
## prjusedrgw1f_mostjobw1i1[4] 2879
## prjhackw1f_mostjobw1i1[1] 2267
## prjhackw1f_mostjobw1i1[2] 2406
## prjhackw1f_mostjobw1i1[3] 3096
## prjhackw1f_mostjobw1i1[4] 2609
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.3.5.5 Prior summary
out.allprjcrime.stjob.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b prjhackw1f
## normal(0, 0.25) b mostjobw1i prjhackw1f
## (flat) b prjharmw1f
## normal(0, 0.25) b mostjobw1i prjharmw1f
## (flat) b prjthfgt5w1f
## normal(0, 0.25) b mostjobw1i prjthfgt5w1f
## (flat) b prjthflt5w1f
## normal(0, 0.25) b mostjobw1i prjthflt5w1f
## (flat) b prjthreatw1f
## normal(0, 0.25) b mostjobw1i prjthreatw1f
## (flat) b prjusedrgw1f
## normal(0, 0.25) b mostjobw1i prjusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept prjhackw1f
## normal(0, 2) Intercept prjharmw1f
## normal(0, 2) Intercept prjthfgt5w1f
## normal(0, 2) Intercept prjthflt5w1f
## normal(0, 2) Intercept prjthreatw1f
## normal(0, 2) Intercept prjusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 prjhackw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 prjharmw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 prjthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 prjthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 prjthreatw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 prjusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.3.6 Bivariate Corr: T1 stthftw1/crim intent
#Bivariate: criminal intent items ~ mo(stthftw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostthftw1i',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostthftw1i1',
resp = prjdv_names))
allprjcrime.stthft.fit <- brm(
mvbind(prjthflt5w1f, prjthfgt5w1f, prjthreatw1f, prjharmw1f, prjusedrgw1f, prjhackw1f) ~ 1 + mo(stthftw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allprjcrime_stthft_fit",
file_refit = "on_change"
)
out.allprjcrime.stthft.fit <- ppchecks(allprjcrime.stthft.fit)4.3.3.6.1 Coefficient plot (intervals)
out.allprjcrime.stthft.fit[[10]]
4.3.3.6.2 Coefficient plot (distributions)
out.allprjcrime.stthft.fit[[9]]
4.3.3.6.3 PPcheck (density)
p1 <- out.allprjcrime.stthft.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.allprjcrime.stthft.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.allprjcrime.stthft.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.allprjcrime.stthft.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.allprjcrime.stthft.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.allprjcrime.stthft.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.3.6.4 Fit summary
out.allprjcrime.stthft.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5w1f ~ 1 + mo(stthftw1i)
## prjthfgt5w1f ~ 1 + mo(stthftw1i)
## prjthreatw1f ~ 1 + mo(stthftw1i)
## prjharmw1f ~ 1 + mo(stthftw1i)
## prjusedrgw1f ~ 1 + mo(stthftw1i)
## prjhackw1f ~ 1 + mo(stthftw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_Intercept -2.37 0.22 -2.83 -1.98 1.00 5040
## prjthfgt5w1f_Intercept -2.63 0.24 -3.11 -2.19 1.00 5098
## prjthreatw1f_Intercept -3.01 0.28 -3.57 -2.50 1.00 4906
## prjharmw1f_Intercept -3.14 0.29 -3.74 -2.59 1.00 4003
## prjusedrgw1f_Intercept -3.19 0.30 -3.83 -2.63 1.00 4766
## prjhackw1f_Intercept -3.42 0.32 -4.07 -2.81 1.00 3739
## prjthflt5w1f_mostthftw1i 0.18 0.11 -0.04 0.39 1.00 4622
## prjthfgt5w1f_mostthftw1i 0.26 0.11 0.05 0.48 1.00 4735
## prjthreatw1f_mostthftw1i 0.16 0.13 -0.12 0.42 1.00 5003
## prjharmw1f_mostthftw1i 0.06 0.15 -0.24 0.35 1.00 4719
## prjusedrgw1f_mostthftw1i -0.02 0.16 -0.33 0.28 1.00 4737
## prjhackw1f_mostthftw1i -0.02 0.17 -0.35 0.30 1.00 4376
## Tail_ESS
## prjthflt5w1f_Intercept 2708
## prjthfgt5w1f_Intercept 2994
## prjthreatw1f_Intercept 2903
## prjharmw1f_Intercept 2896
## prjusedrgw1f_Intercept 2127
## prjhackw1f_Intercept 3110
## prjthflt5w1f_mostthftw1i 3231
## prjthfgt5w1f_mostthftw1i 2725
## prjthreatw1f_mostthftw1i 3428
## prjharmw1f_mostthftw1i 2943
## prjusedrgw1f_mostthftw1i 2306
## prjhackw1f_mostthftw1i 3243
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_mostthftw1i1[1] 0.25 0.14 0.04 0.55 1.00 5444
## prjthflt5w1f_mostthftw1i1[2] 0.23 0.13 0.04 0.53 1.00 5700
## prjthflt5w1f_mostthftw1i1[3] 0.29 0.15 0.05 0.61 1.00 6085
## prjthflt5w1f_mostthftw1i1[4] 0.23 0.13 0.03 0.53 1.00 6068
## prjthfgt5w1f_mostthftw1i1[1] 0.19 0.11 0.03 0.46 1.00 5641
## prjthfgt5w1f_mostthftw1i1[2] 0.28 0.15 0.04 0.60 1.00 5654
## prjthfgt5w1f_mostthftw1i1[3] 0.32 0.16 0.06 0.66 1.00 5617
## prjthfgt5w1f_mostthftw1i1[4] 0.21 0.12 0.03 0.50 1.00 5416
## prjthreatw1f_mostthftw1i1[1] 0.26 0.14 0.04 0.58 1.00 5747
## prjthreatw1f_mostthftw1i1[2] 0.23 0.13 0.03 0.54 1.00 5726
## prjthreatw1f_mostthftw1i1[3] 0.27 0.15 0.04 0.59 1.00 5361
## prjthreatw1f_mostthftw1i1[4] 0.24 0.14 0.04 0.56 1.00 6179
## prjharmw1f_mostthftw1i1[1] 0.25 0.14 0.04 0.57 1.00 4711
## prjharmw1f_mostthftw1i1[2] 0.24 0.14 0.04 0.56 1.00 6023
## prjharmw1f_mostthftw1i1[3] 0.25 0.14 0.04 0.59 1.00 5748
## prjharmw1f_mostthftw1i1[4] 0.26 0.15 0.04 0.58 1.00 5191
## prjusedrgw1f_mostthftw1i1[1] 0.24 0.14 0.04 0.55 1.00 5326
## prjusedrgw1f_mostthftw1i1[2] 0.24 0.14 0.03 0.57 1.00 5958
## prjusedrgw1f_mostthftw1i1[3] 0.25 0.14 0.04 0.58 1.00 5673
## prjusedrgw1f_mostthftw1i1[4] 0.27 0.15 0.04 0.61 1.00 5697
## prjhackw1f_mostthftw1i1[1] 0.25 0.15 0.03 0.59 1.00 5608
## prjhackw1f_mostthftw1i1[2] 0.24 0.14 0.04 0.56 1.00 5641
## prjhackw1f_mostthftw1i1[3] 0.25 0.14 0.04 0.57 1.00 5859
## prjhackw1f_mostthftw1i1[4] 0.26 0.14 0.04 0.57 1.00 5510
## Tail_ESS
## prjthflt5w1f_mostthftw1i1[1] 2291
## prjthflt5w1f_mostthftw1i1[2] 2814
## prjthflt5w1f_mostthftw1i1[3] 2682
## prjthflt5w1f_mostthftw1i1[4] 3024
## prjthfgt5w1f_mostthftw1i1[1] 2341
## prjthfgt5w1f_mostthftw1i1[2] 2224
## prjthfgt5w1f_mostthftw1i1[3] 3005
## prjthfgt5w1f_mostthftw1i1[4] 2878
## prjthreatw1f_mostthftw1i1[1] 2947
## prjthreatw1f_mostthftw1i1[2] 2630
## prjthreatw1f_mostthftw1i1[3] 3175
## prjthreatw1f_mostthftw1i1[4] 2883
## prjharmw1f_mostthftw1i1[1] 1957
## prjharmw1f_mostthftw1i1[2] 2268
## prjharmw1f_mostthftw1i1[3] 2913
## prjharmw1f_mostthftw1i1[4] 3059
## prjusedrgw1f_mostthftw1i1[1] 2565
## prjusedrgw1f_mostthftw1i1[2] 2878
## prjusedrgw1f_mostthftw1i1[3] 3021
## prjusedrgw1f_mostthftw1i1[4] 2989
## prjhackw1f_mostthftw1i1[1] 2609
## prjhackw1f_mostthftw1i1[2] 2512
## prjhackw1f_mostthftw1i1[3] 3138
## prjhackw1f_mostthftw1i1[4] 3278
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.3.6.5 Prior summary
out.allprjcrime.stthft.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b prjhackw1f
## normal(0, 0.25) b mostthftw1i prjhackw1f
## (flat) b prjharmw1f
## normal(0, 0.25) b mostthftw1i prjharmw1f
## (flat) b prjthfgt5w1f
## normal(0, 0.25) b mostthftw1i prjthfgt5w1f
## (flat) b prjthflt5w1f
## normal(0, 0.25) b mostthftw1i prjthflt5w1f
## (flat) b prjthreatw1f
## normal(0, 0.25) b mostthftw1i prjthreatw1f
## (flat) b prjusedrgw1f
## normal(0, 0.25) b mostthftw1i prjusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept prjhackw1f
## normal(0, 2) Intercept prjharmw1f
## normal(0, 2) Intercept prjthfgt5w1f
## normal(0, 2) Intercept prjthflt5w1f
## normal(0, 2) Intercept prjthreatw1f
## normal(0, 2) Intercept prjusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 prjhackw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 prjharmw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 prjthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 prjthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 prjthreatw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 prjusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.3.7 Bivariate Corr: T1 stmugw1/crim intent
#Bivariate: criminal items ~ mo(stmugw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmugw1i',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmugw1i1',
resp = prjdv_names))
allprjcrime.stmug.fit <- brm(
mvbind(prjthflt5w1f, prjthfgt5w1f, prjthreatw1f, prjharmw1f, prjusedrgw1f, prjhackw1f) ~ 1 + mo(stmugw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/allprjcrime_stmug_fit",
file_refit = "on_change"
)
out.allprjcrime.stmug.fit <- ppchecks(allprjcrime.stmug.fit)4.3.3.7.1 Coefficient plot (intervals)
out.allprjcrime.stmug.fit[[10]]
4.3.3.7.2 Coefficient plot (distributions)
out.allprjcrime.stmug.fit[[9]]
4.3.3.7.3 PPcheck (density)
p1 <- out.allprjcrime.stmug.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.allprjcrime.stmug.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.allprjcrime.stmug.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.allprjcrime.stmug.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.allprjcrime.stmug.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.allprjcrime.stmug.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
4.3.3.7.4 Fit summary
out.allprjcrime.stmug.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5w1f ~ 1 + mo(stmugw1i)
## prjthfgt5w1f ~ 1 + mo(stmugw1i)
## prjthreatw1f ~ 1 + mo(stmugw1i)
## prjharmw1f ~ 1 + mo(stmugw1i)
## prjusedrgw1f ~ 1 + mo(stmugw1i)
## prjhackw1f ~ 1 + mo(stmugw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_Intercept -2.18 0.20 -2.60 -1.83 1.00 4278
## prjthfgt5w1f_Intercept -2.40 0.22 -2.85 -1.99 1.00 4421
## prjthreatw1f_Intercept -2.85 0.24 -3.36 -2.41 1.00 4807
## prjharmw1f_Intercept -3.05 0.27 -3.59 -2.55 1.00 4211
## prjusedrgw1f_Intercept -3.36 0.29 -3.95 -2.82 1.00 4246
## prjhackw1f_Intercept -3.46 0.31 -4.09 -2.87 1.00 4358
## prjthflt5w1f_mostmugw1i 0.03 0.13 -0.25 0.28 1.00 3797
## prjthfgt5w1f_mostmugw1i 0.09 0.13 -0.18 0.35 1.00 4262
## prjthreatw1f_mostmugw1i 0.05 0.15 -0.25 0.34 1.00 4473
## prjharmw1f_mostmugw1i -0.01 0.17 -0.35 0.30 1.00 4426
## prjusedrgw1f_mostmugw1i 0.15 0.16 -0.16 0.46 1.00 4302
## prjhackw1f_mostmugw1i 0.02 0.17 -0.33 0.37 1.00 4564
## Tail_ESS
## prjthflt5w1f_Intercept 3283
## prjthfgt5w1f_Intercept 2927
## prjthreatw1f_Intercept 3014
## prjharmw1f_Intercept 2498
## prjusedrgw1f_Intercept 2687
## prjhackw1f_Intercept 3083
## prjthflt5w1f_mostmugw1i 2968
## prjthfgt5w1f_mostmugw1i 2968
## prjthreatw1f_mostmugw1i 3120
## prjharmw1f_mostmugw1i 3238
## prjusedrgw1f_mostmugw1i 2835
## prjhackw1f_mostmugw1i 3035
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5w1f_mostmugw1i1[1] 0.26 0.15 0.03 0.60 1.00 4525
## prjthflt5w1f_mostmugw1i1[2] 0.24 0.14 0.03 0.56 1.00 5148
## prjthflt5w1f_mostmugw1i1[3] 0.24 0.14 0.04 0.57 1.00 5248
## prjthflt5w1f_mostmugw1i1[4] 0.26 0.15 0.04 0.60 1.00 4676
## prjthfgt5w1f_mostmugw1i1[1] 0.27 0.15 0.05 0.60 1.00 4248
## prjthfgt5w1f_mostmugw1i1[2] 0.24 0.14 0.04 0.54 1.00 5361
## prjthfgt5w1f_mostmugw1i1[3] 0.23 0.14 0.04 0.55 1.00 5031
## prjthfgt5w1f_mostmugw1i1[4] 0.26 0.14 0.04 0.57 1.00 5331
## prjthreatw1f_mostmugw1i1[1] 0.25 0.15 0.04 0.57 1.00 5447
## prjthreatw1f_mostmugw1i1[2] 0.24 0.14 0.04 0.55 1.00 5705
## prjthreatw1f_mostmugw1i1[3] 0.25 0.14 0.04 0.57 1.00 5838
## prjthreatw1f_mostmugw1i1[4] 0.26 0.15 0.04 0.59 1.00 4895
## prjharmw1f_mostmugw1i1[1] 0.24 0.14 0.03 0.57 1.00 4802
## prjharmw1f_mostmugw1i1[2] 0.24 0.14 0.03 0.56 1.00 5548
## prjharmw1f_mostmugw1i1[3] 0.25 0.14 0.04 0.58 1.00 5357
## prjharmw1f_mostmugw1i1[4] 0.26 0.15 0.04 0.61 1.00 5081
## prjusedrgw1f_mostmugw1i1[1] 0.26 0.15 0.04 0.60 1.00 4708
## prjusedrgw1f_mostmugw1i1[2] 0.24 0.14 0.03 0.56 1.00 6133
## prjusedrgw1f_mostmugw1i1[3] 0.23 0.14 0.03 0.55 1.00 4968
## prjusedrgw1f_mostmugw1i1[4] 0.28 0.15 0.04 0.62 1.00 5544
## prjhackw1f_mostmugw1i1[1] 0.24 0.14 0.04 0.57 1.00 4814
## prjhackw1f_mostmugw1i1[2] 0.24 0.15 0.03 0.58 1.00 4907
## prjhackw1f_mostmugw1i1[3] 0.25 0.15 0.04 0.58 1.00 4976
## prjhackw1f_mostmugw1i1[4] 0.26 0.15 0.04 0.59 1.00 5089
## Tail_ESS
## prjthflt5w1f_mostmugw1i1[1] 2790
## prjthflt5w1f_mostmugw1i1[2] 2357
## prjthflt5w1f_mostmugw1i1[3] 3349
## prjthflt5w1f_mostmugw1i1[4] 3201
## prjthfgt5w1f_mostmugw1i1[1] 2655
## prjthfgt5w1f_mostmugw1i1[2] 2685
## prjthfgt5w1f_mostmugw1i1[3] 3287
## prjthfgt5w1f_mostmugw1i1[4] 3000
## prjthreatw1f_mostmugw1i1[1] 3073
## prjthreatw1f_mostmugw1i1[2] 2779
## prjthreatw1f_mostmugw1i1[3] 2916
## prjthreatw1f_mostmugw1i1[4] 2657
## prjharmw1f_mostmugw1i1[1] 2545
## prjharmw1f_mostmugw1i1[2] 2750
## prjharmw1f_mostmugw1i1[3] 2788
## prjharmw1f_mostmugw1i1[4] 3077
## prjusedrgw1f_mostmugw1i1[1] 2300
## prjusedrgw1f_mostmugw1i1[2] 2806
## prjusedrgw1f_mostmugw1i1[3] 2930
## prjusedrgw1f_mostmugw1i1[4] 2785
## prjhackw1f_mostmugw1i1[1] 2840
## prjhackw1f_mostmugw1i1[2] 2567
## prjhackw1f_mostmugw1i1[3] 2920
## prjhackw1f_mostmugw1i1[4] 3135
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.3.7.5 Prior summary
out.allprjcrime.stmug.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b prjhackw1f
## normal(0, 0.25) b mostmugw1i prjhackw1f
## (flat) b prjharmw1f
## normal(0, 0.25) b mostmugw1i prjharmw1f
## (flat) b prjthfgt5w1f
## normal(0, 0.25) b mostmugw1i prjthfgt5w1f
## (flat) b prjthflt5w1f
## normal(0, 0.25) b mostmugw1i prjthflt5w1f
## (flat) b prjthreatw1f
## normal(0, 0.25) b mostmugw1i prjthreatw1f
## (flat) b prjusedrgw1f
## normal(0, 0.25) b mostmugw1i prjusedrgw1f
## (flat) Intercept
## normal(0, 2) Intercept prjhackw1f
## normal(0, 2) Intercept prjharmw1f
## normal(0, 2) Intercept prjthfgt5w1f
## normal(0, 2) Intercept prjthflt5w1f
## normal(0, 2) Intercept prjthreatw1f
## normal(0, 2) Intercept prjusedrgw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 prjhackw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 prjharmw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 prjthfgt5w1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 prjthflt5w1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 prjthreatw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 prjusedrgw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.4 “Any” Crime items: Bivariate models
Now we will repeat the above process for “any” past and projected
crime indicators. We did not want to include these items in the
multivariate mvbind() models above for simultaneous
modeling as the items are structurally correlated with the other outcome
variables - i.e., they are composites and hence mathematical functions
of the other binary items. Later, we will need to merge estimates from
the “any” criminal intent models with those from the multivariate
criminal intent item models to display them in Figure 3.
4.3.4.1 Bivariate Corr: T1 ANY past crime/all stress
#Bivariate: past crime items ~ mo(stressvars)
#Create function for repetitive prior settings
setmyprior <- function(mocoefname, simocoefname) {
myprior <- c(
set_prior('normal(0, 2)', class = 'Intercept'),
set_prior('normal(0, 0.25)', class = 'b', coef = mocoefname),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = simocoefname))
}
#Update function to call all ppchecks for bivar any crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit)
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
coord_cartesian(xlim = c(-0.5, 0.5)) + vline_at(0, linetype = 3)
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
coord_cartesian(xlim = c(-0.5, 0.5)) + vline_at(0, linetype = 3)
allchecks <- list(fitsummary, priorsummary, ppcheckdv1,
plotcoefs, plotcoefs2)
return(allchecks)
}
myprior <- setmyprior('mostmonyw1i', 'mostmonyw1i1')
anypstcrime.stmony.fit <- brm(pstanyw1f ~ 1 + mo(stmonyw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anypstcrime_stmony_fit", file_refit = "on_change"
)
out.anypstcrime.stmony.fit <- ppchecks(anypstcrime.stmony.fit)
myprior <- setmyprior('mosttranw1i', 'mosttranw1i1')
anypstcrime.sttran.fit <- brm(pstanyw1f ~ 1 + mo(sttranw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anypstcrime_sttran_fit", file_refit = "on_change"
)
out.anypstcrime.sttran.fit <- ppchecks(anypstcrime.sttran.fit)
myprior <- setmyprior('mostrespw1i', 'mostrespw1i1')
anypstcrime.stresp.fit <- brm(pstanyw1f ~ 1 + mo(strespw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anypstcrime_stresp_fit", file_refit = "on_change"
)
out.anypstcrime.stresp.fit <- ppchecks(anypstcrime.stresp.fit)
myprior <- setmyprior('mostfairw1i', 'mostfairw1i1')
anypstcrime.stfair.fit <- brm(pstanyw1f ~ 1 + mo(stfairw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anypstcrime_stfair_fit", file_refit = "on_change"
)
out.anypstcrime.stfair.fit <- ppchecks(anypstcrime.stfair.fit)
myprior <- setmyprior('mostjobw1i', 'mostjobw1i1')
anypstcrime.stjob.fit <- brm(pstanyw1f ~ 1 + mo(stjobw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anypstcrime_stjob_fit", file_refit = "on_change"
)
out.anypstcrime.stjob.fit <- ppchecks(anypstcrime.stjob.fit)
myprior <- setmyprior('mostthftw1i', 'mostthftw1i1')
anypstcrime.stthft.fit <- brm(pstanyw1f ~ 1 + mo(stthftw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anypstcrime_stthft_fit", file_refit = "on_change"
)
out.anypstcrime.stthft.fit <- ppchecks(anypstcrime.stthft.fit)
myprior <- setmyprior('mostmugw1i', 'mostmugw1i1')
anypstcrime.stmug.fit <- brm(pstanyw1f ~ 1 + mo(stmugw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anypstcrime_stmug_fit", file_refit = "on_change"
)
out.anypstcrime.stmug.fit <- ppchecks(anypstcrime.stmug.fit)4.3.4.1.1 Coefficient plot (intervals)
p1 <- out.anypstcrime.stmony.fit[[5]]
p2 <- out.anypstcrime.sttran.fit[[5]]
p3 <- out.anypstcrime.stresp.fit[[5]]
p4 <- out.anypstcrime.stfair.fit[[5]]
p5 <- out.anypstcrime.stjob.fit[[5]]
p6 <- out.anypstcrime.stthft.fit[[5]]
p7 <- out.anypstcrime.stmug.fit[[5]]
playout <- '
AB
CD
E#
FG
'
p1 + p2 + p3 + p4 + p5 + p6 + p7 +
plot_layout(design = playout) +
plot_annotation(
title = 'Coefficient plot',
subtitle = 'Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80% (thick line), and 95% (thin line) intervals')
4.3.4.1.2 Coefficient plot (distributions)
p1 <- out.anypstcrime.stmony.fit[[4]]
p2 <- out.anypstcrime.sttran.fit[[4]]
p3 <- out.anypstcrime.stresp.fit[[4]]
p4 <- out.anypstcrime.stfair.fit[[4]]
p5 <- out.anypstcrime.stjob.fit[[4]]
p6 <- out.anypstcrime.stthft.fit[[4]]
p7 <- out.anypstcrime.stmug.fit[[4]]
playout <- '
AB
CD
E#
FG
'
p1 + p2 + p3 + p4 + p5 + p6 + p7 +
plot_layout(design = playout) +
plot_annotation(
title = 'Coefficient plot',
subtitle = 'Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% shaded intervals')
4.3.4.1.3 PPcheck (density)
p1 <- out.anypstcrime.stmony.fit[[3]] + labs(title = "Any past crime/stmony (T1)")
p2 <- out.anypstcrime.sttran.fit[[3]] + labs(title = "Any pastcrime/sttran (T1)")
p3 <- out.anypstcrime.stresp.fit[[3]] + labs(title = "Any pastcrime/stresp (T1)")
p4 <- out.anypstcrime.stfair.fit[[3]] + labs(title = "Any pastcrime/stfair (T1)")
p5 <- out.anypstcrime.stjob.fit[[3]] + labs(title = "Any pastcrime/stjob (T1)")
p6 <- out.anypstcrime.stthft.fit[[3]] + labs(title = "Any pastcrime/stthft (T1)")
p7 <- out.anypstcrime.stmug.fit[[3]] + labs(title = "Any pastcrime/stmug (T1)")
(p1 + p2) / (p3 + p4) / (p5 + plot_spacer()) / (p6 + p7)
4.3.4.2 Fit + prior summaries (T1 Any past crime/all stress)
4.3.4.2.1 stmony
out.anypstcrime.stmony.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: pstanyw1f ~ 1 + mo(stmonyw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.34 0.27 -1.85 -0.78 1.00 1991 2113
## mostmonyw1i 0.02 0.11 -0.20 0.24 1.00 1959 2171
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostmonyw1i1[1] 0.25 0.14 0.04 0.58 1.00 3185 2212
## mostmonyw1i1[2] 0.25 0.17 0.03 0.64 1.00 2302 2532
## mostmonyw1i1[3] 0.23 0.14 0.03 0.56 1.00 3225 2870
## mostmonyw1i1[4] 0.27 0.16 0.03 0.62 1.00 2660 2407
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anypstcrime.stmony.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostmonyw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1
## source
## default
## user
## user
## user4.3.4.2.2 sttran
out.anypstcrime.sttran.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: pstanyw1f ~ 1 + mo(sttranw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.37 0.27 -1.87 -0.82 1.00 1623 1975
## mosttranw1i 0.04 0.12 -0.19 0.27 1.00 1606 2168
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mosttranw1i1[1] 0.25 0.14 0.04 0.56 1.00 2905 2170
## mosttranw1i1[2] 0.22 0.15 0.03 0.57 1.00 2130 2380
## mosttranw1i1[3] 0.22 0.13 0.03 0.54 1.00 3029 2045
## mosttranw1i1[4] 0.31 0.18 0.04 0.69 1.00 1874 2235
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anypstcrime.sttran.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mosttranw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mosttranw1i1
## source
## default
## user
## user
## user4.3.4.2.3 stresp
out.anypstcrime.stresp.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: pstanyw1f ~ 1 + mo(strespw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.73 0.24 -2.24 -1.29 1.00 2140 1915
## mostrespw1i 0.17 0.08 0.02 0.33 1.00 2289 1979
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostrespw1i1[1] 0.27 0.14 0.04 0.58 1.00 3046 1975
## mostrespw1i1[2] 0.25 0.14 0.04 0.56 1.00 3244 2452
## mostrespw1i1[3] 0.23 0.14 0.03 0.55 1.00 3610 2721
## mostrespw1i1[4] 0.26 0.14 0.04 0.56 1.00 3189 2247
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anypstcrime.stresp.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostrespw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostrespw1i1
## source
## default
## user
## user
## user4.3.4.2.4 stfair
out.anypstcrime.stfair.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: pstanyw1f ~ 1 + mo(stfairw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.71 0.22 -2.18 -1.29 1.00 2079 1762
## mostfairw1i 0.16 0.07 0.02 0.31 1.00 2128 1941
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostfairw1i1[1] 0.22 0.13 0.03 0.53 1.00 3546 2015
## mostfairw1i1[2] 0.25 0.14 0.04 0.57 1.00 3239 2018
## mostfairw1i1[3] 0.28 0.15 0.04 0.61 1.00 3576 2464
## mostfairw1i1[4] 0.25 0.14 0.04 0.57 1.00 3663 2540
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anypstcrime.stfair.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostfairw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostfairw1i1
## source
## default
## user
## user
## user4.3.4.2.5 stjob
out.anypstcrime.stjob.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: pstanyw1f ~ 1 + mo(stjobw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.69 0.21 -2.14 -1.29 1.00 2281 1968
## mostjobw1i 0.17 0.08 0.01 0.32 1.00 2437 2416
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostjobw1i1[1] 0.20 0.12 0.03 0.49 1.00 3213 2298
## mostjobw1i1[2] 0.19 0.12 0.03 0.47 1.00 2710 1834
## mostjobw1i1[3] 0.32 0.16 0.06 0.65 1.00 3395 2757
## mostjobw1i1[4] 0.29 0.15 0.05 0.61 1.00 3074 2702
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anypstcrime.stjob.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostjobw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostjobw1i1
## source
## default
## user
## user
## user4.3.4.2.6 stthft
out.anypstcrime.stthft.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: pstanyw1f ~ 1 + mo(stthftw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.41 0.15 -1.72 -1.11 1.00 2815 2469
## mostthftw1i 0.07 0.09 -0.10 0.24 1.00 2884 2591
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostthftw1i1[1] 0.24 0.14 0.04 0.56 1.00 4045 2294
## mostthftw1i1[2] 0.24 0.14 0.04 0.55 1.00 4621 2665
## mostthftw1i1[3] 0.26 0.15 0.04 0.60 1.00 4534 2246
## mostthftw1i1[4] 0.26 0.15 0.03 0.59 1.00 4529 2908
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anypstcrime.stthft.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostthftw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostthftw1i1
## source
## default
## user
## user
## user4.3.4.2.7 stmug
out.anypstcrime.stmug.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: pstanyw1f ~ 1 + mo(stmugw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.42 0.15 -1.72 -1.14 1.00 2307 2594
## mostmugw1i 0.10 0.10 -0.10 0.30 1.00 2389 2734
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostmugw1i1[1] 0.26 0.14 0.04 0.57 1.00 3753 2317
## mostmugw1i1[2] 0.22 0.13 0.03 0.53 1.00 3892 2926
## mostmugw1i1[3] 0.24 0.14 0.04 0.56 1.00 4524 2855
## mostmugw1i1[4] 0.29 0.16 0.05 0.63 1.00 3842 2992
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anypstcrime.stmug.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostmugw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostmugw1i1
## source
## default
## user
## user
## user4.3.4.3 Bivariate Corr: T1 ANY crim intent/all stress
#Bivariate: past crime items ~ mo(stressvars)
myprior <- setmyprior('mostmonyw1i', 'mostmonyw1i1')
anyprjcrime.stmony.fit <- brm(prjanyw1f ~ 1 + mo(stmonyw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anyprjcrime_stmony_fit", file_refit = "on_change"
)
out.anyprjcrime.stmony.fit <- ppchecks(anyprjcrime.stmony.fit)
myprior <- setmyprior('mosttranw1i', 'mosttranw1i1')
anyprjcrime.sttran.fit <- brm(prjanyw1f ~ 1 + mo(sttranw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anyprjcrime_sttran_fit", file_refit = "on_change"
)
out.anyprjcrime.sttran.fit <- ppchecks(anyprjcrime.sttran.fit)
myprior <- setmyprior('mostrespw1i', 'mostrespw1i1')
anyprjcrime.stresp.fit <- brm(prjanyw1f ~ 1 + mo(strespw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anyprjcrime_stresp_fit", file_refit = "on_change"
)
out.anyprjcrime.stresp.fit <- ppchecks(anyprjcrime.stresp.fit)
myprior <- setmyprior('mostfairw1i', 'mostfairw1i1')
anyprjcrime.stfair.fit <- brm(prjanyw1f ~ 1 + mo(stfairw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anyprjcrime_stfair_fit", file_refit = "on_change"
)
out.anyprjcrime.stfair.fit <- ppchecks(anyprjcrime.stfair.fit)
myprior <- setmyprior('mostjobw1i', 'mostjobw1i1')
anyprjcrime.stjob.fit <- brm(prjanyw1f ~ 1 + mo(stjobw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anyprjcrime_stjob_fit", file_refit = "on_change"
)
out.anyprjcrime.stjob.fit <- ppchecks(anyprjcrime.stjob.fit)
myprior <- setmyprior('mostthftw1i', 'mostthftw1i1')
anyprjcrime.stthft.fit <- brm(prjanyw1f ~ 1 + mo(stthftw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anyprjcrime_stthft_fit", file_refit = "on_change"
)
out.anyprjcrime.stthft.fit <- ppchecks(anyprjcrime.stthft.fit)
myprior <- setmyprior('mostmugw1i', 'mostmugw1i1')
anyprjcrime.stmug.fit <- brm(prjanyw1f ~ 1 + mo(stmugw1i),
data = stress.wide3, family = "bernoulli", prior = myprior,
cores = nCoresphys,
chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/anyprjcrime_stmug_fit", file_refit = "on_change"
)
out.anyprjcrime.stmug.fit <- ppchecks(anyprjcrime.stmug.fit)4.3.4.3.1 Coefficient plot (intervals)
p1 <- out.anyprjcrime.stmony.fit[[5]]
p2 <- out.anyprjcrime.sttran.fit[[5]]
p3 <- out.anyprjcrime.stresp.fit[[5]]
p4 <- out.anyprjcrime.stfair.fit[[5]]
p5 <- out.anyprjcrime.stjob.fit[[5]]
p6 <- out.anyprjcrime.stthft.fit[[5]]
p7 <- out.anyprjcrime.stmug.fit[[5]]
playout <- '
AB
CD
E#
FG
'
p1 + p2 + p3 + p4 + p5 + p6 + p7 +
plot_layout(design = playout) +
plot_annotation(
title = 'Coefficient plot',
subtitle = 'Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80% (thick line), and 95% (thin line) intervals')
4.3.4.3.2 Coefficient plot (distributions)
p1 <- out.anyprjcrime.stmony.fit[[4]]
p2 <- out.anyprjcrime.sttran.fit[[4]]
p3 <- out.anyprjcrime.stresp.fit[[4]]
p4 <- out.anyprjcrime.stfair.fit[[4]]
p5 <- out.anyprjcrime.stjob.fit[[4]]
p6 <- out.anyprjcrime.stthft.fit[[4]]
p7 <- out.anyprjcrime.stmug.fit[[4]]
playout <- '
AB
CD
E#
FG
'
p1 + p2 + p3 + p4 + p5 + p6 + p7 +
plot_layout(design = playout) +
plot_annotation(
title = 'Coefficient plot',
subtitle = 'Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% shaded intervals')
4.3.4.3.3 PPcheck (density)
p1 <- out.anyprjcrime.stmony.fit[[3]] + labs(title = "Any crime intent/stmony (T1)")
p2 <- out.anyprjcrime.sttran.fit[[3]] + labs(title = "Any crime intent/sttran (T1)")
p3 <- out.anyprjcrime.stresp.fit[[3]] + labs(title = "Any crime intent/stresp (T1)")
p4 <- out.anyprjcrime.stfair.fit[[3]] + labs(title = "Any crime intent/stfair (T1)")
p5 <- out.anyprjcrime.stjob.fit[[3]] + labs(title = "Any crime intent/stjob (T1)")
p6 <- out.anyprjcrime.stthft.fit[[3]] + labs(title = "Any crime intent/stthft (T1)")
p7 <- out.anyprjcrime.stmug.fit[[3]] + labs(title = "Any crime intent/stmug (T1)")
(p1 + p2) / (p3 + p4) / (p5 + plot_spacer()) / (p6 + p7)
4.3.4.4 Fit + prior summaries (T1 Any crime intent/all stress)
4.3.4.4.1 stmony
out.anyprjcrime.stmony.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjanyw1f ~ 1 + mo(stmonyw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.35 0.27 -1.87 -0.83 1.00 2316 2517
## mostmonyw1i -0.12 0.10 -0.31 0.09 1.00 2266 2371
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostmonyw1i1[1] 0.25 0.14 0.04 0.56 1.00 3143 2231
## mostmonyw1i1[2] 0.33 0.17 0.05 0.68 1.00 2627 2236
## mostmonyw1i1[3] 0.20 0.13 0.03 0.51 1.00 3350 2453
## mostmonyw1i1[4] 0.21 0.13 0.03 0.54 1.00 3324 2251
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anyprjcrime.stmony.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostmonyw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1
## source
## default
## user
## user
## user4.3.4.4.2 sttran
out.anyprjcrime.sttran.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjanyw1f ~ 1 + mo(sttranw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.47 0.31 -2.09 -0.87 1.00 1793 2002
## mosttranw1i -0.05 0.13 -0.28 0.22 1.00 1794 1666
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mosttranw1i1[1] 0.26 0.14 0.04 0.57 1.00 2631 2333
## mosttranw1i1[2] 0.28 0.16 0.04 0.62 1.00 2440 2557
## mosttranw1i1[3] 0.23 0.13 0.03 0.54 1.00 2887 2437
## mosttranw1i1[4] 0.23 0.16 0.02 0.61 1.00 2250 2361
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anyprjcrime.sttran.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mosttranw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mosttranw1i1
## source
## default
## user
## user
## user4.3.4.4.3 stresp
out.anyprjcrime.stresp.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjanyw1f ~ 1 + mo(strespw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -2.11 0.25 -2.65 -1.64 1.00 1599 2090
## mostrespw1i 0.20 0.08 0.04 0.37 1.00 1879 2273
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostrespw1i1[1] 0.26 0.14 0.04 0.56 1.00 3002 1762
## mostrespw1i1[2] 0.23 0.13 0.04 0.54 1.00 3591 2295
## mostrespw1i1[3] 0.23 0.13 0.03 0.53 1.00 3640 2597
## mostrespw1i1[4] 0.28 0.14 0.05 0.58 1.00 3615 2708
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anyprjcrime.stresp.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostrespw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostrespw1i1
## source
## default
## user
## user
## user4.3.4.4.4 stfair
out.anyprjcrime.stfair.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjanyw1f ~ 1 + mo(stfairw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -2.18 0.25 -2.69 -1.71 1.00 2012 1918
## mostfairw1i 0.23 0.08 0.07 0.39 1.00 2025 2008
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostfairw1i1[1] 0.21 0.13 0.03 0.51 1.00 3051 1957
## mostfairw1i1[2] 0.23 0.13 0.03 0.52 1.00 3345 2395
## mostfairw1i1[3] 0.23 0.14 0.03 0.55 1.00 3217 2244
## mostfairw1i1[4] 0.32 0.15 0.06 0.63 1.00 3478 2584
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anyprjcrime.stfair.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostfairw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostfairw1i1
## source
## default
## user
## user
## user4.3.4.4.5 stjob
out.anyprjcrime.stjob.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjanyw1f ~ 1 + mo(stjobw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -2.35 0.25 -2.86 -1.87 1.00 2195 1990
## mostjobw1i 0.31 0.08 0.15 0.47 1.00 2364 2031
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostjobw1i1[1] 0.15 0.09 0.02 0.37 1.00 2718 1951
## mostjobw1i1[2] 0.19 0.11 0.03 0.46 1.00 3421 2546
## mostjobw1i1[3] 0.42 0.16 0.11 0.71 1.00 3048 2695
## mostjobw1i1[4] 0.23 0.12 0.04 0.53 1.00 3119 2479
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anyprjcrime.stjob.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostjobw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostjobw1i1
## source
## default
## user
## user
## user4.3.4.4.6 stthft
out.anyprjcrime.stthft.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjanyw1f ~ 1 + mo(stthftw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.75 0.17 -2.09 -1.41 1.00 3022 2660
## mostthftw1i 0.11 0.10 -0.08 0.30 1.00 2888 2558
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostthftw1i1[1] 0.22 0.14 0.03 0.54 1.00 3919 2616
## mostthftw1i1[2] 0.23 0.14 0.03 0.55 1.00 4351 2943
## mostthftw1i1[3] 0.29 0.16 0.04 0.64 1.00 3678 2421
## mostthftw1i1[4] 0.26 0.15 0.04 0.60 1.00 4018 2432
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anyprjcrime.stthft.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostthftw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostthftw1i1
## source
## default
## user
## user
## user4.3.4.4.7 stmug
out.anyprjcrime.stmug.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjanyw1f ~ 1 + mo(stmugw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.67 0.16 -1.99 -1.37 1.00 2591 2553
## mostmugw1i 0.06 0.11 -0.15 0.27 1.00 2764 2817
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostmugw1i1[1] 0.24 0.14 0.03 0.56 1.00 2880 1715
## mostmugw1i1[2] 0.23 0.14 0.03 0.54 1.00 3995 2593
## mostmugw1i1[3] 0.25 0.15 0.04 0.58 1.00 4760 2899
## mostmugw1i1[4] 0.28 0.15 0.05 0.62 1.00 3559 2848
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.anyprjcrime.stmug.fit[[2]]## prior class coef group resp dpar nlpar lb ub
## (flat) b
## normal(0, 0.25) b mostmugw1i
## normal(0, 2) Intercept
## dirichlet(2, 2, 2, 2) simo mostmugw1i1
## source
## default
## user
## user
## user4.3.5 Depressive Symptom Items: Bivariate models
Once again, we will repeat the process of building simple bivariate models for negative emotions by regressing each negative emotion item on each stress indicator. As with past crime, we will specify a Bernoulli distribution with a logit link and use monotonic cumulative ordinal probit thresholds for the stress predictors.
4.3.5.1 Bivariate Corr: T1 stmonyw1/negative emotions
#list of colnames for depressive symptom DVs
depdv_names <- noquote(c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f",
"depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
#Bivariate: depressive symptom items ~ mo(stmonyw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmonyw1i',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmonyw1i1',
resp = depdv_names))
alldepress.stmony.fit <- brm(
mvbind(depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f, depunfairw1f, depmistrtw1f,
depbetrayw1f) ~ 1 + mo(stmonyw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/alldepress_stmony_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar depressive symptom models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="depcantgow1f")
ppcheckdv2 <-pp_check(modelfit, resp="depeffortw1f")
ppcheckdv3 <-pp_check(modelfit, resp="deplonelyw1f")
ppcheckdv4 <-pp_check(modelfit, resp="depbluesw1f")
ppcheckdv5 <-pp_check(modelfit, resp="depunfairw1f")
ppcheckdv6 <-pp_check(modelfit, resp="depmistrtw1f")
ppcheckdv7 <-pp_check(modelfit, resp="depbetrayw1f")
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6, ppcheckdv7,
plotcoefs, plotcoefs2)
return(allchecks)
}
out.alldepress.stmony.fit <- ppchecks(alldepress.stmony.fit)4.3.5.1.1 Coefficient plot (intervals)
out.alldepress.stmony.fit[[11]]
4.3.5.1.2 Coefficient plot (distributions)
out.alldepress.stmony.fit[[10]]
4.3.5.1.3 PPcheck (density)
p1 <- out.alldepress.stmony.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.alldepress.stmony.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.alldepress.stmony.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.alldepress.stmony.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.alldepress.stmony.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.alldepress.stmony.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.alldepress.stmony.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
4.3.5.1.4 Fit summary
out.alldepress.stmony.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgow1f ~ 1 + mo(stmonyw1i)
## depeffortw1f ~ 1 + mo(stmonyw1i)
## deplonelyw1f ~ 1 + mo(stmonyw1i)
## depbluesw1f ~ 1 + mo(stmonyw1i)
## depunfairw1f ~ 1 + mo(stmonyw1i)
## depmistrtw1f ~ 1 + mo(stmonyw1i)
## depbetrayw1f ~ 1 + mo(stmonyw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_Intercept -0.83 0.21 -1.29 -0.46 1.00 3239
## depeffortw1f_Intercept -2.06 0.29 -2.68 -1.52 1.00 3951
## deplonelyw1f_Intercept -1.41 0.23 -1.90 -0.97 1.00 3336
## depbluesw1f_Intercept -1.95 0.31 -2.63 -1.42 1.00 3458
## depunfairw1f_Intercept -2.28 0.31 -2.94 -1.76 1.00 3288
## depmistrtw1f_Intercept -2.22 0.38 -3.10 -1.59 1.00 3229
## depbetrayw1f_Intercept -2.32 0.36 -3.12 -1.73 1.00 2874
## depcantgow1f_mostmonyw1i 0.22 0.08 0.06 0.39 1.00 3339
## depeffortw1f_mostmonyw1i 0.09 0.11 -0.13 0.31 1.00 4078
## deplonelyw1f_mostmonyw1i 0.07 0.10 -0.11 0.26 1.00 3653
## depbluesw1f_mostmonyw1i 0.20 0.11 -0.01 0.41 1.00 3502
## depunfairw1f_mostmonyw1i 0.36 0.10 0.18 0.57 1.00 3250
## depmistrtw1f_mostmonyw1i 0.23 0.12 0.01 0.49 1.00 3340
## depbetrayw1f_mostmonyw1i 0.25 0.12 0.05 0.49 1.00 2967
## Tail_ESS
## depcantgow1f_Intercept 2523
## depeffortw1f_Intercept 2583
## deplonelyw1f_Intercept 1995
## depbluesw1f_Intercept 2372
## depunfairw1f_Intercept 2646
## depmistrtw1f_Intercept 2229
## depbetrayw1f_Intercept 2528
## depcantgow1f_mostmonyw1i 2800
## depeffortw1f_mostmonyw1i 3014
## deplonelyw1f_mostmonyw1i 2840
## depbluesw1f_mostmonyw1i 2899
## depunfairw1f_mostmonyw1i 2597
## depmistrtw1f_mostmonyw1i 2140
## depbetrayw1f_mostmonyw1i 2712
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_mostmonyw1i1[1] 0.23 0.13 0.03 0.51 1.00 4017
## depcantgow1f_mostmonyw1i1[2] 0.16 0.11 0.02 0.42 1.00 5624
## depcantgow1f_mostmonyw1i1[3] 0.32 0.15 0.07 0.64 1.00 5034
## depcantgow1f_mostmonyw1i1[4] 0.28 0.14 0.05 0.59 1.00 4912
## depeffortw1f_mostmonyw1i1[1] 0.27 0.15 0.04 0.59 1.00 4737
## depeffortw1f_mostmonyw1i1[2] 0.22 0.13 0.03 0.53 1.00 4835
## depeffortw1f_mostmonyw1i1[3] 0.27 0.15 0.04 0.59 1.00 4903
## depeffortw1f_mostmonyw1i1[4] 0.25 0.14 0.04 0.57 1.00 5130
## deplonelyw1f_mostmonyw1i1[1] 0.27 0.15 0.04 0.59 1.00 4271
## deplonelyw1f_mostmonyw1i1[2] 0.22 0.13 0.03 0.54 1.00 4444
## deplonelyw1f_mostmonyw1i1[3] 0.24 0.14 0.04 0.55 1.00 5089
## deplonelyw1f_mostmonyw1i1[4] 0.28 0.15 0.04 0.62 1.00 5074
## depbluesw1f_mostmonyw1i1[1] 0.29 0.15 0.04 0.62 1.00 3982
## depbluesw1f_mostmonyw1i1[2] 0.27 0.14 0.05 0.58 1.00 4633
## depbluesw1f_mostmonyw1i1[3] 0.17 0.11 0.02 0.46 1.00 4464
## depbluesw1f_mostmonyw1i1[4] 0.27 0.14 0.05 0.58 1.00 5252
## depunfairw1f_mostmonyw1i1[1] 0.23 0.12 0.04 0.50 1.00 4121
## depunfairw1f_mostmonyw1i1[2] 0.15 0.09 0.02 0.36 1.00 5084
## depunfairw1f_mostmonyw1i1[3] 0.39 0.14 0.13 0.67 1.00 5356
## depunfairw1f_mostmonyw1i1[4] 0.23 0.12 0.04 0.50 1.00 5533
## depmistrtw1f_mostmonyw1i1[1] 0.34 0.16 0.07 0.67 1.00 3544
## depmistrtw1f_mostmonyw1i1[2] 0.22 0.13 0.03 0.52 1.00 4907
## depmistrtw1f_mostmonyw1i1[3] 0.21 0.13 0.03 0.51 1.00 4370
## depmistrtw1f_mostmonyw1i1[4] 0.23 0.13 0.03 0.52 1.00 5010
## depbetrayw1f_mostmonyw1i1[1] 0.33 0.16 0.06 0.65 1.00 3322
## depbetrayw1f_mostmonyw1i1[2] 0.18 0.11 0.03 0.44 1.00 4542
## depbetrayw1f_mostmonyw1i1[3] 0.27 0.14 0.05 0.58 1.00 4524
## depbetrayw1f_mostmonyw1i1[4] 0.22 0.13 0.04 0.50 1.00 4864
## Tail_ESS
## depcantgow1f_mostmonyw1i1[1] 2058
## depcantgow1f_mostmonyw1i1[2] 2601
## depcantgow1f_mostmonyw1i1[3] 2781
## depcantgow1f_mostmonyw1i1[4] 2580
## depeffortw1f_mostmonyw1i1[1] 2674
## depeffortw1f_mostmonyw1i1[2] 2691
## depeffortw1f_mostmonyw1i1[3] 2754
## depeffortw1f_mostmonyw1i1[4] 3335
## deplonelyw1f_mostmonyw1i1[1] 2308
## deplonelyw1f_mostmonyw1i1[2] 2820
## deplonelyw1f_mostmonyw1i1[3] 2935
## deplonelyw1f_mostmonyw1i1[4] 3145
## depbluesw1f_mostmonyw1i1[1] 2596
## depbluesw1f_mostmonyw1i1[2] 3041
## depbluesw1f_mostmonyw1i1[3] 3328
## depbluesw1f_mostmonyw1i1[4] 3393
## depunfairw1f_mostmonyw1i1[1] 2648
## depunfairw1f_mostmonyw1i1[2] 2508
## depunfairw1f_mostmonyw1i1[3] 3110
## depunfairw1f_mostmonyw1i1[4] 2666
## depmistrtw1f_mostmonyw1i1[1] 2793
## depmistrtw1f_mostmonyw1i1[2] 2014
## depmistrtw1f_mostmonyw1i1[3] 3064
## depmistrtw1f_mostmonyw1i1[4] 3106
## depbetrayw1f_mostmonyw1i1[1] 2685
## depbetrayw1f_mostmonyw1i1[2] 2845
## depbetrayw1f_mostmonyw1i1[3] 2678
## depbetrayw1f_mostmonyw1i1[4] 2904
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.5.1.5 Prior summary
out.alldepress.stmony.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b depbetrayw1f
## normal(0, 0.25) b mostmonyw1i depbetrayw1f
## (flat) b depbluesw1f
## normal(0, 0.25) b mostmonyw1i depbluesw1f
## (flat) b depcantgow1f
## normal(0, 0.25) b mostmonyw1i depcantgow1f
## (flat) b depeffortw1f
## normal(0, 0.25) b mostmonyw1i depeffortw1f
## (flat) b deplonelyw1f
## normal(0, 0.25) b mostmonyw1i deplonelyw1f
## (flat) b depmistrtw1f
## normal(0, 0.25) b mostmonyw1i depmistrtw1f
## (flat) b depunfairw1f
## normal(0, 0.25) b mostmonyw1i depunfairw1f
## (flat) Intercept
## normal(0, 2) Intercept depbetrayw1f
## normal(0, 2) Intercept depbluesw1f
## normal(0, 2) Intercept depcantgow1f
## normal(0, 2) Intercept depeffortw1f
## normal(0, 2) Intercept deplonelyw1f
## normal(0, 2) Intercept depmistrtw1f
## normal(0, 2) Intercept depunfairw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 depbetrayw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 depbluesw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 depcantgow1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 depeffortw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 deplonelyw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 depmistrtw1f
## dirichlet(2, 2, 2, 2) simo mostmonyw1i1 depunfairw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.5.2 Bivariate Corr: T1 sttranw1/negative emotions
#Bivariate: depressive symptom items ~ mo(sttranw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mosttranw1i',
resp =depdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mosttranw1i1',
resp =depdv_names))
alldepress.sttran.fit <- brm(
mvbind(depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f, depunfairw1f, depmistrtw1f,
depbetrayw1f) ~ 1 + mo(sttranw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/alldepress_sttran_fit",
file_refit = "on_change"
)
out.alldepress.sttran.fit <- ppchecks(alldepress.sttran.fit)4.3.5.2.1 Coefficient plot (intervals)
out.alldepress.sttran.fit[[11]]
4.3.5.2.2 Coefficient plot (distributions)
out.alldepress.sttran.fit[[10]]
4.3.5.2.3 PPcheck (density)
p1 <- out.alldepress.sttran.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.alldepress.sttran.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.alldepress.sttran.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.alldepress.sttran.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.alldepress.sttran.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.alldepress.sttran.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.alldepress.sttran.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
4.3.5.2.4 Fit summary
out.alldepress.sttran.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgow1f ~ 1 + mo(sttranw1i)
## depeffortw1f ~ 1 + mo(sttranw1i)
## deplonelyw1f ~ 1 + mo(sttranw1i)
## depbluesw1f ~ 1 + mo(sttranw1i)
## depunfairw1f ~ 1 + mo(sttranw1i)
## depmistrtw1f ~ 1 + mo(sttranw1i)
## depbetrayw1f ~ 1 + mo(sttranw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_Intercept -0.82 0.25 -1.37 -0.41 1.00 2297
## depeffortw1f_Intercept -1.94 0.28 -2.52 -1.42 1.00 2633
## deplonelyw1f_Intercept -1.25 0.24 -1.71 -0.75 1.00 2810
## depbluesw1f_Intercept -2.01 0.30 -2.64 -1.47 1.00 2467
## depunfairw1f_Intercept -2.21 0.31 -2.92 -1.72 1.00 2458
## depmistrtw1f_Intercept -1.68 0.28 -2.25 -1.14 1.00 2350
## depbetrayw1f_Intercept -1.82 0.28 -2.40 -1.29 1.00 2433
## depcantgow1f_mosttranw1i 0.20 0.09 0.03 0.39 1.00 2388
## depeffortw1f_mosttranw1i 0.04 0.11 -0.17 0.25 1.00 2714
## deplonelyw1f_mosttranw1i -0.01 0.10 -0.20 0.18 1.00 2786
## depbluesw1f_mosttranw1i 0.22 0.11 0.02 0.44 1.00 2586
## depunfairw1f_mosttranw1i 0.37 0.11 0.17 0.60 1.00 2649
## depmistrtw1f_mosttranw1i 0.02 0.11 -0.20 0.25 1.00 2372
## depbetrayw1f_mosttranw1i 0.05 0.11 -0.16 0.28 1.00 2443
## Tail_ESS
## depcantgow1f_Intercept 2376
## depeffortw1f_Intercept 2578
## deplonelyw1f_Intercept 2461
## depbluesw1f_Intercept 2462
## depunfairw1f_Intercept 2333
## depmistrtw1f_Intercept 2334
## depbetrayw1f_Intercept 2485
## depcantgow1f_mosttranw1i 2433
## depeffortw1f_mosttranw1i 2774
## deplonelyw1f_mosttranw1i 2670
## depbluesw1f_mosttranw1i 2726
## depunfairw1f_mosttranw1i 2817
## depmistrtw1f_mosttranw1i 2411
## depbetrayw1f_mosttranw1i 2795
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_mosttranw1i1[1] 0.27 0.15 0.04 0.58 1.00 3004
## depcantgow1f_mosttranw1i1[2] 0.19 0.12 0.03 0.47 1.00 4442
## depcantgow1f_mosttranw1i1[3] 0.25 0.14 0.04 0.56 1.00 4309
## depcantgow1f_mosttranw1i1[4] 0.28 0.14 0.05 0.59 1.00 4129
## depeffortw1f_mosttranw1i1[1] 0.26 0.15 0.04 0.59 1.00 5113
## depeffortw1f_mosttranw1i1[2] 0.25 0.14 0.04 0.57 1.00 5461
## depeffortw1f_mosttranw1i1[3] 0.24 0.14 0.04 0.57 1.00 5720
## depeffortw1f_mosttranw1i1[4] 0.26 0.14 0.04 0.57 1.00 4893
## deplonelyw1f_mosttranw1i1[1] 0.26 0.15 0.04 0.59 1.00 3676
## deplonelyw1f_mosttranw1i1[2] 0.24 0.14 0.03 0.56 1.00 5743
## deplonelyw1f_mosttranw1i1[3] 0.24 0.14 0.03 0.57 1.00 4723
## deplonelyw1f_mosttranw1i1[4] 0.26 0.15 0.04 0.58 1.00 4627
## depbluesw1f_mosttranw1i1[1] 0.22 0.13 0.03 0.51 1.00 3775
## depbluesw1f_mosttranw1i1[2] 0.34 0.16 0.07 0.66 1.00 4099
## depbluesw1f_mosttranw1i1[3] 0.18 0.11 0.03 0.46 1.00 4655
## depbluesw1f_mosttranw1i1[4] 0.26 0.14 0.04 0.57 1.00 4071
## depunfairw1f_mosttranw1i1[1] 0.22 0.12 0.03 0.48 1.00 3581
## depunfairw1f_mosttranw1i1[2] 0.16 0.10 0.02 0.39 1.00 4767
## depunfairw1f_mosttranw1i1[3] 0.23 0.12 0.04 0.49 1.00 4294
## depunfairw1f_mosttranw1i1[4] 0.39 0.14 0.12 0.67 1.00 4239
## depmistrtw1f_mosttranw1i1[1] 0.27 0.15 0.04 0.60 1.00 4586
## depmistrtw1f_mosttranw1i1[2] 0.24 0.14 0.04 0.55 1.00 4400
## depmistrtw1f_mosttranw1i1[3] 0.23 0.14 0.03 0.56 1.00 4087
## depmistrtw1f_mosttranw1i1[4] 0.26 0.15 0.04 0.62 1.00 3913
## depbetrayw1f_mosttranw1i1[1] 0.27 0.15 0.04 0.59 1.00 4430
## depbetrayw1f_mosttranw1i1[2] 0.24 0.14 0.03 0.56 1.00 5419
## depbetrayw1f_mosttranw1i1[3] 0.23 0.14 0.03 0.56 1.00 4113
## depbetrayw1f_mosttranw1i1[4] 0.26 0.15 0.04 0.60 1.00 4519
## Tail_ESS
## depcantgow1f_mosttranw1i1[1] 2472
## depcantgow1f_mosttranw1i1[2] 2621
## depcantgow1f_mosttranw1i1[3] 2498
## depcantgow1f_mosttranw1i1[4] 2954
## depeffortw1f_mosttranw1i1[1] 2724
## depeffortw1f_mosttranw1i1[2] 2907
## depeffortw1f_mosttranw1i1[3] 3136
## depeffortw1f_mosttranw1i1[4] 2935
## deplonelyw1f_mosttranw1i1[1] 2351
## deplonelyw1f_mosttranw1i1[2] 2859
## deplonelyw1f_mosttranw1i1[3] 3152
## deplonelyw1f_mosttranw1i1[4] 2457
## depbluesw1f_mosttranw1i1[1] 2213
## depbluesw1f_mosttranw1i1[2] 2625
## depbluesw1f_mosttranw1i1[3] 3027
## depbluesw1f_mosttranw1i1[4] 2564
## depunfairw1f_mosttranw1i1[1] 2513
## depunfairw1f_mosttranw1i1[2] 2469
## depunfairw1f_mosttranw1i1[3] 2294
## depunfairw1f_mosttranw1i1[4] 2729
## depmistrtw1f_mosttranw1i1[1] 2254
## depmistrtw1f_mosttranw1i1[2] 2986
## depmistrtw1f_mosttranw1i1[3] 2451
## depmistrtw1f_mosttranw1i1[4] 2985
## depbetrayw1f_mosttranw1i1[1] 2431
## depbetrayw1f_mosttranw1i1[2] 2264
## depbetrayw1f_mosttranw1i1[3] 2695
## depbetrayw1f_mosttranw1i1[4] 3060
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.5.2.5 Prior summary
out.alldepress.sttran.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b depbetrayw1f
## normal(0, 0.25) b mosttranw1i depbetrayw1f
## (flat) b depbluesw1f
## normal(0, 0.25) b mosttranw1i depbluesw1f
## (flat) b depcantgow1f
## normal(0, 0.25) b mosttranw1i depcantgow1f
## (flat) b depeffortw1f
## normal(0, 0.25) b mosttranw1i depeffortw1f
## (flat) b deplonelyw1f
## normal(0, 0.25) b mosttranw1i deplonelyw1f
## (flat) b depmistrtw1f
## normal(0, 0.25) b mosttranw1i depmistrtw1f
## (flat) b depunfairw1f
## normal(0, 0.25) b mosttranw1i depunfairw1f
## (flat) Intercept
## normal(0, 2) Intercept depbetrayw1f
## normal(0, 2) Intercept depbluesw1f
## normal(0, 2) Intercept depcantgow1f
## normal(0, 2) Intercept depeffortw1f
## normal(0, 2) Intercept deplonelyw1f
## normal(0, 2) Intercept depmistrtw1f
## normal(0, 2) Intercept depunfairw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 depbetrayw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 depbluesw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 depcantgow1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 depeffortw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 deplonelyw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 depmistrtw1f
## dirichlet(2, 2, 2, 2) simo mosttranw1i1 depunfairw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.5.3 Bivariate Corr: T1 strespw1/negative emotions
#Bivariate: depressive symptom items ~ mo(strespw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostrespw1i',
resp =depdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostrespw1i1',
resp =depdv_names))
alldepress.stresp.fit <- brm(
mvbind(depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f, depunfairw1f, depmistrtw1f,
depbetrayw1f) ~ 1 + mo(strespw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/alldepress_stresp_fit",
file_refit = "on_change"
)
out.alldepress.stresp.fit <- ppchecks(alldepress.stresp.fit)4.3.5.3.1 Coefficient plot (intervals)
out.alldepress.stresp.fit[[11]]
4.3.5.3.2 Coefficient plot (distributions)
out.alldepress.stresp.fit[[10]]
4.3.5.3.3 PPcheck (density)
p1 <- out.alldepress.stresp.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.alldepress.stresp.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.alldepress.stresp.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.alldepress.stresp.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.alldepress.stresp.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.alldepress.stresp.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.alldepress.stresp.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
4.3.5.3.4 Fit summary
out.alldepress.stresp.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgow1f ~ 1 + mo(strespw1i)
## depeffortw1f ~ 1 + mo(strespw1i)
## deplonelyw1f ~ 1 + mo(strespw1i)
## depbluesw1f ~ 1 + mo(strespw1i)
## depunfairw1f ~ 1 + mo(strespw1i)
## depmistrtw1f ~ 1 + mo(strespw1i)
## depbetrayw1f ~ 1 + mo(strespw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_Intercept -0.45 0.17 -0.79 -0.10 1.00 2751
## depeffortw1f_Intercept -2.32 0.28 -2.92 -1.83 1.00 2781
## deplonelyw1f_Intercept -1.87 0.30 -2.52 -1.34 1.00 2285
## depbluesw1f_Intercept -1.57 0.23 -2.01 -1.12 1.00 3093
## depunfairw1f_Intercept -1.58 0.24 -2.06 -1.08 1.00 2286
## depmistrtw1f_Intercept -2.75 0.34 -3.49 -2.14 1.00 2575
## depbetrayw1f_Intercept -2.70 0.38 -3.47 -2.03 1.00 2447
## depcantgow1f_mostrespw1i 0.02 0.06 -0.11 0.15 1.00 2920
## depeffortw1f_mostrespw1i 0.18 0.09 0.02 0.37 1.00 2876
## deplonelyw1f_mostrespw1i 0.23 0.09 0.06 0.42 1.00 2340
## depbluesw1f_mostrespw1i 0.04 0.09 -0.12 0.21 1.00 2792
## depunfairw1f_mostrespw1i 0.05 0.10 -0.13 0.24 1.00 2140
## depmistrtw1f_mostrespw1i 0.41 0.10 0.23 0.62 1.00 2664
## depbetrayw1f_mostrespw1i 0.35 0.11 0.16 0.57 1.00 2514
## Tail_ESS
## depcantgow1f_Intercept 2809
## depeffortw1f_Intercept 2373
## deplonelyw1f_Intercept 2664
## depbluesw1f_Intercept 2782
## depunfairw1f_Intercept 2251
## depmistrtw1f_Intercept 2108
## depbetrayw1f_Intercept 2719
## depcantgow1f_mostrespw1i 2874
## depeffortw1f_mostrespw1i 2812
## deplonelyw1f_mostrespw1i 2524
## depbluesw1f_mostrespw1i 2794
## depunfairw1f_mostrespw1i 2385
## depmistrtw1f_mostrespw1i 1939
## depbetrayw1f_mostrespw1i 2494
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_mostrespw1i1[1] 0.25 0.15 0.03 0.59 1.00 5087
## depcantgow1f_mostrespw1i1[2] 0.24 0.14 0.03 0.57 1.00 4930
## depcantgow1f_mostrespw1i1[3] 0.26 0.15 0.04 0.60 1.00 4729
## depcantgow1f_mostrespw1i1[4] 0.25 0.14 0.04 0.58 1.00 5625
## depeffortw1f_mostrespw1i1[1] 0.30 0.15 0.05 0.63 1.00 4268
## depeffortw1f_mostrespw1i1[2] 0.23 0.13 0.04 0.54 1.00 5251
## depeffortw1f_mostrespw1i1[3] 0.21 0.13 0.03 0.52 1.00 4363
## depeffortw1f_mostrespw1i1[4] 0.26 0.14 0.04 0.57 1.00 5014
## deplonelyw1f_mostrespw1i1[1] 0.41 0.17 0.09 0.74 1.00 3013
## deplonelyw1f_mostrespw1i1[2] 0.18 0.12 0.02 0.45 1.00 4165
## deplonelyw1f_mostrespw1i1[3] 0.20 0.12 0.03 0.50 1.00 4791
## deplonelyw1f_mostrespw1i1[4] 0.21 0.12 0.04 0.49 1.00 4611
## depbluesw1f_mostrespw1i1[1] 0.24 0.14 0.04 0.57 1.00 4524
## depbluesw1f_mostrespw1i1[2] 0.23 0.14 0.03 0.54 1.00 5549
## depbluesw1f_mostrespw1i1[3] 0.24 0.14 0.04 0.55 1.00 5897
## depbluesw1f_mostrespw1i1[4] 0.30 0.17 0.04 0.66 1.00 3716
## depunfairw1f_mostrespw1i1[1] 0.23 0.15 0.03 0.59 1.00 3578
## depunfairw1f_mostrespw1i1[2] 0.22 0.14 0.03 0.55 1.00 4260
## depunfairw1f_mostrespw1i1[3] 0.23 0.14 0.03 0.55 1.00 5342
## depunfairw1f_mostrespw1i1[4] 0.32 0.19 0.04 0.71 1.00 2545
## depmistrtw1f_mostrespw1i1[1] 0.27 0.13 0.05 0.54 1.00 3837
## depmistrtw1f_mostrespw1i1[2] 0.26 0.14 0.04 0.56 1.00 4543
## depmistrtw1f_mostrespw1i1[3] 0.25 0.13 0.05 0.55 1.00 5017
## depmistrtw1f_mostrespw1i1[4] 0.22 0.11 0.04 0.46 1.00 5771
## depbetrayw1f_mostrespw1i1[1] 0.35 0.16 0.08 0.66 1.00 3231
## depbetrayw1f_mostrespw1i1[2] 0.24 0.13 0.04 0.53 1.00 4543
## depbetrayw1f_mostrespw1i1[3] 0.26 0.13 0.05 0.56 1.00 4385
## depbetrayw1f_mostrespw1i1[4] 0.15 0.09 0.02 0.37 1.00 5321
## Tail_ESS
## depcantgow1f_mostrespw1i1[1] 2916
## depcantgow1f_mostrespw1i1[2] 2573
## depcantgow1f_mostrespw1i1[3] 3417
## depcantgow1f_mostrespw1i1[4] 2882
## depeffortw1f_mostrespw1i1[1] 2568
## depeffortw1f_mostrespw1i1[2] 2685
## depeffortw1f_mostrespw1i1[3] 2799
## depeffortw1f_mostrespw1i1[4] 3170
## deplonelyw1f_mostrespw1i1[1] 2368
## deplonelyw1f_mostrespw1i1[2] 2527
## deplonelyw1f_mostrespw1i1[3] 2708
## deplonelyw1f_mostrespw1i1[4] 2730
## depbluesw1f_mostrespw1i1[1] 2597
## depbluesw1f_mostrespw1i1[2] 2721
## depbluesw1f_mostrespw1i1[3] 2948
## depbluesw1f_mostrespw1i1[4] 3284
## depunfairw1f_mostrespw1i1[1] 2909
## depunfairw1f_mostrespw1i1[2] 2836
## depunfairw1f_mostrespw1i1[3] 2952
## depunfairw1f_mostrespw1i1[4] 2659
## depmistrtw1f_mostrespw1i1[1] 2560
## depmistrtw1f_mostrespw1i1[2] 2165
## depmistrtw1f_mostrespw1i1[3] 2674
## depmistrtw1f_mostrespw1i1[4] 2940
## depbetrayw1f_mostrespw1i1[1] 2327
## depbetrayw1f_mostrespw1i1[2] 2602
## depbetrayw1f_mostrespw1i1[3] 2923
## depbetrayw1f_mostrespw1i1[4] 3136
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.5.3.5 Prior summary
out.alldepress.stresp.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b depbetrayw1f
## normal(0, 0.25) b mostrespw1i depbetrayw1f
## (flat) b depbluesw1f
## normal(0, 0.25) b mostrespw1i depbluesw1f
## (flat) b depcantgow1f
## normal(0, 0.25) b mostrespw1i depcantgow1f
## (flat) b depeffortw1f
## normal(0, 0.25) b mostrespw1i depeffortw1f
## (flat) b deplonelyw1f
## normal(0, 0.25) b mostrespw1i deplonelyw1f
## (flat) b depmistrtw1f
## normal(0, 0.25) b mostrespw1i depmistrtw1f
## (flat) b depunfairw1f
## normal(0, 0.25) b mostrespw1i depunfairw1f
## (flat) Intercept
## normal(0, 2) Intercept depbetrayw1f
## normal(0, 2) Intercept depbluesw1f
## normal(0, 2) Intercept depcantgow1f
## normal(0, 2) Intercept depeffortw1f
## normal(0, 2) Intercept deplonelyw1f
## normal(0, 2) Intercept depmistrtw1f
## normal(0, 2) Intercept depunfairw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 depbetrayw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 depbluesw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 depcantgow1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 depeffortw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 deplonelyw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 depmistrtw1f
## dirichlet(2, 2, 2, 2) simo mostrespw1i1 depunfairw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.5.4 Bivariate Corr: T1 stfairw1/negative emotions
#Bivariate: depressive symptom items ~ mo(stfairw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostfairw1i',
resp =depdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostfairw1i1',
resp =depdv_names))
alldepress.stfair.fit <- brm(
mvbind(depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f, depunfairw1f, depmistrtw1f,
depbetrayw1f) ~ 1 + mo(stfairw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/alldepress_stfair_fit",
file_refit = "on_change"
)
out.alldepress.stfair.fit <- ppchecks(alldepress.stfair.fit)4.3.5.4.1 Coefficient plot (intervals)
out.alldepress.stfair.fit[[11]]
4.3.5.4.2 Coefficient plot (distributions)
out.alldepress.stfair.fit[[10]]
4.3.5.4.3 PPcheck (density)
p1 <- out.alldepress.stfair.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.alldepress.stfair.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.alldepress.stfair.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.alldepress.stfair.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.alldepress.stfair.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.alldepress.stfair.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.alldepress.stfair.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
4.3.5.4.4 Fit summary
out.alldepress.stfair.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgow1f ~ 1 + mo(stfairw1i)
## depeffortw1f ~ 1 + mo(stfairw1i)
## deplonelyw1f ~ 1 + mo(stfairw1i)
## depbluesw1f ~ 1 + mo(stfairw1i)
## depunfairw1f ~ 1 + mo(stfairw1i)
## depmistrtw1f ~ 1 + mo(stfairw1i)
## depbetrayw1f ~ 1 + mo(stfairw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_Intercept -0.53 0.17 -0.86 -0.19 1.00 2885
## depeffortw1f_Intercept -2.29 0.30 -2.91 -1.75 1.00 2934
## deplonelyw1f_Intercept -1.82 0.26 -2.38 -1.36 1.00 2461
## depbluesw1f_Intercept -1.70 0.23 -2.17 -1.26 1.00 2713
## depunfairw1f_Intercept -1.62 0.23 -2.06 -1.17 1.00 3018
## depmistrtw1f_Intercept -2.79 0.36 -3.56 -2.16 1.00 2550
## depbetrayw1f_Intercept -2.90 0.38 -3.68 -2.23 1.00 3024
## depcantgow1f_mostfairw1i 0.06 0.06 -0.06 0.17 1.00 2895
## depeffortw1f_mostfairw1i 0.17 0.09 -0.01 0.35 1.00 3092
## deplonelyw1f_mostfairw1i 0.21 0.08 0.06 0.37 1.00 2614
## depbluesw1f_mostfairw1i 0.09 0.08 -0.07 0.26 1.00 2464
## depunfairw1f_mostfairw1i 0.06 0.08 -0.11 0.21 1.00 3246
## depmistrtw1f_mostfairw1i 0.41 0.10 0.23 0.62 1.00 2667
## depbetrayw1f_mostfairw1i 0.41 0.10 0.22 0.63 1.00 3149
## Tail_ESS
## depcantgow1f_Intercept 2721
## depeffortw1f_Intercept 2557
## deplonelyw1f_Intercept 2037
## depbluesw1f_Intercept 2665
## depunfairw1f_Intercept 2878
## depmistrtw1f_Intercept 2436
## depbetrayw1f_Intercept 2406
## depcantgow1f_mostfairw1i 2404
## depeffortw1f_mostfairw1i 2696
## deplonelyw1f_mostfairw1i 2515
## depbluesw1f_mostfairw1i 2747
## depunfairw1f_mostfairw1i 2693
## depmistrtw1f_mostfairw1i 2378
## depbetrayw1f_mostfairw1i 2470
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_mostfairw1i1[1] 0.23 0.14 0.03 0.56 1.00 4833
## depcantgow1f_mostfairw1i1[2] 0.26 0.15 0.04 0.60 1.00 6028
## depcantgow1f_mostfairw1i1[3] 0.26 0.15 0.04 0.60 1.00 5226
## depcantgow1f_mostfairw1i1[4] 0.24 0.14 0.04 0.56 1.00 5245
## depeffortw1f_mostfairw1i1[1] 0.25 0.14 0.04 0.57 1.00 4405
## depeffortw1f_mostfairw1i1[2] 0.30 0.16 0.05 0.64 1.00 4481
## depeffortw1f_mostfairw1i1[3] 0.23 0.14 0.03 0.55 1.00 5345
## depeffortw1f_mostfairw1i1[4] 0.22 0.13 0.03 0.51 1.00 4574
## deplonelyw1f_mostfairw1i1[1] 0.29 0.15 0.05 0.61 1.00 4493
## deplonelyw1f_mostfairw1i1[2] 0.22 0.13 0.04 0.52 1.00 5061
## deplonelyw1f_mostfairw1i1[3] 0.28 0.15 0.05 0.60 1.00 5749
## deplonelyw1f_mostfairw1i1[4] 0.20 0.12 0.03 0.48 1.00 4709
## depbluesw1f_mostfairw1i1[1] 0.23 0.13 0.03 0.53 1.00 4302
## depbluesw1f_mostfairw1i1[2] 0.20 0.13 0.02 0.52 1.00 4512
## depbluesw1f_mostfairw1i1[3] 0.22 0.13 0.03 0.54 1.00 5617
## depbluesw1f_mostfairw1i1[4] 0.35 0.17 0.05 0.69 1.00 3516
## depunfairw1f_mostfairw1i1[1] 0.23 0.14 0.03 0.55 1.00 4449
## depunfairw1f_mostfairw1i1[2] 0.24 0.14 0.04 0.57 1.00 5520
## depunfairw1f_mostfairw1i1[3] 0.27 0.15 0.04 0.61 1.00 5659
## depunfairw1f_mostfairw1i1[4] 0.27 0.15 0.04 0.59 1.00 5484
## depmistrtw1f_mostfairw1i1[1] 0.25 0.13 0.04 0.54 1.00 3543
## depmistrtw1f_mostfairw1i1[2] 0.30 0.14 0.06 0.59 1.00 3869
## depmistrtw1f_mostfairw1i1[3] 0.24 0.13 0.04 0.54 1.00 4115
## depmistrtw1f_mostfairw1i1[4] 0.21 0.11 0.04 0.46 1.00 5639
## depbetrayw1f_mostfairw1i1[1] 0.28 0.14 0.05 0.58 1.00 4069
## depbetrayw1f_mostfairw1i1[2] 0.38 0.16 0.09 0.70 1.00 4074
## depbetrayw1f_mostfairw1i1[3] 0.19 0.11 0.03 0.45 1.00 4584
## depbetrayw1f_mostfairw1i1[4] 0.15 0.09 0.02 0.36 1.00 5559
## Tail_ESS
## depcantgow1f_mostfairw1i1[1] 2210
## depcantgow1f_mostfairw1i1[2] 2861
## depcantgow1f_mostfairw1i1[3] 3015
## depcantgow1f_mostfairw1i1[4] 2914
## depeffortw1f_mostfairw1i1[1] 2559
## depeffortw1f_mostfairw1i1[2] 2498
## depeffortw1f_mostfairw1i1[3] 3019
## depeffortw1f_mostfairw1i1[4] 2952
## deplonelyw1f_mostfairw1i1[1] 2892
## deplonelyw1f_mostfairw1i1[2] 2404
## deplonelyw1f_mostfairw1i1[3] 2923
## deplonelyw1f_mostfairw1i1[4] 3066
## depbluesw1f_mostfairw1i1[1] 2257
## depbluesw1f_mostfairw1i1[2] 2646
## depbluesw1f_mostfairw1i1[3] 2889
## depbluesw1f_mostfairw1i1[4] 3141
## depunfairw1f_mostfairw1i1[1] 2254
## depunfairw1f_mostfairw1i1[2] 2551
## depunfairw1f_mostfairw1i1[3] 3215
## depunfairw1f_mostfairw1i1[4] 2953
## depmistrtw1f_mostfairw1i1[1] 2039
## depmistrtw1f_mostfairw1i1[2] 2495
## depmistrtw1f_mostfairw1i1[3] 2672
## depmistrtw1f_mostfairw1i1[4] 3358
## depbetrayw1f_mostfairw1i1[1] 2592
## depbetrayw1f_mostfairw1i1[2] 2207
## depbetrayw1f_mostfairw1i1[3] 3093
## depbetrayw1f_mostfairw1i1[4] 3223
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.5.4.5 Prior summary
out.alldepress.stfair.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b depbetrayw1f
## normal(0, 0.25) b mostfairw1i depbetrayw1f
## (flat) b depbluesw1f
## normal(0, 0.25) b mostfairw1i depbluesw1f
## (flat) b depcantgow1f
## normal(0, 0.25) b mostfairw1i depcantgow1f
## (flat) b depeffortw1f
## normal(0, 0.25) b mostfairw1i depeffortw1f
## (flat) b deplonelyw1f
## normal(0, 0.25) b mostfairw1i deplonelyw1f
## (flat) b depmistrtw1f
## normal(0, 0.25) b mostfairw1i depmistrtw1f
## (flat) b depunfairw1f
## normal(0, 0.25) b mostfairw1i depunfairw1f
## (flat) Intercept
## normal(0, 2) Intercept depbetrayw1f
## normal(0, 2) Intercept depbluesw1f
## normal(0, 2) Intercept depcantgow1f
## normal(0, 2) Intercept depeffortw1f
## normal(0, 2) Intercept deplonelyw1f
## normal(0, 2) Intercept depmistrtw1f
## normal(0, 2) Intercept depunfairw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 depbetrayw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 depbluesw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 depcantgow1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 depeffortw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 deplonelyw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 depmistrtw1f
## dirichlet(2, 2, 2, 2) simo mostfairw1i1 depunfairw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.5.5 Bivariate Corr: T1 stjobw1/negative emotions
#Bivariate: depressive symptom items ~ mo(stjobw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostjobw1i',
resp =depdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostjobw1i1',
resp =depdv_names))
alldepress.stjob.fit <- brm(
mvbind(depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f, depunfairw1f, depmistrtw1f,
depbetrayw1f) ~ 1 + mo(stjobw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/alldepress_stjob_fit",
file_refit = "on_change"
)
out.alldepress.stjob.fit <- ppchecks(alldepress.stjob.fit)4.3.5.5.1 Coefficient plot (intervals)
out.alldepress.stjob.fit[[11]]
4.3.5.5.2 Coefficient plot (distributions)
out.alldepress.stjob.fit[[10]]
4.3.5.5.3 PPcheck (density)
p1 <- out.alldepress.stjob.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.alldepress.stjob.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.alldepress.stjob.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.alldepress.stjob.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.alldepress.stjob.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.alldepress.stjob.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.alldepress.stjob.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
4.3.5.5.4 Fit summary
out.alldepress.stjob.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgow1f ~ 1 + mo(stjobw1i)
## depeffortw1f ~ 1 + mo(stjobw1i)
## deplonelyw1f ~ 1 + mo(stjobw1i)
## depbluesw1f ~ 1 + mo(stjobw1i)
## depunfairw1f ~ 1 + mo(stjobw1i)
## depmistrtw1f ~ 1 + mo(stjobw1i)
## depbetrayw1f ~ 1 + mo(stjobw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_Intercept -0.13 0.17 -0.47 0.21 1.00 4391
## depeffortw1f_Intercept -2.12 0.26 -2.67 -1.64 1.00 4507
## deplonelyw1f_Intercept -1.22 0.21 -1.62 -0.81 1.00 4223
## depbluesw1f_Intercept -1.58 0.22 -2.01 -1.18 1.00 4166
## depunfairw1f_Intercept -1.60 0.22 -2.04 -1.18 1.00 3754
## depmistrtw1f_Intercept -1.73 0.24 -2.22 -1.29 1.00 3755
## depbetrayw1f_Intercept -2.05 0.27 -2.62 -1.57 1.00 3540
## depcantgow1f_mostjobw1i -0.12 0.06 -0.24 0.01 1.00 4300
## depeffortw1f_mostjobw1i 0.12 0.10 -0.07 0.31 1.00 4181
## deplonelyw1f_mostjobw1i -0.02 0.08 -0.17 0.14 1.00 4214
## depbluesw1f_mostjobw1i 0.04 0.08 -0.11 0.19 1.00 4234
## depunfairw1f_mostjobw1i 0.05 0.08 -0.10 0.22 1.00 2941
## depmistrtw1f_mostjobw1i 0.04 0.09 -0.13 0.20 1.00 4493
## depbetrayw1f_mostjobw1i 0.14 0.09 -0.03 0.34 1.00 3460
## Tail_ESS
## depcantgow1f_Intercept 3016
## depeffortw1f_Intercept 3191
## deplonelyw1f_Intercept 3002
## depbluesw1f_Intercept 3074
## depunfairw1f_Intercept 3507
## depmistrtw1f_Intercept 2993
## depbetrayw1f_Intercept 2296
## depcantgow1f_mostjobw1i 3135
## depeffortw1f_mostjobw1i 2872
## deplonelyw1f_mostjobw1i 3133
## depbluesw1f_mostjobw1i 3264
## depunfairw1f_mostjobw1i 3386
## depmistrtw1f_mostjobw1i 3110
## depbetrayw1f_mostjobw1i 2557
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_mostjobw1i1[1] 0.18 0.12 0.02 0.46 1.00 7209
## depcantgow1f_mostjobw1i1[2] 0.32 0.16 0.05 0.66 1.00 6678
## depcantgow1f_mostjobw1i1[3] 0.30 0.16 0.05 0.65 1.00 7316
## depcantgow1f_mostjobw1i1[4] 0.20 0.12 0.03 0.50 1.00 6812
## depeffortw1f_mostjobw1i1[1] 0.25 0.14 0.04 0.56 1.00 7920
## depeffortw1f_mostjobw1i1[2] 0.21 0.13 0.03 0.53 1.00 7496
## depeffortw1f_mostjobw1i1[3] 0.23 0.14 0.03 0.53 1.00 7376
## depeffortw1f_mostjobw1i1[4] 0.31 0.16 0.05 0.65 1.00 6658
## deplonelyw1f_mostjobw1i1[1] 0.26 0.15 0.04 0.60 1.00 6826
## deplonelyw1f_mostjobw1i1[2] 0.25 0.15 0.04 0.58 1.00 7311
## deplonelyw1f_mostjobw1i1[3] 0.24 0.14 0.04 0.58 1.00 7840
## deplonelyw1f_mostjobw1i1[4] 0.25 0.15 0.04 0.59 1.00 6821
## depbluesw1f_mostjobw1i1[1] 0.26 0.15 0.04 0.58 1.00 6378
## depbluesw1f_mostjobw1i1[2] 0.25 0.14 0.04 0.57 1.00 6991
## depbluesw1f_mostjobw1i1[3] 0.25 0.15 0.03 0.59 1.00 6846
## depbluesw1f_mostjobw1i1[4] 0.24 0.14 0.03 0.56 1.00 6569
## depunfairw1f_mostjobw1i1[1] 0.24 0.14 0.04 0.56 1.00 8183
## depunfairw1f_mostjobw1i1[2] 0.23 0.13 0.03 0.53 1.00 6426
## depunfairw1f_mostjobw1i1[3] 0.24 0.14 0.03 0.56 1.00 8784
## depunfairw1f_mostjobw1i1[4] 0.30 0.17 0.04 0.65 1.00 5856
## depmistrtw1f_mostjobw1i1[1] 0.27 0.15 0.04 0.61 1.00 7118
## depmistrtw1f_mostjobw1i1[2] 0.23 0.14 0.03 0.56 1.00 5530
## depmistrtw1f_mostjobw1i1[3] 0.24 0.14 0.03 0.58 1.00 7046
## depmistrtw1f_mostjobw1i1[4] 0.26 0.15 0.03 0.59 1.00 6976
## depbetrayw1f_mostjobw1i1[1] 0.30 0.16 0.05 0.65 1.00 5560
## depbetrayw1f_mostjobw1i1[2] 0.19 0.12 0.02 0.49 1.00 6800
## depbetrayw1f_mostjobw1i1[3] 0.21 0.13 0.03 0.52 1.00 6806
## depbetrayw1f_mostjobw1i1[4] 0.30 0.16 0.05 0.65 1.00 5939
## Tail_ESS
## depcantgow1f_mostjobw1i1[1] 2834
## depcantgow1f_mostjobw1i1[2] 2707
## depcantgow1f_mostjobw1i1[3] 3202
## depcantgow1f_mostjobw1i1[4] 3130
## depeffortw1f_mostjobw1i1[1] 2483
## depeffortw1f_mostjobw1i1[2] 2634
## depeffortw1f_mostjobw1i1[3] 2991
## depeffortw1f_mostjobw1i1[4] 3106
## deplonelyw1f_mostjobw1i1[1] 3058
## deplonelyw1f_mostjobw1i1[2] 2649
## deplonelyw1f_mostjobw1i1[3] 2724
## deplonelyw1f_mostjobw1i1[4] 3205
## depbluesw1f_mostjobw1i1[1] 2502
## depbluesw1f_mostjobw1i1[2] 2893
## depbluesw1f_mostjobw1i1[3] 2730
## depbluesw1f_mostjobw1i1[4] 2772
## depunfairw1f_mostjobw1i1[1] 2917
## depunfairw1f_mostjobw1i1[2] 2399
## depunfairw1f_mostjobw1i1[3] 3015
## depunfairw1f_mostjobw1i1[4] 3075
## depmistrtw1f_mostjobw1i1[1] 2838
## depmistrtw1f_mostjobw1i1[2] 3057
## depmistrtw1f_mostjobw1i1[3] 2769
## depmistrtw1f_mostjobw1i1[4] 2794
## depbetrayw1f_mostjobw1i1[1] 2591
## depbetrayw1f_mostjobw1i1[2] 2575
## depbetrayw1f_mostjobw1i1[3] 3292
## depbetrayw1f_mostjobw1i1[4] 3041
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.5.5.5 Prior summary
out.alldepress.stjob.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b depbetrayw1f
## normal(0, 0.25) b mostjobw1i depbetrayw1f
## (flat) b depbluesw1f
## normal(0, 0.25) b mostjobw1i depbluesw1f
## (flat) b depcantgow1f
## normal(0, 0.25) b mostjobw1i depcantgow1f
## (flat) b depeffortw1f
## normal(0, 0.25) b mostjobw1i depeffortw1f
## (flat) b deplonelyw1f
## normal(0, 0.25) b mostjobw1i deplonelyw1f
## (flat) b depmistrtw1f
## normal(0, 0.25) b mostjobw1i depmistrtw1f
## (flat) b depunfairw1f
## normal(0, 0.25) b mostjobw1i depunfairw1f
## (flat) Intercept
## normal(0, 2) Intercept depbetrayw1f
## normal(0, 2) Intercept depbluesw1f
## normal(0, 2) Intercept depcantgow1f
## normal(0, 2) Intercept depeffortw1f
## normal(0, 2) Intercept deplonelyw1f
## normal(0, 2) Intercept depmistrtw1f
## normal(0, 2) Intercept depunfairw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 depbetrayw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 depbluesw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 depcantgow1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 depeffortw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 deplonelyw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 depmistrtw1f
## dirichlet(2, 2, 2, 2) simo mostjobw1i1 depunfairw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.5.6 Bivariate Corr: T1 stthftw1/negative emotions
#Bivariate: depressive symptom items ~ mo(stthftw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostthftw1i',
resp =depdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostthftw1i1',
resp =depdv_names))
alldepress.stthft.fit <- brm(
mvbind(depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f, depunfairw1f, depmistrtw1f,
depbetrayw1f) ~ 1 + mo(stthftw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/alldepress_stthft_fit",
file_refit = "on_change"
)
out.alldepress.stthft.fit <- ppchecks(alldepress.stthft.fit)4.3.5.6.1 Coefficient plot (intervals)
out.alldepress.stthft.fit[[11]]
4.3.5.6.2 Coefficient plot (distributions)
out.alldepress.stthft.fit[[10]]
4.3.5.6.3 PPcheck (density)
p1 <- out.alldepress.stthft.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.alldepress.stthft.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.alldepress.stthft.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.alldepress.stthft.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.alldepress.stthft.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.alldepress.stthft.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.alldepress.stthft.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
4.3.5.6.4 Fit summary
out.alldepress.stthft.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgow1f ~ 1 + mo(stthftw1i)
## depeffortw1f ~ 1 + mo(stthftw1i)
## deplonelyw1f ~ 1 + mo(stthftw1i)
## depbluesw1f ~ 1 + mo(stthftw1i)
## depunfairw1f ~ 1 + mo(stthftw1i)
## depmistrtw1f ~ 1 + mo(stthftw1i)
## depbetrayw1f ~ 1 + mo(stthftw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_Intercept -0.49 0.13 -0.75 -0.23 1.00 5628
## depeffortw1f_Intercept -1.94 0.18 -2.31 -1.59 1.00 5537
## deplonelyw1f_Intercept -1.61 0.17 -1.95 -1.30 1.00 5560
## depbluesw1f_Intercept -1.53 0.17 -1.85 -1.20 1.00 4612
## depunfairw1f_Intercept -1.53 0.17 -1.87 -1.20 1.00 5684
## depmistrtw1f_Intercept -1.94 0.21 -2.38 -1.55 1.00 5364
## depbetrayw1f_Intercept -2.11 0.22 -2.56 -1.71 1.00 5790
## depcantgow1f_mostthftw1i 0.08 0.08 -0.07 0.24 1.00 5314
## depeffortw1f_mostthftw1i 0.07 0.10 -0.14 0.27 1.00 4635
## deplonelyw1f_mostthftw1i 0.29 0.09 0.12 0.47 1.00 6407
## depbluesw1f_mostthftw1i 0.04 0.10 -0.16 0.25 1.00 4218
## depunfairw1f_mostthftw1i 0.04 0.10 -0.16 0.25 1.00 5113
## depmistrtw1f_mostthftw1i 0.19 0.10 0.00 0.38 1.00 5614
## depbetrayw1f_mostthftw1i 0.27 0.09 0.08 0.45 1.00 6153
## Tail_ESS
## depcantgow1f_Intercept 3278
## depeffortw1f_Intercept 2958
## deplonelyw1f_Intercept 2812
## depbluesw1f_Intercept 2843
## depunfairw1f_Intercept 3094
## depmistrtw1f_Intercept 3450
## depbetrayw1f_Intercept 2955
## depcantgow1f_mostthftw1i 3110
## depeffortw1f_mostthftw1i 2852
## deplonelyw1f_mostthftw1i 3089
## depbluesw1f_mostthftw1i 3427
## depunfairw1f_mostthftw1i 3215
## depmistrtw1f_mostthftw1i 2630
## depbetrayw1f_mostthftw1i 3330
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_mostthftw1i1[1] 0.25 0.15 0.04 0.58 1.00 8169
## depcantgow1f_mostthftw1i1[2] 0.24 0.14 0.04 0.56 1.00 7489
## depcantgow1f_mostthftw1i1[3] 0.25 0.14 0.04 0.57 1.00 6620
## depcantgow1f_mostthftw1i1[4] 0.26 0.15 0.04 0.60 1.00 7254
## depeffortw1f_mostthftw1i1[1] 0.23 0.14 0.03 0.58 1.00 6541
## depeffortw1f_mostthftw1i1[2] 0.24 0.14 0.04 0.56 1.00 7741
## depeffortw1f_mostthftw1i1[3] 0.27 0.15 0.04 0.61 1.00 6774
## depeffortw1f_mostthftw1i1[4] 0.26 0.15 0.04 0.60 1.00 7232
## deplonelyw1f_mostthftw1i1[1] 0.19 0.11 0.03 0.45 1.00 7225
## deplonelyw1f_mostthftw1i1[2] 0.23 0.12 0.04 0.52 1.00 7385
## deplonelyw1f_mostthftw1i1[3] 0.35 0.15 0.08 0.66 1.00 7005
## deplonelyw1f_mostthftw1i1[4] 0.23 0.12 0.04 0.49 1.00 6987
## depbluesw1f_mostthftw1i1[1] 0.23 0.14 0.03 0.55 1.00 7504
## depbluesw1f_mostthftw1i1[2] 0.22 0.14 0.03 0.55 1.00 6062
## depbluesw1f_mostthftw1i1[3] 0.28 0.16 0.04 0.63 1.00 4883
## depbluesw1f_mostthftw1i1[4] 0.27 0.15 0.04 0.60 1.00 6327
## depunfairw1f_mostthftw1i1[1] 0.23 0.14 0.03 0.57 1.00 6126
## depunfairw1f_mostthftw1i1[2] 0.23 0.14 0.03 0.56 1.00 7070
## depunfairw1f_mostthftw1i1[3] 0.26 0.15 0.04 0.61 1.00 6558
## depunfairw1f_mostthftw1i1[4] 0.28 0.16 0.04 0.64 1.00 5274
## depmistrtw1f_mostthftw1i1[1] 0.34 0.16 0.06 0.67 1.00 5952
## depmistrtw1f_mostthftw1i1[2] 0.22 0.13 0.03 0.53 1.00 7440
## depmistrtw1f_mostthftw1i1[3] 0.20 0.12 0.03 0.48 1.00 6469
## depmistrtw1f_mostthftw1i1[4] 0.24 0.14 0.03 0.54 1.00 6953
## depbetrayw1f_mostthftw1i1[1] 0.29 0.14 0.05 0.59 1.00 6586
## depbetrayw1f_mostthftw1i1[2] 0.30 0.15 0.06 0.62 1.00 7299
## depbetrayw1f_mostthftw1i1[3] 0.20 0.12 0.03 0.47 1.00 7814
## depbetrayw1f_mostthftw1i1[4] 0.20 0.11 0.03 0.48 1.00 6879
## Tail_ESS
## depcantgow1f_mostthftw1i1[1] 2747
## depcantgow1f_mostthftw1i1[2] 2759
## depcantgow1f_mostthftw1i1[3] 2080
## depcantgow1f_mostthftw1i1[4] 3034
## depeffortw1f_mostthftw1i1[1] 2584
## depeffortw1f_mostthftw1i1[2] 2858
## depeffortw1f_mostthftw1i1[3] 3009
## depeffortw1f_mostthftw1i1[4] 2939
## deplonelyw1f_mostthftw1i1[1] 2845
## deplonelyw1f_mostthftw1i1[2] 2482
## deplonelyw1f_mostthftw1i1[3] 2903
## deplonelyw1f_mostthftw1i1[4] 3007
## depbluesw1f_mostthftw1i1[1] 3144
## depbluesw1f_mostthftw1i1[2] 2403
## depbluesw1f_mostthftw1i1[3] 2669
## depbluesw1f_mostthftw1i1[4] 2979
## depunfairw1f_mostthftw1i1[1] 2800
## depunfairw1f_mostthftw1i1[2] 2467
## depunfairw1f_mostthftw1i1[3] 2808
## depunfairw1f_mostthftw1i1[4] 2546
## depmistrtw1f_mostthftw1i1[1] 2897
## depmistrtw1f_mostthftw1i1[2] 2945
## depmistrtw1f_mostthftw1i1[3] 3059
## depmistrtw1f_mostthftw1i1[4] 2898
## depbetrayw1f_mostthftw1i1[1] 2515
## depbetrayw1f_mostthftw1i1[2] 2822
## depbetrayw1f_mostthftw1i1[3] 2827
## depbetrayw1f_mostthftw1i1[4] 2817
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.5.6.5 Prior summary
out.alldepress.stthft.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b depbetrayw1f
## normal(0, 0.25) b mostthftw1i depbetrayw1f
## (flat) b depbluesw1f
## normal(0, 0.25) b mostthftw1i depbluesw1f
## (flat) b depcantgow1f
## normal(0, 0.25) b mostthftw1i depcantgow1f
## (flat) b depeffortw1f
## normal(0, 0.25) b mostthftw1i depeffortw1f
## (flat) b deplonelyw1f
## normal(0, 0.25) b mostthftw1i deplonelyw1f
## (flat) b depmistrtw1f
## normal(0, 0.25) b mostthftw1i depmistrtw1f
## (flat) b depunfairw1f
## normal(0, 0.25) b mostthftw1i depunfairw1f
## (flat) Intercept
## normal(0, 2) Intercept depbetrayw1f
## normal(0, 2) Intercept depbluesw1f
## normal(0, 2) Intercept depcantgow1f
## normal(0, 2) Intercept depeffortw1f
## normal(0, 2) Intercept deplonelyw1f
## normal(0, 2) Intercept depmistrtw1f
## normal(0, 2) Intercept depunfairw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 depbetrayw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 depbluesw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 depcantgow1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 depeffortw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 deplonelyw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 depmistrtw1f
## dirichlet(2, 2, 2, 2) simo mostthftw1i1 depunfairw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user4.3.5.7 Bivariate Corr: T1 stmugw1/negative emotions
#Bivariate: depressive symptom items ~ mo(stmugw1i)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmugw1i',
resp =depdv_names),
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmugw1i1',
resp =depdv_names))
alldepress.stmug.fit <- brm(
mvbind(depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f, depunfairw1f, depmistrtw1f,
depbetrayw1f) ~ 1 + mo(stmugw1i),
data = stress.wide3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/alldepress_stmug_fit",
file_refit = "on_change"
)
out.alldepress.stmug.fit <- ppchecks(alldepress.stmug.fit)4.3.5.7.1 Coefficient plot (intervals)
out.alldepress.stmug.fit[[11]]
4.3.5.7.2 Coefficient plot (distributions)
out.alldepress.stmug.fit[[10]]
4.3.5.7.3 PPcheck (density)
p1 <- out.alldepress.stmug.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.alldepress.stmug.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.alldepress.stmug.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.alldepress.stmug.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.alldepress.stmug.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.alldepress.stmug.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.alldepress.stmug.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
4.3.5.7.4 Fit summary
out.alldepress.stmug.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgow1f ~ 1 + mo(stmugw1i)
## depeffortw1f ~ 1 + mo(stmugw1i)
## deplonelyw1f ~ 1 + mo(stmugw1i)
## depbluesw1f ~ 1 + mo(stmugw1i)
## depunfairw1f ~ 1 + mo(stmugw1i)
## depmistrtw1f ~ 1 + mo(stmugw1i)
## depbetrayw1f ~ 1 + mo(stmugw1i)
## Data: stress.wide3 (Number of observations: 489)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_Intercept -0.45 0.12 -0.69 -0.22 1.00 5005
## depeffortw1f_Intercept -2.03 0.18 -2.40 -1.69 1.00 5277
## deplonelyw1f_Intercept -1.55 0.15 -1.84 -1.27 1.00 5842
## depbluesw1f_Intercept -1.52 0.15 -1.82 -1.21 1.00 5060
## depunfairw1f_Intercept -1.55 0.15 -1.85 -1.26 1.00 4828
## depmistrtw1f_Intercept -2.37 0.24 -2.86 -1.94 1.00 4394
## depbetrayw1f_Intercept -2.38 0.23 -2.86 -1.94 1.00 4598
## depcantgow1f_mostmugw1i 0.06 0.09 -0.12 0.23 1.00 4123
## depeffortw1f_mostmugw1i 0.18 0.12 -0.04 0.42 1.00 4452
## deplonelyw1f_mostmugw1i 0.32 0.11 0.13 0.54 1.00 4545
## depbluesw1f_mostmugw1i 0.04 0.11 -0.18 0.28 1.00 3715
## depunfairw1f_mostmugw1i 0.08 0.11 -0.13 0.31 1.00 4096
## depmistrtw1f_mostmugw1i 0.50 0.10 0.31 0.70 1.00 5011
## depbetrayw1f_mostmugw1i 0.46 0.10 0.27 0.66 1.00 4796
## Tail_ESS
## depcantgow1f_Intercept 3457
## depeffortw1f_Intercept 2909
## deplonelyw1f_Intercept 3333
## depbluesw1f_Intercept 3140
## depunfairw1f_Intercept 3278
## depmistrtw1f_Intercept 2983
## depbetrayw1f_Intercept 2968
## depcantgow1f_mostmugw1i 3266
## depeffortw1f_mostmugw1i 3039
## deplonelyw1f_mostmugw1i 2725
## depbluesw1f_mostmugw1i 3255
## depunfairw1f_mostmugw1i 2995
## depmistrtw1f_mostmugw1i 3076
## depbetrayw1f_mostmugw1i 3167
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgow1f_mostmugw1i1[1] 0.25 0.14 0.03 0.58 1.00 6296
## depcantgow1f_mostmugw1i1[2] 0.26 0.15 0.04 0.58 1.00 6819
## depcantgow1f_mostmugw1i1[3] 0.23 0.14 0.04 0.56 1.00 7007
## depcantgow1f_mostmugw1i1[4] 0.27 0.15 0.04 0.60 1.00 6006
## depeffortw1f_mostmugw1i1[1] 0.25 0.14 0.04 0.56 1.00 6802
## depeffortw1f_mostmugw1i1[2] 0.20 0.13 0.03 0.50 1.00 7382
## depeffortw1f_mostmugw1i1[3] 0.26 0.14 0.04 0.58 1.00 6612
## depeffortw1f_mostmugw1i1[4] 0.29 0.15 0.05 0.62 1.00 6412
## deplonelyw1f_mostmugw1i1[1] 0.18 0.11 0.03 0.43 1.00 7230
## deplonelyw1f_mostmugw1i1[2] 0.25 0.13 0.04 0.54 1.00 6692
## deplonelyw1f_mostmugw1i1[3] 0.29 0.15 0.06 0.62 1.00 7298
## deplonelyw1f_mostmugw1i1[4] 0.27 0.15 0.05 0.59 1.00 5887
## depbluesw1f_mostmugw1i1[1] 0.22 0.14 0.03 0.56 1.00 5371
## depbluesw1f_mostmugw1i1[2] 0.23 0.14 0.04 0.55 1.00 6510
## depbluesw1f_mostmugw1i1[3] 0.27 0.15 0.04 0.61 1.00 6657
## depbluesw1f_mostmugw1i1[4] 0.27 0.15 0.04 0.62 1.00 6852
## depunfairw1f_mostmugw1i1[1] 0.20 0.14 0.02 0.53 1.00 5047
## depunfairw1f_mostmugw1i1[2] 0.25 0.14 0.04 0.57 1.00 6302
## depunfairw1f_mostmugw1i1[3] 0.27 0.15 0.04 0.61 1.00 6280
## depunfairw1f_mostmugw1i1[4] 0.27 0.15 0.04 0.60 1.00 6642
## depmistrtw1f_mostmugw1i1[1] 0.38 0.12 0.14 0.62 1.00 5343
## depmistrtw1f_mostmugw1i1[2] 0.25 0.12 0.05 0.50 1.00 5756
## depmistrtw1f_mostmugw1i1[3] 0.21 0.10 0.04 0.44 1.00 8034
## depmistrtw1f_mostmugw1i1[4] 0.16 0.09 0.03 0.38 1.00 6288
## depbetrayw1f_mostmugw1i1[1] 0.39 0.13 0.14 0.64 1.00 5235
## depbetrayw1f_mostmugw1i1[2] 0.22 0.12 0.04 0.49 1.00 6298
## depbetrayw1f_mostmugw1i1[3] 0.23 0.12 0.05 0.48 1.00 6223
## depbetrayw1f_mostmugw1i1[4] 0.16 0.09 0.02 0.37 1.00 6743
## Tail_ESS
## depcantgow1f_mostmugw1i1[1] 2347
## depcantgow1f_mostmugw1i1[2] 2965
## depcantgow1f_mostmugw1i1[3] 2649
## depcantgow1f_mostmugw1i1[4] 2105
## depeffortw1f_mostmugw1i1[1] 2880
## depeffortw1f_mostmugw1i1[2] 3445
## depeffortw1f_mostmugw1i1[3] 2704
## depeffortw1f_mostmugw1i1[4] 3096
## deplonelyw1f_mostmugw1i1[1] 2655
## deplonelyw1f_mostmugw1i1[2] 2747
## deplonelyw1f_mostmugw1i1[3] 2909
## deplonelyw1f_mostmugw1i1[4] 2975
## depbluesw1f_mostmugw1i1[1] 2668
## depbluesw1f_mostmugw1i1[2] 2558
## depbluesw1f_mostmugw1i1[3] 2864
## depbluesw1f_mostmugw1i1[4] 3102
## depunfairw1f_mostmugw1i1[1] 2950
## depunfairw1f_mostmugw1i1[2] 2921
## depunfairw1f_mostmugw1i1[3] 2915
## depunfairw1f_mostmugw1i1[4] 2835
## depmistrtw1f_mostmugw1i1[1] 2819
## depmistrtw1f_mostmugw1i1[2] 2789
## depmistrtw1f_mostmugw1i1[3] 3199
## depmistrtw1f_mostmugw1i1[4] 3031
## depbetrayw1f_mostmugw1i1[1] 2800
## depbetrayw1f_mostmugw1i1[2] 2504
## depbetrayw1f_mostmugw1i1[3] 3141
## depbetrayw1f_mostmugw1i1[4] 2873
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).4.3.5.7.5 Prior summary
out.alldepress.stmug.fit[[2]]## prior class coef group resp dpar nlpar lb
## (flat) b
## (flat) b depbetrayw1f
## normal(0, 0.25) b mostmugw1i depbetrayw1f
## (flat) b depbluesw1f
## normal(0, 0.25) b mostmugw1i depbluesw1f
## (flat) b depcantgow1f
## normal(0, 0.25) b mostmugw1i depcantgow1f
## (flat) b depeffortw1f
## normal(0, 0.25) b mostmugw1i depeffortw1f
## (flat) b deplonelyw1f
## normal(0, 0.25) b mostmugw1i deplonelyw1f
## (flat) b depmistrtw1f
## normal(0, 0.25) b mostmugw1i depmistrtw1f
## (flat) b depunfairw1f
## normal(0, 0.25) b mostmugw1i depunfairw1f
## (flat) Intercept
## normal(0, 2) Intercept depbetrayw1f
## normal(0, 2) Intercept depbluesw1f
## normal(0, 2) Intercept depcantgow1f
## normal(0, 2) Intercept depeffortw1f
## normal(0, 2) Intercept deplonelyw1f
## normal(0, 2) Intercept depmistrtw1f
## normal(0, 2) Intercept depunfairw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 depbetrayw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 depbluesw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 depcantgow1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 depeffortw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 deplonelyw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 depmistrtw1f
## dirichlet(2, 2, 2, 2) simo mostmugw1i1 depunfairw1f
## ub source
## default
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## default
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## usersave(stress.wide3, file = here("1_Data_Files/Datasets/stress_wide3.Rdata"))5 Visualizing T1 correlations (RQ2A)
(RMD FILE: BDK_2023_Stress_4_T1corr_viz)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE5.1 Visualizing “bivariate” T1 associations
Now that we have estimated all these bivariate models, we need a way
to efficiently visualize all these associations. Let’s start by
visualizing the association between one of the seven stress items and
all six past crime items. We will start by grabbing and plotting a full
range of posterior predicted expected crime probabilities for each
stress category, using the stmonyw1i item as an example.
For a good description of differences across the types of posterior
predictions you can request, see Andrew
Heiss’s blog.
5.1.1 PP Exp Past Crime Probability Distributions by Stress Category
#load stress.wide (from Fig1 Rmd)
load(here("1_Data_Files/Datasets/stress_wide.Rdata"))
stress.wide <- zap_labels(stress.wide)
stress.wide <- zap_label(stress.wide)
load(here("1_Data_Files/Datasets/stress_wide2.Rdata"))
load(here("1_Data_Files/Datasets/stress_wide3.Rdata"))
#load bivariate brms model fits
#past crime
allpstcrime.stmony.fit <- readRDS(here("Models/allpstcrime_stmony_fit.rds"))
allpstcrime.sttran.fit <- readRDS(here("Models/allpstcrime_sttran_fit.rds"))
allpstcrime.stresp.fit <- readRDS(here("Models/allpstcrime_stresp_fit.rds"))
allpstcrime.stfair.fit <- readRDS(here("Models/allpstcrime_stfair_fit.rds"))
allpstcrime.stjob.fit <- readRDS(here("Models/allpstcrime_stjob_fit.rds"))
allpstcrime.stthft.fit <- readRDS(here("Models/allpstcrime_stthft_fit.rds"))
allpstcrime.stmug.fit <- readRDS(here("Models/allpstcrime_stmug_fit.rds"))
anypstcrime.stmony.fit <- readRDS(here("Models/anypstcrime_stmony_fit.rds"))
anypstcrime.sttran.fit <- readRDS(here("Models/anypstcrime_sttran_fit.rds"))
anypstcrime.stresp.fit <- readRDS(here("Models/anypstcrime_stresp_fit.rds"))
anypstcrime.stfair.fit <- readRDS(here("Models/anypstcrime_stfair_fit.rds"))
anypstcrime.stjob.fit <- readRDS(here("Models/anypstcrime_stjob_fit.rds"))
anypstcrime.stthft.fit <- readRDS(here("Models/anypstcrime_stthft_fit.rds"))
anypstcrime.stmug.fit <- readRDS(here("Models/anypstcrime_stmug_fit.rds"))
#criminal intent
allprjcrime.stmony.fit <- readRDS(here("Models/allprjcrime_stmony_fit.rds"))
allprjcrime.sttran.fit <- readRDS(here("Models/allprjcrime_sttran_fit.rds"))
allprjcrime.stresp.fit <- readRDS(here("Models/allprjcrime_stresp_fit.rds"))
allprjcrime.stfair.fit <- readRDS(here("Models/allprjcrime_stfair_fit.rds"))
allprjcrime.stjob.fit <- readRDS(here("Models/allprjcrime_stjob_fit.rds"))
allprjcrime.stthft.fit <- readRDS(here("Models/allprjcrime_stthft_fit.rds"))
allprjcrime.stmug.fit <- readRDS(here("Models/allprjcrime_stmug_fit.rds"))
anyprjcrime.stmony.fit <- readRDS(here("Models/anyprjcrime_stmony_fit.rds"))
anyprjcrime.sttran.fit <- readRDS(here("Models/anyprjcrime_sttran_fit.rds"))
anyprjcrime.stresp.fit <- readRDS(here("Models/anyprjcrime_stresp_fit.rds"))
anyprjcrime.stfair.fit <- readRDS(here("Models/anyprjcrime_stfair_fit.rds"))
anyprjcrime.stjob.fit <- readRDS(here("Models/anyprjcrime_stjob_fit.rds"))
anyprjcrime.stthft.fit <- readRDS(here("Models/anyprjcrime_stthft_fit.rds"))
anyprjcrime.stmug.fit <- readRDS(here("Models/anyprjcrime_stmug_fit.rds"))
#negative emotions
alldepress.stmony.fit <- readRDS(here("Models/alldepress_stmony_fit.rds"))
alldepress.sttran.fit <- readRDS(here("Models/alldepress_sttran_fit.rds"))
alldepress.stresp.fit <- readRDS(here("Models/alldepress_stresp_fit.rds"))
alldepress.stfair.fit <- readRDS(here("Models/alldepress_stfair_fit.rds"))
alldepress.stjob.fit <- readRDS(here("Models/alldepress_stjob_fit.rds"))
alldepress.stthft.fit <- readRDS(here("Models/alldepress_stthft_fit.rds"))
alldepress.stmug.fit <- readRDS(here("Models/alldepress_stmug_fit.rds"))# http://mjskay.github.io/tidybayes/articles/tidy-brms.html
theme_set(theme_tidybayes())
epred.stmony <- stress.wide3 %>%
data_grid(stmonyw1i) %>%
add_epred_draws(allpstcrime.stmony.fit)
#epred
ggplot(data=epred.stmony, aes(y= .epred, x= stmonyw1i)) +
facet_wrap(~ .category, nrow=3, ncol=2, ) +
scale_x_continuous(breaks=c(1,2,3,4,5), name="Stress: Money") +
scale_y_continuous(name="Posterior expected probability of crime")+
stat_halfeye(.width=.95, size=.5) +
coord_flip(ylim=c(0,.3))
This is certainly an informative summary of the association between
stress about money and each of the six binary past crime items.
However, we have six more stress items, multiplied by two more dependent
variables, and again by each of our four community groups. That is a lot
of figures - just to summarize bivariate associations at T1! Then, we
have change correlations and other analyses. We need a more efficient
way to summarize these six associations. We can start by extracting the
most pertinent information from these predictive distributions - the
fitted (median) posterior probability estimates and uncertainty
intervals, then plot them using a simpler stat interval format.
5.1.2 PP Exp Past Crime Probability Est/Intervals by Stress Category
#function to find & drop leading zeroes (used for x-axis label)
dropLeadingZero <- function(l){
str_replace(l, '0(?=.)', '')
}
#fitted median_qi
fitted.stmony <- stress.wide3 %>%
data_grid(stmonyw1i) %>%
add_epred_draws(allpstcrime.stmony.fit) %>%
median_qi(.epred)
pstcrimelabs <- c("pstthflt5w1f"="Theft <5BAM",
"pstthfgt5w1f"="Theft >5BAM",
"pstthreatw1f"="Threaten",
"pstharmw1f"="Phys. harm",
"pstusedrgw1f"="Use drugs",
"psthackw1f"="Hack info.")
ggplot(data=fitted.stmony,
aes(x= .epred, y= stmonyw1i, xmin=.lower, xmax=.upper)) +
facet_wrap(~ .category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
scale_y_continuous(breaks=c(1,2,3,4,5),
name="Stress: Money") +
scale_x_continuous(breaks=c(0,.1,.2,.3,.4),
labels = dropLeadingZero, name="Posterior expected probability of past crime at T1") +
geom_pointinterval() +
coord_cartesian(xlim=c(0,.4))
#fitted values in figure below replicate those plotted in the posterior expectation predictions aboveThat condenses our summary plots while still retaining most of the useful information. But again, we have a LOT more information to add. Let’s try dropping a couple categories from our stress item to further condense things.
5.1.3 PP Exp Past Crime Probability Est/Intervals by Stress Category (Condensed)
stress3catlabs <- c("1"="Never",
"3"="Sometimes",
"5"="Very often")
ggplot(data=droplevels(subset(fitted.stmony, stmonyw1i %in% c(1,3,5))),
aes(x= .epred, y= stmonyw1i, xmin=.lower, xmax=.upper)) +
facet_wrap(~ .category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
scale_y_continuous(breaks=c(1,3,5),
labels = stress3catlabs, name="Stress: Money") +
scale_x_continuous(breaks=c(0,.1,.2,.3,.4),
labels = dropLeadingZero, name="Posterior expected probability of past crime at T1") +
geom_pointinterval() +
coord_cartesian(xlim=c(0,.4))
Even better. We could combine six more plots like this - one for each
of our stress items - into a single plot to summarize the associations
between stress and crime at T1 in the full sample. But we also have to
estimate these associations in our community subsamples. Perhaps we can
condense things even further. Instead of plotting distributions of
predicted values for specific stress categories, we could instead plot
something akin to a regression beta coefficient, as is found in a
typical forest plot of effect sizes. However, rather than simply
visualizing the model’s outputted b for each stress item
(like we did above after building each model), which summarizes
associations with the latent DV on the log odds scale, we will want to
do some transformations to maximize the meaningfulness or interpretive
utility of our summary plots.
One benefit of specifying the cumulative ordinal distribution for our
IV with brms built-in mo() function is that
the beta coefficient from these brms ordinal IV models is comparable in
interpretation to the beta from a classic logistic regression model - it
represents the average increase in the log odds of the latent DV
associated with a one threshold unit increase in the ordinal IV (see
Buerkner). From this beta, we can calculate a “maximum” effect size by
multiplying the beta coefficient by the number of ordinal IV thresholds
(see McElreath; Kurz), which represents the estimated increase or
decrease in Y associated with increasing from the lowest IV response
category (minimum subjective stress, or Never) to the highest
IV response category (maximum subjective stress, or Very
often). This “max” estimate is useful for interpreting the
theoretical “maximum effect” of a predictor that can be expected given
the data and priors; it is also useful for comparing estimated effects
of variables with different response scales. Such a max transformation
would also be useful as a normalized counterfactual comparison in
interpreting longitudinal change score associations representing the
predicted effect of increasing (decreasing) from minimum to maximum (or
vice versa) stress on changes in the probability of crime. However, as a
counterfactual comparison, it is important to consider whether such a
comparison might be based on implausible out-of-sample predictions -
that is, we may not observe any “maximum” changes from the lowest
(highest) stress response category to the highest (lowest) stress
response category in our two-year panel data.
For now, let’s try it to get a sense of how we might examine a maximum contrast. We’ll start by estimating the “maximum” effect of stress about money at T1 (stmonyw1) on past theft less than 5BAM at T1 (pstthflt5w1), followed by the same maximum contrast estimate for stress about assault (stmugw1).
5.1.4 PP “Maximum” Effect Size of Stress on Past Crime (beta)
Maximum contrast estimate for stress about money at T1 (stmonyw1) and past theft less than 5BAM at T1 (pstthflt5w1).
#https://bookdown.org/content/4857/monsters-and-mixtures.html#ordered-categorical-predictors
#From Kurz:
#"To clarify, the b in this section is what we’re calling βE in the current example and Bürkner and Charpentier used ζi in place of McElreath’s δj. The upshot of all this is that if we’d like to compare the summary of our b12.6 to the results McElreath reported for his m12.6, we’ll need to multiply our moedu_new by 7."
# posterior_samples(allpstcrime.stmony.fit) %>%
# glimpse()
as_draws_df(allpstcrime.stmony.fit) %>%
transmute(bE = bsp_pstthflt5w1f_mostmonyw1i * 4) %>%
median_qi(.width = .95) %>%
mutate_if(is.double, round, digits = 2)## # A tibble: 1 × 6
## bE .lower .upper .width .point .interval
## <dbl> <dbl> <dbl> <dbl> <chr> <chr>
## 1 0.15 -0.7 1.09 0.95 median qiEstimated maximum effect for stress about assault at T1 (stmugw1) and past theft less than 5BAM at T1 (pstthflt5w1).
as_draws_df(allpstcrime.stmug.fit) %>%
transmute(bE = bsp_pstthflt5w1f_mostmugw1i * 4) %>%
median_qi(.width = .95) %>%
mutate_if(is.double, round, digits = 2)## # A tibble: 1 × 6
## bE .lower .upper .width .point .interval
## <dbl> <dbl> <dbl> <dbl> <chr> <chr>
## 1 0.33 -0.61 1.24 0.95 median qi5.1.5 PP “Maximum” Effect Size of Stress on Past Crime (5-1 category contrast pred margins)
As above, we could continue to generate posterior samples for every stress/crime item combination, then calculate predicted probabilities and their differences for certain contrasts (e.g., stress = 1 versus = 5).
However, another way to approach summarizing each association would
be to estimate and plot the predictive margins, or the predicted
probabilities for each crime item, at all response levels of each stress
item, from the current model fits. We can use the
marginal_effects function from brms (and
tidybayes) to do this.
Note that the name is a bit misleading. This function indeed
calculates predictive margins on the outcome scale by stress category in
a bivariate model like ours. However, when additional covariates are
included in the model, the marginal effects function does
not output marginalized predictions used to calculate average
marginal effects (AME) - that is, predictions are not marginalized
over all values of other covariates. Rather, the function outputs
conditional predicted probabilities, or conditional predictive margins,
which are used to calculate the simpler marginal effects at
representative values (MER) instead. (For more information, see Andrew
Heiss’s excellent explanation.) So, when we add additional
predictors to our model, we will want to change to an alternative
approach for estimating average marginal effects, such as with the
“marginaleffects” package.
#marginal_effects command technically outputs conditional effects,
# http://paul-buerkner.github.io/brms/reference/marginal_effects.html
# aka marginal effects at representative values (MERs)
# not an issue in bivariate model w/o covariates
# BUT, when those are added, ideally we would marginalize out effects of covars
# by averaging effect of IV "A" on y (E[y|A] across all values of covars B, C, etc.
# See also: https://github.com/paul-buerkner/brms/issues/552
# http://htmlpreview.github.io/?https://github.com/mjskay/uncertainty-examples/blob/master/marginal-effects_categorical-predictor.html
## plot all marginal effects
p.mer.mony.pstcrim <- plot(conditional_effects(allpstcrime.stmony.fit), ask = FALSE)
p.mer.mony.pstcrim$pstthflt5w1f.pstthflt5w1f_stmonyw1i +
p.mer.mony.pstcrim$pstthfgt5w1f.pstthfgt5w1f_stmonyw1i +
p.mer.mony.pstcrim$pstthreatw1f.pstthreatw1f_stmonyw1i +
p.mer.mony.pstcrim$pstharmw1f.pstharmw1f_stmonyw1i +
p.mer.mony.pstcrim$pstusedrgw1f.pstusedrgw1f_stmonyw1i +
p.mer.mony.pstcrim$psthackw1f.psthackw1f_stmonyw1i +
plot_layout(ncol = 3)
plot_annotation(
title = 'MER, Stress abt Money on Past Crime')## $title
## [1] "MER, Stress abt Money on Past Crime"
##
## $subtitle
## NULL
##
## $caption
## NULL
##
## $tag_levels
## NULL
##
## $tag_prefix
## NULL
##
## $tag_suffix
## NULL
##
## $tag_sep
## NULL
##
## $theme
## Named list()
## - attr(*, "class")= chr [1:2] "theme" "gg"
## - attr(*, "complete")= logi FALSE
## - attr(*, "validate")= logi TRUE
##
## attr(,"class")
## [1] "plot_annotation"# For AMEs intead (equivalent w/out covariates), try:
# allpstcrime.stmony.mfx <- marginaleffects(allpstcrime.stmony.fit)
# summary(allpstcrime.stmony.mfx)
# save posterior draws for plotting
# allpstcrime.stmony.mfx <- marginaleffects(allpstcrime.stmony.fit) %>%
# posteriordraws()
# or for AME manual approach with tidybayes compare_levels, see:
# https://htmlpreview.github.io/?https://github.com/mjskay/uncertainty-examples/blob/master/marginal-effects_categorical-predictor.html#differences-in-amesAlternatively, we could calculate and display the posterior distribution for the difference in expected proportions between two specific categories of stress. To do this, we might begin by plotting the posterior distribution of expected values for the probability of each crime item, or the posterior distributions for predictive margins, for each stress response category. Then, as mentioned before, we might narrow our focus to estimating and visualizing the “maximum” marginal effect by contrasting expected probabilities from the lowest (1=never) and highest (5=very often) stress categories. We can follow by taking the posterior difference distribution for these two contrasts.
First, let’s plot the posterior predictions for all five stress response categories and all six past crime items. These plots are inspired by Matthew Kay’s impressive efforts here.
plotpredmarg_stmony_pstcrim_T1 = stress.wide3 %>%
data_grid(stmonyw1i) %>% # condition on everything
add_epred_draws(allpstcrime.stmony.fit, value = "E[y|stmonyw1i]") %>% # get conditional expectations
ggplot(aes(y = paste0("E[y|stmonyw1i = ", stmonyw1i, "]"), x = `E[y|stmonyw1i]`, fill = stmonyw1i)) +
stat_halfeye() +
xlim(0, .35) +
facet_grid(rows=vars(.category), cols=NULL,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
ylab(NULL) +
xlab("E[y|stmonyw1i], bivariate") +
geom_vline(xintercept = c(0, 1))
plotpredmarg_stmony_pstcrim_T1
As the plot shows, the predicted probability of past crime does not appear to vary much across the stress about money response categories. As such, we would expect the contrasts, or marginal effects, to be very small, with posterior predictive difference distributions centered close to and spanning “zero” difference in predicted probability of crime. Let’s begin by plotting all possible contrasts (e.g., 1=never vs. 2=rarely; 1=never* vs. 3=sometimes, etc.). Then we can narrow in on specific meaningful contrasts, such as 1=never vs. 5=very often representing the “maximum marginal effect” of the stress about money item on past crime.
#updated tidybayes to 3.0 & redone using add_epred_draws instead of add_fitted_draws
#use efficient & transparent caching
# https://bookdown.org/yihui/rmarkdown-cookbook/cache-rds.html
predmarg_stmony_pstcrim_T1 = stress.wide3 %>%
data_grid(stmonyw1i) %>% # get draws for each stress item response cat
# add_fitted_draws(allpstcrime.stmony.fit) %>% # add_fitted_draws depricated in tidybayes 3.0
add_epred_draws(allpstcrime.stmony.fit) %>% # getting expected values by stress cat
group_by(.category, stmonyw1i, .draw) # grouping by crime type (multivar DVs)
# we can use compare_levels to calculate the mean difference
# (changed .value to .epred after updating to add_epred_draws)
margeffs_stmony_pstcrim_T1 = xfun::cache_rds({
predmarg_stmony_pstcrim_T1 %>%
compare_levels(.epred, by = stmonyw1i) %>% # pairwise differences in `E[y|A]`, by levels of A
rename(`difference in E[y|stress]` = `.epred`) # give this column a more accurate name
}, file = "cache_4_1a")
# plot all stress response category contrasts (10)
margeffs_stmony_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = stmonyw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
As expected, these marginal effect contrasts all center around zero.
Focusing on the maximum marginal effect, or the “5-1” contrast, seemed like a good idea for interpretive reasons. After all, if we want to ensure that stress process theories are given the benefit of doubt, then the maximum marginal effect estimate seems like a good place to start. However, as the figure above shows, this contrast also has the widest uncertainty intervals. This should be expected considering how uncertainty bands around the predictive margins increased at the ends of the stress response distribution in our “MER” line plots above.
On its face, a disperse uncertainty interval around the maximum marginal effect, hereafter “MaxME”, essentially tells us that a wide range of predictive differences in crime probabilities are plausible - in this case, including differences in negative and positive directions - between those who reportedly never versus very often stress about money. This is perhaps unsurprising given the relative rarity of our crime outcomes; it might also be a quite appropriate, informative, and parsimonious summary of the very weak or null magnitudes observed for these stress/past crime associations. For now, let’s focus on the “MaxME” contrast (in our case, the “5-1” contrast) as planned. We can revisit this decision if we deem the contrast to be problematic (e.g., exhibiting exaggerated uncertainty relative to other contrasts) or otherwise non-representative for the associations we are summarizing and then consider a shift to more complex alternatives.
#MME short for MaxME
MME_stmony_pstcrim_T1 = xfun::cache_rds({
predmarg_stmony_pstcrim_T1 %>%
filter(stmonyw1i=="1" | stmonyw1i=="5") %>%
compare_levels(.epred, by = stmonyw1i) %>% # pairwise differences in `E[y|A]`, by levels of A
rename(`difference in E[y|stmonyw1i]` = `.epred`, 'MME' = 'stmonyw1i') # rename cols
}, file="cache_4_2")
#NOTE: may need t explicitly call dplyr::filter
#see: https://hackernoon.com/silly-r-errors-and-the-silly-reasons-im-probably-getting-them-c6bd9ada59c
# plot only maximum stress response category contrast, or MME (5-1)
MME_stmony_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stmonyw1i]`, y = 'MME')) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.25, .25) +
ylab("Stress:\nMoney") +
coord_cartesian(xlim=c(-.15,.15))
# alt plot - stat_lineinterval (change label: 'MME' = `5 - 1`)
# MME_stmony_pstcrim_T1 %>%
# ggplot(aes(x = `difference in E[y|stmonyw1i]`, y = 'MME')) +
# facet_wrap(~.category, nrow=1,
# labeller = labeller(.category= as_labeller(pstcrimelabs))) +
# stat_pointinterval(.width=.95) +
# geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.25, .25) +
# ylab("Stress:\nMoney") +
# coord_cartesian(xlim=c(-.15,.15)) This is what a MaxME effect estimate looks like with our ordinal stress items. Again, this is a parsimonious summary of the “maximum” marginal association between an ordinal (stress) IV and a binary outcome.
We could add additional rows for each of the remaining stress items. However, earlier, I expressed reservations about the utility of this effect estimate in our particular application. Thinking ahead, this effect estimate might be especially problematic when we examine change correlations from T1 to T2. After all, as Figure 1 shows, within-person increases or decreases in stress from T1 to T2 (across a 2-yr interval) are rarely as large or larger than two categories (e.g., from 1=never to 3=sometimes, or from 2=rarely to 4=often, or from 3=sometimes to 5=very often). Rather, most people report the same level of stress or a one-threshold response category change, which can be seen in Figure 1 above and in the bar charts below.
p1 <- ggplot(stress.wide, aes(stmonych12)) +
geom_bar()
p2 <- ggplot(stress.wide, aes(sttranch12)) +
geom_bar()
p3 <- ggplot(stress.wide, aes(strespch12)) +
geom_bar()
p4 <- ggplot(stress.wide, aes(stfairch12)) +
geom_bar()
p5 <- ggplot(stress.wide, aes(stjobch12)) +
geom_bar()
p6 <- ggplot(stress.wide, aes(stthftch12)) +
geom_bar()
p7 <- ggplot(stress.wide, aes(stmugch12)) +
geom_bar()
(p1 + p2) / (p3 + p4) / (p5 + plot_spacer()) / (p6 + p7)
5.2 Practically Large Marginal Effects (PLME)
Given this, rather than focusing on “maximum” effects, it might be wiser to select instead an effect estimate that observes such plausible real-world limits to the types of large yet actually-occurring changes that might be observed in nature. Since, in our particular case, anything larger than a one-category change in subjective stress across waves appears uncommon regardless of stress type, then attempts to estimate “maximum” effects or other large-difference category contrasts may result in especially noisy, uncertain, or incomparable estimates when we move to modeling change later. Rather, an alternative standard one-category contrast such as “4-3” might be better for our purposes, since the overwhelming majority of changes in stress from T1 to T2 actually occurring in our data are one-category changes. Meanwhile, a two-category change such as “4-2” might represent a “practically large” stress contrast, given that this is the largest change observed for most stress items across waves. Even better, we could calculate a combined/average posterior predictive difference distribution for multiple contrasts - for instance, an average of all two-unit response category jumps, including “5-3”, “4-2”, and “3-1” - to estimate an average difference in expected crime probability across stress differences of a magnitude that are in line with the largest within-person changes in stress observed from T1 to T2 in these data.
We might refer to the process of determining such an estimated average difference in probabilities across all one-unit or two-unit contrasts as akin to identifying an appropriate “practically large marginal effect” (PLME), as it represents an estimate of the potential (theoretically causal) effect of increasing or decreasing an IV (stress) up or down a large amount relative to change magnitudes that are typically observed in nature.
In our case, we will start with the average of one-category contrasts to permit comparison with a two-category PLME. The average one-category contrast corresponds to typical regression-style interpretations, since it is interpreted as the average difference in the posterior predicted probability of crime across all one-category contrasts in stress. We will then move to the average two-category PLME estimates. Note that when modeling other theorized causes, the identified PLME contrast might be different; for instance, if three-category changes in another predictor variable representing a theorized cause were more commonly observed in one’s data, then an average of all three-category contrasts might be identified as an appropriate “practically large marginal effect” for that variable.
In addition to its interpretive appeal, a “PLME” estimate might have better or more robust distributional properties across all the stress/outcome item contrasts we will be making (e.g., six more stress items; criminal intent & negative emotions; repeating all with T2-T1 change correlations), as it might be less noisy than “maximum” contrasts where X or Y item distributions or joint XY distributions might result in noisy estimates due, for instance, to low joint cell frequencies. Let’s give it a shot.
First, we will generate data containing all 10 posterior contrasts.
Also, before plotting these together, remember that “any” past crime item? We also want to pull contrasts from those models and combine them with contrasts from the individual criminal intent outcome models so we can add them to Figure 3 later. Let’s do that now as well.
5.2.1 Predictive Marginal Contrasts (Past Crime)
#Get predictive margins data
#could not get a function to work, so typing all separately :/
#NOTE: stmonyw1i draws saved above, but redone to add all past crime
predmarg_stmony_pstcrim_T1 = stress.wide3 %>%
data_grid(stmonyw1i) %>%
add_epred_draws(allpstcrime.stmony.fit) %>%
group_by(.category, stmonyw1i, .draw)
predmarg_sttran_pstcrim_T1 = stress.wide3 %>%
data_grid(sttranw1i) %>%
add_epred_draws(allpstcrime.sttran.fit) %>%
group_by(.category, sttranw1i, .draw)
predmarg_stresp_pstcrim_T1 = stress.wide3 %>%
data_grid(strespw1i) %>%
add_epred_draws(allpstcrime.stresp.fit) %>%
group_by(.category, strespw1i, .draw)
predmarg_stfair_pstcrim_T1 = stress.wide3 %>%
data_grid(stfairw1i) %>%
add_epred_draws(allpstcrime.stfair.fit) %>%
group_by(.category, stfairw1i, .draw)
predmarg_stjob_pstcrim_T1 = stress.wide3 %>%
data_grid(stjobw1i) %>%
add_epred_draws(allpstcrime.stjob.fit) %>%
group_by(.category, stjobw1i, .draw)
predmarg_stthft_pstcrim_T1 = stress.wide3 %>%
data_grid(stthftw1i) %>%
add_epred_draws(allpstcrime.stthft.fit) %>%
group_by(.category, stthftw1i, .draw)
predmarg_stmug_pstcrim_T1 = stress.wide3 %>%
data_grid(stmugw1i) %>%
add_epred_draws(allpstcrime.stmug.fit) %>%
group_by(.category, stmugw1i, .draw)
#function to get epred draws from "any past crime" models
epred_anycrm <- function(mydata, mystressvar, myoutcomevar, mymodelfit) {
mydata %>%
data_grid({{mystressvar}}) %>%
mutate(.category=myoutcomevar) %>%
add_epred_draws(mymodelfit) %>%
group_by({{mystressvar}}, .draw)
}
#get "any past crime" epred draws & merge w/individual outcome epred data
predmarg_stmony_anypstcrim_T1 =epred_anycrm(stress.wide3, stmonyw1i,
"pstanyw1f", anypstcrime.stmony.fit)
predmarg_stmony_pstcrim_T1 <- bind_rows(predmarg_stmony_pstcrim_T1,
predmarg_stmony_anypstcrim_T1)
rm(predmarg_stmony_anypstcrim_T1) #clean environment
predmarg_sttran_anypstcrim_T1 =epred_anycrm(stress.wide3, sttranw1i,
"pstanyw1f", anypstcrime.sttran.fit)
predmarg_sttran_pstcrim_T1 <- bind_rows(predmarg_sttran_pstcrim_T1,
predmarg_sttran_anypstcrim_T1)
rm(predmarg_sttran_anypstcrim_T1)
predmarg_stresp_anypstcrim_T1 =epred_anycrm(stress.wide3, strespw1i,
"pstanyw1f", anypstcrime.stresp.fit)
predmarg_stresp_pstcrim_T1 <- bind_rows(predmarg_stresp_pstcrim_T1,
predmarg_stresp_anypstcrim_T1)
rm(predmarg_stresp_anypstcrim_T1)
predmarg_stfair_anypstcrim_T1 =epred_anycrm(stress.wide3, stfairw1i,
"pstanyw1f", anypstcrime.stfair.fit)
predmarg_stfair_pstcrim_T1 <- bind_rows(predmarg_stfair_pstcrim_T1,
predmarg_stfair_anypstcrim_T1)
rm(predmarg_stfair_anypstcrim_T1)
predmarg_stjob_anypstcrim_T1 =epred_anycrm(stress.wide3, stjobw1i,
"pstanyw1f", anypstcrime.stjob.fit)
predmarg_stjob_pstcrim_T1 <- bind_rows(predmarg_stjob_pstcrim_T1,
predmarg_stjob_anypstcrim_T1)
rm(predmarg_stjob_anypstcrim_T1)
predmarg_stthft_anypstcrim_T1 =epred_anycrm(stress.wide3, stthftw1i,
"pstanyw1f", anypstcrime.stthft.fit)
predmarg_stthft_pstcrim_T1 <- bind_rows(predmarg_stthft_pstcrim_T1,
predmarg_stthft_anypstcrim_T1)
rm(predmarg_stthft_anypstcrim_T1)
predmarg_stmug_anypstcrim_T1 =epred_anycrm(stress.wide3, stmugw1i,
"pstanyw1f", anypstcrime.stmug.fit)
predmarg_stmug_pstcrim_T1 <- bind_rows(predmarg_stmug_pstcrim_T1,
predmarg_stmug_anypstcrim_T1)
rm(predmarg_stmug_anypstcrim_T1)
#function to calculate mean difference contrast (marginal effect contrasts, or MEs)
calc_MEs <- function(predmarg_data, xitem) {
predmarg_data %>%
compare_levels(.epred, by = xitem) %>%
rename(`difference in E[y|stress]` = `.epred`)
}
#Warning - takes a while to generate these ME contrasts...
margeffs_stmony_pstcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stmony_pstcrim_T1, "stmonyw1i")}, file="cache_4_1")
margeffs_sttran_pstcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_sttran_pstcrim_T1, "sttranw1i")}, file="cache_4_3")
margeffs_stresp_pstcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stresp_pstcrim_T1, "strespw1i")}, file="cache_4_4")
margeffs_stfair_pstcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stfair_pstcrim_T1, "stfairw1i")}, file="cache_4_5")
margeffs_stjob_pstcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stjob_pstcrim_T1, "stjobw1i")}, file="cache_4_6")
margeffs_stthft_pstcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stthft_pstcrim_T1, "stthftw1i")}, file="cache_4_7")
margeffs_stmug_pstcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stmug_pstcrim_T1, "stmugw1i")}, file="cache_4_8")Now, let’s examine the plots for all ten possible effect contrasts again for stmonyw1 and the remaining six stress items.
5.2.2 Visualize all contrasts
5.2.2.0.1 Stress about money
# plot all stress response category contrasts (10)
pstcrimelabs <- c("pstthflt5w1f"="Theft <5BAM",
"pstthfgt5w1f"="Theft >5BAM",
"pstthreatw1f"="Threaten",
"pstharmw1f"="Phys. harm",
"pstusedrgw1f"="Use drugs",
"psthackw1f"="Hack info.",
"pstanyw1f"="Any crime")
#Money - all 10 contrasts
margeffs_stmony_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = stmonyw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
5.2.2.0.2 Stress about transportation
#Transportation - all 10 contrasts
margeffs_sttran_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = sttranw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
5.2.2.0.3 Stress about respect
#Respect - all 10 contrasts
margeffs_stresp_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = strespw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
5.2.2.0.4 Stress about fair treatment
#Fair - all 10 contrasts
margeffs_stfair_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = stfairw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
5.2.2.0.5 Stress about job
#Job - all 10 contrasts
margeffs_stjob_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = stjobw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
5.2.2.0.6 Stress about theft
#Theft - all 10 contrasts
margeffs_stthft_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = stthftw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
5.2.2.0.7 Stress about assault
#Assault - all 10 contrasts
margeffs_stmug_pstcrim_T1 %>%
ggplot(aes(x = `difference in E[y|stress]`, y = stmugw1i)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabs))) +
stat_halfeye() +
geom_vline(xintercept = 0, linetype = "dashed") +
xlim(-.5, .5) +
coord_cartesian(xlim=c(-.15,.15))
These plots give us a full sense of the associations between each stress item and each past crime item by showing us the predicted differences in outcome probabilities across all possible stress item response contrasts. Remember, the goal is to get an informative yet more parsimonious summary effect estimate on the response scale (like these). We also want a summary that will permit consistent interpretation across these T1 associations and T2-T1 change associations. Finally, with respect to change correlations, we want a counterfactual effect estimate that respects real-world boundaries on the types of within-person changes that we can and do observe.
Still, this is too much information for our task; we need a meaningful yet more parsimonious summary of the predicted marginal effect contrasts. Toward this end, we start by plotting the averaged posterior predicted marginal effect for all one-category contrasts (5-4; 4-3; 3-2; 2-1) below.
5.2.3 ME1 - Limit Predictive Margins to 1-Category Contrasts
#Create diff data for each stress item (note, run everything before the "slice" command to check logic)
#stmony
PLME1_stmony_pstcrim_T1 = margeffs_stmony_pstcrim_T1 %>%
filter(stmonyw1i=="5 - 4" | stmonyw1i=="4 - 3" | stmonyw1i=="3 - 2" | stmonyw1i=="2 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_onedif = mean(`difference in E[y|stress]`), #average selected contrasts
stress_var = "Stress:\nMoney",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>% # should have four identical average values for all contrasts, slice to keep first line only
ungroup() %>%
dplyr::select(-c(stmonyw1i, `difference in E[y|stress]`)) #drop first contrast, keep ave (mean_onediff)
#sttran
PLME1_sttran_pstcrim_T1 = margeffs_sttran_pstcrim_T1 %>%
filter(sttranw1i=="5 - 4" | sttranw1i=="4 - 3" | sttranw1i=="3 - 2" | sttranw1i=="2 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_onedif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTransport",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(sttranw1i, `difference in E[y|stress]`))
#stresp
PLME1_stresp_pstcrim_T1 = margeffs_stresp_pstcrim_T1 %>%
filter(strespw1i=="5 - 4" | strespw1i=="4 - 3" | strespw1i=="3 - 2" | strespw1i=="2 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_onedif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nRespect",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(strespw1i, `difference in E[y|stress]`))
#stfair
PLME1_stfair_pstcrim_T1 = margeffs_stfair_pstcrim_T1 %>%
filter(stfairw1i=="5 - 4" | stfairw1i=="4 - 3" | stfairw1i=="3 - 2" | stfairw1i=="2 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_onedif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nFair Trtmt",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stfairw1i, `difference in E[y|stress]`))
#stjob
PLME1_stjob_pstcrim_T1 = margeffs_stjob_pstcrim_T1 %>%
filter(stjobw1i=="5 - 4" | stjobw1i=="4 - 3" | stjobw1i=="3 - 2" | stjobw1i=="2 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_onedif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nJob",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stjobw1i, `difference in E[y|stress]`))
#stthft
PLME1_stthft_pstcrim_T1 = margeffs_stthft_pstcrim_T1 %>%
filter(stthftw1i=="5 - 4" | stthftw1i=="4 - 3" | stthftw1i=="3 - 2" | stthftw1i=="2 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_onedif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTheft Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stthftw1i, `difference in E[y|stress]`))
#stmug
PLME1_stmug_pstcrim_T1 = margeffs_stmug_pstcrim_T1 %>%
filter(stmugw1i=="5 - 4" | stmugw1i=="4 - 3" | stmugw1i=="3 - 2" | stmugw1i=="2 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_onedif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nAssault Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmugw1i, `difference in E[y|stress]`))
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
onedif_combined <- bind_rows(
PLME1_stmony_pstcrim_T1,
PLME1_sttran_pstcrim_T1,
PLME1_stresp_pstcrim_T1,
PLME1_stfair_pstcrim_T1,
PLME1_stjob_pstcrim_T1,
PLME1_stthft_pstcrim_T1,
PLME1_stmug_pstcrim_T1)
#transform stress_var into reverse-ordered factor
onedif_combined2<- onedif_combined %>%
mutate(
stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney"))
)
#calculate row & col averages of P(mean_onedif) > 0
#Rows: P(mean_onedif | Stress item) > 0
#aka P(PLME>0|stress)
#View marginal probabilities
# onedif_combined2 %>%
# group_by(stress_varf) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_onedif>0),
# p_gt0 = n_gt0 / n_ests)
# Stress:\nMoney 0.54
# Stress:\nTransport 0.56
# Stress:\nRespect 0.90
# Stress:\nFair Trtmt 0.93
# Stress:\nJob 0.89
# Stress:\nTheft Vctm 0.78
# Stress:\nAssault Vctm 0.85
#Cols: P(mean_onedif | Crime item) > 0
#aka P(PLME>0|crime)
#View marginal probabilities
# onedif_combined2 %>%
# group_by(.category) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_onedif>0),
# p_gt0 = n_gt0 / n_ests)
# pstthflt5w1f 0.82
# pstthfgt5w1f 0.74
# pstthreatw1f 0.85
# pstharmw1f 0.71
# pstusedrgw1f 0.86
# psthackw1f 0.66
# pstanyw1f 0.83
#Add these posterior probabilities into variable name
pstcrimelabsPT1 <- c(
"pstthflt5w1f"="Theft <5BAM\nP(ME>0|col)\n=.82",
"pstthfgt5w1f"="Theft >5BAM\nP(ME>0|col)\n=.74",
"pstthreatw1f"="Threaten\nP(ME>0|col)\n=.85",
"pstharmw1f"="Phys. harm\nP(ME>0|col)\n=.71",
"pstusedrgw1f"="Use drugs\nP(ME>0|col)\n=.86",
"psthackw1f"="Hack info.\nP(ME>0|col)\n=.66",
"pstanyw1f"="Any crime\nP(ME>0|col)\n=.83")
stress_varlabsPT1 <- c(
"Stress:\nMoney" = "Stress: Money\nP(ME>0|row)\n=.54",
"Stress:\nTransport" = "Stress: Transport\nP(ME>0|row)\n=.55",
"Stress:\nRespect" = "Stress: Respect\nP(ME>0|row)=.89",
"Stress:\nFair Trtmt" = "Stress: Fair Trtmt\nP(ME>0|row)\n=.92",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.87",
"Stress:\nTheft Vctm" = "Stress: Theft\nP(ME>0|row)\n=0.78",
"Stress:\nAssault Vctm" = "Stress: Assault\nP(ME>0|row)\n=0.85")
# levels(twodif_combined2$stress_varf)
#Plot:
Fig3vME1 <- ggplot(onedif_combined2, aes(x = mean_onedif, y = stress_varf,
fill=after_stat(x > 0))) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabsPT1))) +
stat_halfeye(.width = .95) +
geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.5, .5) +
coord_cartesian(xlim=c(-.11,.11)) +
scale_x_continuous(breaks=c(-.1,0,.1)) +
xlab("Posterior Predicted Difference in E[y|stress]") +
scale_y_discrete(labels=stress_varlabsPT1) +
scale_fill_manual(values = c("gray80", "#C34C4A")) +
plot_annotation(
title = 'FIGURE 3vME1: \"Marginal Effect\" of 1-Category Difference in Stress on Past Crime at T1, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves.') &
theme(axis.title.y = element_blank(),
legend.position = "none",
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0))
Fig3vME1 
That is a good start. We managed to get all 42 posterior predicted difference distributions on one figure, and yet it remains relatively easy to read. But we also need to examine and communicate differences in these differences across communities. We will turn to that soon. First, let’s compare this plot with an alternative plot that presents our “practically large marginal effect” or PLME, estimated as the average expected difference in crime across all two-category contrasts (i.e., 5-3, 4-2, & 3-1) instead of all one-category contrasts.
Of course, since the contrast is larger (i.e. a two-unit difference instead of one-unit difference in stress), then plotted associations between stress and crime should display larger point estimates or expected differences in predicted probabilities of past crime. But what about the uncertainty intervals? How might those change? What about the marginal probabilities, or the average proportion of the posterior distributions that are greater than zero? We can answer these questions by extracting, averaging, and plotting all two-category contrasts instead.
5.2.4 ME2 - Combine Predictive Margins from 2-Category Contrasts (“PLME”)
#Create diff data for each stress item (note, run everything before the "slice" command to check logic)
#stmony
PLME2_stmony_pstcrim_T1 = margeffs_stmony_pstcrim_T1 %>%
filter(stmonyw1i=="5 - 3" | stmonyw1i=="4 - 2" | stmonyw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nMoney",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmonyw1i, `difference in E[y|stress]`))
#sttran
PLME2_sttran_pstcrim_T1 = margeffs_sttran_pstcrim_T1 %>%
filter(sttranw1i=="5 - 3" | sttranw1i=="4 - 2" | sttranw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTransport",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(sttranw1i, `difference in E[y|stress]`))
#stresp
PLME2_stresp_pstcrim_T1 = margeffs_stresp_pstcrim_T1 %>%
filter(strespw1i=="5 - 3" | strespw1i=="4 - 2" | strespw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nRespect",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(strespw1i, `difference in E[y|stress]`))
#stfair
PLME2_stfair_pstcrim_T1 = margeffs_stfair_pstcrim_T1 %>%
filter(stfairw1i=="5 - 3" | stfairw1i=="4 - 2" | stfairw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nFair Trtmt",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stfairw1i, `difference in E[y|stress]`))
#stjob
PLME2_stjob_pstcrim_T1 = margeffs_stjob_pstcrim_T1 %>%
filter(stjobw1i=="5 - 3" | stjobw1i=="4 - 2" | stjobw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nJob",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stjobw1i, `difference in E[y|stress]`))
#stthft
PLME2_stthft_pstcrim_T1 = margeffs_stthft_pstcrim_T1 %>%
filter(stthftw1i=="5 - 3" | stthftw1i=="4 - 2" | stthftw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTheft Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stthftw1i, `difference in E[y|stress]`))
#stmug
PLME2_stmug_pstcrim_T1 = margeffs_stmug_pstcrim_T1 %>%
filter(stmugw1i=="5 - 3" | stmugw1i=="4 - 2" | stmugw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nAssault Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmugw1i, `difference in E[y|stress]`))
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
twodif_combined <- bind_rows(
PLME2_stmony_pstcrim_T1,
PLME2_sttran_pstcrim_T1,
PLME2_stresp_pstcrim_T1,
PLME2_stfair_pstcrim_T1,
PLME2_stjob_pstcrim_T1,
PLME2_stthft_pstcrim_T1,
PLME2_stmug_pstcrim_T1)
#transform stress_var into reverse-ordered factor
twodif_combined2<- twodif_combined %>%
mutate(
stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney")),
.category = factor(.category,
levels=c("pstthflt5w1f", "pstthfgt5w1f", "pstthreatw1f",
"pstharmw1f", "pstusedrgw1f", "psthackw1f",
"pstanyw1f"))
)
#calculate row & col averages of P(mean_twodif) > 0
#Rows: P(mean_twodif | Stress item) > 0
#aka P(PLME>0|stress)
#View marginal contrasts
# twodif_combined2 %>%
# group_by(stress_varf) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_twodif>0),
# p_gt0 = n_gt0 / n_ests)
# Stress:\nMoney 0.54
# Stress:\nTransport 0.56
# Stress:\nRespect 0.90
# Stress:\nFair Trtmt 0.93
# Stress:\nJob 0.89
# Stress:\nTheft Vctm 0.78
# Stress:\nAssault Vctm 0.85
#Cols: P(mean_twodif | Crime item) > 0
#aka P(PLME>0|crime)
# twodif_combined2 %>%
# group_by(.category) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_twodif>0),
# p_gt0 = n_gt0 / n_ests)
# pstthflt5w1f 0.82
# pstthfgt5w1f 0.74
# pstthreatw1f 0.85
# pstharmw1f 0.71
# pstusedrgw1f 0.86
# psthackw1f 0.66
# pstanyw1f 0.83
#Add these posterior probabilities into variable name
pstcrimelabsPT1 <- c(
"pstthflt5w1f"="Theft <5BAM\nP(ME>0|col)\n=.82",
"pstthfgt5w1f"="Theft >5BAM\nP(ME>0|col)\n=.74",
"pstthreatw1f"="Threaten\nP(ME>0|col)\n=.85",
"pstharmw1f"="Phys. harm\nP(ME>0|col)\n=.71",
"pstusedrgw1f"="Use drugs\nP(ME>0|col)\n=.86",
"psthackw1f"="Hack info.\nP(ME>0|col)\n=.66",
"pstanyw1f"="Any crime\nP(ME>0|col)\n=.83")
stress_varlabsPT1 <- c(
"Stress:\nMoney" = "Stress: Money\nP(ME>0|row)\n=.54",
"Stress:\nTransport" = "Stress: Transport\nP(ME>0|row)\n=.56",
"Stress:\nRespect" = "Stress: Respect\nP(ME>0|row)=.90",
"Stress:\nFair Trtmt" = "Stress: Fair Trtmt\nP(ME>0|row)\n=.93",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.89",
"Stress:\nTheft Vctm" = "Stress: Theft\nP(ME>0|row)\n=0.78",
"Stress:\nAssault Vctm" = "Stress: Assault\nP(ME>0|row)\n=0.85")
# levels(twodif_combined2$stress_varf)
#Plot:
Fig3vPLME2 <- ggplot(twodif_combined2, aes(x = mean_twodif, y = stress_varf, fill=after_stat(x > 0))) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabsPT1))) +
stat_halfeye(.width = .95) +
geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.5, .5) +
coord_cartesian(xlim=c(-.11,.11)) +
scale_x_continuous(breaks=c(-.1,0,.1)) +
xlab("Posterior Predicted Difference in E[y|stress]") +
scale_y_discrete(labels=stress_varlabsPT1) +
scale_fill_manual(values = c("gray80", "#C34C4A")) +
plot_annotation(
title = 'FIGURE 3v2: \"Practically Large Marginal Effect\" of 2-Category Difference in Stress on Past Crime at T1, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves.') &
theme(axis.title.y = element_blank(),
legend.position = "none",
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0))
Fig3vPLME2
As expected, the two-category PLME contrasts generated larger estimates of the difference in predicted probabilities of crime (i.e., point estimates further to the right of the dashed line at zero). However, these estimates also are proportionally more uncertain as well, as the row and column marginal posterior probabilities (that PLME2|stress item]>0 or PLME2|crime item]>0 ) remain the same in this case.
Now, let’s build upon this two-category PLME plot (Fig3vPLME2) and modify it a bit by adding shading to help communicate the magnitude of these differences using an alpha transparency scale linked to the probability that the difference is greater than zero for each posterior (i.e., the proportion of the posterior density to the right of zero).
5.3 Supp. Figure 2A: PLME Stress/Past Crime T1
#First, add p_gt0 variable used above to each stress/item combo in data (i.e. to each plot)
#Create alpha_scale variable =1 if p_gt0 < 0 (i.e., to be fully opaque) & =p_gt0 if p_gt0 > 1
twodif_combined3 <- twodif_combined2 %>%
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(mean_twodif>0),
p_gt0 = n_gt0 / n_ests,
gt0 = ifelse(mean_twodif>0, 1, 0),
alpha_scale = ifelse(mean_twodif <=0, 1, p_gt0),
rural.ses.med = as.factor("0")) #add to combine figures 3 & 4 later
#Using sequential color palette (scio = lajolla)
#Also note I'm using ggnewscale to add multiple fill palettes
SuppFigure2A <- ggplot(data = twodif_combined3, mapping = aes(x = mean_twodif, y = stress_varf)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(pstcrimelabsPT1))) +
stat_slab(mapping = aes(fill = p_gt0), .width = .95) +
scale_fill_scico(palette = "lajolla", begin = .8, end = .3, #tells it where to start and end palette
name = "P(PLME > 0)", breaks = c(.1, .5, .9)) +
new_scale_fill() + #from ggnewscale
stat_halfeye(mapping = aes(fill = after_stat(x > 0)), .width = .95, show.legend=FALSE) +
scale_fill_manual(values = c("grey80", "NA")) +
geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.5, .5) +
coord_cartesian(xlim=c(-.11,.16)) +
scale_x_continuous(breaks=c(-.1,0,.1)) +
xlab("Posterior Predicted Difference in E[y|stress]") +
scale_y_discrete(labels=stress_varlabsPT1) +
labs(
title = 'SUPPLEMENTAL FIGURE 2A\nMarginal Effect of 2-Category Difference in Stress on Past Crime at T1, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. "PLME" refers to "Practically Large Marginal Effect" of 2-category difference in stress on outcome. Darker shaded\nregions indicate larger proportion of posterior effect estimates are greater than zero (positive effect is more probable); grey shading indicates portion of posterior distribution of effect\nestimates equal to or less than zero.') +
theme(axis.title.y = element_blank(),
axis.text.y = element_text(size=10),
strip.text.x = element_text(size=10),
legend.position = "bottom",
strip.background = element_blank(),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
plot.title.position = "plot",
plot.caption.position = "plot")
SuppFigure2A
#Export to image
ggsave("SuppFigure2A.jpeg", width=9, height=6.5, path=here("Output"))At the survey design stage, we included two indicators each for subjective stress from financial (money; transportation), interpersonal (respect; fair treatment), and victimization (theft; assault) sources, along with an item tapping job-related stress. Likewise, these T1 stress/past crime item correlations largely show comparable patterns for the item pairs tapping financial, interpersonal, and victimization stress, respectively.
Overall, these correlational patterns at T1 suggest modest associations between interpersonal stress items and five of six past crime items, excluding identity fraud. Job stress is modestly associated with past minor theft, verbal assault (threaten harm), drug use, and identity fraud (hack info). In contrast to theoretical expectations, victimization stress displays relatively weak and inconsistent associations with past crime at T1. Finally, in contrast to classic strain-based theorizing (cf. Merton; inflation-based arguments) and popular perceptions about the stress/crime relationship, financial stress is not associated with past crime at T1 in this sample.
Before examining whether these associations vary by community characteristics (rurality & SES), let’s repeat this process and generate a comparable figure for criminal intent.
5.3.1 Predictive Marginal Contrasts (Crim Intent)
#Get predictive margins data
predmarg_stmony_prjcrim_T1 = stress.wide3 %>%
data_grid(stmonyw1i) %>%
add_epred_draws(allprjcrime.stmony.fit) %>%
group_by(.category, stmonyw1i, .draw)
predmarg_sttran_prjcrim_T1 = stress.wide3 %>%
data_grid(sttranw1i) %>%
add_epred_draws(allprjcrime.sttran.fit) %>%
group_by(.category, sttranw1i, .draw)
predmarg_stresp_prjcrim_T1 = stress.wide3 %>%
data_grid(strespw1i) %>%
add_epred_draws(allprjcrime.stresp.fit) %>%
group_by(.category, strespw1i, .draw)
predmarg_stfair_prjcrim_T1 = stress.wide3 %>%
data_grid(stfairw1i) %>%
add_epred_draws(allprjcrime.stfair.fit) %>%
group_by(.category, stfairw1i, .draw)
predmarg_stjob_prjcrim_T1 = stress.wide3 %>%
data_grid(stjobw1i) %>%
add_epred_draws(allprjcrime.stjob.fit) %>%
group_by(.category, stjobw1i, .draw)
predmarg_stthft_prjcrim_T1 = stress.wide3 %>%
data_grid(stthftw1i) %>%
add_epred_draws(allprjcrime.stthft.fit) %>%
group_by(.category, stthftw1i, .draw)
predmarg_stmug_prjcrim_T1 = stress.wide3 %>%
data_grid(stmugw1i) %>%
add_epred_draws(allprjcrime.stmug.fit) %>%
group_by(.category, stmugw1i, .draw)
#function to get epred draws from "any crim intent" models
epred_anycrm <- function(mydata, mystressvar, myoutcomevar, mymodelfit) {
mydata %>%
data_grid({{mystressvar}}) %>%
mutate(.category=myoutcomevar) %>%
add_epred_draws(mymodelfit) %>%
group_by({{mystressvar}}, .draw)
}
#get "any crim intent" epred draws & merge w/individual outcome epred data
predmarg_stmony_anyprjcrim_T1 =epred_anycrm(stress.wide3, stmonyw1i,
"prjanyw1f", anyprjcrime.stmony.fit)
predmarg_stmony_prjcrim_T1 <- bind_rows(predmarg_stmony_prjcrim_T1,
predmarg_stmony_anyprjcrim_T1)
rm(predmarg_stmony_anyprjcrim_T1) #clean environment
predmarg_sttran_anyprjcrim_T1 =epred_anycrm(stress.wide3, sttranw1i,
"prjanyw1f", anyprjcrime.sttran.fit)
predmarg_sttran_prjcrim_T1 <- bind_rows(predmarg_sttran_prjcrim_T1,
predmarg_sttran_anyprjcrim_T1)
rm(predmarg_sttran_anyprjcrim_T1)
predmarg_stresp_anyprjcrim_T1 =epred_anycrm(stress.wide3, strespw1i,
"prjanyw1f", anyprjcrime.stresp.fit)
predmarg_stresp_prjcrim_T1 <- bind_rows(predmarg_stresp_prjcrim_T1,
predmarg_stresp_anyprjcrim_T1)
rm(predmarg_stresp_anyprjcrim_T1)
predmarg_stfair_anyprjcrim_T1 =epred_anycrm(stress.wide3, stfairw1i,
"prjanyw1f", anyprjcrime.stfair.fit)
predmarg_stfair_prjcrim_T1 <- bind_rows(predmarg_stfair_prjcrim_T1,
predmarg_stfair_anyprjcrim_T1)
rm(predmarg_stfair_anyprjcrim_T1)
predmarg_stjob_anyprjcrim_T1 =epred_anycrm(stress.wide3, stjobw1i,
"prjanyw1f", anyprjcrime.stjob.fit)
predmarg_stjob_prjcrim_T1 <- bind_rows(predmarg_stjob_prjcrim_T1,
predmarg_stjob_anyprjcrim_T1)
rm(predmarg_stjob_anyprjcrim_T1)
predmarg_stthft_anyprjcrim_T1 =epred_anycrm(stress.wide3, stthftw1i,
"prjanyw1f", anyprjcrime.stthft.fit)
predmarg_stthft_prjcrim_T1 <- bind_rows(predmarg_stthft_prjcrim_T1,
predmarg_stthft_anyprjcrim_T1)
rm(predmarg_stthft_anyprjcrim_T1)
predmarg_stmug_anyprjcrim_T1 =epred_anycrm(stress.wide3, stmugw1i,
"prjanyw1f", anyprjcrime.stmug.fit)
predmarg_stmug_prjcrim_T1 <- bind_rows(predmarg_stmug_prjcrim_T1,
predmarg_stmug_anyprjcrim_T1)
rm(predmarg_stmug_anyprjcrim_T1)
#function to calculate mean difference contrast (marginal effect contrasts, or MEs)
calc_MEs <- function(predmarg_data, xitem) {
predmarg_data %>%
compare_levels(.epred, by = xitem) %>%
rename(`difference in E[y|stress]` = `.epred`)
}
#Warning - takes a while to generate these ME contrasts...
margeffs_stmony_prjcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stmony_prjcrim_T1, "stmonyw1i")}, file="cache_4_9")
margeffs_sttran_prjcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_sttran_prjcrim_T1, "sttranw1i")}, file="cache_4_10")
margeffs_stresp_prjcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stresp_prjcrim_T1, "strespw1i")}, file="cache_4_11")
margeffs_stfair_prjcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stfair_prjcrim_T1, "stfairw1i")}, file="cache_4_12")
margeffs_stjob_prjcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stjob_prjcrim_T1, "stjobw1i")}, file="cache_4_13")
margeffs_stthft_prjcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stthft_prjcrim_T1, "stthftw1i")}, file="cache_4_14")
margeffs_stmug_prjcrim_T1 = xfun::cache_rds({calc_MEs(predmarg_stmug_prjcrim_T1, "stmugw1i")}, file="cache_4_15")5.4 Supp. Figure 2B: PLME Stress/Criminal Intent T1
#Create diff data for each stress item (note, run everything before the "slice" command to check logic)
#stmony
PLME2_stmony_prjcrim_T1 = margeffs_stmony_prjcrim_T1 %>%
filter(stmonyw1i=="5 - 3" | stmonyw1i=="4 - 2" | stmonyw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nMoney",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmonyw1i, `difference in E[y|stress]`))
#sttran
PLME2_sttran_prjcrim_T1 = margeffs_sttran_prjcrim_T1 %>%
filter(sttranw1i=="5 - 3" | sttranw1i=="4 - 2" | sttranw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTransport",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(sttranw1i, `difference in E[y|stress]`))
#stresp
PLME2_stresp_prjcrim_T1 = margeffs_stresp_prjcrim_T1 %>%
filter(strespw1i=="5 - 3" | strespw1i=="4 - 2" | strespw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nRespect",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(strespw1i, `difference in E[y|stress]`))
#stfair
PLME2_stfair_prjcrim_T1 = margeffs_stfair_prjcrim_T1 %>%
filter(stfairw1i=="5 - 3" | stfairw1i=="4 - 2" | stfairw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nFair Trtmt",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stfairw1i, `difference in E[y|stress]`))
#stjob
PLME2_stjob_prjcrim_T1 = margeffs_stjob_prjcrim_T1 %>%
filter(stjobw1i=="5 - 3" | stjobw1i=="4 - 2" | stjobw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nJob",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stjobw1i, `difference in E[y|stress]`))
#stthft
PLME2_stthft_prjcrim_T1 = margeffs_stthft_prjcrim_T1 %>%
filter(stthftw1i=="5 - 3" | stthftw1i=="4 - 2" | stthftw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTheft Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stthftw1i, `difference in E[y|stress]`))
#stmug
PLME2_stmug_prjcrim_T1 = margeffs_stmug_prjcrim_T1 %>%
filter(stmugw1i=="5 - 3" | stmugw1i=="4 - 2" | stmugw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nAssault Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmugw1i, `difference in E[y|stress]`))
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
twodif_combinedprj <- bind_rows(
PLME2_stmony_prjcrim_T1,
PLME2_sttran_prjcrim_T1,
PLME2_stresp_prjcrim_T1,
PLME2_stfair_prjcrim_T1,
PLME2_stjob_prjcrim_T1,
PLME2_stthft_prjcrim_T1,
PLME2_stmug_prjcrim_T1)
#transform stress_var into reverse-ordered factor
twodif_combinedprj2<- twodif_combinedprj %>%
mutate(
stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney"))
)
#calculate row & col averages of P(mean_twodif) > 0
#Rows: P(mean_twodif | Stress item) > 0
#aka P(PLME>0|stress)
#View marginal probabilities
# twodif_combinedprj2 %>%
# group_by(stress_varf) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_twodif>0),
# p_gt0 = n_gt0 / n_ests)
# Stress:\nMoney 0.28
# Stress:\nTransport 0.36
# Stress:\nRespect 0.87
# Stress:\nFair Trtmt 0.91
# Stress:\nJob 0.96
# Stress:\nTheft Vctm 0.76
# Stress:\nAssault Vctm 0.65
#Cols: P(mean_twodif | Crime item) > 0
#aka P(PLME>0|crime)
#View marginal probabilities
# twodif_combinedprj2 %>%
# group_by(.category) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_twodif>0),
# p_gt0 = n_gt0 / n_ests)
# prjthflt5w1f 0.77
# prjthfgt5w1f 0.84
# prjthreatw1f 0.71
# prjharmw1f 0.51
# prjusedrgw1f 0.63
# prjhackw1f 0.62
# prjanyw1f 0.72
#Add these posterior probabilities into variable name
prjcrimelabsPT1 <- c(
"prjthflt5w1f"="Theft <5BAM\nP(ME>0|col)\n=.77",
"prjthfgt5w1f"="Theft >5BAM\nP(ME>0|col)\n=.84",
"prjthreatw1f"="Threaten\nP(ME>0|col)\n=.71",
"prjharmw1f"="Phys. harm\nP(ME>0|col)\n=.51",
"prjusedrgw1f"="Use drugs\nP(ME>0|col)\n=.63",
"prjhackw1f"="Hack info.\nP(ME>0|col)\n=.62",
"prjanyw1f"="Any crime\nP(ME>0|col)\n=.72")
stress_varlabsPT1 <- c(
"Stress:\nMoney" = "Stress: Money\nP(ME>0|row)\n=.28",
"Stress:\nTransport" = "Stress: Transport\nP(ME>0|row)\n=.36",
"Stress:\nRespect" = "Stress: Respect\nP(ME>0|row)=.87",
"Stress:\nFair Trtmt" = "Stress: Fair Trtmt\nP(ME>0|row)\n=.91",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.96",
"Stress:\nTheft Vctm" = "Stress: Theft\nP(ME>0|row)\n=0.76",
"Stress:\nAssault Vctm" = "Stress: Assault\nP(ME>0|row)\n=0.65")
#First, add p_gt0 variable used above to each stress/item combo in data (i.e. to each plot)
#Create alpha_scale variable =1 if p_gt0 < 0 (i.e., to be fully opaque) & =p_gt0 if p_gt0 > 1
twodif_combinedprj3 <- twodif_combinedprj2 %>%
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(mean_twodif>0),
p_gt0 = n_gt0 / n_ests,
gt0 = ifelse(mean_twodif>0, 1, 0),
alpha_scale = ifelse(mean_twodif <=0, 1, p_gt0),
rural.ses.med = as.factor("0")) %>% #add to combine overall & community-specific figures later
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5w1f", "prjthfgt5w1f", "prjthreatw1f",
"prjharmw1f", "prjusedrgw1f", "prjhackw1f",
"prjanyw1f"))
)
#Using sequential color palette (scio = lajolla)
#Also using ggnewscale to add multiple fill palettes
SuppFigure2B <- ggplot(data = twodif_combinedprj3, mapping = aes(x = mean_twodif, y = stress_varf)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(prjcrimelabsPT1))) +
stat_slab(mapping = aes(fill = p_gt0), .width = .95) +
scale_fill_scico(palette = "lajolla", begin = .8, end = .3, #tells it where to start and end palette
name = "P(PLME > 0)", breaks = c(.1, .5, .9)) +
new_scale_fill() + #from ggnewscale
stat_halfeye(mapping = aes(fill = after_stat(x > 0)), .width = .95, show.legend=FALSE) +
scale_fill_manual(values = c("grey80", "NA")) +
geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.5, .5) +
coord_cartesian(xlim=c(-.11,.16)) +
scale_x_continuous(breaks=c(-.1,0,.1)) +
xlab("Posterior Predicted Difference in E[y|stress]") +
scale_y_discrete(labels=stress_varlabsPT1) +
labs(
title = 'SUPPLEMENTAL FIGURE 2B\nMarginal Effect of 2-Category Difference in Stress on Criminal Intent at T1, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. "PLME" refers to "Practically Large Marginal Effect" of 2-category difference in stress on outcome. Darker shaded\nregions indicate larger proportion of posterior effect estimates are greater than zero (positive effect is more probable); grey shading indicates portion of posterior distribution of effect\nestimates equal to or less than zero.') +
theme(axis.title.y = element_blank(),
axis.text.y = element_text(size=10),
strip.text.x = element_text(size=10),
legend.position = "bottom",
strip.background = element_blank(),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
plot.title.position = "plot",
plot.caption.position = "plot")
SuppFigure2B
# Export to image
ggsave("SuppFigure2B.jpeg", width=9, height=6.5, path=here("Output"))Now we repeat this process yet again to generate a comparable bivariate association plot for stress and negative emotions.
5.4.1 Predictive Marginal Contrasts (negative emotions)
#Get predictive margins data
predmarg_stmony_depress_T1 = stress.wide3 %>%
data_grid(stmonyw1i) %>%
add_epred_draws(alldepress.stmony.fit) %>%
group_by(.category, stmonyw1i, .draw)
predmarg_sttran_depress_T1 = stress.wide3 %>%
data_grid(sttranw1i) %>%
add_epred_draws(alldepress.sttran.fit) %>%
group_by(.category, sttranw1i, .draw)
predmarg_stresp_depress_T1 = stress.wide3 %>%
data_grid(strespw1i) %>%
add_epred_draws(alldepress.stresp.fit) %>%
group_by(.category, strespw1i, .draw)
predmarg_stfair_depress_T1 = stress.wide3 %>%
data_grid(stfairw1i) %>%
add_epred_draws(alldepress.stfair.fit) %>%
group_by(.category, stfairw1i, .draw)
predmarg_stjob_depress_T1 = stress.wide3 %>%
data_grid(stjobw1i) %>%
add_epred_draws(alldepress.stjob.fit) %>%
group_by(.category, stjobw1i, .draw)
predmarg_stthft_depress_T1 = stress.wide3 %>%
data_grid(stthftw1i) %>%
add_epred_draws(alldepress.stthft.fit) %>%
group_by(.category, stthftw1i, .draw)
predmarg_stmug_depress_T1 = stress.wide3 %>%
data_grid(stmugw1i) %>%
add_epred_draws(alldepress.stmug.fit) %>%
group_by(.category, stmugw1i, .draw)
#function to calculate mean difference contrast (marginal effect contrasts, or MEs)
calc_MEs <- function(predmarg_data, xitem) {
predmarg_data %>%
compare_levels(.epred, by = xitem) %>%
rename(`difference in E[y|stress]` = `.epred`)
}
#Warning - takes a while to generate these ME contrasts...
margeffs_stmony_depress_T1 = xfun::cache_rds({calc_MEs(predmarg_stmony_depress_T1, "stmonyw1i")}, file="cache_4_16")
margeffs_sttran_depress_T1 = xfun::cache_rds({calc_MEs(predmarg_sttran_depress_T1, "sttranw1i")}, file="cache_4_17")
margeffs_stresp_depress_T1 = xfun::cache_rds({calc_MEs(predmarg_stresp_depress_T1, "strespw1i")}, file="cache_4_18")
margeffs_stfair_depress_T1 = xfun::cache_rds({calc_MEs(predmarg_stfair_depress_T1, "stfairw1i")}, file="cache_4_19")
margeffs_stjob_depress_T1 = xfun::cache_rds({calc_MEs(predmarg_stjob_depress_T1, "stjobw1i")}, file="cache_4_20")
margeffs_stthft_depress_T1 = xfun::cache_rds({calc_MEs(predmarg_stthft_depress_T1, "stthftw1i")}, file="cache_4_21")
margeffs_stmug_depress_T1 = xfun::cache_rds({calc_MEs(predmarg_stmug_depress_T1, "stmugw1i")}, file="cache_4_22")5.5 Supp. Figure 2C: PLME Stress/negative emotions T1
#Create diff data for each stress item (note, run everything before the "slice" command to check logic)
#stmony
PLME2_stmony_depress_T1 = margeffs_stmony_depress_T1 %>%
filter(stmonyw1i=="5 - 3" | stmonyw1i=="4 - 2" | stmonyw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nMoney",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmonyw1i, `difference in E[y|stress]`))
#sttran
PLME2_sttran_depress_T1 = margeffs_sttran_depress_T1 %>%
filter(sttranw1i=="5 - 3" | sttranw1i=="4 - 2" | sttranw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTransport",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(sttranw1i, `difference in E[y|stress]`))
#stresp
PLME2_stresp_depress_T1 = margeffs_stresp_depress_T1 %>%
filter(strespw1i=="5 - 3" | strespw1i=="4 - 2" | strespw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nRespect",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(strespw1i, `difference in E[y|stress]`))
#stfair
PLME2_stfair_depress_T1 = margeffs_stfair_depress_T1 %>%
filter(stfairw1i=="5 - 3" | stfairw1i=="4 - 2" | stfairw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nFair Trtmt",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stfairw1i, `difference in E[y|stress]`))
#stjob
PLME2_stjob_depress_T1 = margeffs_stjob_depress_T1 %>%
filter(stjobw1i=="5 - 3" | stjobw1i=="4 - 2" | stjobw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nJob",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stjobw1i, `difference in E[y|stress]`))
#stthft
PLME2_stthft_depress_T1 = margeffs_stthft_depress_T1 %>%
filter(stthftw1i=="5 - 3" | stthftw1i=="4 - 2" | stthftw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nTheft Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stthftw1i, `difference in E[y|stress]`))
#stmug
PLME2_stmug_depress_T1 = margeffs_stmug_depress_T1 %>%
filter(stmugw1i=="5 - 3" | stmugw1i=="4 - 2" | stmugw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nAssault Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmugw1i, `difference in E[y|stress]`))
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
twodif_combineddep <- bind_rows(
PLME2_stmony_depress_T1,
PLME2_sttran_depress_T1,
PLME2_stresp_depress_T1,
PLME2_stfair_depress_T1,
PLME2_stjob_depress_T1,
PLME2_stthft_depress_T1,
PLME2_stmug_depress_T1)
#transform stress_var into reverse-ordered factor
twodif_combineddep2<- twodif_combineddep %>%
mutate(
stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney"))
)
#calculate row & col averages of P(mean_twodif) > 0
#Rows: P(mean_twodif | Stress item) > 0
#aka P(PLME>0|stress)
#View marginal probabilities
# twodif_combineddep2 %>%
# group_by(stress_varf) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_twodif>0),
# p_gt0 = n_gt0 / n_ests)
# Stress:\nMoney 0.93
# Stress:\nTransport 0.76
# Stress:\nRespect 0.86
# Stress:\nFair Trtmt 0.92
# Stress:\nJob 0.62
# Stress:\nTheft Vctm 0.84
# Stress:\nAssault Vctm 0.87
#Cols: P(mean_twodif | Crime item) > 0
#aka P(PLME>0|crime)
#View marginal probabilities
# twodif_combineddep2 %>%
# group_by(.category) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(mean_twodif>0),
# p_gt0 = n_gt0 / n_ests)
# depcantgow1f 0.73
# depeffortw1f 0.86
# deplonelyw1f 0.81
# depbluesw1f 0.78
# depunfairw1f 0.80
# depmistrtw1f 0.88
# depbetrayw1f 0.94
#Add these posterior probabilities into variable name
depresslabsPT1 <- c(
"depcantgow1f"="Can't go\nP(ME>0|col)\n=.73",
"depeffortw1f"="Effort\nP(ME>0|col)\n=.86",
"deplonelyw1f"="Lonely\nP(ME>0|col)\n=.81",
"depbluesw1f"="Blues\nP(ME>0|col)\n=.78",
"depunfairw1f"="Unfair\nP(ME>0|col)\n=.80",
"depmistrtw1f"="Mistreated\nP(ME>0|col)\n=.88",
"depbetrayw1f"="Betrayed\nP(ME>0|col)\n=.94")
stress_varlabsPT1 <- c(
"Stress:\nMoney" = "Stress: Money\nP(ME>0|row)\n=.31",
"Stress:\nTransport" = "Stress: Transport\nP(ME>0|row)\n=.37",
"Stress:\nRespect" = "Stress: Respect\nP(ME>0|row)=.85",
"Stress:\nFair Trtmt" = "Stress: Fair Trtmt\nP(ME>0|row)\n=.89",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.96",
"Stress:\nTheft Vctm" = "Stress: Theft\nP(ME>0|row)\n=0.74",
"Stress:\nAssault Vctm" = "Stress: Assault\nP(ME>0|row)\n=0.65")
#First, add p_gt0 variable used above to each stress/item combo in data (i.e. to each plot)
#Create alpha_scale variable =1 if p_gt0 < 0 (i.e., to be fully opaque) & =p_gt0 if p_gt0 > 1
twodif_combineddep3 <- twodif_combineddep2 %>%
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(mean_twodif>0),
p_gt0 = n_gt0 / n_ests,
gt0 = ifelse(mean_twodif>0, 1, 0),
alpha_scale = ifelse(mean_twodif <=0, 1, p_gt0),
rural.ses.med = as.factor("0")) #add to combine figures 3 & 4 later
#Using sequential color palette (scio = lajolla)
#Also using ggnewscale to add multiple fill palettes
SuppFigure2C <- ggplot(data = twodif_combineddep3, mapping = aes(x = mean_twodif, y = stress_varf)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(depresslabsPT1))) +
stat_slab(mapping = aes(fill = p_gt0), .width = .95) +
scale_fill_scico(palette = "lajolla", begin = .8, end = .3, #tells it where to start and end palette
name = "P(PLME > 0)", breaks = c(.1, .5, .9)) +
new_scale_fill() + #from ggnewscale
stat_halfeye(mapping = aes(fill = after_stat(x > 0)), .width = .95, show.legend=FALSE) +
scale_fill_manual(values = c("grey80", "NA")) +
geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.5, .5) +
coord_cartesian(xlim=c(-.25,.25)) +
scale_x_continuous(breaks=c(-.2,0,.2)) +
xlab("Posterior Predicted Difference in E[y|stress]") +
scale_y_discrete(labels=stress_varlabsPT1) +
labs(
title = 'SUPPLEMENTAL FIGURE 2C\nMarginal Effect of 2-Category Difference in Stress on Negative Emotions at T1, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. "PLME" refers to "Practically Large Marginal Effect" of 2-category difference in stress on outcome. Darker shaded\nregions indicate larger proportion of posterior effect estimates are greater than zero (positive effect is more probable); grey shading indicates portion of posterior distribution of effect\nestimates equal to or less than zero.') +
theme(axis.title.y = element_blank(),
axis.text.y = element_text(size=10),
strip.text.x = element_text(size=10),
legend.position = "bottom",
strip.background = element_blank(),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
plot.title.position = "plot",
plot.caption.position = "plot")
SuppFigure2C
#Export to image
ggsave("SuppFigure2C.jpeg", width=9, height=6.5, path=here("Output"))Note the x-axis scales are different for the depressive symptom bivariate association plot. This is to accommodate the larger differences in predicted outcome probabilities associated with a two-category change in stress in these plots relative to the crime plots. This suggests stronger associations between stress and negative emotions versus crime items, which may in part also reflect the fact that negative emotions are more prevalent (higher overall outcome probabilities).
As we mentioned above, at the survey design stage, we included two indicators each for subjective stress from financial (money; transportation), relational (respect; fair treatment), and victimization (theft; assault) sources, along with an item tapping job-related stress. Once again, these T1 stress/past crime item correlations largely show comparable patterns for the item pairs tapping financial, interpersonal, and victimization stress, respectively.
Additionally, patterns are largely consistent across conceptually similar crime item pairs (theft; violence). Hence, to make the presentation of results more manageable, we could try reducing these results to a single plot by averaging the PLME (2-category) contrasts for the conceptually similar stress and crime item pairs. We could also collapse marginal contrasts for the first four classic CESD depressive symptom items together (can’t get going; effort; lonely; blues), as well as collapse marginal contrasts for the last three negative emotions that are theorized as especially criminogenic (unfair; mistreated; betrayed). However, lacking strong theoretical rationale for doing so, we pursue these possibilities below in supplementary figures, while cautioning that they are merely an exploratory exercise in descriptive data reduction.
Specifically, in supplementary figures below, we simply average the marginal posterior probability expectations. For an alternative approach (e.g., with more items), we could consider building a multilevel model where item is a nesting or grouping level. Alternatively, since our posterior predictive averaging approach is akin to posterior predictive stacking (see also here) with equal or uniform model weights specified (i.e., an equal number of predictive draws per model), we could instead consider weighting the number of posterior draws for each model’s predictive distribution by a vector of model-specific weights that optimizes the (loo-)predictive average logged score of the stacked prediction.
Again, we lack strong theoretical or empirical reasons for assuming the collapsed items reflect a common data generating process that can be modeled appropriately with joint parameters in a multilevel framework. Also, at this stage, our aim is not to maximize predictive performance; rather, it is simply to simplify presentation of item-level stress/outcome associations by collapsing or combining similar stress/outcome pairings with the goal of assessing whether correlational patterns in the data are consistent with or contradict theoretical expectations. Hence, a simple stacking approach with equal model weights effectively communicates estimate magnitudes and the degree of uncertainty surrounding the collapsed estimates. For instance, the displayed posterior distribution of marginal effect contrasts will be unimodal wherever model contrasts of differences in predicted probabilities are similar across collapsed stress/outcome pairings. In contrast, multimodal posterior distributions imply that estimates of the marginal effect contrast varies across the collapsed stress/outcome item pairs. Think of it like presenting an unweighted meta-analytic average estimate (and distribution of estimates) across all models or a select group of models.
Finally, patterns to this point are quite similar across past crime and criminal intent plots, and criminal intent establishes more appropriate causal ordering and thus provides much stronger grounds for causal inference (i.e., past crime occurred before and may have caused presently reported stress). Hence, from this point forward, we will will focus on results from models predicting criminal intent and negative emotions.
5.6 Supp. Figure 2X: PLME “Stacked” Stress/Outcomes T1
# Compare to naive Bayes classifier or Bayesian model averaging (BMA)
# w/multiple modes over various x models
# See last comment here:
# https://statmodeling.stat.columbia.edu/2021/12/18/use-stacking-rather-than-bayesian-model-averaging/
# Akin to "stacking" posterior pred dist but specifying equal model weights
# Consider if BMA w/predictive weights is more desirable for aims:
# https://www.ajordannafa.com/blog/2022/05/24/bma-ames/
# also includes example of way more efficient modeling (see "Estimation")
# combine multiple datasets w/diff stress vars, outcome, etc. in list
# then apply same formula (or diff formula) to list of datasets!
#begin caching
# xfun::cache_rds({
#Create diff data for each stress item (note, run everything before the "slice" command to check logic)
#stmony models - drop "any" outcome, then collapse like outcomes & stress_var (financial)
tempdata <- margeffs_stmony_prjcrim_T1 %>%
dplyr::filter(.category != "prjanyw1f") %>%
mutate(
.category = as.factor(.category)
)
PLME_stmony_outcomes_T1 <-
rbind(tempdata, margeffs_stmony_depress_T1) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheftw1" = c("prjthflt5w1f", "prjthfgt5w1f"),
"prjviolw1" = c("prjthreatw1f", "prjharmw1f"),
"prjusedrgw1f" = "prjusedrgw1f",
"prjhackw1f" = "prjhackw1f",
"depsymw1" = c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f"),
"negemow1" = c("depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
) %>%
filter(stmonyw1i=="5 - 3" | stmonyw1i=="4 - 2" | stmonyw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nFinancial",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmonyw1i, `difference in E[y|stress]`))
#sttran models - drop "any" outcome, collapse outcomes & stress var (financial)
tempdata <- margeffs_sttran_prjcrim_T1 %>%
dplyr::filter(.category != "prjanyw1f") %>%
mutate(
.category = as.factor(.category)
)
PLME_sttran_outcomes_T1 <-
rbind(tempdata, margeffs_sttran_depress_T1) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheftw1" = c("prjthflt5w1f", "prjthfgt5w1f"),
"prjviolw1" = c("prjthreatw1f", "prjharmw1f"),
"prjusedrgw1f" = "prjusedrgw1f",
"prjhackw1f" = "prjhackw1f",
"depsymw1" = c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f"),
"negemow1" = c("depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
) %>%
filter(sttranw1i=="5 - 3" | sttranw1i=="4 - 2" | sttranw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nFinancial",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(sttranw1i, `difference in E[y|stress]`))
#stresp - drop "any" outcome, collapse outcomes & stress var (financial)
tempdata <- margeffs_stresp_prjcrim_T1 %>%
dplyr::filter(.category != "prjanyw1f") %>%
mutate(
.category = as.factor(.category)
)
PLME_stresp_outcomes_T1 <-
rbind(tempdata, margeffs_stresp_depress_T1) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheftw1" = c("prjthflt5w1f", "prjthfgt5w1f"),
"prjviolw1" = c("prjthreatw1f", "prjharmw1f"),
"prjusedrgw1f" = "prjusedrgw1f",
"prjhackw1f" = "prjhackw1f",
"depsymw1" = c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f"),
"negemow1" = c("depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
) %>%
filter(strespw1i=="5 - 3" | strespw1i=="4 - 2" | strespw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nRelational",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(strespw1i, `difference in E[y|stress]`))
#stfair - drop "any" outcome, collapse outcomes & stress var (financial)
tempdata <- margeffs_stfair_prjcrim_T1 %>%
dplyr::filter(.category != "prjanyw1f") %>%
mutate(
.category = as.factor(.category)
)
PLME_stfair_outcomes_T1 <-
rbind(tempdata, margeffs_stfair_depress_T1) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheftw1" = c("prjthflt5w1f", "prjthfgt5w1f"),
"prjviolw1" = c("prjthreatw1f", "prjharmw1f"),
"prjusedrgw1f" = "prjusedrgw1f",
"prjhackw1f" = "prjhackw1f",
"depsymw1" = c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f"),
"negemow1" = c("depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
) %>%
filter(stfairw1i=="5 - 3" | stfairw1i=="4 - 2" | stfairw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nRelational",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stfairw1i, `difference in E[y|stress]`))
#stjob - drop "any" outcome, collapse outcomes
tempdata <- margeffs_stjob_prjcrim_T1 %>%
dplyr::filter(.category != "prjanyw1f") %>%
mutate(
.category = as.factor(.category)
)
PLME_stjob_outcomes_T1 <-
rbind(tempdata, margeffs_stjob_depress_T1) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheftw1" = c("prjthflt5w1f", "prjthfgt5w1f"),
"prjviolw1" = c("prjthreatw1f", "prjharmw1f"),
"prjusedrgw1f" = "prjusedrgw1f",
"prjhackw1f" = "prjhackw1f",
"depsymw1" = c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f"),
"negemow1" = c("depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
) %>%
filter(stjobw1i=="5 - 3" | stjobw1i=="4 - 2" | stjobw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nJob",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stjobw1i, `difference in E[y|stress]`))
#stthft - drop "any" outcome, collapse outcomes & stress var (financial)
tempdata <- margeffs_stthft_prjcrim_T1 %>%
dplyr::filter(.category != "prjanyw1f") %>%
mutate(
.category = as.factor(.category)
)
PLME_stthft_outcomes_T1 <-
rbind(tempdata, margeffs_stthft_depress_T1) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheftw1" = c("prjthflt5w1f", "prjthfgt5w1f"),
"prjviolw1" = c("prjthreatw1f", "prjharmw1f"),
"prjusedrgw1f" = "prjusedrgw1f",
"prjhackw1f" = "prjhackw1f",
"depsymw1" = c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f"),
"negemow1" = c("depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
) %>%
filter(stthftw1i=="5 - 3" | stthftw1i=="4 - 2" | stthftw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nCrime Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stthftw1i, `difference in E[y|stress]`))
#stmug - drop "any" outcome, collapse outcomes & stress var (financial)
tempdata <- margeffs_stmug_prjcrim_T1 %>%
dplyr::filter(.category != "prjanyw1f") %>%
mutate(
.category = as.factor(.category)
)
PLME_stmug_outcomes_T1 <-
rbind(tempdata, margeffs_stmug_depress_T1) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheftw1" = c("prjthflt5w1f", "prjthfgt5w1f"),
"prjviolw1" = c("prjthreatw1f", "prjharmw1f"),
"prjusedrgw1f" = "prjusedrgw1f",
"prjhackw1f" = "prjhackw1f",
"depsymw1" = c("depcantgow1f", "depeffortw1f", "deplonelyw1f", "depbluesw1f"),
"negemow1" = c("depunfairw1f", "depmistrtw1f", "depbetrayw1f"))
) %>%
filter(stmugw1i=="5 - 3" | stmugw1i=="4 - 2" | stmugw1i=="3 - 1") %>%
group_by(.category, .draw) %>%
mutate(mean_twodif = mean(`difference in E[y|stress]`),
stress_var = "Stress:\nCrime Vctm",
dif_label = "difference in E[y|stress]") %>%
slice_head() %>%
ungroup() %>%
dplyr::select(-c(stmugw1i, `difference in E[y|stress]`))
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
twodif_combineddvs <- bind_rows(
PLME_stmony_outcomes_T1,
PLME_sttran_outcomes_T1,
PLME_stresp_outcomes_T1,
PLME_stfair_outcomes_T1,
PLME_stjob_outcomes_T1,
PLME_stthft_outcomes_T1,
PLME_stmug_outcomes_T1)
#transform stress_var into reverse-ordered factor
twodif_combineddvs2<- twodif_combineddvs %>%
mutate(
stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nCrime Vctm",
"Stress:\nJob",
"Stress:\nRelational",
"Stress:\nFinancial"))
)
#calculate row & col averages of P(mean_twodif) > 0
#Rows: P(mean_twodif | Stress item) > 0
#aka P(PLME>0|stress)
twodif_combineddvs2 %>%
group_by(stress_varf) %>%
summarise(n_ests = n(),
n_gt0 = sum(mean_twodif>0),
p_gt0 = n_gt0 / n_ests) ## # A tibble: 4 × 4
## stress_varf n_ests n_gt0 p_gt0
## <ord> <int> <int> <dbl>
## 1 "Stress:\nCrime Vctm" 48000 38362 0.799
## 2 "Stress:\nJob" 24000 20601 0.858
## 3 "Stress:\nRelational" 48000 44630 0.930
## 4 "Stress:\nFinancial" 48000 25660 0.535# Stress:\nCrime Vctm 0.80
# Stress:\nJob 24000 0.86
# Stress:\nRelational 0.93
# Stress:\nFinancial 0.54
#Cols: P(mean_twodif | Crime item) > 0
#aka P(PLME>0|crime)
twodif_combineddvs2 %>%
group_by(.category) %>%
summarise(n_ests = n(),
n_gt0 = sum(mean_twodif>0),
p_gt0 = n_gt0 / n_ests) ## # A tibble: 6 × 4
## .category n_ests n_gt0 p_gt0
## <fct> <int> <int> <dbl>
## 1 prjhackw1f 28000 17447 0.623
## 2 prjviolw1 28000 18502 0.661
## 3 prjtheftw1 28000 23183 0.828
## 4 prjusedrgw1f 28000 17448 0.623
## 5 depsymw1 28000 25029 0.894
## 6 negemow1 28000 27644 0.987# prjtheftw1 0.83
# prjviolw1 0.66
# prjusedrgw1f 0.63
# prjhackw1f 0.62
# depsymw1 0.89
# negemow1 0.99
#Add these posterior probabilities into variable name
outcomelabsPT1 <- c(
"prjtheftw1"="Theft\nIntent\nP(ME>0|col)\n=.83",
"prjviolw1"="Violence\nIntent\nP(ME>0|col)\n=.66",
"prjusedrgw1f"="Use drugs\nIntent\nP(ME>0|col)\n=.63",
"prjhackw1f"="Hack info.\nIntent\nP(ME>0|col)\n=.62",
"depsymw1"="Depressive\nSymptoms\nP(ME>0|col)\n=.89",
"negemow1"="Criminogenic\nEmotions\nP(ME>0|col)\n=.99")
stress_varlabsPT1 <- c(
"Stress:\nFinancial" = "Stress: Financial\nP(ME>0|row)\n=.54",
"Stress:\nRelational" = "Stress: Personal\nP(ME>0|row)\n=.93",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.86",
"Stress:\nCrime Vctm" = "Stress: Crime Vctm\nP(ME>0|row)\n=0.80")
#First, add p_gt0 variable used above to each stress/item combo in data (i.e. to each plot)
#Create alpha_scale variable =1 if p_gt0 < 0 (i.e., to be fully opaque) & =p_gt0 if p_gt0 > 1
twodif_combineddvs3 <- twodif_combineddvs2 %>%
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(mean_twodif>0),
p_gt0 = n_gt0 / n_ests,
gt0 = ifelse(mean_twodif>0, 1, 0),
alpha_scale = ifelse(mean_twodif <=0, 1, p_gt0),
rural.ses.med = as.factor("0")) #add to combine figures 3 & 4 later
# }, file="cache_4_23") #end caching
#Using sequential color palette (scio = lajolla)
#using ggnewscale to add multiple fill palettes (could also use ggh4x)
#using ggh4x to set custom xlim coords for each facet
SuppFigure2XX <- ggplot(data = twodif_combineddvs3, mapping = aes(x = mean_twodif, y = stress_varf)) +
facet_wrap(~.category, nrow=1, scales="free_x",
labeller = labeller(.category= as_labeller(outcomelabsPT1))) +
stat_slab(mapping = aes(fill = p_gt0), .width = .95) +
scale_fill_scico(palette = "lajolla", begin = .8, end = .3, #tells it where to start and end palette
name = "P(PLME > 0)", breaks = c(.1, .5, .9), labels = dropLeadingZero) +
new_scale_fill() + #from ggnewscale
stat_halfeye(mapping = aes(fill = after_stat(x > 0)), .width = .95, show.legend=FALSE) +
scale_fill_manual(values = c("grey80", "NA")) +
geom_vline(xintercept = 0, linetype = "dashed") +
xlab("Posterior Predicted Difference in E[y|stress]") +
scale_y_discrete(labels=stress_varlabsPT1) +
facetted_pos_scales(
x = list(
.category %in% c("prjtheftw1", "prjviolw1", "prjusedrgw1f", "prjhackw1f") ~
scale_x_continuous(breaks=c(-.1,-.05,0,.05,.1),
limits = c(-.11,.11), labels = dropLeadingZero),
.category %in% c("depsymw1", "negemow1") ~
scale_x_continuous(breaks=c(-.05,0,.05,.1, .15),
limits = c(-.06,.16), labels = dropLeadingZero)
) ) +
labs(
title = 'SUPP. FIGURE 2D: Marginal Effects of 2-Category Stress Difference on "Stacked" Outcome Probabilities at T1, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. Estimates derived from 14 multivariate Bayesian logistic regression models, with each of seven \ncombined criminal intent outcomes (using `brms::mvbind()`) and each of seven combined negative emotion outcomes separately specifying a single T1 stress \ntype as a predictor. Stress items were specified as monotonic ordinal predictors with a cumulative probit link function. Models were estimated in brms with \nfour chains and 4000 total post-warmup posterior draws. Practically Large Marginal Effect (PLME) parameter distributions were first estimated from the \nexpectation of the posterior predictive distribution for each model by averaging predicted probability difference distributions for all 2-category stress contrasts \n(5-3; 4-2; 3-1) for each bivariate item pair. The final displayed PLME posterior parameter estimate distributions were then generated by "stacking" to \naverage posterior PLME estimates across models with conceptually similar outcome (e.g., Depressive Symptoms) and stress (e.g., Financial) item groupings.') +
theme(axis.title.y = element_blank(),
axis.text.y = element_text(size=10, hjust=1),
legend.position = "bottom",
strip.background = element_blank(),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0))
SuppFigure2XX 
#Export to image
# ggsave("SuppFigure2XX.jpeg", width=9, height=6.5, path=here("Output"))The above figure essentially displays results of simple Bayesian model averaging (BMA) of posterior probabilities across conceptually similar models. Typical BMA approaches move beyond simple averages to weight model-specific posterior probabilities by indicators of the model’s relative fit or predictive performance (e.g., stacking; loo; waic).
However, again, our goal here is not to select the “best” model; rather, it is to describe correlational patterns between various stress and outcome types while communicating uncertainty in estimates. Here, we essentially assigned equal weights to each model when averaging posterior probabilities, which requires strong independence assumptions and assumes all models are equally probable.
Let’s now turn to within-person change models and “fixed effects” estimates.
save(twodif_combineddvs3, file = here("1_Data_Files/Datasets/twodif_combineddvs3.Rdata"))
save(twodif_combinedprj3, file = here("1_Data_Files/Datasets/twodif_combinedprj3.Rdata")) #NEED TO RUN
save(twodif_combineddep3, file = here("1_Data_Files/Datasets/twodif_combineddep3.Rdata")) #NEED TO RUN6 Modeling change (T2-T1) correlations (RQ2B)
(RMD FILE: BDK_2023_Stress_5_Chgcorr_mods)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE6.1 Between-person vs within-person estimates
Descriptively, the cross-sectional T1 effect estimates above reveal between-person differences in expected outcome probabilities across participants reporting different ordinal responses to stress items. Such between-person descriptive differences are not reliable estimates of causal effects of stress on crime. For instance, as causal effect estimates, they are potentially confounded by unmeasured sources of heterogeneity; they may also reflect selection or reverse causality processes. Though causal modeling with observational data is possible, it requires a precise theory that comprehensively identifies confounders and colliders as well as mediators of the focal causal process.
Meanwhile, within-person (T2-T1 or “fixed”) effects describe average correlations between within-person changes in stress and outcome responses. These estimates essentially control for time-stable (e.g., between-person) sources of confounding are blocked in estimation, which has made them a popular alternative for generating causal inferences from observational data. Additionally, within-person correlations can be estimated simultaneously and compared with between-person estimates using a “between-within” or “hybrid modeling approach.
However, such estimates do not rule out time-varying sources of confounding, nor do they necessarily isolate selection or reverse causality processes; additionally, such estimates assess average within-person change correlations over a very specific time lag (in this case, changes over a two-year interval). Thus, though they have clear advantages to strictly estimating between-person associations, within-person change correlations from observational data should also be interpreted descriptively and should not be used to make strong causal inferences. Like cross-sectional analyses of observational data, within-person change results that are consistent with theoretical expectations may have been generated by the theorized causal process, and/or they may have been generated by one or more alternative and possibly contradictory processes.
Given our rich description aims, we estimate both between-person difference and within-person change associations. Doing so will allow us to compare them and assess whether researchers might draw different inferences from these different sets of estimates.
6.2 RQ2B: Stress-Outcome Change Correlations from T1 to T2: Within-Person Method
RQ2B: (Stress deficit; Within-person): Are within-person increases in subjective stress from T1 to T2 (i.e., T2-T1) correlated with within-person increases (T2-T1) in the probability of reporting criminal intent or negative emotions?
6.2.1 Transform data from wide to long format
#load stress.wide (from Fig1 Rmd)
load(here("1_Data_Files/Datasets/stress_wide.Rdata"))
stress.wide <- zap_labels(stress.wide)
stress.wide <- zap_label(stress.wide)
load(here("1_Data_Files/Datasets/stress_wide2.Rdata"))
load(here("1_Data_Files/Datasets/stress_wide3.Rdata"))
#T1 posterior PLME contrasts
load(here("1_Data_Files/Datasets/twodif_combineddvs3.Rdata"))
load(here("1_Data_Files/Datasets/twodif_combinedprj3.Rdata"))
load(here("1_Data_Files/Datasets/twodif_combineddep3.Rdata"))First, we need to manipulate key variables from the second wave of data. For instance, outcome variables should be transformed into factors, while stress items must be integers for ordinal predictor specification in brms. Then we select variables to retain and reshape the data from wide to long format (two rows per person - one for each wave) in anticipation of building multilevel within/between models. With everything we will be doing, this is the place where we will restrict focus to the more temporally appropriate criminal intent items (alongside depressive symptoms) and ignore past crime items.
#First, need to recode T2 outcome variables as factors & T2 stress items as integers for mo()
stress.wide4 <- stress.wide3 %>%
mutate(
pstthflt5w2f = factor(pstthflt5w2di, ordered=TRUE, levels = c(0,1)),
pstthfgt5w2f = factor(pstthfgt5w2di, ordered=TRUE, levels = c(0,1)),
pstthreatw2f = factor(pstthreatw2di, ordered=TRUE, levels = c(0,1)),
pstharmw2f = factor(pstharmw2di, ordered=TRUE, levels = c(0,1)),
pstusedrgw2f = factor(pstusedrgw2di, ordered=TRUE, levels = c(0,1)),
psthackw2f = factor(psthackw2di, ordered=TRUE, levels = c(0,1)),
prjthflt5w2f = factor(prjthflt5w2di, ordered=TRUE, levels = c(0,1)),
prjthfgt5w2f = factor(prjthfgt5w2di, ordered=TRUE, levels = c(0,1)),
prjthreatw2f = factor(prjthreatw2di, ordered=TRUE, levels = c(0,1)),
prjharmw2f = factor(prjharmw2di, ordered=TRUE, levels = c(0,1)),
prjusedrgw2f = factor(prjusedrgw2di, ordered=TRUE, levels = c(0,1)),
prjhackw2f = factor(prjhackw2di, ordered=TRUE, levels = c(0,1)),
prjanyw2f = factor(if_else(prjthflt5w2di == 1 | prjthfgt5w2di == 1 | prjthreatw2di == 1 |
prjharmw2di == 1 | prjusedrgw2di == 1 | prjhackw2di == 1, 1, 0),
ordered=TRUE, levels = c(0,1)),
depcantgow2di = if_else(depcantgow2 %in% c(4,5), 1, 0),
depcantgow2f = factor(depcantgow2di, ordered=TRUE, levels = c(0,1)),
depeffortw2di = if_else(depeffortw2 %in% c(4,5), 1, 0),
depeffortw2f = factor(depeffortw2di, ordered=TRUE, levels = c(0,1)),
deplonelyw2di = if_else(deplonelyw2 %in% c(4,5), 1, 0),
deplonelyw2f = factor(deplonelyw2di, ordered=TRUE, levels = c(0,1)),
depbluesw2di = if_else(depbluesw2 %in% c(4,5), 1, 0),
depbluesw2f = factor(depbluesw2di, ordered=TRUE, levels = c(0,1)),
depunfairw2di = if_else(depunfairw2 %in% c(4,5), 1, 0),
depunfairw2f = factor(depunfairw2di, ordered=TRUE, levels = c(0,1)),
depmistrtw2di = if_else(depmistrtw2 %in% c(4,5), 1, 0),
depmistrtw2f = factor(depmistrtw2di, ordered=TRUE, levels = c(0,1)),
depbetrayw2di = if_else(depbetrayw2 %in% c(4,5), 1, 0),
depbetrayw2f = factor(depbetrayw2di, ordered=TRUE, levels = c(0,1)),
stmonyw2f = factor(stmonyw2, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
sttranw2f = factor(sttranw2, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
strespw2f = factor(strespw2, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stfairw2f = factor(stfairw2, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stjobw2f = factor(stjobw2, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stthftw2f = factor(stthftw2, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stmugw2f = factor(stmugw2, ordered=TRUE, levels = c(1, 2, 3, 4, 5)),
stmonyw2i = recode(stmonyw2f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
sttranw2i = recode(sttranw2f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
strespw2i = recode(strespw2f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stfairw2i = recode(stfairw2f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stjobw2i = recode(stjobw2f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stthftw2i = recode(stthftw2f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
stmugw2i = recode(stmugw2f,
"1" = 1, "2" = 2, "3" = 3, "4" = 4, "5" = 5) %>%
as.integer(),
fnhdav12 = (osfinaw1 + osfinaw2)/2,
fnhddvw1i = osfinaw1 - fnhdav12,
fnhddvw2i = osfinaw2 - fnhdav12
) %>%
rename(
depunfairordw1 = depunfairw1,
depmistrtordw1 = depmistrtw1,
depbetrayordw1 = depbetrayw1,
depunfairordw2 = depunfairw2,
depmistrtordw2 = depmistrtw2,
depbetrayordw2 = depbetrayw2
) %>%
dplyr::select(c(
id, female, rural, area_id, agew1, educw1, kidsw1, marriedw1,
fnhdav12, fnhddvw1i, fnhddvw2i, L2sesw1,
rural.ses.avg, rural.ses.med, rural.ses.med2,
stmonyw1i, sttranw1i, strespw1i, stfairw1i, stjobw1i, stthftw1i, stmugw1i,
stmonyw2i, sttranw2i, strespw2i, stfairw2i, stjobw2i, stthftw2i, stmugw2i,
pstthflt5w1f, pstthfgt5w1f, pstthreatw1f,
pstharmw1f, pstusedrgw1f, psthackw1f,
pstthflt5w2f, pstthfgt5w2f, pstthreatw2f,
pstharmw2f, pstusedrgw2f, psthackw2f,
prjthflt5w1f, prjthfgt5w1f, prjthreatw1f,
prjharmw1f, prjusedrgw1f, prjhackw1f, prjanyw1f,
prjthflt5w2f, prjthfgt5w2f, prjthreatw2f,
prjharmw2f, prjusedrgw2f, prjhackw2f, prjanyw2f,
depcantgow1f, depeffortw1f, deplonelyw1f, depbluesw1f,
depunfairw1f, depmistrtw1f, depbetrayw1f,
depcantgow2f, depeffortw2f, deplonelyw2f, depbluesw2f,
depunfairw2f, depmistrtw2f, depbetrayw2f,
depunfairordw1, depmistrtordw1, depbetrayordw1,
depunfairordw2, depmistrtordw2, depbetrayordw2))
scale2x <- function(x, na.rm = FALSE) (x*2)
stress.long <- stress.wide4 %>%
pivot_longer(
cols = !c(id, female, rural, area_id, agew1, educw1, kidsw1, marriedw1,
fnhdav12, L2sesw1,
rural.ses.avg, rural.ses.med, rural.ses.med2),
names_to = c(".value","year"), #splits each varname into new var + year
names_pattern = "(.*?)w(\\d)" #use regex + names_pattern to split varname
) %>% #newvar = text up to 'w' (drop w; not in group), year is next digit
group_by(id) %>% #create cross-time ave stress variables (av12)
mutate(across(stmony:stmug, mean, .names = "{col}_av12")) %>%
ungroup() %>%
mutate(across(stmony_av12:stmug_av12, scale2x, .names = "{col}x2")) %>% #mult 2x av12 vars to make whole integers
mutate(across(stmony_av12x2:stmug_av12x2, as.integer)) %>% #convert av12 vars to integers for mo()
mutate(
stmony_dev = stmony - stmony_av12,
sttran_dev = sttran - sttran_av12,
stresp_dev = stresp - stresp_av12,
stfair_dev = stfair - stfair_av12,
stjob_dev = stjob - stjob_av12,
stthft_dev = stthft - stthft_av12,
stmug_dev = stmug - stmug_av12
) %>% #create stress deviation scores from within-person cross-time avg
mutate(across(stmony_dev:stmug_dev, scale2x, .names = "{col}x2")) %>% #mult 2x dev vars to make whole integers (representing +/- unit stress change)
mutate(across(stmony_devx2:stmug_devx2, ~ifelse(.<=-2, -2, .))) %>% #recode to cap 2x changes at +/-2
mutate(across(stmony_devx2:stmug_devx2, ~ifelse(.>=2, 2, .))) %>% #recode to cap 2x changes at +/-2
mutate(across(stmony_devx2:stmug_devx2, as.integer)) #convert devx2 vars to integers for mo()
#NOTE: very few ids had stress changes <-2 or >2, so collapsed max changes to +/- 2
# stress_devx2 scores have 5 levels so 4 ordered thresholds
# table(stress.long$stmony_devx2)
# table(stress.long$sttran_devx2)
# table(stress.long$stresp_devx2)
# table(stress.long$stfair_devx2)
# table(stress.long$stjob_devx2)
# table(stress.long$stthft_devx2)
# table(stress.long$stmug_devx2)
# stress_av12x2 scores have 9 levels so 9 ordered thresholds
# table(stress.long$stmony_av12x2)
# table(stress.long$sttran_av12x2)
# table(stress.long$stresp_av12x2)
# table(stress.long$stfair_av12x2)
# table(stress.long$stjob_av12x2)
# table(stress.long$stthft_av12x2)
# table(stress.long$stmug_av12x2)
#view amount of change in specific crim intent & "any" crim intent
tempdata <- stress.wide %>%
mutate(
prjthflt5ch12 = prjthflt5w2di - prjthflt5w1di,
prjthfgt5ch12 = prjthfgt5w2di - prjthfgt5w1di,
prjthreatch12 = prjthreatw2di - prjthreatw1di,
prjharmch12 = prjharmw2di - prjharmw1di,
prjusedrgch12 = prjusedrgw2di - prjusedrgw1di,
prjhackch12 = prjhackw2di - prjhackw1di)
p1 <- ggplot(tempdata, aes(prjthflt5ch12)) + geom_bar(fill="#E99D53")
p2 <- ggplot(tempdata, aes(prjthfgt5ch12)) + geom_bar(fill="#E99D53")
p3 <- ggplot(tempdata, aes(prjthreatch12)) + geom_bar(fill="#E99D53")
p4 <- ggplot(tempdata, aes(prjharmch12)) + geom_bar(fill="#E99D53")
p5 <- ggplot(tempdata, aes(prjusedrgch12)) + geom_bar(fill="#E99D53")
p6 <- ggplot(tempdata, aes(prjhackch12)) + geom_bar(fill="#E99D53")
# table(tempdata$prjthflt5ch12)
# table(tempdata$prjthfgt5ch12)
# table(tempdata$prjthreatch12)
# table(tempdata$prjharmch12)
# table(tempdata$prjusedrgch12)
# table(tempdata$prjhackch12)
tempdata <- stress.wide4 %>%
mutate(
prjanyw1 = as.integer(prjanyw1f),
prjanyw2 = as.integer(prjanyw2f),
prjchg12 = prjanyw2 - prjanyw1
)
p7 <- ggplot(tempdata, aes(prjchg12)) + geom_bar(fill="#E99D53")
# table(tempdata$prjchg12)
#NOTE - change in criminal intent across waves is very rare (as is "1" on crim intent)
(p1 | p2 | p3 | p4) / (p5 | p6 | p7 | plot_spacer())
tempdata <- stress.wide %>%
mutate(
prjthflt5ch12 = prjthflt5w2di - prjthflt5w1di,
prjthfgt5ch12 = prjthfgt5w2di - prjthfgt5w1di,
prjthreatch12 = prjthreatw2di - prjthreatw1di,
prjharmch12 = prjharmw2di - prjharmw1di,
prjusedrgch12 = prjusedrgw2di - prjusedrgw1di,
prjhackch12 = prjhackw2di - prjhackw1di)
p1b <- ggplot(tempdata, aes(x=prjthflt5ch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p2b <- ggplot(tempdata, aes(x=prjthfgt5ch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p3b <- ggplot(tempdata, aes(x=prjthreatch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p4b <- ggplot(tempdata, aes(x=prjharmch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p5b <- ggplot(tempdata, aes(x=prjusedrgch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p6b <- ggplot(tempdata, aes(x=prjhackch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
tempdata <- stress.wide4 %>%
mutate(
prjanyw1 = as.integer(prjanyw1f),
prjanyw2 = as.integer(prjanyw2f),
prjchg12 = prjanyw2 - prjanyw1
)
p7b <- ggplot(tempdata, aes(x=prjchg12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
# table(tempdata$prjchg12)
#NOTE - change in criminal intent across waves is very rare (as is "1" on crim intent)
pdesign <- "1234 \n 5678"
p1b + p2b + p3b + p4b + p5b + p6b + p7b + guide_area() +
plot_layout(design = pdesign, guides='collect')
As expected, there is relatively little variation in criminal intent changes, which means we should be wary when interpreting results due to potentially low signal-to-noise ratio. As we hoped, the “any crime” item shows a bit more variability overall and among both urban and rural residents.
Before moving to modeling, let’s take a quick look at stress changes by rural/urban residence as well.
tempdata <- stress.wide %>%
mutate(
stmonych12 = stmonyw2 - stmonyw1,
sttranch12 = sttranw2 - sttranw1,
strespch12 = strespw2 - strespw1,
stfairch12 = stfairw2 - stfairw1,
stjobch12 = stjobw2 - stjobw1,
stthftch12 = stthftw2 - stthftw1,
stmugch12 = stmugw2 - stmugw1)
p1s <- ggplot(tempdata, aes(x=stmonych12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p2s <- ggplot(tempdata, aes(x=sttranch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p3s <- ggplot(tempdata, aes(x=strespch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p4s <- ggplot(tempdata, aes(x=stfairch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p5s <- ggplot(tempdata, aes(x=stjobch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p6s <- ggplot(tempdata, aes(x=stthftch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
p7s <- ggplot(tempdata, aes(x=stmugch12, y= ..prop.., fill=as.factor(rural))) +
geom_bar(position = "dodge") +
scale_fill_scico_d(palette = "lajolla", , begin = .8, end = .3)
# table(tempdata$prjchg12)
#NOTE - change in criminal intent across waves is very rare (as is "1" on crim intent)
pdesign <- "1234 \n 5678"
p1s + p2s + p3s + p4s + p5s + p6s + p7s + guide_area() +
plot_layout(design = pdesign, guides='collect')
There is much more variability in subjective stress change items than in criminal intent items, overall and separately among both rural and urban residents. So, perhaps unsurprisingly, subjective stress reports change across waves but criminal intent is much more rare and is more stable over time.
6.2.2 Criminal intent: Bivariate Change models
Refer to Kurz’s similar application with a binary IV.
As we noted earlier, very few respondents reported stress reductions or increases across the two waves greater than two units (on Likert-type stress items ranging from 1 to 5). In fact, even two-unit changes in either direction across the two waves were quite rare. So, as part of the data wrangling above, the few changes greater than two units were collapsed to a maximum category of (+/-)2.
Now we are ready to build the basic within/between models. As before, we will specify multivariate logistic brms models w/monotonic ordinal predictors. However, these will be multilevel “between/within” models with a random intercept, which allows outcome probabilities to vary across individuals. We do not add a fixed effect for time (i.e., latent growth curve) to avoid inappropriately partialling out any causal effects of systematic changes in stress across waves that might be absorbed by an estimator for population-level outcome trends over time (although inclusion only very slightly attenuates estimated change associations).
For more information about different models for two-wave data, see Kurz’s excellent posts here and here. For more information about estimating between/within models in R, see here and here.
The “between” part of a between/within multilevel model indicates that we will include a person-level average stress score (\(\overline{X_i}\), or cross-time average X value for individual i) as a L2 predictor to estimate between-person stress/outcome correlations. The “within” part indicates that we will also include the within-person deviation from the person level mean (\((X_{ij} - \overline{Xi})\), or difference between X value for individual i at time j and individual i’s cross-time X value) as a L1 predictor. This within-person deviation score estimates the association between changes in X (stress) and changes in an outcome; also known as a “fixed effects” estimator, this estimate differences out all time stable effects of time-invariant (observed or unobserved) confounders.
Both the within and between variables were specified as monotonic
ordinal predictors with brms built-in cumulative probit link function
(mo()). To do this, recall our predictors need to be
formatted as integers. Hence, in data wrangling above, we multiplied
stress L2 and L1 predictors by two and then transformed them to
integers. Hence, rather than being scaled from ‘1’ to ‘5’ with 0.5-unit
increments, the cross-time average L2 stress items now range from ‘2’ to
‘10’ with 1-unit increments. Likewise, after rescaling, a 1-unit stress
change on the original scale (e.g., increase from 2 to 3) is represented
by -1 at year 1 and 1 at year 2 (rather than
-.5 and .5). The scaling of the units is
inconsequential since the estimated effects of ordinal increases are
modeled across latent thresholds anyway.
6.2.2.1 B/W Change Corr: stmony/crim intent
#Bivariate Change: criminal intent items ~ mo(stmony)
#Vectorize priors:
#list of colnames for projected crime DVs
prjdv_names <- noquote(c("prjthflt5", "prjthfgt5", "prjthreat", "prjharm",
"prjusedrg", "prjhack"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmony_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmony_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmony_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmony_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.stmony.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stmony_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="prjthflt5")
ppcheckdv2 <-pp_check(modelfit, resp="prjthfgt5")
ppcheckdv3 <-pp_check(modelfit, resp="prjthreat")
ppcheckdv4 <-pp_check(modelfit, resp="prjharm")
ppcheckdv5 <-pp_check(modelfit, resp="prjusedrg")
ppcheckdv6 <-pp_check(modelfit, resp="prjhack")
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6,
plotcoefs, plotcoefs2)
return(allchecks)
}
out.chg.prjcrime.stmony.fit <- ppchecks(chg.prjcrime.stmony.fit)6.2.2.1.1 Coefficient plot (intervals)
out.chg.prjcrime.stmony.fit[[10]]
6.2.2.1.2 Coefficient plot (distributions)
out.chg.prjcrime.stmony.fit[[9]]
6.2.2.1.3 PPcheck (density)
p1 <- out.chg.prjcrime.stmony.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.chg.prjcrime.stmony.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.chg.prjcrime.stmony.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.chg.prjcrime.stmony.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.chg.prjcrime.stmony.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.chg.prjcrime.stmony.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
6.2.2.1.4 Fit summary
out.chg.prjcrime.stmony.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## prjthfgt5 ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## prjthreat ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## prjharm ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## prjusedrg ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## prjhack ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.59 0.64 3.44 5.89 1.00 1491
## sd(prjthfgt5_Intercept) 3.83 0.56 2.84 5.02 1.00 1012
## sd(prjthreat_Intercept) 3.57 0.59 2.53 4.86 1.00 1523
## sd(prjharm_Intercept) 2.98 0.55 2.01 4.16 1.00 1053
## sd(prjusedrg_Intercept) 3.25 0.56 2.23 4.43 1.00 1640
## sd(prjhack_Intercept) 0.87 0.55 0.04 2.05 1.01 411
## Tail_ESS
## sd(prjthflt5_Intercept) 1984
## sd(prjthfgt5_Intercept) 1722
## sd(prjthreat_Intercept) 2427
## sd(prjharm_Intercept) 2437
## sd(prjusedrg_Intercept) 2214
## sd(prjhack_Intercept) 1178
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_Intercept -5.87 0.85 -7.62 -4.29 1.01 1982
## prjthfgt5_Intercept -5.65 0.80 -7.29 -4.16 1.00 1928
## prjthreat_Intercept -5.91 0.93 -7.86 -4.21 1.00 2330
## prjharm_Intercept -5.36 0.90 -7.23 -3.74 1.00 1705
## prjusedrg_Intercept -5.84 0.92 -7.79 -4.09 1.00 2490
## prjhack_Intercept -4.10 0.69 -5.63 -2.86 1.00 1545
## prjthflt5_mostmony_devx2 0.20 0.17 -0.16 0.52 1.00 4336
## prjthflt5_mostmony_av12x2 -0.04 0.10 -0.23 0.15 1.00 2714
## prjthfgt5_mostmony_devx2 0.27 0.16 -0.06 0.58 1.00 4547
## prjthfgt5_mostmony_av12x2 -0.01 0.09 -0.19 0.16 1.00 3047
## prjthreat_mostmony_devx2 -0.07 0.19 -0.45 0.31 1.00 5597
## prjthreat_mostmony_av12x2 -0.09 0.10 -0.28 0.11 1.00 3780
## prjharm_mostmony_devx2 -0.07 0.19 -0.47 0.29 1.00 4985
## prjharm_mostmony_av12x2 -0.12 0.09 -0.29 0.06 1.00 4373
## prjusedrg_mostmony_devx2 -0.15 0.20 -0.52 0.23 1.00 6376
## prjusedrg_mostmony_av12x2 -0.06 0.10 -0.24 0.13 1.00 3620
## prjhack_mostmony_devx2 0.06 0.19 -0.34 0.42 1.00 4748
## prjhack_mostmony_av12x2 -0.04 0.08 -0.21 0.12 1.00 5592
## Tail_ESS
## prjthflt5_Intercept 2534
## prjthfgt5_Intercept 2921
## prjthreat_Intercept 2395
## prjharm_Intercept 2841
## prjusedrg_Intercept 2835
## prjhack_Intercept 1978
## prjthflt5_mostmony_devx2 2045
## prjthflt5_mostmony_av12x2 3001
## prjthfgt5_mostmony_devx2 2946
## prjthfgt5_mostmony_av12x2 3119
## prjthreat_mostmony_devx2 2868
## prjthreat_mostmony_av12x2 2584
## prjharm_mostmony_devx2 2294
## prjharm_mostmony_av12x2 2831
## prjusedrg_mostmony_devx2 3288
## prjusedrg_mostmony_av12x2 2815
## prjhack_mostmony_devx2 2378
## prjhack_mostmony_av12x2 2998
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_mostmony_devx21[1] 0.22 0.14 0.03 0.54 1.00
## prjthflt5_mostmony_devx21[2] 0.26 0.14 0.04 0.58 1.00
## prjthflt5_mostmony_devx21[3] 0.29 0.16 0.04 0.63 1.00
## prjthflt5_mostmony_devx21[4] 0.23 0.14 0.03 0.55 1.00
## prjthflt5_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## prjthflt5_mostmony_av12x21[2] 0.13 0.08 0.02 0.34 1.00
## prjthflt5_mostmony_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjthflt5_mostmony_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mostmony_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mostmony_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mostmony_av12x21[7] 0.12 0.08 0.01 0.32 1.00
## prjthflt5_mostmony_av12x21[8] 0.12 0.08 0.02 0.30 1.00
## prjthfgt5_mostmony_devx21[1] 0.20 0.13 0.03 0.49 1.00
## prjthfgt5_mostmony_devx21[2] 0.30 0.15 0.05 0.62 1.00
## prjthfgt5_mostmony_devx21[3] 0.28 0.15 0.05 0.59 1.00
## prjthfgt5_mostmony_devx21[4] 0.22 0.13 0.03 0.53 1.00
## prjthfgt5_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## prjthfgt5_mostmony_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## prjthfgt5_mostmony_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjthfgt5_mostmony_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjthfgt5_mostmony_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mostmony_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mostmony_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mostmony_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjthreat_mostmony_devx21[2] 0.24 0.14 0.03 0.56 1.00
## prjthreat_mostmony_devx21[3] 0.24 0.14 0.04 0.57 1.00
## prjthreat_mostmony_devx21[4] 0.25 0.15 0.04 0.59 1.00
## prjthreat_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## prjthreat_mostmony_av12x21[2] 0.12 0.08 0.02 0.32 1.00
## prjthreat_mostmony_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjthreat_mostmony_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mostmony_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mostmony_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mostmony_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## prjharm_mostmony_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjharm_mostmony_devx21[2] 0.24 0.14 0.03 0.57 1.00
## prjharm_mostmony_devx21[3] 0.23 0.14 0.03 0.56 1.00
## prjharm_mostmony_devx21[4] 0.26 0.15 0.04 0.60 1.00
## prjharm_mostmony_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## prjharm_mostmony_av12x21[2] 0.12 0.08 0.01 0.32 1.00
## prjharm_mostmony_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjharm_mostmony_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## prjharm_mostmony_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## prjharm_mostmony_av12x21[6] 0.14 0.09 0.02 0.35 1.00
## prjharm_mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjharm_mostmony_av12x21[8] 0.11 0.07 0.02 0.29 1.00
## prjusedrg_mostmony_devx21[1] 0.27 0.15 0.04 0.61 1.00
## prjusedrg_mostmony_devx21[2] 0.25 0.15 0.04 0.58 1.00
## prjusedrg_mostmony_devx21[3] 0.23 0.13 0.03 0.54 1.00
## prjusedrg_mostmony_devx21[4] 0.25 0.15 0.04 0.58 1.00
## prjusedrg_mostmony_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## prjusedrg_mostmony_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostmony_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjusedrg_mostmony_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## prjusedrg_mostmony_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mostmony_av12x21[6] 0.12 0.08 0.02 0.32 1.00
## prjusedrg_mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mostmony_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## prjhack_mostmony_devx21[1] 0.25 0.14 0.04 0.58 1.00
## prjhack_mostmony_devx21[2] 0.24 0.14 0.04 0.55 1.00
## prjhack_mostmony_devx21[3] 0.26 0.14 0.04 0.58 1.00
## prjhack_mostmony_devx21[4] 0.25 0.14 0.04 0.57 1.00
## prjhack_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## prjhack_mostmony_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## prjhack_mostmony_av12x21[3] 0.13 0.08 0.01 0.33 1.00
## prjhack_mostmony_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjhack_mostmony_av12x21[5] 0.12 0.08 0.01 0.30 1.00
## prjhack_mostmony_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## prjhack_mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjhack_mostmony_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_mostmony_devx21[1] 4940 2069
## prjthflt5_mostmony_devx21[2] 4958 2353
## prjthflt5_mostmony_devx21[3] 4346 2559
## prjthflt5_mostmony_devx21[4] 5674 2411
## prjthflt5_mostmony_av12x21[1] 5639 2488
## prjthflt5_mostmony_av12x21[2] 5022 2561
## prjthflt5_mostmony_av12x21[3] 6539 2758
## prjthflt5_mostmony_av12x21[4] 6718 2355
## prjthflt5_mostmony_av12x21[5] 7046 2625
## prjthflt5_mostmony_av12x21[6] 4234 2310
## prjthflt5_mostmony_av12x21[7] 5660 2839
## prjthflt5_mostmony_av12x21[8] 5002 3127
## prjthfgt5_mostmony_devx21[1] 7068 3215
## prjthfgt5_mostmony_devx21[2] 4678 2776
## prjthfgt5_mostmony_devx21[3] 5041 2769
## prjthfgt5_mostmony_devx21[4] 6731 2790
## prjthfgt5_mostmony_av12x21[1] 5319 2487
## prjthfgt5_mostmony_av12x21[2] 6610 2693
## prjthfgt5_mostmony_av12x21[3] 5287 2216
## prjthfgt5_mostmony_av12x21[4] 5870 2165
## prjthfgt5_mostmony_av12x21[5] 6238 2416
## prjthfgt5_mostmony_av12x21[6] 5453 2779
## prjthfgt5_mostmony_av12x21[7] 5508 2509
## prjthfgt5_mostmony_av12x21[8] 6166 2977
## prjthreat_mostmony_devx21[1] 5739 2493
## prjthreat_mostmony_devx21[2] 7266 2072
## prjthreat_mostmony_devx21[3] 6132 2871
## prjthreat_mostmony_devx21[4] 6892 2986
## prjthreat_mostmony_av12x21[1] 6119 2632
## prjthreat_mostmony_av12x21[2] 6249 2572
## prjthreat_mostmony_av12x21[3] 4460 2337
## prjthreat_mostmony_av12x21[4] 5565 2505
## prjthreat_mostmony_av12x21[5] 5752 2261
## prjthreat_mostmony_av12x21[6] 5070 2855
## prjthreat_mostmony_av12x21[7] 5215 2742
## prjthreat_mostmony_av12x21[8] 5178 2468
## prjharm_mostmony_devx21[1] 6258 2519
## prjharm_mostmony_devx21[2] 5892 2512
## prjharm_mostmony_devx21[3] 4650 2398
## prjharm_mostmony_devx21[4] 5089 2594
## prjharm_mostmony_av12x21[1] 5000 1879
## prjharm_mostmony_av12x21[2] 6111 2297
## prjharm_mostmony_av12x21[3] 6402 2171
## prjharm_mostmony_av12x21[4] 6620 2899
## prjharm_mostmony_av12x21[5] 5990 2712
## prjharm_mostmony_av12x21[6] 6090 2645
## prjharm_mostmony_av12x21[7] 6688 3073
## prjharm_mostmony_av12x21[8] 6351 2694
## prjusedrg_mostmony_devx21[1] 5728 2436
## prjusedrg_mostmony_devx21[2] 6212 2589
## prjusedrg_mostmony_devx21[3] 5025 2658
## prjusedrg_mostmony_devx21[4] 6039 2813
## prjusedrg_mostmony_av12x21[1] 5792 2165
## prjusedrg_mostmony_av12x21[2] 6063 2709
## prjusedrg_mostmony_av12x21[3] 5607 2727
## prjusedrg_mostmony_av12x21[4] 4863 2286
## prjusedrg_mostmony_av12x21[5] 4984 2261
## prjusedrg_mostmony_av12x21[6] 5710 2817
## prjusedrg_mostmony_av12x21[7] 5293 3042
## prjusedrg_mostmony_av12x21[8] 5817 2972
## prjhack_mostmony_devx21[1] 6384 2228
## prjhack_mostmony_devx21[2] 5713 2598
## prjhack_mostmony_devx21[3] 4777 2593
## prjhack_mostmony_devx21[4] 4461 2381
## prjhack_mostmony_av12x21[1] 6203 2637
## prjhack_mostmony_av12x21[2] 6448 2866
## prjhack_mostmony_av12x21[3] 5061 1903
## prjhack_mostmony_av12x21[4] 6829 2611
## prjhack_mostmony_av12x21[5] 5258 1834
## prjhack_mostmony_av12x21[6] 5805 2763
## prjhack_mostmony_av12x21[7] 5737 2855
## prjhack_mostmony_av12x21[8] 5432 3024
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.2.1.5 Prior summary
out.chg.prjcrime.stmony.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b prjhack
## normal(0, 0.125) b mostmony_av12x2 prjhack
## normal(0, 0.25) b mostmony_devx2 prjhack
## (flat) b prjharm
## normal(0, 0.125) b mostmony_av12x2 prjharm
## normal(0, 0.25) b mostmony_devx2 prjharm
## (flat) b prjthfgt5
## normal(0, 0.125) b mostmony_av12x2 prjthfgt5
## normal(0, 0.25) b mostmony_devx2 prjthfgt5
## (flat) b prjthflt5
## normal(0, 0.125) b mostmony_av12x2 prjthflt5
## normal(0, 0.25) b mostmony_devx2 prjthflt5
## (flat) b prjthreat
## normal(0, 0.125) b mostmony_av12x2 prjthreat
## normal(0, 0.25) b mostmony_devx2 prjthreat
## (flat) b prjusedrg
## normal(0, 0.125) b mostmony_av12x2 prjusedrg
## normal(0, 0.25) b mostmony_devx2 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 prjhack
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 prjhack
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 prjharm
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 prjharm
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 prjthfgt5
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 prjthfgt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 prjthflt5
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 prjthflt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 prjthreat
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 prjthreat
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 prjusedrg
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 prjusedrg
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.2.2 B/W Change Corr: sttran/crim intent
#Bivariate Change: criminal intent items ~ mo(sttran)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mosttran_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mosttran_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mosttran_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mosttran_av12x21',
resp = prjdv_names)
)
chg.prjcrime.sttran.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_sttran_fit",
file_refit = "on_change"
)
out.chg.prjcrime.sttran.fit <- ppchecks(chg.prjcrime.sttran.fit)6.2.2.2.1 Coefficient plot (intervals)
out.chg.prjcrime.sttran.fit[[10]]
6.2.2.2.2 Coefficient plot (distributions)
out.chg.prjcrime.sttran.fit[[9]]
6.2.2.2.3 PPcheck (density)
p1 <- out.chg.prjcrime.sttran.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.chg.prjcrime.sttran.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.chg.prjcrime.sttran.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.chg.prjcrime.sttran.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.chg.prjcrime.sttran.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.chg.prjcrime.sttran.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
6.2.2.2.4 Fit summary
out.chg.prjcrime.sttran.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## prjthfgt5 ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## prjthreat ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## prjharm ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## prjusedrg ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## prjhack ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.57 0.65 3.45 6.00 1.00 1496
## sd(prjthfgt5_Intercept) 3.81 0.55 2.86 5.00 1.00 1111
## sd(prjthreat_Intercept) 3.59 0.59 2.56 4.90 1.00 1614
## sd(prjharm_Intercept) 3.05 0.53 2.11 4.20 1.00 1720
## sd(prjusedrg_Intercept) 3.26 0.58 2.26 4.49 1.00 1565
## sd(prjhack_Intercept) 0.92 0.56 0.04 2.09 1.00 675
## Tail_ESS
## sd(prjthflt5_Intercept) 2124
## sd(prjthfgt5_Intercept) 2379
## sd(prjthreat_Intercept) 2140
## sd(prjharm_Intercept) 2572
## sd(prjusedrg_Intercept) 2195
## sd(prjhack_Intercept) 1477
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_Intercept -5.25 0.84 -7.09 -3.69 1.00 2118
## prjthfgt5_Intercept -5.03 0.75 -6.60 -3.63 1.00 1640
## prjthreat_Intercept -5.86 0.96 -7.86 -4.11 1.00 2283
## prjharm_Intercept -4.78 0.91 -6.66 -3.05 1.00 2287
## prjusedrg_Intercept -5.62 0.96 -7.66 -3.91 1.00 1956
## prjhack_Intercept -3.67 0.73 -5.20 -2.30 1.00 1873
## prjthflt5_mosttran_devx2 0.01 0.17 -0.34 0.34 1.00 3645
## prjthflt5_mosttran_av12x2 -0.10 0.10 -0.30 0.10 1.00 2865
## prjthfgt5_mosttran_devx2 0.02 0.17 -0.33 0.36 1.00 4867
## prjthfgt5_mosttran_av12x2 -0.03 0.10 -0.23 0.15 1.00 2472
## prjthreat_mosttran_devx2 -0.21 0.18 -0.58 0.15 1.00 5342
## prjthreat_mosttran_av12x2 -0.04 0.10 -0.24 0.16 1.00 3338
## prjharm_mosttran_devx2 -0.34 0.19 -0.73 0.04 1.00 5086
## prjharm_mosttran_av12x2 -0.15 0.09 -0.33 0.04 1.00 3801
## prjusedrg_mosttran_devx2 -0.25 0.20 -0.63 0.13 1.00 5810
## prjusedrg_mosttran_av12x2 -0.07 0.10 -0.26 0.14 1.00 4254
## prjhack_mosttran_devx2 -0.24 0.19 -0.62 0.13 1.00 5013
## prjhack_mosttran_av12x2 -0.01 0.09 -0.18 0.16 1.00 4614
## Tail_ESS
## prjthflt5_Intercept 2121
## prjthfgt5_Intercept 2972
## prjthreat_Intercept 2504
## prjharm_Intercept 2310
## prjusedrg_Intercept 2329
## prjhack_Intercept 2313
## prjthflt5_mosttran_devx2 2294
## prjthflt5_mosttran_av12x2 2594
## prjthfgt5_mosttran_devx2 2880
## prjthfgt5_mosttran_av12x2 2347
## prjthreat_mosttran_devx2 3012
## prjthreat_mosttran_av12x2 2961
## prjharm_mosttran_devx2 2773
## prjharm_mosttran_av12x2 3044
## prjusedrg_mosttran_devx2 2733
## prjusedrg_mosttran_av12x2 3131
## prjhack_mosttran_devx2 2940
## prjhack_mosttran_av12x2 2774
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_mosttran_devx21[1] 0.26 0.15 0.04 0.60 1.00
## prjthflt5_mosttran_devx21[2] 0.24 0.14 0.04 0.56 1.00
## prjthflt5_mosttran_devx21[3] 0.24 0.14 0.03 0.56 1.00
## prjthflt5_mosttran_devx21[4] 0.26 0.15 0.03 0.59 1.00
## prjthflt5_mosttran_av12x21[1] 0.13 0.08 0.02 0.34 1.00
## prjthflt5_mosttran_av12x21[2] 0.14 0.09 0.02 0.34 1.00
## prjthflt5_mosttran_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjthflt5_mosttran_av12x21[4] 0.13 0.08 0.02 0.31 1.00
## prjthflt5_mosttran_av12x21[5] 0.12 0.08 0.02 0.32 1.00
## prjthflt5_mosttran_av12x21[6] 0.12 0.08 0.02 0.32 1.00
## prjthflt5_mosttran_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mosttran_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## prjthfgt5_mosttran_devx21[1] 0.26 0.14 0.04 0.59 1.00
## prjthfgt5_mosttran_devx21[2] 0.23 0.14 0.03 0.55 1.00
## prjthfgt5_mosttran_devx21[3] 0.24 0.14 0.03 0.57 1.00
## prjthfgt5_mosttran_devx21[4] 0.26 0.15 0.04 0.58 1.00
## prjthfgt5_mosttran_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## prjthfgt5_mosttran_av12x21[2] 0.13 0.08 0.02 0.34 1.00
## prjthfgt5_mosttran_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjthfgt5_mosttran_av12x21[4] 0.13 0.08 0.02 0.32 1.00
## prjthfgt5_mosttran_av12x21[5] 0.12 0.07 0.02 0.30 1.00
## prjthfgt5_mosttran_av12x21[6] 0.12 0.08 0.02 0.32 1.00
## prjthfgt5_mosttran_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mosttran_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## prjthreat_mosttran_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjthreat_mosttran_devx21[2] 0.26 0.14 0.04 0.58 1.00
## prjthreat_mosttran_devx21[3] 0.23 0.14 0.04 0.55 1.00
## prjthreat_mosttran_devx21[4] 0.24 0.14 0.04 0.56 1.00
## prjthreat_mosttran_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mosttran_av12x21[2] 0.13 0.08 0.01 0.33 1.00
## prjthreat_mosttran_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mosttran_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjthreat_mosttran_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mosttran_av12x21[6] 0.12 0.08 0.01 0.32 1.00
## prjthreat_mosttran_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mosttran_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## prjharm_mosttran_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjharm_mosttran_devx21[2] 0.24 0.14 0.03 0.56 1.00
## prjharm_mosttran_devx21[3] 0.27 0.15 0.04 0.60 1.00
## prjharm_mosttran_devx21[4] 0.22 0.13 0.03 0.54 1.00
## prjharm_mosttran_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## prjharm_mosttran_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## prjharm_mosttran_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## prjharm_mosttran_av12x21[4] 0.14 0.09 0.02 0.35 1.00
## prjharm_mosttran_av12x21[5] 0.14 0.08 0.02 0.34 1.00
## prjharm_mosttran_av12x21[6] 0.12 0.07 0.02 0.30 1.00
## prjharm_mosttran_av12x21[7] 0.11 0.07 0.01 0.29 1.00
## prjharm_mosttran_av12x21[8] 0.11 0.07 0.02 0.28 1.00
## prjusedrg_mosttran_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjusedrg_mosttran_devx21[2] 0.25 0.14 0.04 0.57 1.00
## prjusedrg_mosttran_devx21[3] 0.25 0.14 0.04 0.57 1.00
## prjusedrg_mosttran_devx21[4] 0.23 0.14 0.04 0.54 1.00
## prjusedrg_mosttran_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## prjusedrg_mosttran_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mosttran_av12x21[3] 0.13 0.09 0.02 0.33 1.00
## prjusedrg_mosttran_av12x21[4] 0.13 0.08 0.02 0.34 1.00
## prjusedrg_mosttran_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mosttran_av12x21[6] 0.12 0.08 0.02 0.30 1.00
## prjusedrg_mosttran_av12x21[7] 0.12 0.08 0.01 0.30 1.00
## prjusedrg_mosttran_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## prjhack_mosttran_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjhack_mosttran_devx21[2] 0.26 0.15 0.04 0.60 1.00
## prjhack_mosttran_devx21[3] 0.23 0.14 0.03 0.55 1.00
## prjhack_mosttran_devx21[4] 0.24 0.14 0.03 0.55 1.00
## prjhack_mosttran_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## prjhack_mosttran_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## prjhack_mosttran_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjhack_mosttran_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjhack_mosttran_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjhack_mosttran_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## prjhack_mosttran_av12x21[7] 0.13 0.08 0.02 0.34 1.00
## prjhack_mosttran_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_mosttran_devx21[1] 4854 2478
## prjthflt5_mosttran_devx21[2] 5554 2431
## prjthflt5_mosttran_devx21[3] 5300 3123
## prjthflt5_mosttran_devx21[4] 5527 3095
## prjthflt5_mosttran_av12x21[1] 4463 1744
## prjthflt5_mosttran_av12x21[2] 4789 2228
## prjthflt5_mosttran_av12x21[3] 5414 2774
## prjthflt5_mosttran_av12x21[4] 4924 2310
## prjthflt5_mosttran_av12x21[5] 5945 2544
## prjthflt5_mosttran_av12x21[6] 5684 2909
## prjthflt5_mosttran_av12x21[7] 5863 3012
## prjthflt5_mosttran_av12x21[8] 5104 2500
## prjthfgt5_mosttran_devx21[1] 5116 2618
## prjthfgt5_mosttran_devx21[2] 5365 2681
## prjthfgt5_mosttran_devx21[3] 5012 3035
## prjthfgt5_mosttran_devx21[4] 5111 2482
## prjthfgt5_mosttran_av12x21[1] 5472 2363
## prjthfgt5_mosttran_av12x21[2] 5667 2541
## prjthfgt5_mosttran_av12x21[3] 5096 2219
## prjthfgt5_mosttran_av12x21[4] 4880 2372
## prjthfgt5_mosttran_av12x21[5] 3964 2129
## prjthfgt5_mosttran_av12x21[6] 4654 2478
## prjthfgt5_mosttran_av12x21[7] 5147 2909
## prjthfgt5_mosttran_av12x21[8] 5592 2863
## prjthreat_mosttran_devx21[1] 5920 2128
## prjthreat_mosttran_devx21[2] 5819 2677
## prjthreat_mosttran_devx21[3] 5523 2914
## prjthreat_mosttran_devx21[4] 5395 2498
## prjthreat_mosttran_av12x21[1] 5466 2497
## prjthreat_mosttran_av12x21[2] 5260 2084
## prjthreat_mosttran_av12x21[3] 5191 2287
## prjthreat_mosttran_av12x21[4] 5249 2298
## prjthreat_mosttran_av12x21[5] 5224 1911
## prjthreat_mosttran_av12x21[6] 5197 2508
## prjthreat_mosttran_av12x21[7] 5690 3110
## prjthreat_mosttran_av12x21[8] 6208 2606
## prjharm_mosttran_devx21[1] 5678 2309
## prjharm_mosttran_devx21[2] 4946 2240
## prjharm_mosttran_devx21[3] 5892 2557
## prjharm_mosttran_devx21[4] 4962 2656
## prjharm_mosttran_av12x21[1] 5642 2260
## prjharm_mosttran_av12x21[2] 5048 2407
## prjharm_mosttran_av12x21[3] 4791 1988
## prjharm_mosttran_av12x21[4] 5429 2753
## prjharm_mosttran_av12x21[5] 4569 2855
## prjharm_mosttran_av12x21[6] 4920 2575
## prjharm_mosttran_av12x21[7] 5425 2904
## prjharm_mosttran_av12x21[8] 5347 3039
## prjusedrg_mosttran_devx21[1] 6687 2375
## prjusedrg_mosttran_devx21[2] 6636 3366
## prjusedrg_mosttran_devx21[3] 6129 2734
## prjusedrg_mosttran_devx21[4] 5663 2442
## prjusedrg_mosttran_av12x21[1] 5542 2183
## prjusedrg_mosttran_av12x21[2] 5241 2447
## prjusedrg_mosttran_av12x21[3] 5604 2640
## prjusedrg_mosttran_av12x21[4] 4791 2481
## prjusedrg_mosttran_av12x21[5] 4632 2535
## prjusedrg_mosttran_av12x21[6] 4703 2474
## prjusedrg_mosttran_av12x21[7] 5130 2804
## prjusedrg_mosttran_av12x21[8] 5020 2560
## prjhack_mosttran_devx21[1] 4333 2166
## prjhack_mosttran_devx21[2] 5979 2495
## prjhack_mosttran_devx21[3] 5627 2808
## prjhack_mosttran_devx21[4] 5347 3115
## prjhack_mosttran_av12x21[1] 4974 2083
## prjhack_mosttran_av12x21[2] 6043 2260
## prjhack_mosttran_av12x21[3] 5375 2363
## prjhack_mosttran_av12x21[4] 4694 2374
## prjhack_mosttran_av12x21[5] 5679 2756
## prjhack_mosttran_av12x21[6] 6127 2999
## prjhack_mosttran_av12x21[7] 5170 2699
## prjhack_mosttran_av12x21[8] 6062 2940
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.2.2.5 Prior summary
out.chg.prjcrime.sttran.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b prjhack
## normal(0, 0.125) b mosttran_av12x2 prjhack
## normal(0, 0.25) b mosttran_devx2 prjhack
## (flat) b prjharm
## normal(0, 0.125) b mosttran_av12x2 prjharm
## normal(0, 0.25) b mosttran_devx2 prjharm
## (flat) b prjthfgt5
## normal(0, 0.125) b mosttran_av12x2 prjthfgt5
## normal(0, 0.25) b mosttran_devx2 prjthfgt5
## (flat) b prjthflt5
## normal(0, 0.125) b mosttran_av12x2 prjthflt5
## normal(0, 0.25) b mosttran_devx2 prjthflt5
## (flat) b prjthreat
## normal(0, 0.125) b mosttran_av12x2 prjthreat
## normal(0, 0.25) b mosttran_devx2 prjthreat
## (flat) b prjusedrg
## normal(0, 0.125) b mosttran_av12x2 prjusedrg
## normal(0, 0.25) b mosttran_devx2 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 prjhack
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 prjhack
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 prjharm
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 prjharm
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 prjthfgt5
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 prjthfgt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 prjthflt5
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 prjthflt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 prjthreat
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 prjthreat
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 prjusedrg
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 prjusedrg
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.2.3 B/W Change Corr: stresp/crim intent
#Bivariate Change: criminal intent items ~ mo(stresp)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostresp_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostresp_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostresp_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostresp_av12x21',
resp = prjdv_names)
)
chg.prjcrime.stresp.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stresp_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stresp.fit <- ppchecks(chg.prjcrime.stresp.fit)6.2.2.3.1 Coefficient plot (intervals)
out.chg.prjcrime.stresp.fit[[10]]
6.2.2.3.2 Coefficient plot (distributions)
out.chg.prjcrime.stresp.fit[[9]]
6.2.2.3.3 PPcheck (density)
p1 <- out.chg.prjcrime.stresp.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.chg.prjcrime.stresp.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.chg.prjcrime.stresp.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.chg.prjcrime.stresp.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.chg.prjcrime.stresp.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.chg.prjcrime.stresp.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
6.2.2.3.4 Fit summary
out.chg.prjcrime.stresp.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## prjthfgt5 ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## prjthreat ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## prjharm ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## prjusedrg ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## prjhack ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.35 0.62 3.27 5.66 1.00 1350
## sd(prjthfgt5_Intercept) 3.77 0.56 2.76 4.98 1.00 1157
## sd(prjthreat_Intercept) 3.23 0.54 2.28 4.37 1.00 1574
## sd(prjharm_Intercept) 2.89 0.53 1.94 4.02 1.00 1233
## sd(prjusedrg_Intercept) 3.01 0.55 2.05 4.18 1.00 1483
## sd(prjhack_Intercept) 0.85 0.54 0.05 1.96 1.01 597
## Tail_ESS
## sd(prjthflt5_Intercept) 2213
## sd(prjthfgt5_Intercept) 1809
## sd(prjthreat_Intercept) 2219
## sd(prjharm_Intercept) 2340
## sd(prjusedrg_Intercept) 2322
## sd(prjhack_Intercept) 1582
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_Intercept -5.64 0.79 -7.29 -4.16 1.00 2257
## prjthfgt5_Intercept -5.18 0.78 -6.76 -3.72 1.00 1506
## prjthreat_Intercept -6.24 0.87 -8.02 -4.60 1.00 2322
## prjharm_Intercept -5.74 0.86 -7.57 -4.19 1.00 2022
## prjusedrg_Intercept -6.12 0.84 -7.90 -4.54 1.00 2146
## prjhack_Intercept -4.16 0.70 -5.65 -2.82 1.00 1889
## prjthflt5_mostresp_devx2 -0.22 0.17 -0.56 0.10 1.00 4307
## prjthflt5_mostresp_av12x2 0.13 0.08 -0.04 0.29 1.00 2455
## prjthfgt5_mostresp_devx2 -0.18 0.17 -0.53 0.13 1.00 4504
## prjthfgt5_mostresp_av12x2 0.09 0.08 -0.06 0.24 1.00 2757
## prjthreat_mostresp_devx2 -0.30 0.18 -0.67 0.05 1.00 4797
## prjthreat_mostresp_av12x2 0.17 0.08 0.01 0.34 1.00 3202
## prjharm_mostresp_devx2 -0.24 0.18 -0.62 0.12 1.00 4241
## prjharm_mostresp_av12x2 0.07 0.08 -0.10 0.23 1.00 3702
## prjusedrg_mostresp_devx2 -0.34 0.18 -0.70 0.00 1.00 4610
## prjusedrg_mostresp_av12x2 0.15 0.08 -0.02 0.31 1.00 3473
## prjhack_mostresp_devx2 -0.25 0.19 -0.62 0.12 1.00 3625
## prjhack_mostresp_av12x2 0.11 0.07 -0.03 0.25 1.00 4500
## Tail_ESS
## prjthflt5_Intercept 2694
## prjthfgt5_Intercept 2516
## prjthreat_Intercept 2524
## prjharm_Intercept 2501
## prjusedrg_Intercept 2439
## prjhack_Intercept 1924
## prjthflt5_mostresp_devx2 3083
## prjthflt5_mostresp_av12x2 3058
## prjthfgt5_mostresp_devx2 2879
## prjthfgt5_mostresp_av12x2 2773
## prjthreat_mostresp_devx2 2900
## prjthreat_mostresp_av12x2 3012
## prjharm_mostresp_devx2 2813
## prjharm_mostresp_av12x2 3133
## prjusedrg_mostresp_devx2 2988
## prjusedrg_mostresp_av12x2 3021
## prjhack_mostresp_devx2 3065
## prjhack_mostresp_av12x2 2807
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_mostresp_devx21[1] 0.29 0.16 0.05 0.63 1.00
## prjthflt5_mostresp_devx21[2] 0.24 0.14 0.04 0.56 1.00
## prjthflt5_mostresp_devx21[3] 0.21 0.13 0.03 0.53 1.00
## prjthflt5_mostresp_devx21[4] 0.25 0.15 0.04 0.58 1.00
## prjthflt5_mostresp_av12x21[1] 0.12 0.08 0.02 0.32 1.00
## prjthflt5_mostresp_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mostresp_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjthflt5_mostresp_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mostresp_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mostresp_av12x21[6] 0.13 0.08 0.02 0.34 1.00
## prjthflt5_mostresp_av12x21[7] 0.13 0.08 0.02 0.33 1.00
## prjthflt5_mostresp_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## prjthfgt5_mostresp_devx21[1] 0.29 0.16 0.04 0.64 1.00
## prjthfgt5_mostresp_devx21[2] 0.24 0.14 0.04 0.55 1.00
## prjthfgt5_mostresp_devx21[3] 0.21 0.14 0.03 0.55 1.00
## prjthfgt5_mostresp_devx21[4] 0.26 0.15 0.03 0.60 1.00
## prjthfgt5_mostresp_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## prjthfgt5_mostresp_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mostresp_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mostresp_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjthfgt5_mostresp_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mostresp_av12x21[6] 0.13 0.08 0.02 0.34 1.00
## prjthfgt5_mostresp_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## prjthfgt5_mostresp_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mostresp_devx21[1] 0.28 0.15 0.04 0.62 1.00
## prjthreat_mostresp_devx21[2] 0.30 0.15 0.05 0.62 1.00
## prjthreat_mostresp_devx21[3] 0.20 0.13 0.03 0.50 1.00
## prjthreat_mostresp_devx21[4] 0.22 0.13 0.03 0.53 1.00
## prjthreat_mostresp_av12x21[1] 0.11 0.07 0.01 0.29 1.00
## prjthreat_mostresp_av12x21[2] 0.11 0.07 0.01 0.29 1.00
## prjthreat_mostresp_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mostresp_av12x21[4] 0.13 0.08 0.01 0.33 1.00
## prjthreat_mostresp_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mostresp_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mostresp_av12x21[7] 0.14 0.09 0.02 0.36 1.00
## prjthreat_mostresp_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## prjharm_mostresp_devx21[1] 0.28 0.15 0.04 0.60 1.00
## prjharm_mostresp_devx21[2] 0.27 0.15 0.04 0.60 1.00
## prjharm_mostresp_devx21[3] 0.21 0.13 0.03 0.53 1.00
## prjharm_mostresp_devx21[4] 0.24 0.14 0.03 0.54 1.00
## prjharm_mostresp_av12x21[1] 0.12 0.08 0.02 0.32 1.00
## prjharm_mostresp_av12x21[2] 0.13 0.08 0.02 0.31 1.00
## prjharm_mostresp_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjharm_mostresp_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjharm_mostresp_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## prjharm_mostresp_av12x21[6] 0.12 0.08 0.02 0.32 1.00
## prjharm_mostresp_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## prjharm_mostresp_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mostresp_devx21[1] 0.25 0.14 0.04 0.56 1.00
## prjusedrg_mostresp_devx21[2] 0.29 0.15 0.05 0.61 1.00
## prjusedrg_mostresp_devx21[3] 0.26 0.14 0.04 0.58 1.00
## prjusedrg_mostresp_devx21[4] 0.20 0.12 0.03 0.50 1.00
## prjusedrg_mostresp_av12x21[1] 0.11 0.07 0.01 0.29 1.00
## prjusedrg_mostresp_av12x21[2] 0.11 0.07 0.01 0.29 1.00
## prjusedrg_mostresp_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mostresp_av12x21[4] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostresp_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## prjusedrg_mostresp_av12x21[6] 0.13 0.08 0.02 0.34 1.00
## prjusedrg_mostresp_av12x21[7] 0.13 0.08 0.02 0.34 1.00
## prjusedrg_mostresp_av12x21[8] 0.14 0.08 0.02 0.34 1.00
## prjhack_mostresp_devx21[1] 0.30 0.16 0.05 0.64 1.00
## prjhack_mostresp_devx21[2] 0.24 0.14 0.03 0.56 1.00
## prjhack_mostresp_devx21[3] 0.22 0.13 0.03 0.53 1.00
## prjhack_mostresp_devx21[4] 0.24 0.14 0.04 0.56 1.00
## prjhack_mostresp_av12x21[1] 0.12 0.08 0.02 0.30 1.00
## prjhack_mostresp_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## prjhack_mostresp_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## prjhack_mostresp_av12x21[4] 0.13 0.08 0.02 0.32 1.00
## prjhack_mostresp_av12x21[5] 0.13 0.08 0.02 0.32 1.00
## prjhack_mostresp_av12x21[6] 0.13 0.08 0.02 0.32 1.00
## prjhack_mostresp_av12x21[7] 0.12 0.08 0.01 0.31 1.00
## prjhack_mostresp_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_mostresp_devx21[1] 4602 2316
## prjthflt5_mostresp_devx21[2] 5270 2846
## prjthflt5_mostresp_devx21[3] 4659 2836
## prjthflt5_mostresp_devx21[4] 5018 2578
## prjthflt5_mostresp_av12x21[1] 4400 2295
## prjthflt5_mostresp_av12x21[2] 4950 2304
## prjthflt5_mostresp_av12x21[3] 4643 2370
## prjthflt5_mostresp_av12x21[4] 4609 2540
## prjthflt5_mostresp_av12x21[5] 5402 2349
## prjthflt5_mostresp_av12x21[6] 4988 2594
## prjthflt5_mostresp_av12x21[7] 5186 2818
## prjthflt5_mostresp_av12x21[8] 4756 2636
## prjthfgt5_mostresp_devx21[1] 5045 2615
## prjthfgt5_mostresp_devx21[2] 5290 2655
## prjthfgt5_mostresp_devx21[3] 4817 2761
## prjthfgt5_mostresp_devx21[4] 5062 2407
## prjthfgt5_mostresp_av12x21[1] 4724 2218
## prjthfgt5_mostresp_av12x21[2] 5619 2838
## prjthfgt5_mostresp_av12x21[3] 5814 2599
## prjthfgt5_mostresp_av12x21[4] 4962 2091
## prjthfgt5_mostresp_av12x21[5] 5491 2345
## prjthfgt5_mostresp_av12x21[6] 5528 2565
## prjthfgt5_mostresp_av12x21[7] 4036 2492
## prjthfgt5_mostresp_av12x21[8] 4909 2686
## prjthreat_mostresp_devx21[1] 5007 2363
## prjthreat_mostresp_devx21[2] 5009 2576
## prjthreat_mostresp_devx21[3] 4668 3217
## prjthreat_mostresp_devx21[4] 5189 2765
## prjthreat_mostresp_av12x21[1] 4334 2053
## prjthreat_mostresp_av12x21[2] 5111 2052
## prjthreat_mostresp_av12x21[3] 5791 2383
## prjthreat_mostresp_av12x21[4] 5631 2324
## prjthreat_mostresp_av12x21[5] 4952 2150
## prjthreat_mostresp_av12x21[6] 5229 2959
## prjthreat_mostresp_av12x21[7] 4523 2893
## prjthreat_mostresp_av12x21[8] 5467 2811
## prjharm_mostresp_devx21[1] 5046 2483
## prjharm_mostresp_devx21[2] 5544 2889
## prjharm_mostresp_devx21[3] 4937 3152
## prjharm_mostresp_devx21[4] 5134 2690
## prjharm_mostresp_av12x21[1] 4811 2390
## prjharm_mostresp_av12x21[2] 5712 2222
## prjharm_mostresp_av12x21[3] 5137 2628
## prjharm_mostresp_av12x21[4] 4574 2056
## prjharm_mostresp_av12x21[5] 4740 2170
## prjharm_mostresp_av12x21[6] 4225 2666
## prjharm_mostresp_av12x21[7] 5096 3176
## prjharm_mostresp_av12x21[8] 4518 2577
## prjusedrg_mostresp_devx21[1] 5132 2572
## prjusedrg_mostresp_devx21[2] 4121 2848
## prjusedrg_mostresp_devx21[3] 4920 2819
## prjusedrg_mostresp_devx21[4] 4811 2357
## prjusedrg_mostresp_av12x21[1] 4748 2102
## prjusedrg_mostresp_av12x21[2] 5172 2059
## prjusedrg_mostresp_av12x21[3] 3982 2057
## prjusedrg_mostresp_av12x21[4] 5958 2103
## prjusedrg_mostresp_av12x21[5] 5092 2663
## prjusedrg_mostresp_av12x21[6] 4955 2584
## prjusedrg_mostresp_av12x21[7] 5686 2820
## prjusedrg_mostresp_av12x21[8] 4791 2810
## prjhack_mostresp_devx21[1] 4959 2381
## prjhack_mostresp_devx21[2] 4833 2525
## prjhack_mostresp_devx21[3] 5036 2985
## prjhack_mostresp_devx21[4] 4744 2976
## prjhack_mostresp_av12x21[1] 5307 2175
## prjhack_mostresp_av12x21[2] 5627 2238
## prjhack_mostresp_av12x21[3] 4637 1944
## prjhack_mostresp_av12x21[4] 5006 2211
## prjhack_mostresp_av12x21[5] 4737 2430
## prjhack_mostresp_av12x21[6] 4646 2567
## prjhack_mostresp_av12x21[7] 5004 2745
## prjhack_mostresp_av12x21[8] 4925 2228
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.2.3.5 Prior summary
out.chg.prjcrime.stresp.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b prjhack
## normal(0, 0.125) b mostresp_av12x2 prjhack
## normal(0, 0.25) b mostresp_devx2 prjhack
## (flat) b prjharm
## normal(0, 0.125) b mostresp_av12x2 prjharm
## normal(0, 0.25) b mostresp_devx2 prjharm
## (flat) b prjthfgt5
## normal(0, 0.125) b mostresp_av12x2 prjthfgt5
## normal(0, 0.25) b mostresp_devx2 prjthfgt5
## (flat) b prjthflt5
## normal(0, 0.125) b mostresp_av12x2 prjthflt5
## normal(0, 0.25) b mostresp_devx2 prjthflt5
## (flat) b prjthreat
## normal(0, 0.125) b mostresp_av12x2 prjthreat
## normal(0, 0.25) b mostresp_devx2 prjthreat
## (flat) b prjusedrg
## normal(0, 0.125) b mostresp_av12x2 prjusedrg
## normal(0, 0.25) b mostresp_devx2 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 prjhack
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 prjhack
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 prjharm
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 prjharm
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 prjthfgt5
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 prjthfgt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 prjthflt5
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 prjthflt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 prjthreat
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 prjthreat
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 prjusedrg
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 prjusedrg
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.2.4 B/W Change Corr: stfair/crim intent
#Bivariate Change: criminal intent items ~ mo(stfair)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostfair_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostfair_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostfair_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostfair_av12x21',
resp = prjdv_names)
)
chg.prjcrime.stfair.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stfair_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stfair.fit <- ppchecks(chg.prjcrime.stfair.fit)6.2.2.4.1 Coefficient plot (intervals)
out.chg.prjcrime.stfair.fit[[10]]
6.2.2.4.2 Coefficient plot (distributions)
out.chg.prjcrime.stfair.fit[[9]]
6.2.2.4.3 PPcheck (density)
p1 <- out.chg.prjcrime.stfair.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.chg.prjcrime.stfair.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.chg.prjcrime.stfair.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.chg.prjcrime.stfair.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.chg.prjcrime.stfair.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.chg.prjcrime.stfair.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
6.2.2.4.4 Fit summary
out.chg.prjcrime.stfair.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## prjthfgt5 ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## prjthreat ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## prjharm ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## prjusedrg ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## prjhack ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.15 0.59 3.16 5.42 1.00 1561
## sd(prjthfgt5_Intercept) 3.67 0.53 2.74 4.80 1.00 1544
## sd(prjthreat_Intercept) 3.32 0.56 2.35 4.54 1.00 1592
## sd(prjharm_Intercept) 2.99 0.54 2.02 4.13 1.01 1511
## sd(prjusedrg_Intercept) 3.11 0.56 2.11 4.27 1.00 1532
## sd(prjhack_Intercept) 0.80 0.52 0.04 1.90 1.00 1144
## Tail_ESS
## sd(prjthflt5_Intercept) 2651
## sd(prjthfgt5_Intercept) 2481
## sd(prjthreat_Intercept) 2528
## sd(prjharm_Intercept) 2619
## sd(prjusedrg_Intercept) 2157
## sd(prjhack_Intercept) 1650
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_Intercept -6.16 0.78 -7.77 -4.75 1.00 3330
## prjthfgt5_Intercept -5.73 0.74 -7.21 -4.31 1.00 3281
## prjthreat_Intercept -6.32 0.90 -8.11 -4.66 1.00 2980
## prjharm_Intercept -5.70 0.88 -7.46 -4.07 1.00 2596
## prjusedrg_Intercept -6.20 0.91 -8.05 -4.51 1.00 2771
## prjhack_Intercept -4.30 0.67 -5.71 -3.09 1.00 2796
## prjthflt5_mostfair_devx2 0.03 0.18 -0.34 0.36 1.00 6884
## prjthflt5_mostfair_av12x2 0.17 0.08 0.01 0.33 1.00 3266
## prjthfgt5_mostfair_devx2 0.10 0.17 -0.26 0.42 1.00 6303
## prjthfgt5_mostfair_av12x2 0.11 0.08 -0.04 0.26 1.00 3415
## prjthreat_mostfair_devx2 -0.25 0.19 -0.63 0.11 1.00 6558
## prjthreat_mostfair_av12x2 0.14 0.08 -0.03 0.30 1.00 4742
## prjharm_mostfair_devx2 -0.24 0.19 -0.63 0.13 1.00 6230
## prjharm_mostfair_av12x2 0.03 0.08 -0.14 0.19 1.00 4747
## prjusedrg_mostfair_devx2 -0.22 0.20 -0.62 0.17 1.00 6495
## prjusedrg_mostfair_av12x2 0.09 0.08 -0.08 0.24 1.00 4770
## prjhack_mostfair_devx2 -0.22 0.19 -0.60 0.15 1.00 7087
## prjhack_mostfair_av12x2 0.12 0.07 -0.01 0.26 1.00 7155
## Tail_ESS
## prjthflt5_Intercept 3096
## prjthfgt5_Intercept 2972
## prjthreat_Intercept 2760
## prjharm_Intercept 2658
## prjusedrg_Intercept 3089
## prjhack_Intercept 2430
## prjthflt5_mostfair_devx2 3133
## prjthflt5_mostfair_av12x2 3215
## prjthfgt5_mostfair_devx2 3123
## prjthfgt5_mostfair_av12x2 3199
## prjthreat_mostfair_devx2 2967
## prjthreat_mostfair_av12x2 3454
## prjharm_mostfair_devx2 3191
## prjharm_mostfair_av12x2 3274
## prjusedrg_mostfair_devx2 2877
## prjusedrg_mostfair_av12x2 3320
## prjhack_mostfair_devx2 2814
## prjhack_mostfair_av12x2 3030
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_mostfair_devx21[1] 0.26 0.15 0.04 0.61 1.00
## prjthflt5_mostfair_devx21[2] 0.24 0.14 0.03 0.56 1.00
## prjthflt5_mostfair_devx21[3] 0.24 0.15 0.03 0.58 1.00
## prjthflt5_mostfair_devx21[4] 0.26 0.15 0.04 0.59 1.00
## prjthflt5_mostfair_av12x21[1] 0.12 0.08 0.01 0.30 1.00
## prjthflt5_mostfair_av12x21[2] 0.12 0.07 0.02 0.30 1.00
## prjthflt5_mostfair_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## prjthflt5_mostfair_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## prjthflt5_mostfair_av12x21[5] 0.12 0.07 0.02 0.30 1.00
## prjthflt5_mostfair_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## prjthflt5_mostfair_av12x21[7] 0.14 0.08 0.02 0.34 1.00
## prjthflt5_mostfair_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## prjthfgt5_mostfair_devx21[1] 0.24 0.14 0.03 0.57 1.00
## prjthfgt5_mostfair_devx21[2] 0.24 0.15 0.03 0.58 1.00
## prjthfgt5_mostfair_devx21[3] 0.27 0.15 0.04 0.59 1.00
## prjthfgt5_mostfair_devx21[4] 0.25 0.14 0.04 0.57 1.00
## prjthfgt5_mostfair_av12x21[1] 0.13 0.09 0.02 0.34 1.00
## prjthfgt5_mostfair_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## prjthfgt5_mostfair_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## prjthfgt5_mostfair_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## prjthfgt5_mostfair_av12x21[5] 0.12 0.08 0.01 0.30 1.00
## prjthfgt5_mostfair_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## prjthfgt5_mostfair_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## prjthfgt5_mostfair_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## prjthreat_mostfair_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjthreat_mostfair_devx21[2] 0.25 0.14 0.04 0.57 1.00
## prjthreat_mostfair_devx21[3] 0.25 0.14 0.04 0.56 1.00
## prjthreat_mostfair_devx21[4] 0.23 0.14 0.03 0.55 1.00
## prjthreat_mostfair_av12x21[1] 0.12 0.08 0.02 0.30 1.00
## prjthreat_mostfair_av12x21[2] 0.11 0.07 0.02 0.28 1.00
## prjthreat_mostfair_av12x21[3] 0.12 0.08 0.02 0.30 1.00
## prjthreat_mostfair_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mostfair_av12x21[5] 0.12 0.08 0.02 0.32 1.00
## prjthreat_mostfair_av12x21[6] 0.13 0.08 0.02 0.31 1.00
## prjthreat_mostfair_av12x21[7] 0.13 0.09 0.02 0.33 1.00
## prjthreat_mostfair_av12x21[8] 0.14 0.09 0.02 0.34 1.00
## prjharm_mostfair_devx21[1] 0.26 0.15 0.04 0.59 1.00
## prjharm_mostfair_devx21[2] 0.27 0.15 0.04 0.59 1.00
## prjharm_mostfair_devx21[3] 0.23 0.14 0.04 0.55 1.00
## prjharm_mostfair_devx21[4] 0.24 0.14 0.03 0.55 1.00
## prjharm_mostfair_av12x21[1] 0.13 0.08 0.02 0.31 1.00
## prjharm_mostfair_av12x21[2] 0.12 0.08 0.02 0.32 1.00
## prjharm_mostfair_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## prjharm_mostfair_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## prjharm_mostfair_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## prjharm_mostfair_av12x21[6] 0.12 0.08 0.02 0.32 1.00
## prjharm_mostfair_av12x21[7] 0.12 0.08 0.01 0.31 1.00
## prjharm_mostfair_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostfair_devx21[1] 0.27 0.15 0.04 0.60 1.00
## prjusedrg_mostfair_devx21[2] 0.26 0.14 0.04 0.58 1.00
## prjusedrg_mostfair_devx21[3] 0.23 0.13 0.04 0.54 1.00
## prjusedrg_mostfair_devx21[4] 0.24 0.14 0.04 0.56 1.00
## prjusedrg_mostfair_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mostfair_av12x21[2] 0.12 0.08 0.02 0.30 1.00
## prjusedrg_mostfair_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostfair_av12x21[4] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostfair_av12x21[5] 0.12 0.08 0.02 0.33 1.00
## prjusedrg_mostfair_av12x21[6] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostfair_av12x21[7] 0.13 0.08 0.02 0.33 1.00
## prjusedrg_mostfair_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## prjhack_mostfair_devx21[1] 0.25 0.14 0.04 0.57 1.00
## prjhack_mostfair_devx21[2] 0.28 0.14 0.05 0.59 1.00
## prjhack_mostfair_devx21[3] 0.23 0.13 0.03 0.54 1.00
## prjhack_mostfair_devx21[4] 0.24 0.14 0.04 0.55 1.00
## prjhack_mostfair_av12x21[1] 0.11 0.07 0.02 0.30 1.00
## prjhack_mostfair_av12x21[2] 0.11 0.07 0.02 0.30 1.00
## prjhack_mostfair_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjhack_mostfair_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## prjhack_mostfair_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## prjhack_mostfair_av12x21[6] 0.13 0.08 0.02 0.32 1.00
## prjhack_mostfair_av12x21[7] 0.14 0.09 0.02 0.35 1.00
## prjhack_mostfair_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_mostfair_devx21[1] 7336 2401
## prjthflt5_mostfair_devx21[2] 7231 2442
## prjthflt5_mostfair_devx21[3] 6693 3268
## prjthflt5_mostfair_devx21[4] 6561 2750
## prjthflt5_mostfair_av12x21[1] 6450 2516
## prjthflt5_mostfair_av12x21[2] 6665 2585
## prjthflt5_mostfair_av12x21[3] 6714 2647
## prjthflt5_mostfair_av12x21[4] 6950 2721
## prjthflt5_mostfair_av12x21[5] 6215 2761
## prjthflt5_mostfair_av12x21[6] 7372 2276
## prjthflt5_mostfair_av12x21[7] 6887 2418
## prjthflt5_mostfair_av12x21[8] 6056 3192
## prjthfgt5_mostfair_devx21[1] 6068 2143
## prjthfgt5_mostfair_devx21[2] 8710 2919
## prjthfgt5_mostfair_devx21[3] 6349 3114
## prjthfgt5_mostfair_devx21[4] 6522 2884
## prjthfgt5_mostfair_av12x21[1] 7706 2727
## prjthfgt5_mostfair_av12x21[2] 7453 2553
## prjthfgt5_mostfair_av12x21[3] 6666 2380
## prjthfgt5_mostfair_av12x21[4] 7902 2690
## prjthfgt5_mostfair_av12x21[5] 7653 2986
## prjthfgt5_mostfair_av12x21[6] 7116 2403
## prjthfgt5_mostfair_av12x21[7] 7410 3357
## prjthfgt5_mostfair_av12x21[8] 6570 3244
## prjthreat_mostfair_devx21[1] 7575 2498
## prjthreat_mostfair_devx21[2] 7117 2986
## prjthreat_mostfair_devx21[3] 7333 2659
## prjthreat_mostfair_devx21[4] 7202 2444
## prjthreat_mostfair_av12x21[1] 7582 2472
## prjthreat_mostfair_av12x21[2] 7171 2611
## prjthreat_mostfair_av12x21[3] 6808 2467
## prjthreat_mostfair_av12x21[4] 7427 2491
## prjthreat_mostfair_av12x21[5] 6643 2569
## prjthreat_mostfair_av12x21[6] 6902 2989
## prjthreat_mostfair_av12x21[7] 5957 2740
## prjthreat_mostfair_av12x21[8] 6993 2927
## prjharm_mostfair_devx21[1] 8287 2693
## prjharm_mostfair_devx21[2] 6890 2993
## prjharm_mostfair_devx21[3] 5944 2781
## prjharm_mostfair_devx21[4] 7384 2608
## prjharm_mostfair_av12x21[1] 7642 2316
## prjharm_mostfair_av12x21[2] 7001 2029
## prjharm_mostfair_av12x21[3] 6844 2376
## prjharm_mostfair_av12x21[4] 6889 2541
## prjharm_mostfair_av12x21[5] 7121 2426
## prjharm_mostfair_av12x21[6] 8360 3025
## prjharm_mostfair_av12x21[7] 7521 2281
## prjharm_mostfair_av12x21[8] 6629 2652
## prjusedrg_mostfair_devx21[1] 8845 2356
## prjusedrg_mostfair_devx21[2] 7628 3116
## prjusedrg_mostfair_devx21[3] 7193 3114
## prjusedrg_mostfair_devx21[4] 7044 2773
## prjusedrg_mostfair_av12x21[1] 7642 2931
## prjusedrg_mostfair_av12x21[2] 6148 2410
## prjusedrg_mostfair_av12x21[3] 8072 2589
## prjusedrg_mostfair_av12x21[4] 7342 2457
## prjusedrg_mostfair_av12x21[5] 6792 2409
## prjusedrg_mostfair_av12x21[6] 6221 2987
## prjusedrg_mostfair_av12x21[7] 7765 2936
## prjusedrg_mostfair_av12x21[8] 6578 2837
## prjhack_mostfair_devx21[1] 6770 2785
## prjhack_mostfair_devx21[2] 7604 2312
## prjhack_mostfair_devx21[3] 6501 2981
## prjhack_mostfair_devx21[4] 7805 2710
## prjhack_mostfair_av12x21[1] 6544 2552
## prjhack_mostfair_av12x21[2] 7768 2876
## prjhack_mostfair_av12x21[3] 8069 2226
## prjhack_mostfair_av12x21[4] 6435 2328
## prjhack_mostfair_av12x21[5] 7735 2770
## prjhack_mostfair_av12x21[6] 7025 2538
## prjhack_mostfair_av12x21[7] 7469 2584
## prjhack_mostfair_av12x21[8] 6677 2969
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.2.4.5 Prior summary
out.chg.prjcrime.stfair.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b prjhack
## normal(0, 0.125) b mostfair_av12x2 prjhack
## normal(0, 0.25) b mostfair_devx2 prjhack
## (flat) b prjharm
## normal(0, 0.125) b mostfair_av12x2 prjharm
## normal(0, 0.25) b mostfair_devx2 prjharm
## (flat) b prjthfgt5
## normal(0, 0.125) b mostfair_av12x2 prjthfgt5
## normal(0, 0.25) b mostfair_devx2 prjthfgt5
## (flat) b prjthflt5
## normal(0, 0.125) b mostfair_av12x2 prjthflt5
## normal(0, 0.25) b mostfair_devx2 prjthflt5
## (flat) b prjthreat
## normal(0, 0.125) b mostfair_av12x2 prjthreat
## normal(0, 0.25) b mostfair_devx2 prjthreat
## (flat) b prjusedrg
## normal(0, 0.125) b mostfair_av12x2 prjusedrg
## normal(0, 0.25) b mostfair_devx2 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 prjhack
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 prjhack
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 prjharm
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 prjharm
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 prjthfgt5
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 prjthfgt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 prjthflt5
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 prjthflt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 prjthreat
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 prjthreat
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 prjusedrg
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 prjusedrg
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.2.5 B/W Change Corr: stjob/crim intent
#Bivariate Change: criminal intent items ~ mo(stjob)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostjob_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostjob_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostjob_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostjob_av12x21',
resp = prjdv_names)
)
chg.prjcrime.stjob.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stjob_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stjob.fit <- ppchecks(chg.prjcrime.stjob.fit)6.2.2.5.1 Coefficient plot (intervals)
out.chg.prjcrime.stjob.fit[[10]]
6.2.2.5.2 Coefficient plot (distributions)
out.chg.prjcrime.stjob.fit[[9]]
6.2.2.5.3 PPcheck (density)
p1 <- out.chg.prjcrime.stjob.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.chg.prjcrime.stjob.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.chg.prjcrime.stjob.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.chg.prjcrime.stjob.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.chg.prjcrime.stjob.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.chg.prjcrime.stjob.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
6.2.2.5.4 Fit summary
out.chg.prjcrime.stjob.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## prjthfgt5 ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## prjthreat ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## prjharm ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## prjusedrg ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## prjhack ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.20 0.59 3.16 5.47 1.00 1179
## sd(prjthfgt5_Intercept) 3.51 0.51 2.62 4.62 1.01 1051
## sd(prjthreat_Intercept) 3.02 0.52 2.16 4.17 1.00 1275
## sd(prjharm_Intercept) 2.82 0.52 1.90 3.94 1.00 1363
## sd(prjusedrg_Intercept) 2.79 0.52 1.88 3.90 1.00 1191
## sd(prjhack_Intercept) 0.58 0.40 0.02 1.52 1.00 971
## Tail_ESS
## sd(prjthflt5_Intercept) 2150
## sd(prjthfgt5_Intercept) 1979
## sd(prjthreat_Intercept) 2132
## sd(prjharm_Intercept) 2111
## sd(prjusedrg_Intercept) 2069
## sd(prjhack_Intercept) 1566
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_Intercept -5.98 0.77 -7.55 -4.53 1.00 1706
## prjthfgt5_Intercept -5.74 0.72 -7.27 -4.42 1.00 1815
## prjthreat_Intercept -6.67 0.83 -8.43 -5.10 1.00 2328
## prjharm_Intercept -6.12 0.86 -7.90 -4.51 1.00 2148
## prjusedrg_Intercept -6.56 0.88 -8.40 -4.94 1.00 1681
## prjhack_Intercept -5.36 0.67 -6.73 -4.12 1.00 2487
## prjthflt5_mostjob_devx2 -0.03 0.18 -0.40 0.31 1.00 3930
## prjthflt5_mostjob_av12x2 0.16 0.08 -0.01 0.32 1.00 2083
## prjthfgt5_mostjob_devx2 0.06 0.18 -0.30 0.40 1.00 4051
## prjthfgt5_mostjob_av12x2 0.17 0.07 0.02 0.31 1.00 2548
## prjthreat_mostjob_devx2 -0.05 0.20 -0.45 0.32 1.00 4262
## prjthreat_mostjob_av12x2 0.22 0.08 0.05 0.37 1.00 2692
## prjharm_mostjob_devx2 -0.08 0.19 -0.47 0.29 1.00 4371
## prjharm_mostjob_av12x2 0.10 0.08 -0.06 0.26 1.00 3598
## prjusedrg_mostjob_devx2 -0.06 0.20 -0.49 0.31 1.00 4089
## prjusedrg_mostjob_av12x2 0.19 0.08 0.02 0.34 1.00 3116
## prjhack_mostjob_devx2 -0.00 0.19 -0.40 0.37 1.00 4063
## prjhack_mostjob_av12x2 0.28 0.07 0.14 0.42 1.00 4350
## Tail_ESS
## prjthflt5_Intercept 2474
## prjthfgt5_Intercept 2539
## prjthreat_Intercept 2889
## prjharm_Intercept 2633
## prjusedrg_Intercept 2421
## prjhack_Intercept 2443
## prjthflt5_mostjob_devx2 2957
## prjthflt5_mostjob_av12x2 2202
## prjthfgt5_mostjob_devx2 2792
## prjthfgt5_mostjob_av12x2 2964
## prjthreat_mostjob_devx2 2589
## prjthreat_mostjob_av12x2 2785
## prjharm_mostjob_devx2 2827
## prjharm_mostjob_av12x2 3148
## prjusedrg_mostjob_devx2 2664
## prjusedrg_mostjob_av12x2 2684
## prjhack_mostjob_devx2 2575
## prjhack_mostjob_av12x2 2715
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_mostjob_devx21[1] 0.27 0.15 0.04 0.59 1.00 5099
## prjthflt5_mostjob_devx21[2] 0.23 0.14 0.03 0.55 1.00 4848
## prjthflt5_mostjob_devx21[3] 0.24 0.14 0.04 0.57 1.00 4336
## prjthflt5_mostjob_devx21[4] 0.26 0.15 0.04 0.60 1.00 5592
## prjthflt5_mostjob_av12x21[1] 0.11 0.07 0.01 0.28 1.00 4731
## prjthflt5_mostjob_av12x21[2] 0.11 0.07 0.01 0.27 1.00 4168
## prjthflt5_mostjob_av12x21[3] 0.11 0.07 0.02 0.28 1.00 4983
## prjthflt5_mostjob_av12x21[4] 0.12 0.07 0.02 0.30 1.00 5839
## prjthflt5_mostjob_av12x21[5] 0.13 0.08 0.02 0.33 1.00 5172
## prjthflt5_mostjob_av12x21[6] 0.15 0.09 0.02 0.37 1.00 5198
## prjthflt5_mostjob_av12x21[7] 0.14 0.08 0.02 0.34 1.00 4514
## prjthflt5_mostjob_av12x21[8] 0.14 0.08 0.02 0.34 1.00 5839
## prjthfgt5_mostjob_devx21[1] 0.25 0.15 0.04 0.59 1.00 5072
## prjthfgt5_mostjob_devx21[2] 0.25 0.15 0.03 0.59 1.00 4473
## prjthfgt5_mostjob_devx21[3] 0.23 0.14 0.04 0.56 1.00 5855
## prjthfgt5_mostjob_devx21[4] 0.26 0.15 0.04 0.59 1.00 6068
## prjthfgt5_mostjob_av12x21[1] 0.11 0.07 0.01 0.27 1.00 4948
## prjthfgt5_mostjob_av12x21[2] 0.11 0.08 0.01 0.30 1.00 4866
## prjthfgt5_mostjob_av12x21[3] 0.11 0.07 0.01 0.29 1.00 4945
## prjthfgt5_mostjob_av12x21[4] 0.12 0.08 0.02 0.31 1.00 4359
## prjthfgt5_mostjob_av12x21[5] 0.13 0.08 0.02 0.33 1.00 4992
## prjthfgt5_mostjob_av12x21[6] 0.15 0.09 0.02 0.36 1.00 4322
## prjthfgt5_mostjob_av12x21[7] 0.13 0.08 0.02 0.32 1.00 5535
## prjthfgt5_mostjob_av12x21[8] 0.15 0.09 0.02 0.36 1.00 4849
## prjthreat_mostjob_devx21[1] 0.26 0.15 0.04 0.59 1.00 4985
## prjthreat_mostjob_devx21[2] 0.24 0.14 0.03 0.57 1.00 5611
## prjthreat_mostjob_devx21[3] 0.24 0.14 0.03 0.55 1.00 4806
## prjthreat_mostjob_devx21[4] 0.26 0.15 0.04 0.60 1.00 6292
## prjthreat_mostjob_av12x21[1] 0.10 0.07 0.01 0.26 1.00 5422
## prjthreat_mostjob_av12x21[2] 0.11 0.07 0.02 0.27 1.00 6010
## prjthreat_mostjob_av12x21[3] 0.11 0.07 0.01 0.28 1.00 4952
## prjthreat_mostjob_av12x21[4] 0.11 0.07 0.01 0.30 1.00 4669
## prjthreat_mostjob_av12x21[5] 0.12 0.08 0.02 0.30 1.00 4783
## prjthreat_mostjob_av12x21[6] 0.14 0.09 0.02 0.35 1.00 5168
## prjthreat_mostjob_av12x21[7] 0.17 0.10 0.02 0.39 1.00 4755
## prjthreat_mostjob_av12x21[8] 0.15 0.09 0.02 0.35 1.00 5603
## prjharm_mostjob_devx21[1] 0.27 0.15 0.04 0.60 1.00 5206
## prjharm_mostjob_devx21[2] 0.24 0.14 0.04 0.57 1.00 5196
## prjharm_mostjob_devx21[3] 0.23 0.14 0.03 0.55 1.00 4964
## prjharm_mostjob_devx21[4] 0.26 0.15 0.04 0.59 1.00 5910
## prjharm_mostjob_av12x21[1] 0.12 0.08 0.02 0.31 1.00 4750
## prjharm_mostjob_av12x21[2] 0.12 0.08 0.01 0.32 1.00 5279
## prjharm_mostjob_av12x21[3] 0.12 0.08 0.02 0.31 1.00 5591
## prjharm_mostjob_av12x21[4] 0.13 0.08 0.02 0.32 1.00 5412
## prjharm_mostjob_av12x21[5] 0.12 0.08 0.02 0.31 1.00 5602
## prjharm_mostjob_av12x21[6] 0.13 0.09 0.02 0.34 1.00 4882
## prjharm_mostjob_av12x21[7] 0.13 0.09 0.02 0.34 1.00 5272
## prjharm_mostjob_av12x21[8] 0.12 0.08 0.01 0.30 1.00 5272
## prjusedrg_mostjob_devx21[1] 0.27 0.15 0.04 0.61 1.00 5290
## prjusedrg_mostjob_devx21[2] 0.24 0.14 0.04 0.58 1.00 5360
## prjusedrg_mostjob_devx21[3] 0.24 0.14 0.03 0.57 1.00 5531
## prjusedrg_mostjob_devx21[4] 0.26 0.14 0.04 0.58 1.00 5154
## prjusedrg_mostjob_av12x21[1] 0.10 0.07 0.01 0.27 1.00 5866
## prjusedrg_mostjob_av12x21[2] 0.11 0.07 0.02 0.29 1.00 5177
## prjusedrg_mostjob_av12x21[3] 0.12 0.08 0.02 0.30 1.00 5454
## prjusedrg_mostjob_av12x21[4] 0.11 0.07 0.02 0.29 1.00 5009
## prjusedrg_mostjob_av12x21[5] 0.13 0.08 0.01 0.33 1.00 6000
## prjusedrg_mostjob_av12x21[6] 0.14 0.09 0.02 0.35 1.00 3991
## prjusedrg_mostjob_av12x21[7] 0.16 0.10 0.02 0.40 1.00 4721
## prjusedrg_mostjob_av12x21[8] 0.12 0.08 0.02 0.31 1.00 5507
## prjhack_mostjob_devx21[1] 0.26 0.15 0.04 0.59 1.00 4924
## prjhack_mostjob_devx21[2] 0.24 0.14 0.03 0.55 1.00 6635
## prjhack_mostjob_devx21[3] 0.24 0.14 0.04 0.57 1.00 4414
## prjhack_mostjob_devx21[4] 0.26 0.14 0.04 0.57 1.00 5279
## prjhack_mostjob_av12x21[1] 0.09 0.06 0.01 0.24 1.00 4669
## prjhack_mostjob_av12x21[2] 0.09 0.06 0.01 0.24 1.00 4722
## prjhack_mostjob_av12x21[3] 0.11 0.07 0.01 0.26 1.00 4197
## prjhack_mostjob_av12x21[4] 0.10 0.07 0.01 0.26 1.00 5162
## prjhack_mostjob_av12x21[5] 0.12 0.08 0.02 0.31 1.00 5771
## prjhack_mostjob_av12x21[6] 0.15 0.09 0.02 0.36 1.00 4407
## prjhack_mostjob_av12x21[7] 0.20 0.11 0.03 0.43 1.00 4974
## prjhack_mostjob_av12x21[8] 0.15 0.08 0.02 0.35 1.00 4889
## Tail_ESS
## prjthflt5_mostjob_devx21[1] 2926
## prjthflt5_mostjob_devx21[2] 2400
## prjthflt5_mostjob_devx21[3] 2944
## prjthflt5_mostjob_devx21[4] 2821
## prjthflt5_mostjob_av12x21[1] 2165
## prjthflt5_mostjob_av12x21[2] 2283
## prjthflt5_mostjob_av12x21[3] 2600
## prjthflt5_mostjob_av12x21[4] 2822
## prjthflt5_mostjob_av12x21[5] 2603
## prjthflt5_mostjob_av12x21[6] 2932
## prjthflt5_mostjob_av12x21[7] 2995
## prjthflt5_mostjob_av12x21[8] 2981
## prjthfgt5_mostjob_devx21[1] 2472
## prjthfgt5_mostjob_devx21[2] 2106
## prjthfgt5_mostjob_devx21[3] 2812
## prjthfgt5_mostjob_devx21[4] 2699
## prjthfgt5_mostjob_av12x21[1] 2692
## prjthfgt5_mostjob_av12x21[2] 2083
## prjthfgt5_mostjob_av12x21[3] 2619
## prjthfgt5_mostjob_av12x21[4] 2445
## prjthfgt5_mostjob_av12x21[5] 2381
## prjthfgt5_mostjob_av12x21[6] 2473
## prjthfgt5_mostjob_av12x21[7] 3008
## prjthfgt5_mostjob_av12x21[8] 2957
## prjthreat_mostjob_devx21[1] 2930
## prjthreat_mostjob_devx21[2] 2292
## prjthreat_mostjob_devx21[3] 3020
## prjthreat_mostjob_devx21[4] 3038
## prjthreat_mostjob_av12x21[1] 2367
## prjthreat_mostjob_av12x21[2] 2527
## prjthreat_mostjob_av12x21[3] 2089
## prjthreat_mostjob_av12x21[4] 2652
## prjthreat_mostjob_av12x21[5] 2421
## prjthreat_mostjob_av12x21[6] 2614
## prjthreat_mostjob_av12x21[7] 2782
## prjthreat_mostjob_av12x21[8] 3177
## prjharm_mostjob_devx21[1] 2458
## prjharm_mostjob_devx21[2] 2185
## prjharm_mostjob_devx21[3] 2888
## prjharm_mostjob_devx21[4] 3014
## prjharm_mostjob_av12x21[1] 2352
## prjharm_mostjob_av12x21[2] 1992
## prjharm_mostjob_av12x21[3] 2689
## prjharm_mostjob_av12x21[4] 2614
## prjharm_mostjob_av12x21[5] 2687
## prjharm_mostjob_av12x21[6] 2668
## prjharm_mostjob_av12x21[7] 2392
## prjharm_mostjob_av12x21[8] 3008
## prjusedrg_mostjob_devx21[1] 2395
## prjusedrg_mostjob_devx21[2] 2587
## prjusedrg_mostjob_devx21[3] 2913
## prjusedrg_mostjob_devx21[4] 2786
## prjusedrg_mostjob_av12x21[1] 2532
## prjusedrg_mostjob_av12x21[2] 2603
## prjusedrg_mostjob_av12x21[3] 2707
## prjusedrg_mostjob_av12x21[4] 2476
## prjusedrg_mostjob_av12x21[5] 2614
## prjusedrg_mostjob_av12x21[6] 2698
## prjusedrg_mostjob_av12x21[7] 2916
## prjusedrg_mostjob_av12x21[8] 3089
## prjhack_mostjob_devx21[1] 2650
## prjhack_mostjob_devx21[2] 2951
## prjhack_mostjob_devx21[3] 2920
## prjhack_mostjob_devx21[4] 2746
## prjhack_mostjob_av12x21[1] 2188
## prjhack_mostjob_av12x21[2] 2520
## prjhack_mostjob_av12x21[3] 2288
## prjhack_mostjob_av12x21[4] 2287
## prjhack_mostjob_av12x21[5] 2368
## prjhack_mostjob_av12x21[6] 2287
## prjhack_mostjob_av12x21[7] 2932
## prjhack_mostjob_av12x21[8] 2732
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.2.5.5 Prior summary
out.chg.prjcrime.stjob.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b prjhack
## normal(0, 0.125) b mostjob_av12x2 prjhack
## normal(0, 0.25) b mostjob_devx2 prjhack
## (flat) b prjharm
## normal(0, 0.125) b mostjob_av12x2 prjharm
## normal(0, 0.25) b mostjob_devx2 prjharm
## (flat) b prjthfgt5
## normal(0, 0.125) b mostjob_av12x2 prjthfgt5
## normal(0, 0.25) b mostjob_devx2 prjthfgt5
## (flat) b prjthflt5
## normal(0, 0.125) b mostjob_av12x2 prjthflt5
## normal(0, 0.25) b mostjob_devx2 prjthflt5
## (flat) b prjthreat
## normal(0, 0.125) b mostjob_av12x2 prjthreat
## normal(0, 0.25) b mostjob_devx2 prjthreat
## (flat) b prjusedrg
## normal(0, 0.125) b mostjob_av12x2 prjusedrg
## normal(0, 0.25) b mostjob_devx2 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 prjhack
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 prjhack
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 prjharm
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 prjharm
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 prjthfgt5
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 prjthfgt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 prjthflt5
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 prjthflt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 prjthreat
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 prjthreat
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 prjusedrg
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 prjusedrg
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.2.6 B/W Change Corr: stthft/crim intent
#Bivariate Change: criminal intent items ~ mo(stthft)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostthft_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostthft_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostthft_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostthft_av12x21',
resp = prjdv_names)
)
chg.prjcrime.stthft.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stthft_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stthft.fit <- ppchecks(chg.prjcrime.stthft.fit)6.2.2.6.1 Coefficient plot (intervals)
out.chg.prjcrime.stthft.fit[[10]]
6.2.2.6.2 Coefficient plot (distributions)
out.chg.prjcrime.stthft.fit[[9]]
6.2.2.6.3 PPcheck (density)
p1 <- out.chg.prjcrime.stthft.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.chg.prjcrime.stthft.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.chg.prjcrime.stthft.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.chg.prjcrime.stthft.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.chg.prjcrime.stthft.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.chg.prjcrime.stthft.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
6.2.2.6.4 Fit summary
out.chg.prjcrime.stthft.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## prjthfgt5 ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## prjthreat ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## prjharm ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## prjusedrg ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## prjhack ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.40 0.62 3.30 5.74 1.00 1311
## sd(prjthfgt5_Intercept) 3.60 0.52 2.68 4.71 1.00 1245
## sd(prjthreat_Intercept) 3.36 0.57 2.36 4.59 1.00 1557
## sd(prjharm_Intercept) 2.99 0.55 2.04 4.17 1.00 1222
## sd(prjusedrg_Intercept) 3.22 0.56 2.19 4.40 1.00 1639
## sd(prjhack_Intercept) 0.91 0.56 0.04 2.08 1.00 1105
## Tail_ESS
## sd(prjthflt5_Intercept) 2389
## sd(prjthfgt5_Intercept) 2481
## sd(prjthreat_Intercept) 2059
## sd(prjharm_Intercept) 2254
## sd(prjusedrg_Intercept) 2138
## sd(prjhack_Intercept) 2049
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_Intercept -5.53 0.77 -7.12 -4.12 1.00 2103
## prjthfgt5_Intercept -5.26 0.71 -6.69 -3.92 1.00 2475
## prjthreat_Intercept -5.96 0.87 -7.74 -4.35 1.00 2846
## prjharm_Intercept -5.44 0.87 -7.15 -3.75 1.00 2459
## prjusedrg_Intercept -6.00 0.90 -7.83 -4.34 1.00 2401
## prjhack_Intercept -3.84 0.70 -5.37 -2.56 1.00 2277
## prjthflt5_mostthft_devx2 -0.12 0.18 -0.48 0.22 1.00 6388
## prjthflt5_mostthft_av12x2 0.11 0.09 -0.07 0.29 1.00 3073
## prjthfgt5_mostthft_devx2 -0.06 0.18 -0.41 0.30 1.00 5419
## prjthfgt5_mostthft_av12x2 0.17 0.08 0.01 0.34 1.00 3873
## prjthreat_mostthft_devx2 -0.26 0.20 -0.66 0.11 1.00 6454
## prjthreat_mostthft_av12x2 0.13 0.09 -0.06 0.31 1.00 4625
## prjharm_mostthft_devx2 -0.36 0.19 -0.74 0.01 1.00 6401
## prjharm_mostthft_av12x2 0.07 0.09 -0.11 0.24 1.00 4966
## prjusedrg_mostthft_devx2 -0.13 0.21 -0.55 0.27 1.00 6403
## prjusedrg_mostthft_av12x2 -0.03 0.09 -0.22 0.15 1.00 5010
## prjhack_mostthft_devx2 -0.19 0.20 -0.57 0.21 1.00 6871
## prjhack_mostthft_av12x2 0.02 0.08 -0.14 0.17 1.00 6406
## Tail_ESS
## prjthflt5_Intercept 3025
## prjthfgt5_Intercept 2754
## prjthreat_Intercept 2956
## prjharm_Intercept 2778
## prjusedrg_Intercept 2727
## prjhack_Intercept 2174
## prjthflt5_mostthft_devx2 2645
## prjthflt5_mostthft_av12x2 2917
## prjthfgt5_mostthft_devx2 2847
## prjthfgt5_mostthft_av12x2 3106
## prjthreat_mostthft_devx2 3155
## prjthreat_mostthft_av12x2 3371
## prjharm_mostthft_devx2 3251
## prjharm_mostthft_av12x2 3019
## prjusedrg_mostthft_devx2 2570
## prjusedrg_mostthft_av12x2 3269
## prjhack_mostthft_devx2 2947
## prjhack_mostthft_av12x2 3045
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_mostthft_devx21[1] 0.27 0.15 0.04 0.62 1.00
## prjthflt5_mostthft_devx21[2] 0.25 0.14 0.04 0.57 1.00
## prjthflt5_mostthft_devx21[3] 0.23 0.14 0.03 0.55 1.00
## prjthflt5_mostthft_devx21[4] 0.25 0.15 0.03 0.58 1.00
## prjthflt5_mostthft_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## prjthflt5_mostthft_av12x21[2] 0.12 0.07 0.02 0.30 1.00
## prjthflt5_mostthft_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjthflt5_mostthft_av12x21[4] 0.12 0.08 0.01 0.32 1.00
## prjthflt5_mostthft_av12x21[5] 0.14 0.09 0.02 0.36 1.00
## prjthflt5_mostthft_av12x21[6] 0.14 0.08 0.02 0.34 1.00
## prjthflt5_mostthft_av12x21[7] 0.12 0.07 0.02 0.29 1.00
## prjthflt5_mostthft_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## prjthfgt5_mostthft_devx21[1] 0.27 0.15 0.04 0.62 1.00
## prjthfgt5_mostthft_devx21[2] 0.24 0.14 0.04 0.57 1.00
## prjthfgt5_mostthft_devx21[3] 0.23 0.14 0.03 0.56 1.00
## prjthfgt5_mostthft_devx21[4] 0.26 0.15 0.04 0.60 1.00
## prjthfgt5_mostthft_av12x21[1] 0.11 0.07 0.01 0.30 1.00
## prjthfgt5_mostthft_av12x21[2] 0.11 0.07 0.01 0.30 1.00
## prjthfgt5_mostthft_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjthfgt5_mostthft_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## prjthfgt5_mostthft_av12x21[5] 0.15 0.09 0.02 0.36 1.00
## prjthfgt5_mostthft_av12x21[6] 0.14 0.09 0.02 0.34 1.00
## prjthfgt5_mostthft_av12x21[7] 0.12 0.08 0.02 0.30 1.00
## prjthfgt5_mostthft_av12x21[8] 0.11 0.07 0.02 0.29 1.00
## prjthreat_mostthft_devx21[1] 0.28 0.16 0.04 0.62 1.00
## prjthreat_mostthft_devx21[2] 0.28 0.15 0.05 0.60 1.00
## prjthreat_mostthft_devx21[3] 0.20 0.13 0.02 0.50 1.00
## prjthreat_mostthft_devx21[4] 0.24 0.14 0.03 0.55 1.00
## prjthreat_mostthft_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mostthft_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mostthft_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## prjthreat_mostthft_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjthreat_mostthft_av12x21[5] 0.14 0.09 0.02 0.35 1.00
## prjthreat_mostthft_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## prjthreat_mostthft_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## prjthreat_mostthft_av12x21[8] 0.11 0.07 0.01 0.29 1.00
## prjharm_mostthft_devx21[1] 0.26 0.15 0.04 0.60 1.00
## prjharm_mostthft_devx21[2] 0.30 0.15 0.05 0.63 1.00
## prjharm_mostthft_devx21[3] 0.22 0.13 0.03 0.52 1.00
## prjharm_mostthft_devx21[4] 0.22 0.13 0.03 0.51 1.00
## prjharm_mostthft_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## prjharm_mostthft_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## prjharm_mostthft_av12x21[3] 0.12 0.08 0.02 0.30 1.00
## prjharm_mostthft_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## prjharm_mostthft_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## prjharm_mostthft_av12x21[6] 0.12 0.08 0.01 0.32 1.00
## prjharm_mostthft_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## prjharm_mostthft_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## prjusedrg_mostthft_devx21[1] 0.28 0.16 0.04 0.63 1.00
## prjusedrg_mostthft_devx21[2] 0.24 0.14 0.04 0.56 1.00
## prjusedrg_mostthft_devx21[3] 0.23 0.14 0.03 0.54 1.00
## prjusedrg_mostthft_devx21[4] 0.25 0.15 0.04 0.59 1.00
## prjusedrg_mostthft_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostthft_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## prjusedrg_mostthft_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjusedrg_mostthft_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mostthft_av12x21[5] 0.12 0.08 0.02 0.32 1.00
## prjusedrg_mostthft_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## prjusedrg_mostthft_av12x21[7] 0.13 0.08 0.02 0.33 1.00
## prjusedrg_mostthft_av12x21[8] 0.13 0.09 0.02 0.34 1.00
## prjhack_mostthft_devx21[1] 0.27 0.15 0.04 0.61 1.00
## prjhack_mostthft_devx21[2] 0.26 0.14 0.04 0.58 1.00
## prjhack_mostthft_devx21[3] 0.24 0.14 0.04 0.55 1.00
## prjhack_mostthft_devx21[4] 0.23 0.14 0.03 0.57 1.00
## prjhack_mostthft_av12x21[1] 0.12 0.08 0.02 0.32 1.00
## prjhack_mostthft_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## prjhack_mostthft_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## prjhack_mostthft_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## prjhack_mostthft_av12x21[5] 0.12 0.08 0.02 0.32 1.00
## prjhack_mostthft_av12x21[6] 0.13 0.08 0.02 0.31 1.00
## prjhack_mostthft_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## prjhack_mostthft_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_mostthft_devx21[1] 6288 2939
## prjthflt5_mostthft_devx21[2] 5984 2872
## prjthflt5_mostthft_devx21[3] 5756 2563
## prjthflt5_mostthft_devx21[4] 5679 2452
## prjthflt5_mostthft_av12x21[1] 6797 2696
## prjthflt5_mostthft_av12x21[2] 6238 2466
## prjthflt5_mostthft_av12x21[3] 7090 2644
## prjthflt5_mostthft_av12x21[4] 6209 2253
## prjthflt5_mostthft_av12x21[5] 5725 2318
## prjthflt5_mostthft_av12x21[6] 5920 2609
## prjthflt5_mostthft_av12x21[7] 6289 2561
## prjthflt5_mostthft_av12x21[8] 6163 3108
## prjthfgt5_mostthft_devx21[1] 7824 2972
## prjthfgt5_mostthft_devx21[2] 8102 2462
## prjthfgt5_mostthft_devx21[3] 5622 2979
## prjthfgt5_mostthft_devx21[4] 7338 2689
## prjthfgt5_mostthft_av12x21[1] 5881 2182
## prjthfgt5_mostthft_av12x21[2] 7160 2704
## prjthfgt5_mostthft_av12x21[3] 6575 2526
## prjthfgt5_mostthft_av12x21[4] 6470 2704
## prjthfgt5_mostthft_av12x21[5] 5689 2471
## prjthfgt5_mostthft_av12x21[6] 6240 2915
## prjthfgt5_mostthft_av12x21[7] 5793 2707
## prjthfgt5_mostthft_av12x21[8] 6224 2848
## prjthreat_mostthft_devx21[1] 5830 2859
## prjthreat_mostthft_devx21[2] 6296 2913
## prjthreat_mostthft_devx21[3] 5822 3428
## prjthreat_mostthft_devx21[4] 7452 2600
## prjthreat_mostthft_av12x21[1] 6878 2538
## prjthreat_mostthft_av12x21[2] 6017 2511
## prjthreat_mostthft_av12x21[3] 6948 2646
## prjthreat_mostthft_av12x21[4] 5806 2493
## prjthreat_mostthft_av12x21[5] 6655 2377
## prjthreat_mostthft_av12x21[6] 7058 2862
## prjthreat_mostthft_av12x21[7] 7508 3175
## prjthreat_mostthft_av12x21[8] 6183 2878
## prjharm_mostthft_devx21[1] 6972 2886
## prjharm_mostthft_devx21[2] 5980 2486
## prjharm_mostthft_devx21[3] 6095 3042
## prjharm_mostthft_devx21[4] 6071 2918
## prjharm_mostthft_av12x21[1] 6166 2675
## prjharm_mostthft_av12x21[2] 6780 2147
## prjharm_mostthft_av12x21[3] 5511 2385
## prjharm_mostthft_av12x21[4] 7399 2951
## prjharm_mostthft_av12x21[5] 5975 2495
## prjharm_mostthft_av12x21[6] 5639 2006
## prjharm_mostthft_av12x21[7] 5267 2723
## prjharm_mostthft_av12x21[8] 6585 3030
## prjusedrg_mostthft_devx21[1] 6998 3103
## prjusedrg_mostthft_devx21[2] 6428 2403
## prjusedrg_mostthft_devx21[3] 6766 2887
## prjusedrg_mostthft_devx21[4] 6617 2959
## prjusedrg_mostthft_av12x21[1] 6340 2467
## prjusedrg_mostthft_av12x21[2] 6720 2321
## prjusedrg_mostthft_av12x21[3] 6272 2849
## prjusedrg_mostthft_av12x21[4] 6186 2431
## prjusedrg_mostthft_av12x21[5] 5418 2058
## prjusedrg_mostthft_av12x21[6] 6425 2648
## prjusedrg_mostthft_av12x21[7] 5814 2744
## prjusedrg_mostthft_av12x21[8] 6745 2908
## prjhack_mostthft_devx21[1] 5963 2094
## prjhack_mostthft_devx21[2] 6397 2443
## prjhack_mostthft_devx21[3] 6883 2806
## prjhack_mostthft_devx21[4] 6775 3012
## prjhack_mostthft_av12x21[1] 6336 2550
## prjhack_mostthft_av12x21[2] 6234 2595
## prjhack_mostthft_av12x21[3] 6787 2733
## prjhack_mostthft_av12x21[4] 6213 2346
## prjhack_mostthft_av12x21[5] 5921 2484
## prjhack_mostthft_av12x21[6] 6638 2715
## prjhack_mostthft_av12x21[7] 6289 2520
## prjhack_mostthft_av12x21[8] 6810 2662
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.2.6.5 Prior summary
out.chg.prjcrime.stthft.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b prjhack
## normal(0, 0.125) b mostthft_av12x2 prjhack
## normal(0, 0.25) b mostthft_devx2 prjhack
## (flat) b prjharm
## normal(0, 0.125) b mostthft_av12x2 prjharm
## normal(0, 0.25) b mostthft_devx2 prjharm
## (flat) b prjthfgt5
## normal(0, 0.125) b mostthft_av12x2 prjthfgt5
## normal(0, 0.25) b mostthft_devx2 prjthfgt5
## (flat) b prjthflt5
## normal(0, 0.125) b mostthft_av12x2 prjthflt5
## normal(0, 0.25) b mostthft_devx2 prjthflt5
## (flat) b prjthreat
## normal(0, 0.125) b mostthft_av12x2 prjthreat
## normal(0, 0.25) b mostthft_devx2 prjthreat
## (flat) b prjusedrg
## normal(0, 0.125) b mostthft_av12x2 prjusedrg
## normal(0, 0.25) b mostthft_devx2 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 prjhack
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 prjhack
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 prjharm
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 prjharm
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 prjthfgt5
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 prjthfgt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 prjthflt5
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 prjthflt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 prjthreat
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 prjthreat
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 prjusedrg
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 prjusedrg
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.2.7 B/W Change Corr: stmug/crim intent
#Bivariate Change: criminal intent items ~ mo(stmug)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmug_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmug_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmug_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmug_av12x21',
resp = prjdv_names)
)
chg.prjcrime.stmug.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stmug_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stmug.fit <- ppchecks(chg.prjcrime.stmug.fit)6.2.2.7.1 Coefficient plot (intervals)
out.chg.prjcrime.stmug.fit[[10]]
6.2.2.7.2 Coefficient plot (distributions)
out.chg.prjcrime.stmug.fit[[9]]
6.2.2.7.3 PPcheck (density)
p1 <- out.chg.prjcrime.stmug.fit[[3]] + labs(title = "Theft <5BAM Intent (T1)")
p2 <- out.chg.prjcrime.stmug.fit[[4]] + labs(title = "Theft >5BAM Intent (T1)")
p3 <- out.chg.prjcrime.stmug.fit[[5]] + labs(title = "Threat Intent (T1)")
p4 <- out.chg.prjcrime.stmug.fit[[6]] + labs(title = "Harm Intent (T1)")
p5 <- out.chg.prjcrime.stmug.fit[[7]] + labs(title = "Use Drugs Intent (T1)")
p6 <- out.chg.prjcrime.stmug.fit[[8]] + labs(title = "Hack Intent (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
6.2.2.7.4 Fit summary
out.chg.prjcrime.stmug.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## prjthfgt5 ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## prjthreat ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## prjharm ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## prjusedrg ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## prjhack ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.54 0.66 3.40 6.00 1.00 1115
## sd(prjthfgt5_Intercept) 3.80 0.56 2.82 4.99 1.01 898
## sd(prjthreat_Intercept) 3.55 0.57 2.55 4.72 1.00 1348
## sd(prjharm_Intercept) 2.96 0.52 2.02 4.07 1.00 1316
## sd(prjusedrg_Intercept) 3.15 0.55 2.15 4.34 1.00 1152
## sd(prjhack_Intercept) 0.85 0.54 0.04 2.00 1.00 557
## Tail_ESS
## sd(prjthflt5_Intercept) 1901
## sd(prjthfgt5_Intercept) 1726
## sd(prjthreat_Intercept) 2303
## sd(prjharm_Intercept) 2261
## sd(prjusedrg_Intercept) 1594
## sd(prjhack_Intercept) 1309
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_Intercept -5.47 0.81 -7.20 -3.94 1.00 1239
## prjthfgt5_Intercept -5.18 0.73 -6.67 -3.79 1.00 1283
## prjthreat_Intercept -5.96 0.89 -7.79 -4.25 1.01 1515
## prjharm_Intercept -5.81 0.86 -7.60 -4.27 1.00 1600
## prjusedrg_Intercept -6.27 0.91 -8.14 -4.57 1.00 1333
## prjhack_Intercept -4.14 0.64 -5.49 -2.94 1.00 1079
## prjthflt5_mostmug_devx2 -0.06 0.19 -0.44 0.32 1.00 3060
## prjthflt5_mostmug_av12x2 -0.02 0.10 -0.22 0.17 1.00 2312
## prjthfgt5_mostmug_devx2 0.03 0.18 -0.33 0.38 1.00 3155
## prjthfgt5_mostmug_av12x2 -0.01 0.09 -0.19 0.18 1.00 2146
## prjthreat_mostmug_devx2 -0.21 0.20 -0.62 0.18 1.00 3449
## prjthreat_mostmug_av12x2 -0.01 0.10 -0.20 0.19 1.00 2944
## prjharm_mostmug_devx2 -0.08 0.20 -0.50 0.32 1.00 3217
## prjharm_mostmug_av12x2 0.01 0.09 -0.17 0.19 1.00 3322
## prjusedrg_mostmug_devx2 0.02 0.22 -0.45 0.44 1.00 2501
## prjusedrg_mostmug_av12x2 0.01 0.09 -0.18 0.19 1.00 2853
## prjhack_mostmug_devx2 -0.02 0.21 -0.42 0.37 1.00 2068
## prjhack_mostmug_av12x2 0.02 0.08 -0.14 0.18 1.00 3201
## Tail_ESS
## prjthflt5_Intercept 1618
## prjthfgt5_Intercept 2190
## prjthreat_Intercept 2532
## prjharm_Intercept 2055
## prjusedrg_Intercept 1553
## prjhack_Intercept 1406
## prjthflt5_mostmug_devx2 2534
## prjthflt5_mostmug_av12x2 2312
## prjthfgt5_mostmug_devx2 3017
## prjthfgt5_mostmug_av12x2 2787
## prjthreat_mostmug_devx2 2730
## prjthreat_mostmug_av12x2 2619
## prjharm_mostmug_devx2 2776
## prjharm_mostmug_av12x2 3251
## prjusedrg_mostmug_devx2 2762
## prjusedrg_mostmug_av12x2 2873
## prjhack_mostmug_devx2 2315
## prjhack_mostmug_av12x2 2600
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## prjthflt5_mostmug_devx21[1] 0.26 0.15 0.04 0.60 1.00 4071
## prjthflt5_mostmug_devx21[2] 0.25 0.14 0.04 0.57 1.00 4602
## prjthflt5_mostmug_devx21[3] 0.23 0.14 0.03 0.57 1.00 3158
## prjthflt5_mostmug_devx21[4] 0.26 0.15 0.04 0.60 1.00 4207
## prjthflt5_mostmug_av12x21[1] 0.12 0.08 0.02 0.32 1.00 4314
## prjthflt5_mostmug_av12x21[2] 0.12 0.08 0.02 0.31 1.00 4129
## prjthflt5_mostmug_av12x21[3] 0.12 0.08 0.02 0.32 1.00 4106
## prjthflt5_mostmug_av12x21[4] 0.13 0.08 0.02 0.33 1.00 4995
## prjthflt5_mostmug_av12x21[5] 0.13 0.08 0.01 0.32 1.00 3806
## prjthflt5_mostmug_av12x21[6] 0.13 0.08 0.02 0.31 1.00 5002
## prjthflt5_mostmug_av12x21[7] 0.13 0.08 0.02 0.33 1.00 4353
## prjthflt5_mostmug_av12x21[8] 0.13 0.08 0.02 0.32 1.00 4647
## prjthfgt5_mostmug_devx21[1] 0.25 0.15 0.03 0.60 1.00 4226
## prjthfgt5_mostmug_devx21[2] 0.23 0.14 0.03 0.57 1.00 4844
## prjthfgt5_mostmug_devx21[3] 0.26 0.15 0.04 0.60 1.00 3582
## prjthfgt5_mostmug_devx21[4] 0.26 0.14 0.04 0.57 1.00 4255
## prjthfgt5_mostmug_av12x21[1] 0.12 0.08 0.02 0.32 1.00 3860
## prjthfgt5_mostmug_av12x21[2] 0.12 0.08 0.02 0.31 1.00 4383
## prjthfgt5_mostmug_av12x21[3] 0.12 0.08 0.02 0.31 1.00 3691
## prjthfgt5_mostmug_av12x21[4] 0.12 0.08 0.02 0.32 1.00 4746
## prjthfgt5_mostmug_av12x21[5] 0.13 0.08 0.02 0.32 1.00 4460
## prjthfgt5_mostmug_av12x21[6] 0.13 0.08 0.02 0.33 1.00 3847
## prjthfgt5_mostmug_av12x21[7] 0.13 0.08 0.02 0.32 1.00 5091
## prjthfgt5_mostmug_av12x21[8] 0.13 0.08 0.02 0.32 1.00 4124
## prjthreat_mostmug_devx21[1] 0.29 0.16 0.05 0.62 1.00 3732
## prjthreat_mostmug_devx21[2] 0.26 0.14 0.04 0.57 1.00 4746
## prjthreat_mostmug_devx21[3] 0.21 0.13 0.03 0.52 1.00 3416
## prjthreat_mostmug_devx21[4] 0.25 0.14 0.03 0.58 1.00 5217
## prjthreat_mostmug_av12x21[1] 0.12 0.08 0.02 0.31 1.00 4253
## prjthreat_mostmug_av12x21[2] 0.12 0.08 0.02 0.32 1.00 4295
## prjthreat_mostmug_av12x21[3] 0.12 0.08 0.01 0.32 1.00 5206
## prjthreat_mostmug_av12x21[4] 0.12 0.08 0.02 0.32 1.00 4671
## prjthreat_mostmug_av12x21[5] 0.13 0.08 0.02 0.32 1.00 4681
## prjthreat_mostmug_av12x21[6] 0.13 0.08 0.02 0.32 1.00 3825
## prjthreat_mostmug_av12x21[7] 0.13 0.08 0.02 0.32 1.00 4181
## prjthreat_mostmug_av12x21[8] 0.13 0.08 0.02 0.32 1.00 3691
## prjharm_mostmug_devx21[1] 0.27 0.15 0.04 0.62 1.00 3616
## prjharm_mostmug_devx21[2] 0.25 0.14 0.04 0.56 1.00 3831
## prjharm_mostmug_devx21[3] 0.23 0.14 0.03 0.56 1.00 3708
## prjharm_mostmug_devx21[4] 0.25 0.15 0.03 0.60 1.00 3617
## prjharm_mostmug_av12x21[1] 0.12 0.08 0.02 0.31 1.00 3854
## prjharm_mostmug_av12x21[2] 0.13 0.08 0.02 0.32 1.00 4282
## prjharm_mostmug_av12x21[3] 0.13 0.08 0.02 0.32 1.00 4342
## prjharm_mostmug_av12x21[4] 0.12 0.08 0.02 0.31 1.00 4047
## prjharm_mostmug_av12x21[5] 0.12 0.08 0.02 0.32 1.00 5068
## prjharm_mostmug_av12x21[6] 0.12 0.08 0.02 0.32 1.00 4147
## prjharm_mostmug_av12x21[7] 0.13 0.08 0.02 0.33 1.00 4480
## prjharm_mostmug_av12x21[8] 0.13 0.08 0.02 0.32 1.00 4591
## prjusedrg_mostmug_devx21[1] 0.25 0.16 0.03 0.62 1.00 2919
## prjusedrg_mostmug_devx21[2] 0.23 0.14 0.03 0.57 1.00 3545
## prjusedrg_mostmug_devx21[3] 0.27 0.16 0.03 0.63 1.00 2886
## prjusedrg_mostmug_devx21[4] 0.25 0.15 0.04 0.58 1.00 5175
## prjusedrg_mostmug_av12x21[1] 0.12 0.08 0.02 0.32 1.00 4631
## prjusedrg_mostmug_av12x21[2] 0.12 0.08 0.02 0.31 1.00 4506
## prjusedrg_mostmug_av12x21[3] 0.13 0.08 0.02 0.32 1.00 3722
## prjusedrg_mostmug_av12x21[4] 0.12 0.08 0.02 0.31 1.00 4601
## prjusedrg_mostmug_av12x21[5] 0.13 0.08 0.02 0.32 1.00 3654
## prjusedrg_mostmug_av12x21[6] 0.13 0.08 0.02 0.32 1.00 4634
## prjusedrg_mostmug_av12x21[7] 0.13 0.08 0.02 0.32 1.00 4037
## prjusedrg_mostmug_av12x21[8] 0.13 0.08 0.02 0.33 1.00 5404
## prjhack_mostmug_devx21[1] 0.26 0.15 0.04 0.61 1.00 3638
## prjhack_mostmug_devx21[2] 0.24 0.14 0.03 0.57 1.00 4506
## prjhack_mostmug_devx21[3] 0.24 0.14 0.03 0.57 1.00 3655
## prjhack_mostmug_devx21[4] 0.26 0.14 0.04 0.58 1.00 3793
## prjhack_mostmug_av12x21[1] 0.12 0.08 0.02 0.31 1.00 4584
## prjhack_mostmug_av12x21[2] 0.12 0.08 0.02 0.30 1.00 4278
## prjhack_mostmug_av12x21[3] 0.12 0.08 0.02 0.31 1.00 4097
## prjhack_mostmug_av12x21[4] 0.12 0.08 0.02 0.31 1.00 4308
## prjhack_mostmug_av12x21[5] 0.13 0.08 0.02 0.31 1.00 4002
## prjhack_mostmug_av12x21[6] 0.13 0.08 0.02 0.32 1.00 4025
## prjhack_mostmug_av12x21[7] 0.13 0.08 0.02 0.33 1.00 3389
## prjhack_mostmug_av12x21[8] 0.13 0.08 0.02 0.33 1.00 5108
## Tail_ESS
## prjthflt5_mostmug_devx21[1] 2319
## prjthflt5_mostmug_devx21[2] 3062
## prjthflt5_mostmug_devx21[3] 2141
## prjthflt5_mostmug_devx21[4] 2348
## prjthflt5_mostmug_av12x21[1] 1969
## prjthflt5_mostmug_av12x21[2] 2566
## prjthflt5_mostmug_av12x21[3] 2120
## prjthflt5_mostmug_av12x21[4] 2504
## prjthflt5_mostmug_av12x21[5] 1800
## prjthflt5_mostmug_av12x21[6] 2811
## prjthflt5_mostmug_av12x21[7] 2736
## prjthflt5_mostmug_av12x21[8] 2737
## prjthfgt5_mostmug_devx21[1] 2235
## prjthfgt5_mostmug_devx21[2] 2766
## prjthfgt5_mostmug_devx21[3] 3011
## prjthfgt5_mostmug_devx21[4] 2972
## prjthfgt5_mostmug_av12x21[1] 1876
## prjthfgt5_mostmug_av12x21[2] 1619
## prjthfgt5_mostmug_av12x21[3] 1574
## prjthfgt5_mostmug_av12x21[4] 2213
## prjthfgt5_mostmug_av12x21[5] 2492
## prjthfgt5_mostmug_av12x21[6] 2257
## prjthfgt5_mostmug_av12x21[7] 2370
## prjthfgt5_mostmug_av12x21[8] 2682
## prjthreat_mostmug_devx21[1] 2689
## prjthreat_mostmug_devx21[2] 2629
## prjthreat_mostmug_devx21[3] 2585
## prjthreat_mostmug_devx21[4] 2486
## prjthreat_mostmug_av12x21[1] 2343
## prjthreat_mostmug_av12x21[2] 2180
## prjthreat_mostmug_av12x21[3] 2339
## prjthreat_mostmug_av12x21[4] 1875
## prjthreat_mostmug_av12x21[5] 2713
## prjthreat_mostmug_av12x21[6] 2112
## prjthreat_mostmug_av12x21[7] 3070
## prjthreat_mostmug_av12x21[8] 2393
## prjharm_mostmug_devx21[1] 2295
## prjharm_mostmug_devx21[2] 2430
## prjharm_mostmug_devx21[3] 2483
## prjharm_mostmug_devx21[4] 2402
## prjharm_mostmug_av12x21[1] 2049
## prjharm_mostmug_av12x21[2] 2360
## prjharm_mostmug_av12x21[3] 2130
## prjharm_mostmug_av12x21[4] 2241
## prjharm_mostmug_av12x21[5] 2771
## prjharm_mostmug_av12x21[6] 2637
## prjharm_mostmug_av12x21[7] 3080
## prjharm_mostmug_av12x21[8] 2869
## prjusedrg_mostmug_devx21[1] 2342
## prjusedrg_mostmug_devx21[2] 2608
## prjusedrg_mostmug_devx21[3] 2796
## prjusedrg_mostmug_devx21[4] 2524
## prjusedrg_mostmug_av12x21[1] 2162
## prjusedrg_mostmug_av12x21[2] 2450
## prjusedrg_mostmug_av12x21[3] 2167
## prjusedrg_mostmug_av12x21[4] 2385
## prjusedrg_mostmug_av12x21[5] 2170
## prjusedrg_mostmug_av12x21[6] 2447
## prjusedrg_mostmug_av12x21[7] 2698
## prjusedrg_mostmug_av12x21[8] 2820
## prjhack_mostmug_devx21[1] 2676
## prjhack_mostmug_devx21[2] 2757
## prjhack_mostmug_devx21[3] 2621
## prjhack_mostmug_devx21[4] 2330
## prjhack_mostmug_av12x21[1] 2616
## prjhack_mostmug_av12x21[2] 2296
## prjhack_mostmug_av12x21[3] 1637
## prjhack_mostmug_av12x21[4] 2181
## prjhack_mostmug_av12x21[5] 2062
## prjhack_mostmug_av12x21[6] 2601
## prjhack_mostmug_av12x21[7] 2415
## prjhack_mostmug_av12x21[8] 2935
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.2.7.5 Prior summary
out.chg.prjcrime.stmug.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b prjhack
## normal(0, 0.125) b mostmug_av12x2 prjhack
## normal(0, 0.25) b mostmug_devx2 prjhack
## (flat) b prjharm
## normal(0, 0.125) b mostmug_av12x2 prjharm
## normal(0, 0.25) b mostmug_devx2 prjharm
## (flat) b prjthfgt5
## normal(0, 0.125) b mostmug_av12x2 prjthfgt5
## normal(0, 0.25) b mostmug_devx2 prjthfgt5
## (flat) b prjthflt5
## normal(0, 0.125) b mostmug_av12x2 prjthflt5
## normal(0, 0.25) b mostmug_devx2 prjthflt5
## (flat) b prjthreat
## normal(0, 0.125) b mostmug_av12x2 prjthreat
## normal(0, 0.25) b mostmug_devx2 prjthreat
## (flat) b prjusedrg
## normal(0, 0.125) b mostmug_av12x2 prjusedrg
## normal(0, 0.25) b mostmug_devx2 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 prjhack
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 prjhack
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 prjharm
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 prjharm
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 prjthfgt5
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 prjthfgt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 prjthflt5
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 prjthflt5
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 prjthreat
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 prjthreat
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 prjusedrg
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 prjusedrg
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.2.8 B/W Change Corr: stress & “any” crim intent
#Bivariate Change: "any" criminal intent ~ mo(stress)
#Create function for repetitive prior settings
setmyprior <- function(mochgcoefname, moavcoefname,
simochgcoefname, simoavcoefname) {
c(
set_prior('normal(0, 2)', class = 'Intercept'),
set_prior('normal(0, 0.25)', class = 'b', coef = mochgcoefname), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = moavcoefname), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = simochgcoefname),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = simoavcoefname))
}
#Update function to call all ppchecks for bivar any crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit)
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95)
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95)
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, plotcoefs, plotcoefs2)
return(allchecks)
}
myprior <- setmyprior('mostmony_devx2', 'mostmony_av12x2',
'mostmony_devx21', 'mostmony_av12x21')
chg.anyprjcrime.stmony.fit <- brm(prjany ~ 1 +
mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stmony_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stmony.fit <- ppchecks(chg.anyprjcrime.stmony.fit)
myprior <- setmyprior('mosttran_devx2', 'mosttran_av12x2',
'mosttran_devx21', 'mosttran_av12x21')
chg.anyprjcrime.sttran.fit <- brm(prjany ~ 1 +
mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_sttran_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.sttran.fit <- ppchecks(chg.anyprjcrime.sttran.fit)
myprior <- setmyprior('mostresp_devx2', 'mostresp_av12x2',
'mostresp_devx21', 'mostresp_av12x21')
chg.anyprjcrime.stresp.fit <- brm(prjany ~ 1 +
mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stresp_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stresp.fit <- ppchecks(chg.anyprjcrime.stresp.fit)
myprior <- setmyprior('mostfair_devx2', 'mostfair_av12x2',
'mostfair_devx21', 'mostfair_av12x21')
chg.anyprjcrime.stfair.fit <- brm(prjany ~ 1 +
mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stfair_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stfair.fit <- ppchecks(chg.anyprjcrime.stfair.fit)
myprior <- setmyprior('mostjob_devx2', 'mostjob_av12x2',
'mostjob_devx21', 'mostjob_av12x21')
chg.anyprjcrime.stjob.fit <- brm(prjany ~ 1 +
mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stjob_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stjob.fit <- ppchecks(chg.anyprjcrime.stjob.fit)
myprior <- setmyprior('mostthft_devx2', 'mostthft_av12x2',
'mostthft_devx21', 'mostthft_av12x21')
chg.anyprjcrime.stthft.fit <- brm(prjany ~ 1 +
mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stthft_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stthft.fit <- ppchecks(chg.anyprjcrime.stthft.fit)
myprior <- setmyprior('mostmug_devx2', 'mostmug_av12x2',
'mostmug_devx21', 'mostmug_av12x21')
chg.anyprjcrime.stmug.fit <- brm(prjany ~ 1 +
mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stmug_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stmug.fit <- ppchecks(chg.anyprjcrime.stmug.fit)6.2.2.8.1 Coefficient plot (intervals)
p1 <- out.chg.anyprjcrime.stmony.fit[[5]]
p2 <- out.chg.anyprjcrime.sttran.fit[[5]]
p3 <- out.chg.anyprjcrime.stresp.fit[[5]]
p4 <- out.chg.anyprjcrime.stfair.fit[[5]]
p5 <- out.chg.anyprjcrime.stjob.fit[[5]]
p6 <- out.chg.anyprjcrime.stthft.fit[[5]]
p7 <- out.chg.anyprjcrime.stmug.fit[[5]]
playout <- '
AB
CD
E#
FG
'
p1 + p2 + p3 + p4 + p5 + p6 + p7 +
plot_layout(design = playout) +
plot_annotation(
title = 'Coefficient plot',
subtitle = 'Posterior intervals for monotonic ordinal within-person ("devx2") and between-person ("av12x2") stress\ncoefficients predicting "any criminal intent" with medians, 80% (thick line), and 95% (thin line) intervals')
6.2.2.8.2 Coefficient plot (distributions)
p1 <- out.chg.anyprjcrime.stmony.fit[[4]]
p2 <- out.chg.anyprjcrime.sttran.fit[[4]]
p3 <- out.chg.anyprjcrime.stresp.fit[[4]]
p4 <- out.chg.anyprjcrime.stfair.fit[[4]]
p5 <- out.chg.anyprjcrime.stjob.fit[[4]]
p6 <- out.chg.anyprjcrime.stthft.fit[[4]]
p7 <- out.chg.anyprjcrime.stmug.fit[[4]]
playout <- '
AB
CD
E#
FG
'
p1 + p2 + p3 + p4 + p5 + p6 + p7 +
plot_layout(design = playout) +
plot_annotation(
title = 'Coefficient plot',
subtitle = 'Posterior distributions for monotonic ordinal within-person ("devx2") and between-person ("av12x2") stress\ncoefficients predicting "any criminal intent" with medians, 80% (thick line), and 95% (thin line) intervals')
6.2.2.8.3 PPcheck (density)
p1 <- out.chg.anyprjcrime.stmony.fit[[3]] + labs(title = "Any crime intent/stmony (T1)")
p2 <- out.chg.anyprjcrime.sttran.fit[[3]] + labs(title = "Any crime intent/sttran (T1)")
p3 <- out.chg.anyprjcrime.stresp.fit[[3]] + labs(title = "Any crime intent/stresp (T1)")
p4 <- out.chg.anyprjcrime.stfair.fit[[3]] + labs(title = "Any crime intent/stfair (T1)")
p5 <- out.chg.anyprjcrime.stjob.fit[[3]] + labs(title = "Any crime intent/stjob (T1)")
p6 <- out.chg.anyprjcrime.stthft.fit[[3]] + labs(title = "Any crime intent/stthft (T1)")
p7 <- out.chg.anyprjcrime.stmug.fit[[3]] + labs(title = "Any crime intent/stmug (T1)")
(p1 + p2) / (p3 + p4) / (p5 + plot_spacer()) / (p6 + p7)
6.2.2.9 Fit + prior summaries (T1 Any past crime/all stress)
6.2.2.9.1 stmony
out.chg.anyprjcrime.stmony.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.62 0.46 2.81 4.60 1.00 1230 1798
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -3.91 0.64 -5.23 -2.69 1.00 1931 2510
## mostmony_devx2 0.25 0.15 -0.04 0.54 1.00 5262 3052
## mostmony_av12x2 -0.04 0.08 -0.21 0.12 1.00 2404 3060
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostmony_devx21[1] 0.22 0.13 0.03 0.53 1.00 6062 2461
## mostmony_devx21[2] 0.25 0.14 0.04 0.57 1.00 6032 2982
## mostmony_devx21[3] 0.29 0.15 0.05 0.60 1.00 4894 2551
## mostmony_devx21[4] 0.23 0.13 0.04 0.52 1.00 6290 2626
## mostmony_av12x21[1] 0.13 0.08 0.02 0.33 1.00 9162 3011
## mostmony_av12x21[2] 0.13 0.08 0.02 0.33 1.00 5332 2426
## mostmony_av12x21[3] 0.13 0.08 0.02 0.34 1.00 7006 2783
## mostmony_av12x21[4] 0.13 0.08 0.02 0.33 1.00 6077 2662
## mostmony_av12x21[5] 0.12 0.08 0.02 0.31 1.00 7429 2911
## mostmony_av12x21[6] 0.12 0.07 0.02 0.30 1.00 6271 2350
## mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00 6668 2775
## mostmony_av12x21[8] 0.12 0.08 0.02 0.32 1.00 6828 2533
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stmony.fit[[2]]## prior class coef group resp dpar
## (flat) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## nlpar lb ub source
## default
## user
## user
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## user
## user6.2.2.9.2 sttran
out.chg.anyprjcrime.sttran.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.61 0.47 2.79 4.66 1.00 1280 2162
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -3.38 0.63 -4.67 -2.20 1.00 2083 2711
## mosttran_devx2 0.06 0.15 -0.25 0.34 1.00 5009 2880
## mosttran_av12x2 -0.07 0.09 -0.24 0.09 1.00 2119 2980
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mosttran_devx21[1] 0.25 0.15 0.04 0.57 1.00 6171 2362
## mosttran_devx21[2] 0.23 0.14 0.03 0.56 1.00 6690 2888
## mosttran_devx21[3] 0.24 0.14 0.03 0.56 1.00 5804 2771
## mosttran_devx21[4] 0.27 0.15 0.04 0.61 1.00 6831 3029
## mosttran_av12x21[1] 0.13 0.08 0.02 0.32 1.00 6438 2775
## mosttran_av12x21[2] 0.13 0.09 0.02 0.34 1.00 6587 2224
## mosttran_av12x21[3] 0.13 0.08 0.02 0.33 1.00 6766 2502
## mosttran_av12x21[4] 0.14 0.09 0.02 0.35 1.00 6507 2969
## mosttran_av12x21[5] 0.12 0.08 0.02 0.32 1.00 6963 2725
## mosttran_av12x21[6] 0.11 0.08 0.02 0.30 1.00 6634 2738
## mosttran_av12x21[7] 0.11 0.08 0.02 0.30 1.00 5505 2723
## mosttran_av12x21[8] 0.12 0.08 0.02 0.31 1.00 5915 2834
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.sttran.fit[[2]]## prior class coef group resp dpar
## (flat) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## nlpar lb ub source
## default
## user
## user
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## user
## user6.2.2.9.3 stresp
out.chg.anyprjcrime.stresp.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.38 0.44 2.59 4.30 1.00 1122 1614
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -4.08 0.59 -5.24 -2.97 1.00 2325 2180
## mostresp_devx2 -0.18 0.15 -0.50 0.10 1.00 5429 3083
## mostresp_av12x2 0.20 0.07 0.06 0.34 1.00 2646 2991
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostresp_devx21[1] 0.27 0.15 0.04 0.59 1.00 6606 3102
## mostresp_devx21[2] 0.23 0.13 0.04 0.55 1.00 6482 2639
## mostresp_devx21[3] 0.23 0.14 0.03 0.54 1.00 5957 2797
## mostresp_devx21[4] 0.27 0.15 0.04 0.60 1.00 5863 2570
## mostresp_av12x21[1] 0.12 0.08 0.02 0.31 1.00 7180 2583
## mostresp_av12x21[2] 0.11 0.07 0.02 0.28 1.00 6408 2687
## mostresp_av12x21[3] 0.12 0.08 0.01 0.31 1.00 5387 2141
## mostresp_av12x21[4] 0.13 0.08 0.02 0.32 1.00 6496 2799
## mostresp_av12x21[5] 0.12 0.08 0.02 0.30 1.00 6742 2956
## mostresp_av12x21[6] 0.14 0.09 0.02 0.35 1.00 5401 2322
## mostresp_av12x21[7] 0.13 0.08 0.02 0.33 1.00 6269 3041
## mostresp_av12x21[8] 0.12 0.08 0.02 0.30 1.00 6730 2850
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stresp.fit[[2]]## prior class coef group resp dpar
## (flat) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## nlpar lb ub source
## default
## user
## user
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## user
## user6.2.2.9.4 stfair
out.chg.anyprjcrime.stfair.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.31 0.43 2.56 4.23 1.00 1499 2499
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -4.49 0.58 -5.68 -3.39 1.00 2452 2728
## mostfair_devx2 0.05 0.15 -0.25 0.35 1.00 4017 3082
## mostfair_av12x2 0.20 0.07 0.07 0.33 1.00 2374 3310
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostfair_devx21[1] 0.26 0.15 0.04 0.59 1.00 6686 2114
## mostfair_devx21[2] 0.23 0.14 0.03 0.54 1.00 7239 2809
## mostfair_devx21[3] 0.25 0.14 0.04 0.58 1.00 6041 2338
## mostfair_devx21[4] 0.26 0.15 0.04 0.59 1.00 6829 2914
## mostfair_av12x21[1] 0.12 0.08 0.01 0.31 1.00 5428 2000
## mostfair_av12x21[2] 0.11 0.07 0.02 0.28 1.00 6006 2476
## mostfair_av12x21[3] 0.14 0.09 0.02 0.35 1.00 6149 2179
## mostfair_av12x21[4] 0.12 0.08 0.02 0.30 1.00 6815 3052
## mostfair_av12x21[5] 0.11 0.07 0.01 0.28 1.00 5981 2408
## mostfair_av12x21[6] 0.12 0.08 0.02 0.32 1.00 6755 2923
## mostfair_av12x21[7] 0.15 0.09 0.02 0.37 1.00 5995 2298
## mostfair_av12x21[8] 0.14 0.08 0.02 0.33 1.00 5131 2765
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stfair.fit[[2]]## prior class coef group resp dpar
## (flat) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## nlpar lb ub source
## default
## user
## user
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## user
## user6.2.2.9.5 stjob
out.chg.anyprjcrime.stjob.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.27 0.42 2.54 4.17 1.00 1317 2359
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -4.56 0.58 -5.74 -3.48 1.00 2348 2826
## mostjob_devx2 0.09 0.15 -0.22 0.38 1.00 4901 2915
## mostjob_av12x2 0.22 0.07 0.09 0.35 1.00 2514 3053
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostjob_devx21[1] 0.25 0.15 0.03 0.61 1.00 8130 2378
## mostjob_devx21[2] 0.26 0.15 0.04 0.60 1.00 6390 3118
## mostjob_devx21[3] 0.23 0.13 0.04 0.54 1.00 6188 3004
## mostjob_devx21[4] 0.26 0.15 0.04 0.58 1.00 7428 2668
## mostjob_av12x21[1] 0.09 0.06 0.01 0.24 1.00 6083 2484
## mostjob_av12x21[2] 0.10 0.06 0.01 0.25 1.00 5698 2557
## mostjob_av12x21[3] 0.10 0.07 0.01 0.26 1.00 5039 2129
## mostjob_av12x21[4] 0.11 0.07 0.01 0.28 1.00 6467 2550
## mostjob_av12x21[5] 0.12 0.08 0.02 0.32 1.00 7108 2817
## mostjob_av12x21[6] 0.18 0.10 0.03 0.41 1.00 5164 2797
## mostjob_av12x21[7] 0.17 0.10 0.02 0.40 1.00 5569 3029
## mostjob_av12x21[8] 0.14 0.09 0.02 0.34 1.00 6219 3433
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stjob.fit[[2]]## prior class coef group resp dpar
## (flat) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## nlpar lb ub source
## default
## user
## user
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## user
## user6.2.2.9.6 stthft
out.chg.anyprjcrime.stthft.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.53 0.45 2.72 4.49 1.00 1248 2362
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -3.57 0.56 -4.70 -2.50 1.00 1976 2699
## mostthft_devx2 -0.12 0.16 -0.43 0.20 1.00 4146 2603
## mostthft_av12x2 0.11 0.08 -0.05 0.26 1.00 2636 2970
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostthft_devx21[1] 0.27 0.15 0.04 0.61 1.00 5943 2444
## mostthft_devx21[2] 0.26 0.15 0.04 0.60 1.00 6152 2966
## mostthft_devx21[3] 0.22 0.14 0.03 0.54 1.00 6099 3020
## mostthft_devx21[4] 0.25 0.14 0.04 0.58 1.00 6237 3137
## mostthft_av12x21[1] 0.12 0.08 0.02 0.30 1.00 6027 2551
## mostthft_av12x21[2] 0.11 0.07 0.02 0.30 1.00 6806 2223
## mostthft_av12x21[3] 0.12 0.08 0.01 0.31 1.00 7291 2118
## mostthft_av12x21[4] 0.12 0.08 0.02 0.31 1.00 6822 3121
## mostthft_av12x21[5] 0.15 0.09 0.02 0.36 1.00 5315 2852
## mostthft_av12x21[6] 0.14 0.09 0.02 0.36 1.00 6619 2891
## mostthft_av12x21[7] 0.12 0.08 0.02 0.31 1.00 6267 2879
## mostthft_av12x21[8] 0.12 0.08 0.02 0.30 1.00 6758 2848
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stthft.fit[[2]]## prior class coef group resp dpar
## (flat) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## nlpar lb ub source
## default
## user
## user
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## user
## user6.2.2.9.7 stmug
out.chg.anyprjcrime.stmug.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.60 0.48 2.74 4.62 1.00 1433 2116
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -3.60 0.56 -4.75 -2.56 1.00 2520 2786
## mostmug_devx2 0.01 0.17 -0.32 0.34 1.00 4899 3347
## mostmug_av12x2 -0.01 0.09 -0.18 0.15 1.00 3712 3327
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## mostmug_devx21[1] 0.26 0.15 0.04 0.60 1.00 6920 2513
## mostmug_devx21[2] 0.24 0.14 0.03 0.58 1.00 6063 3018
## mostmug_devx21[3] 0.24 0.15 0.03 0.59 1.00 6151 2570
## mostmug_devx21[4] 0.26 0.15 0.04 0.59 1.00 5249 2911
## mostmug_av12x21[1] 0.12 0.08 0.02 0.30 1.00 6402 2308
## mostmug_av12x21[2] 0.12 0.08 0.02 0.32 1.00 6170 2446
## mostmug_av12x21[3] 0.12 0.08 0.02 0.32 1.00 8512 2433
## mostmug_av12x21[4] 0.12 0.08 0.02 0.32 1.00 5462 2351
## mostmug_av12x21[5] 0.13 0.08 0.02 0.33 1.00 7304 2386
## mostmug_av12x21[6] 0.13 0.08 0.02 0.32 1.00 7675 2600
## mostmug_av12x21[7] 0.13 0.08 0.02 0.32 1.00 6434 2780
## mostmug_av12x21[8] 0.13 0.08 0.02 0.31 1.00 6620 2630
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stmug.fit[[2]]## prior class coef group resp dpar
## (flat) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## nlpar lb ub source
## default
## user
## user
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## user
## user6.2.3 Negative Emotions: Bivariate Change models
Let’s turn to multilevel models predicting negative emotions outcome probabilities.
6.2.3.1 B/W Change Corr: stmony/neg emotions
#Bivariate Change: negative emotions items ~ mo(stmony)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmony_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmony_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmony_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmony_av12x21',
resp = depdv_names)
)
chg.alldepress.stmony.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stmony_fit",
file_refit = "on_change"
)
##Update function to call all ppchecks for bivar depressive symptom chg models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="depcantgo")
ppcheckdv2 <-pp_check(modelfit, resp="depeffort")
ppcheckdv3 <-pp_check(modelfit, resp="deplonely")
ppcheckdv4 <-pp_check(modelfit, resp="depblues")
ppcheckdv5 <-pp_check(modelfit, resp="depunfair")
ppcheckdv6 <-pp_check(modelfit, resp="depmistrt")
ppcheckdv7 <-pp_check(modelfit, resp="depbetray")
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2, ppcheckdv3,
ppcheckdv4, ppcheckdv5, ppcheckdv6, ppcheckdv7,
plotcoefs, plotcoefs2)
return(allchecks)
}
out.chg.alldepress.stmony.fit <- ppchecks(chg.alldepress.stmony.fit)6.2.3.1.1 Coefficient plot (intervals)
out.chg.alldepress.stmony.fit[[11]]
6.2.3.1.2 Coefficient plot (distributions)
out.chg.alldepress.stmony.fit[[10]]
6.2.3.1.3 PPcheck (density)
p1 <- out.chg.alldepress.stmony.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.chg.alldepress.stmony.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.chg.alldepress.stmony.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.chg.alldepress.stmony.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.chg.alldepress.stmony.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.chg.alldepress.stmony.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.chg.alldepress.stmony.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
6.2.3.1.4 Fit summary
out.chg.alldepress.stmony.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## depeffort ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## deplonely ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## depblues ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## depunfair ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## depmistrt ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## depbetray ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.30 0.19 0.01 0.70 1.01 642
## sd(depeffort_Intercept) 0.45 0.27 0.02 0.99 1.01 621
## sd(deplonely_Intercept) 0.43 0.24 0.03 0.89 1.01 584
## sd(depblues_Intercept) 0.67 0.31 0.06 1.24 1.02 368
## sd(depunfair_Intercept) 0.24 0.17 0.01 0.62 1.00 835
## sd(depmistrt_Intercept) 0.31 0.21 0.01 0.76 1.01 805
## sd(depbetray_Intercept) 0.43 0.26 0.02 0.96 1.01 528
## Tail_ESS
## sd(depcantgo_Intercept) 1458
## sd(depeffort_Intercept) 1281
## sd(deplonely_Intercept) 1186
## sd(depblues_Intercept) 616
## sd(depunfair_Intercept) 2150
## sd(depmistrt_Intercept) 1899
## sd(depbetray_Intercept) 1169
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_Intercept -1.02 0.29 -1.67 -0.50 1.00 4177
## depeffort_Intercept -2.03 0.36 -2.77 -1.36 1.00 2920
## deplonely_Intercept -1.26 0.31 -1.91 -0.67 1.00 2988
## depblues_Intercept -1.99 0.40 -2.78 -1.17 1.00 2672
## depunfair_Intercept -1.92 0.29 -2.57 -1.41 1.00 4280
## depmistrt_Intercept -2.30 0.33 -3.00 -1.70 1.00 3202
## depbetray_Intercept -2.40 0.37 -3.18 -1.75 1.00 2819
## depcantgo_mostmony_devx2 0.22 0.11 0.03 0.46 1.00 4028
## depcantgo_mostmony_av12x2 0.05 0.03 -0.01 0.11 1.00 5901
## depeffort_mostmony_devx2 0.13 0.12 -0.11 0.37 1.00 3896
## depeffort_mostmony_av12x2 0.01 0.04 -0.07 0.09 1.00 4811
## deplonely_mostmony_devx2 0.13 0.11 -0.09 0.33 1.00 3136
## deplonely_mostmony_av12x2 -0.01 0.03 -0.08 0.05 1.00 6082
## depblues_mostmony_devx2 -0.12 0.13 -0.40 0.12 1.00 4534
## depblues_mostmony_av12x2 0.07 0.05 -0.02 0.16 1.00 5218
## depunfair_mostmony_devx2 0.28 0.09 0.11 0.46 1.00 5010
## depunfair_mostmony_av12x2 0.09 0.04 0.02 0.17 1.00 5391
## depmistrt_mostmony_devx2 0.14 0.12 -0.08 0.38 1.00 4107
## depmistrt_mostmony_av12x2 0.11 0.04 0.03 0.20 1.00 4979
## depbetray_mostmony_devx2 0.04 0.12 -0.19 0.29 1.00 4377
## depbetray_mostmony_av12x2 0.16 0.05 0.07 0.25 1.00 5089
## Tail_ESS
## depcantgo_Intercept 2990
## depeffort_Intercept 2710
## deplonely_Intercept 2777
## depblues_Intercept 2717
## depunfair_Intercept 3063
## depmistrt_Intercept 2743
## depbetray_Intercept 2147
## depcantgo_mostmony_devx2 3022
## depcantgo_mostmony_av12x2 3357
## depeffort_mostmony_devx2 2578
## depeffort_mostmony_av12x2 2923
## deplonely_mostmony_devx2 2583
## deplonely_mostmony_av12x2 2495
## depblues_mostmony_devx2 2939
## depblues_mostmony_av12x2 3024
## depunfair_mostmony_devx2 2699
## depunfair_mostmony_av12x2 3121
## depmistrt_mostmony_devx2 2897
## depmistrt_mostmony_av12x2 3365
## depbetray_mostmony_devx2 2807
## depbetray_mostmony_av12x2 2730
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_mostmony_devx21[1] 0.24 0.13 0.04 0.54 1.00
## depcantgo_mostmony_devx21[2] 0.27 0.14 0.05 0.57 1.00
## depcantgo_mostmony_devx21[3] 0.20 0.12 0.03 0.47 1.00
## depcantgo_mostmony_devx21[4] 0.29 0.15 0.04 0.61 1.00
## depcantgo_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## depcantgo_mostmony_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## depcantgo_mostmony_av12x21[3] 0.13 0.08 0.02 0.31 1.00
## depcantgo_mostmony_av12x21[4] 0.14 0.09 0.02 0.35 1.00
## depcantgo_mostmony_av12x21[5] 0.12 0.08 0.02 0.30 1.00
## depcantgo_mostmony_av12x21[6] 0.12 0.08 0.02 0.30 1.00
## depcantgo_mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mostmony_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostmony_devx21[1] 0.25 0.15 0.03 0.58 1.00
## depeffort_mostmony_devx21[2] 0.27 0.14 0.05 0.59 1.00
## depeffort_mostmony_devx21[3] 0.22 0.13 0.03 0.53 1.00
## depeffort_mostmony_devx21[4] 0.26 0.15 0.04 0.59 1.00
## depeffort_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## depeffort_mostmony_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostmony_av12x21[3] 0.12 0.08 0.01 0.33 1.00
## depeffort_mostmony_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostmony_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostmony_av12x21[6] 0.13 0.08 0.02 0.30 1.00
## depeffort_mostmony_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## depeffort_mostmony_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## deplonely_mostmony_devx21[1] 0.24 0.14 0.03 0.54 1.00
## deplonely_mostmony_devx21[2] 0.34 0.16 0.06 0.66 1.00
## deplonely_mostmony_devx21[3] 0.20 0.12 0.03 0.49 1.00
## deplonely_mostmony_devx21[4] 0.23 0.14 0.03 0.54 1.00
## deplonely_mostmony_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## deplonely_mostmony_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## deplonely_mostmony_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## deplonely_mostmony_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## deplonely_mostmony_av12x21[5] 0.12 0.08 0.02 0.30 1.00
## deplonely_mostmony_av12x21[6] 0.12 0.08 0.02 0.32 1.00
## deplonely_mostmony_av12x21[7] 0.12 0.08 0.02 0.32 1.00
## deplonely_mostmony_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## depblues_mostmony_devx21[1] 0.28 0.16 0.04 0.62 1.00
## depblues_mostmony_devx21[2] 0.24 0.14 0.04 0.55 1.00
## depblues_mostmony_devx21[3] 0.22 0.13 0.03 0.52 1.00
## depblues_mostmony_devx21[4] 0.26 0.15 0.04 0.60 1.00
## depblues_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## depblues_mostmony_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## depblues_mostmony_av12x21[3] 0.12 0.07 0.02 0.30 1.00
## depblues_mostmony_av12x21[4] 0.13 0.08 0.02 0.32 1.00
## depblues_mostmony_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depblues_mostmony_av12x21[6] 0.11 0.07 0.02 0.29 1.00
## depblues_mostmony_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## depblues_mostmony_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## depunfair_mostmony_devx21[1] 0.19 0.11 0.03 0.45 1.00
## depunfair_mostmony_devx21[2] 0.25 0.12 0.06 0.51 1.00
## depunfair_mostmony_devx21[3] 0.36 0.13 0.11 0.63 1.00
## depunfair_mostmony_devx21[4] 0.20 0.12 0.03 0.46 1.00
## depunfair_mostmony_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## depunfair_mostmony_av12x21[2] 0.12 0.08 0.02 0.30 1.00
## depunfair_mostmony_av12x21[3] 0.10 0.07 0.01 0.26 1.00
## depunfair_mostmony_av12x21[4] 0.10 0.06 0.01 0.25 1.00
## depunfair_mostmony_av12x21[5] 0.13 0.08 0.01 0.32 1.00
## depunfair_mostmony_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## depunfair_mostmony_av12x21[7] 0.15 0.09 0.02 0.36 1.00
## depunfair_mostmony_av12x21[8] 0.16 0.09 0.02 0.37 1.00
## depmistrt_mostmony_devx21[1] 0.24 0.14 0.04 0.56 1.00
## depmistrt_mostmony_devx21[2] 0.20 0.12 0.03 0.49 1.00
## depmistrt_mostmony_devx21[3] 0.29 0.15 0.05 0.63 1.00
## depmistrt_mostmony_devx21[4] 0.26 0.15 0.04 0.58 1.00
## depmistrt_mostmony_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## depmistrt_mostmony_av12x21[2] 0.14 0.08 0.02 0.33 1.00
## depmistrt_mostmony_av12x21[3] 0.13 0.08 0.02 0.31 1.00
## depmistrt_mostmony_av12x21[4] 0.12 0.07 0.01 0.30 1.00
## depmistrt_mostmony_av12x21[5] 0.10 0.06 0.01 0.25 1.00
## depmistrt_mostmony_av12x21[6] 0.10 0.06 0.01 0.26 1.00
## depmistrt_mostmony_av12x21[7] 0.17 0.10 0.03 0.39 1.00
## depmistrt_mostmony_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## depbetray_mostmony_devx21[1] 0.27 0.15 0.04 0.61 1.00
## depbetray_mostmony_devx21[2] 0.23 0.14 0.03 0.55 1.00
## depbetray_mostmony_devx21[3] 0.23 0.14 0.04 0.55 1.00
## depbetray_mostmony_devx21[4] 0.26 0.15 0.04 0.60 1.00
## depbetray_mostmony_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## depbetray_mostmony_av12x21[2] 0.12 0.08 0.02 0.30 1.00
## depbetray_mostmony_av12x21[3] 0.11 0.07 0.02 0.27 1.00
## depbetray_mostmony_av12x21[4] 0.11 0.07 0.01 0.29 1.00
## depbetray_mostmony_av12x21[5] 0.11 0.07 0.01 0.28 1.00
## depbetray_mostmony_av12x21[6] 0.10 0.07 0.01 0.26 1.00
## depbetray_mostmony_av12x21[7] 0.17 0.09 0.03 0.37 1.00
## depbetray_mostmony_av12x21[8] 0.16 0.09 0.03 0.37 1.00
## Bulk_ESS Tail_ESS
## depcantgo_mostmony_devx21[1] 5830 2612
## depcantgo_mostmony_devx21[2] 5283 2276
## depcantgo_mostmony_devx21[3] 5743 3306
## depcantgo_mostmony_devx21[4] 5850 2822
## depcantgo_mostmony_av12x21[1] 6929 2558
## depcantgo_mostmony_av12x21[2] 7472 2694
## depcantgo_mostmony_av12x21[3] 6001 2638
## depcantgo_mostmony_av12x21[4] 7741 2425
## depcantgo_mostmony_av12x21[5] 7583 3047
## depcantgo_mostmony_av12x21[6] 7040 2754
## depcantgo_mostmony_av12x21[7] 7030 3053
## depcantgo_mostmony_av12x21[8] 6631 3167
## depeffort_mostmony_devx21[1] 6080 2365
## depeffort_mostmony_devx21[2] 5955 2903
## depeffort_mostmony_devx21[3] 6420 2608
## depeffort_mostmony_devx21[4] 7106 2659
## depeffort_mostmony_av12x21[1] 6800 2431
## depeffort_mostmony_av12x21[2] 6946 2690
## depeffort_mostmony_av12x21[3] 6067 2311
## depeffort_mostmony_av12x21[4] 7033 2497
## depeffort_mostmony_av12x21[5] 6778 2539
## depeffort_mostmony_av12x21[6] 7429 2742
## depeffort_mostmony_av12x21[7] 6090 2157
## depeffort_mostmony_av12x21[8] 6977 2537
## deplonely_mostmony_devx21[1] 6612 2811
## deplonely_mostmony_devx21[2] 4565 2365
## deplonely_mostmony_devx21[3] 5460 2897
## deplonely_mostmony_devx21[4] 5331 2540
## deplonely_mostmony_av12x21[1] 6945 2365
## deplonely_mostmony_av12x21[2] 6762 2401
## deplonely_mostmony_av12x21[3] 6349 2479
## deplonely_mostmony_av12x21[4] 7597 2694
## deplonely_mostmony_av12x21[5] 6035 2647
## deplonely_mostmony_av12x21[6] 5591 2637
## deplonely_mostmony_av12x21[7] 6822 2831
## deplonely_mostmony_av12x21[8] 7535 3132
## depblues_mostmony_devx21[1] 6245 3058
## depblues_mostmony_devx21[2] 6675 2873
## depblues_mostmony_devx21[3] 5461 2935
## depblues_mostmony_devx21[4] 7682 2837
## depblues_mostmony_av12x21[1] 6414 2255
## depblues_mostmony_av12x21[2] 7286 2249
## depblues_mostmony_av12x21[3] 7327 2643
## depblues_mostmony_av12x21[4] 8549 2522
## depblues_mostmony_av12x21[5] 6529 2361
## depblues_mostmony_av12x21[6] 6593 2659
## depblues_mostmony_av12x21[7] 6101 3154
## depblues_mostmony_av12x21[8] 5970 2309
## depunfair_mostmony_devx21[1] 6142 2666
## depunfair_mostmony_devx21[2] 5541 2222
## depunfair_mostmony_devx21[3] 5475 2423
## depunfair_mostmony_devx21[4] 6360 2689
## depunfair_mostmony_av12x21[1] 6089 2620
## depunfair_mostmony_av12x21[2] 7101 2523
## depunfair_mostmony_av12x21[3] 7539 2513
## depunfair_mostmony_av12x21[4] 6079 2500
## depunfair_mostmony_av12x21[5] 6381 2353
## depunfair_mostmony_av12x21[6] 7112 3049
## depunfair_mostmony_av12x21[7] 6364 2542
## depunfair_mostmony_av12x21[8] 5791 3248
## depmistrt_mostmony_devx21[1] 7349 2144
## depmistrt_mostmony_devx21[2] 6300 2590
## depmistrt_mostmony_devx21[3] 4910 2547
## depmistrt_mostmony_devx21[4] 6132 2672
## depmistrt_mostmony_av12x21[1] 6172 2443
## depmistrt_mostmony_av12x21[2] 6735 2443
## depmistrt_mostmony_av12x21[3] 6338 1979
## depmistrt_mostmony_av12x21[4] 8145 2608
## depmistrt_mostmony_av12x21[5] 6264 2857
## depmistrt_mostmony_av12x21[6] 6815 2858
## depmistrt_mostmony_av12x21[7] 6205 2589
## depmistrt_mostmony_av12x21[8] 6271 2980
## depbetray_mostmony_devx21[1] 5223 2382
## depbetray_mostmony_devx21[2] 7367 2948
## depbetray_mostmony_devx21[3] 7172 3109
## depbetray_mostmony_devx21[4] 6335 2795
## depbetray_mostmony_av12x21[1] 6276 2580
## depbetray_mostmony_av12x21[2] 6770 2468
## depbetray_mostmony_av12x21[3] 6337 2135
## depbetray_mostmony_av12x21[4] 6553 2566
## depbetray_mostmony_av12x21[5] 6043 2328
## depbetray_mostmony_av12x21[6] 7063 2583
## depbetray_mostmony_av12x21[7] 5995 2727
## depbetray_mostmony_av12x21[8] 6400 2703
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.3.1.5 Prior summary
out.chg.alldepress.stmony.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b depbetray
## normal(0, 0.125) b mostmony_av12x2 depbetray
## normal(0, 0.25) b mostmony_devx2 depbetray
## (flat) b depblues
## normal(0, 0.125) b mostmony_av12x2 depblues
## normal(0, 0.25) b mostmony_devx2 depblues
## (flat) b depcantgo
## normal(0, 0.125) b mostmony_av12x2 depcantgo
## normal(0, 0.25) b mostmony_devx2 depcantgo
## (flat) b depeffort
## normal(0, 0.125) b mostmony_av12x2 depeffort
## normal(0, 0.25) b mostmony_devx2 depeffort
## (flat) b deplonely
## normal(0, 0.125) b mostmony_av12x2 deplonely
## normal(0, 0.25) b mostmony_devx2 deplonely
## (flat) b depmistrt
## normal(0, 0.125) b mostmony_av12x2 depmistrt
## normal(0, 0.25) b mostmony_devx2 depmistrt
## (flat) b depunfair
## normal(0, 0.125) b mostmony_av12x2 depunfair
## normal(0, 0.25) b mostmony_devx2 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 depbetray
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 depbetray
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 depblues
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 depblues
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 depcantgo
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 depcantgo
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 depeffort
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 depeffort
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 deplonely
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 deplonely
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 depmistrt
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 depmistrt
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21 depunfair
## dirichlet(2, 2, 2, 2) simo mostmony_devx21 depunfair
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.3.2 B/W Change Corr: sttran/neg emotions
#Bivariate Change: negative emotions items ~ mo(sttran)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mosttran_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mosttran_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mosttran_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mosttran_av12x21',
resp = depdv_names)
)
chg.alldepress.sttran.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_sttran_fit",
file_refit = "on_change"
)
out.chg.alldepress.sttran.fit <- ppchecks(chg.alldepress.sttran.fit)6.2.3.2.1 Coefficient plot (intervals)
out.chg.alldepress.sttran.fit[[11]]
6.2.3.2.2 Coefficient plot (distributions)
out.chg.alldepress.sttran.fit[[10]]
6.2.3.2.3 PPcheck (density)
p1 <- out.chg.alldepress.sttran.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.chg.alldepress.sttran.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.chg.alldepress.sttran.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.chg.alldepress.sttran.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.chg.alldepress.sttran.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.chg.alldepress.sttran.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.chg.alldepress.sttran.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
6.2.3.2.4 Fit summary
out.chg.alldepress.sttran.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## depeffort ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## deplonely ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## depblues ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## depunfair ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## depmistrt ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## depbetray ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.30 0.19 0.02 0.69 1.00 527
## sd(depeffort_Intercept) 0.43 0.27 0.02 0.99 1.00 489
## sd(deplonely_Intercept) 0.41 0.24 0.03 0.89 1.00 457
## sd(depblues_Intercept) 0.65 0.31 0.07 1.24 1.01 482
## sd(depunfair_Intercept) 0.26 0.18 0.01 0.66 1.00 725
## sd(depmistrt_Intercept) 0.33 0.22 0.02 0.80 1.01 750
## sd(depbetray_Intercept) 0.51 0.28 0.03 1.06 1.01 447
## Tail_ESS
## sd(depcantgo_Intercept) 1010
## sd(depeffort_Intercept) 877
## sd(deplonely_Intercept) 1026
## sd(depblues_Intercept) 1230
## sd(depunfair_Intercept) 1385
## sd(depmistrt_Intercept) 1676
## sd(depbetray_Intercept) 593
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_Intercept -0.95 0.27 -1.53 -0.44 1.00 2907
## depeffort_Intercept -1.71 0.34 -2.39 -1.02 1.00 2760
## deplonely_Intercept -1.17 0.27 -1.72 -0.64 1.00 3090
## depblues_Intercept -2.01 0.38 -2.76 -1.28 1.00 2421
## depunfair_Intercept -2.07 0.30 -2.71 -1.54 1.00 4069
## depmistrt_Intercept -2.02 0.37 -2.76 -1.26 1.00 2503
## depbetray_Intercept -2.08 0.35 -2.83 -1.40 1.00 3043
## depcantgo_mosttran_devx2 0.19 0.09 0.03 0.38 1.00 3802
## depcantgo_mosttran_av12x2 0.04 0.03 -0.02 0.11 1.00 4077
## depeffort_mosttran_devx2 0.03 0.12 -0.22 0.28 1.00 4203
## depeffort_mosttran_av12x2 -0.01 0.04 -0.10 0.07 1.00 4821
## deplonely_mosttran_devx2 0.13 0.10 -0.06 0.34 1.00 3260
## deplonely_mosttran_av12x2 -0.01 0.04 -0.09 0.06 1.00 5107
## depblues_mosttran_devx2 -0.12 0.13 -0.37 0.14 1.00 3871
## depblues_mosttran_av12x2 0.06 0.05 -0.03 0.16 1.00 4606
## depunfair_mosttran_devx2 0.32 0.09 0.16 0.50 1.00 5352
## depunfair_mosttran_av12x2 0.10 0.04 0.03 0.18 1.00 5161
## depmistrt_mosttran_devx2 0.06 0.13 -0.24 0.28 1.00 3065
## depmistrt_mosttran_av12x2 0.08 0.05 -0.01 0.19 1.00 3323
## depbetray_mosttran_devx2 -0.01 0.12 -0.25 0.23 1.00 4082
## depbetray_mosttran_av12x2 0.11 0.05 0.02 0.20 1.00 4865
## Tail_ESS
## depcantgo_Intercept 2709
## depeffort_Intercept 2706
## deplonely_Intercept 2308
## depblues_Intercept 2561
## depunfair_Intercept 2526
## depmistrt_Intercept 2140
## depbetray_Intercept 2647
## depcantgo_mosttran_devx2 2492
## depcantgo_mosttran_av12x2 3142
## depeffort_mosttran_devx2 2918
## depeffort_mosttran_av12x2 3097
## deplonely_mosttran_devx2 2408
## deplonely_mosttran_av12x2 2782
## depblues_mosttran_devx2 2696
## depblues_mosttran_av12x2 3061
## depunfair_mosttran_devx2 2749
## depunfair_mosttran_av12x2 3338
## depmistrt_mosttran_devx2 1906
## depmistrt_mosttran_av12x2 2402
## depbetray_mosttran_devx2 2665
## depbetray_mosttran_av12x2 2888
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_mosttran_devx21[1] 0.23 0.13 0.03 0.53 1.00
## depcantgo_mosttran_devx21[2] 0.30 0.14 0.07 0.59 1.00
## depcantgo_mosttran_devx21[3] 0.24 0.12 0.05 0.52 1.00
## depcantgo_mosttran_devx21[4] 0.23 0.13 0.03 0.51 1.00
## depcantgo_mosttran_av12x21[1] 0.12 0.08 0.02 0.30 1.00
## depcantgo_mosttran_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## depcantgo_mosttran_av12x21[3] 0.11 0.07 0.01 0.29 1.00
## depcantgo_mosttran_av12x21[4] 0.12 0.07 0.02 0.29 1.00
## depcantgo_mosttran_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mosttran_av12x21[6] 0.12 0.07 0.02 0.30 1.00
## depcantgo_mosttran_av12x21[7] 0.14 0.08 0.02 0.34 1.00
## depcantgo_mosttran_av12x21[8] 0.14 0.09 0.02 0.35 1.00
## depeffort_mosttran_devx21[1] 0.26 0.15 0.04 0.61 1.00
## depeffort_mosttran_devx21[2] 0.24 0.14 0.03 0.56 1.00
## depeffort_mosttran_devx21[3] 0.23 0.13 0.04 0.55 1.00
## depeffort_mosttran_devx21[4] 0.27 0.15 0.04 0.61 1.00
## depeffort_mosttran_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## depeffort_mosttran_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## depeffort_mosttran_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## depeffort_mosttran_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## depeffort_mosttran_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depeffort_mosttran_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## depeffort_mosttran_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## depeffort_mosttran_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## deplonely_mosttran_devx21[1] 0.22 0.13 0.03 0.52 1.00
## deplonely_mosttran_devx21[2] 0.21 0.12 0.03 0.51 1.00
## deplonely_mosttran_devx21[3] 0.31 0.15 0.06 0.62 1.00
## deplonely_mosttran_devx21[4] 0.26 0.15 0.04 0.59 1.00
## deplonely_mosttran_av12x21[1] 0.13 0.09 0.02 0.34 1.00
## deplonely_mosttran_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## deplonely_mosttran_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## deplonely_mosttran_av12x21[4] 0.12 0.08 0.01 0.32 1.00
## deplonely_mosttran_av12x21[5] 0.11 0.08 0.01 0.30 1.00
## deplonely_mosttran_av12x21[6] 0.12 0.08 0.02 0.30 1.00
## deplonely_mosttran_av12x21[7] 0.12 0.08 0.02 0.32 1.00
## deplonely_mosttran_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## depblues_mosttran_devx21[1] 0.25 0.14 0.04 0.57 1.00
## depblues_mosttran_devx21[2] 0.22 0.13 0.03 0.53 1.00
## depblues_mosttran_devx21[3] 0.27 0.15 0.04 0.59 1.00
## depblues_mosttran_devx21[4] 0.26 0.15 0.04 0.58 1.00
## depblues_mosttran_av12x21[1] 0.12 0.08 0.02 0.33 1.00
## depblues_mosttran_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## depblues_mosttran_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## depblues_mosttran_av12x21[4] 0.14 0.09 0.02 0.35 1.00
## depblues_mosttran_av12x21[5] 0.11 0.07 0.01 0.29 1.00
## depblues_mosttran_av12x21[6] 0.11 0.07 0.02 0.29 1.00
## depblues_mosttran_av12x21[7] 0.14 0.09 0.02 0.35 1.00
## depblues_mosttran_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## depunfair_mosttran_devx21[1] 0.15 0.09 0.02 0.37 1.00
## depunfair_mosttran_devx21[2] 0.33 0.12 0.11 0.59 1.00
## depunfair_mosttran_devx21[3] 0.33 0.12 0.11 0.59 1.00
## depunfair_mosttran_devx21[4] 0.19 0.11 0.03 0.42 1.00
## depunfair_mosttran_av12x21[1] 0.12 0.07 0.02 0.29 1.00
## depunfair_mosttran_av12x21[2] 0.11 0.07 0.01 0.29 1.00
## depunfair_mosttran_av12x21[3] 0.10 0.07 0.01 0.26 1.00
## depunfair_mosttran_av12x21[4] 0.11 0.07 0.02 0.29 1.00
## depunfair_mosttran_av12x21[5] 0.11 0.07 0.01 0.28 1.00
## depunfair_mosttran_av12x21[6] 0.13 0.08 0.02 0.31 1.00
## depunfair_mosttran_av12x21[7] 0.19 0.10 0.03 0.42 1.00
## depunfair_mosttran_av12x21[8] 0.14 0.08 0.02 0.33 1.00
## depmistrt_mosttran_devx21[1] 0.25 0.15 0.04 0.59 1.00
## depmistrt_mosttran_devx21[2] 0.21 0.13 0.03 0.53 1.00
## depmistrt_mosttran_devx21[3] 0.28 0.16 0.03 0.63 1.00
## depmistrt_mosttran_devx21[4] 0.26 0.14 0.04 0.58 1.00
## depmistrt_mosttran_av12x21[1] 0.13 0.09 0.02 0.34 1.00
## depmistrt_mosttran_av12x21[2] 0.14 0.09 0.02 0.35 1.00
## depmistrt_mosttran_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## depmistrt_mosttran_av12x21[4] 0.11 0.07 0.01 0.28 1.00
## depmistrt_mosttran_av12x21[5] 0.09 0.07 0.01 0.26 1.00
## depmistrt_mosttran_av12x21[6] 0.10 0.06 0.01 0.25 1.00
## depmistrt_mosttran_av12x21[7] 0.16 0.09 0.03 0.38 1.00
## depmistrt_mosttran_av12x21[8] 0.14 0.08 0.02 0.34 1.00
## depbetray_mosttran_devx21[1] 0.27 0.15 0.04 0.61 1.00
## depbetray_mosttran_devx21[2] 0.23 0.14 0.03 0.54 1.00
## depbetray_mosttran_devx21[3] 0.23 0.14 0.04 0.55 1.00
## depbetray_mosttran_devx21[4] 0.27 0.15 0.04 0.61 1.00
## depbetray_mosttran_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## depbetray_mosttran_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## depbetray_mosttran_av12x21[3] 0.11 0.07 0.02 0.28 1.00
## depbetray_mosttran_av12x21[4] 0.11 0.07 0.02 0.28 1.00
## depbetray_mosttran_av12x21[5] 0.10 0.06 0.01 0.26 1.00
## depbetray_mosttran_av12x21[6] 0.12 0.08 0.01 0.31 1.00
## depbetray_mosttran_av12x21[7] 0.19 0.10 0.03 0.42 1.00
## depbetray_mosttran_av12x21[8] 0.12 0.08 0.02 0.30 1.00
## Bulk_ESS Tail_ESS
## depcantgo_mosttran_devx21[1] 5232 2617
## depcantgo_mosttran_devx21[2] 5435 2844
## depcantgo_mosttran_devx21[3] 5410 2450
## depcantgo_mosttran_devx21[4] 5019 2768
## depcantgo_mosttran_av12x21[1] 5139 2137
## depcantgo_mosttran_av12x21[2] 6179 2793
## depcantgo_mosttran_av12x21[3] 5838 2451
## depcantgo_mosttran_av12x21[4] 4829 2645
## depcantgo_mosttran_av12x21[5] 4588 2082
## depcantgo_mosttran_av12x21[6] 5025 2517
## depcantgo_mosttran_av12x21[7] 4937 2809
## depcantgo_mosttran_av12x21[8] 5957 2870
## depeffort_mosttran_devx21[1] 5698 2456
## depeffort_mosttran_devx21[2] 6194 2434
## depeffort_mosttran_devx21[3] 4834 2432
## depeffort_mosttran_devx21[4] 5496 3019
## depeffort_mosttran_av12x21[1] 5777 2150
## depeffort_mosttran_av12x21[2] 5915 2120
## depeffort_mosttran_av12x21[3] 5630 2455
## depeffort_mosttran_av12x21[4] 7066 2767
## depeffort_mosttran_av12x21[5] 5997 2567
## depeffort_mosttran_av12x21[6] 5753 2635
## depeffort_mosttran_av12x21[7] 4418 3055
## depeffort_mosttran_av12x21[8] 4813 2489
## deplonely_mosttran_devx21[1] 4709 2383
## deplonely_mosttran_devx21[2] 5659 2489
## deplonely_mosttran_devx21[3] 4155 3158
## deplonely_mosttran_devx21[4] 5610 2809
## deplonely_mosttran_av12x21[1] 5971 2342
## deplonely_mosttran_av12x21[2] 6369 2563
## deplonely_mosttran_av12x21[3] 5751 2241
## deplonely_mosttran_av12x21[4] 5909 2263
## deplonely_mosttran_av12x21[5] 6368 2791
## deplonely_mosttran_av12x21[6] 5194 2493
## deplonely_mosttran_av12x21[7] 5802 2587
## deplonely_mosttran_av12x21[8] 5551 2270
## depblues_mosttran_devx21[1] 6159 2869
## depblues_mosttran_devx21[2] 6282 2730
## depblues_mosttran_devx21[3] 5326 3242
## depblues_mosttran_devx21[4] 5801 2917
## depblues_mosttran_av12x21[1] 6239 2774
## depblues_mosttran_av12x21[2] 4920 2156
## depblues_mosttran_av12x21[3] 6817 2695
## depblues_mosttran_av12x21[4] 5458 2218
## depblues_mosttran_av12x21[5] 6492 2775
## depblues_mosttran_av12x21[6] 6696 2869
## depblues_mosttran_av12x21[7] 5975 2609
## depblues_mosttran_av12x21[8] 5342 2885
## depunfair_mosttran_devx21[1] 4707 2529
## depunfair_mosttran_devx21[2] 5635 2810
## depunfair_mosttran_devx21[3] 4498 3000
## depunfair_mosttran_devx21[4] 5773 2775
## depunfair_mosttran_av12x21[1] 6364 2815
## depunfair_mosttran_av12x21[2] 6376 2375
## depunfair_mosttran_av12x21[3] 5778 2006
## depunfair_mosttran_av12x21[4] 4997 2065
## depunfair_mosttran_av12x21[5] 6198 3113
## depunfair_mosttran_av12x21[6] 5917 2470
## depunfair_mosttran_av12x21[7] 4984 3122
## depunfair_mosttran_av12x21[8] 6010 2953
## depmistrt_mosttran_devx21[1] 3926 2557
## depmistrt_mosttran_devx21[2] 5924 3133
## depmistrt_mosttran_devx21[3] 3813 3086
## depmistrt_mosttran_devx21[4] 6593 3141
## depmistrt_mosttran_av12x21[1] 5248 2187
## depmistrt_mosttran_av12x21[2] 4635 2561
## depmistrt_mosttran_av12x21[3] 4725 2342
## depmistrt_mosttran_av12x21[4] 5597 2132
## depmistrt_mosttran_av12x21[5] 5717 2648
## depmistrt_mosttran_av12x21[6] 5037 2537
## depmistrt_mosttran_av12x21[7] 5619 2707
## depmistrt_mosttran_av12x21[8] 5719 3019
## depbetray_mosttran_devx21[1] 4902 2675
## depbetray_mosttran_devx21[2] 5544 2548
## depbetray_mosttran_devx21[3] 5461 2887
## depbetray_mosttran_devx21[4] 5430 2864
## depbetray_mosttran_av12x21[1] 5263 1923
## depbetray_mosttran_av12x21[2] 5292 2624
## depbetray_mosttran_av12x21[3] 5281 2815
## depbetray_mosttran_av12x21[4] 5873 2822
## depbetray_mosttran_av12x21[5] 5550 2539
## depbetray_mosttran_av12x21[6] 6182 2531
## depbetray_mosttran_av12x21[7] 6382 2661
## depbetray_mosttran_av12x21[8] 6448 2831
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.3.2.5 Prior summary
out.chg.alldepress.sttran.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b depbetray
## normal(0, 0.125) b mosttran_av12x2 depbetray
## normal(0, 0.25) b mosttran_devx2 depbetray
## (flat) b depblues
## normal(0, 0.125) b mosttran_av12x2 depblues
## normal(0, 0.25) b mosttran_devx2 depblues
## (flat) b depcantgo
## normal(0, 0.125) b mosttran_av12x2 depcantgo
## normal(0, 0.25) b mosttran_devx2 depcantgo
## (flat) b depeffort
## normal(0, 0.125) b mosttran_av12x2 depeffort
## normal(0, 0.25) b mosttran_devx2 depeffort
## (flat) b deplonely
## normal(0, 0.125) b mosttran_av12x2 deplonely
## normal(0, 0.25) b mosttran_devx2 deplonely
## (flat) b depmistrt
## normal(0, 0.125) b mosttran_av12x2 depmistrt
## normal(0, 0.25) b mosttran_devx2 depmistrt
## (flat) b depunfair
## normal(0, 0.125) b mosttran_av12x2 depunfair
## normal(0, 0.25) b mosttran_devx2 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 depbetray
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 depbetray
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 depblues
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 depblues
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 depcantgo
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 depcantgo
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 depeffort
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 depeffort
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 deplonely
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 deplonely
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 depmistrt
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 depmistrt
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21 depunfair
## dirichlet(2, 2, 2, 2) simo mosttran_devx21 depunfair
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.3.3 B/W Change Corr: stresp/neg emotions
#Bivariate Change: negative emotions items ~ mo(stresp)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostresp_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostresp_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostresp_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostresp_av12x21',
resp = depdv_names)
)
chg.alldepress.stresp.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stresp_fit",
file_refit = "on_change"
)
out.chg.alldepress.stresp.fit <- ppchecks(chg.alldepress.stresp.fit)6.2.3.3.1 Coefficient plot (intervals)
out.chg.alldepress.stresp.fit[[11]]
6.2.3.3.2 Coefficient plot (distributions)
out.chg.alldepress.stresp.fit[[10]]
6.2.3.3.3 PPcheck (density)
p1 <- out.chg.alldepress.stresp.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.chg.alldepress.stresp.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.chg.alldepress.stresp.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.chg.alldepress.stresp.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.chg.alldepress.stresp.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.chg.alldepress.stresp.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.chg.alldepress.stresp.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
6.2.3.3.4 Fit summary
out.chg.alldepress.stresp.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## depeffort ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## deplonely ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## depblues ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## depunfair ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## depmistrt ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## depbetray ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.29 0.19 0.01 0.69 1.00 692
## sd(depeffort_Intercept) 0.44 0.26 0.03 0.98 1.01 642
## sd(deplonely_Intercept) 0.44 0.24 0.03 0.90 1.01 712
## sd(depblues_Intercept) 0.64 0.32 0.04 1.24 1.01 551
## sd(depunfair_Intercept) 0.25 0.17 0.01 0.65 1.00 1059
## sd(depmistrt_Intercept) 0.29 0.20 0.01 0.75 1.00 997
## sd(depbetray_Intercept) 0.48 0.27 0.03 1.01 1.00 692
## Tail_ESS
## sd(depcantgo_Intercept) 1541
## sd(depeffort_Intercept) 1226
## sd(deplonely_Intercept) 1247
## sd(depblues_Intercept) 1245
## sd(depunfair_Intercept) 1718
## sd(depmistrt_Intercept) 1612
## sd(depbetray_Intercept) 1585
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_Intercept -0.39 0.24 -0.88 0.06 1.00 4518
## depeffort_Intercept -2.23 0.34 -2.95 -1.63 1.00 3668
## deplonely_Intercept -1.27 0.26 -1.80 -0.78 1.00 3703
## depblues_Intercept -1.93 0.33 -2.56 -1.27 1.00 3054
## depunfair_Intercept -1.80 0.25 -2.34 -1.35 1.00 5345
## depmistrt_Intercept -2.36 0.31 -2.99 -1.76 1.00 4159
## depbetray_Intercept -2.36 0.32 -3.01 -1.78 1.00 2997
## depcantgo_mostresp_devx2 0.09 0.09 -0.09 0.28 1.00 5519
## depcantgo_mostresp_av12x2 -0.03 0.03 -0.08 0.02 1.00 7236
## depeffort_mostresp_devx2 0.14 0.12 -0.08 0.39 1.00 5737
## depeffort_mostresp_av12x2 0.05 0.03 -0.01 0.12 1.00 7387
## deplonely_mostresp_devx2 0.08 0.10 -0.10 0.28 1.00 4922
## deplonely_mostresp_av12x2 0.02 0.03 -0.04 0.08 1.00 6063
## depblues_mostresp_devx2 -0.07 0.11 -0.29 0.15 1.00 5397
## depblues_mostresp_av12x2 0.03 0.04 -0.05 0.10 1.00 6779
## depunfair_mostresp_devx2 0.26 0.09 0.10 0.45 1.00 5632
## depunfair_mostresp_av12x2 0.06 0.03 0.01 0.12 1.00 7463
## depmistrt_mostresp_devx2 0.08 0.11 -0.14 0.31 1.00 4862
## depmistrt_mostresp_av12x2 0.13 0.03 0.07 0.20 1.00 7845
## depbetray_mostresp_devx2 0.08 0.11 -0.14 0.31 1.00 5818
## depbetray_mostresp_av12x2 0.12 0.03 0.05 0.18 1.00 5226
## Tail_ESS
## depcantgo_Intercept 3005
## depeffort_Intercept 2804
## deplonely_Intercept 3012
## depblues_Intercept 3115
## depunfair_Intercept 3063
## depmistrt_Intercept 3034
## depbetray_Intercept 2945
## depcantgo_mostresp_devx2 3195
## depcantgo_mostresp_av12x2 3235
## depeffort_mostresp_devx2 3054
## depeffort_mostresp_av12x2 3537
## deplonely_mostresp_devx2 3013
## deplonely_mostresp_av12x2 3149
## depblues_mostresp_devx2 2984
## depblues_mostresp_av12x2 3237
## depunfair_mostresp_devx2 2973
## depunfair_mostresp_av12x2 2447
## depmistrt_mostresp_devx2 2932
## depmistrt_mostresp_av12x2 3121
## depbetray_mostresp_devx2 2948
## depbetray_mostresp_av12x2 2800
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_mostresp_devx21[1] 0.27 0.15 0.04 0.59 1.00
## depcantgo_mostresp_devx21[2] 0.20 0.13 0.03 0.51 1.00
## depcantgo_mostresp_devx21[3] 0.28 0.15 0.04 0.60 1.00
## depcantgo_mostresp_devx21[4] 0.26 0.14 0.04 0.58 1.00
## depcantgo_mostresp_av12x21[1] 0.13 0.08 0.02 0.34 1.00
## depcantgo_mostresp_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## depcantgo_mostresp_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mostresp_av12x21[4] 0.12 0.08 0.01 0.30 1.00
## depcantgo_mostresp_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mostresp_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mostresp_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mostresp_av12x21[8] 0.15 0.09 0.02 0.36 1.00
## depeffort_mostresp_devx21[1] 0.27 0.15 0.04 0.58 1.00
## depeffort_mostresp_devx21[2] 0.26 0.15 0.04 0.58 1.00
## depeffort_mostresp_devx21[3] 0.20 0.12 0.03 0.49 1.00
## depeffort_mostresp_devx21[4] 0.27 0.15 0.04 0.59 1.00
## depeffort_mostresp_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostresp_av12x21[2] 0.13 0.08 0.02 0.34 1.00
## depeffort_mostresp_av12x21[3] 0.12 0.08 0.02 0.30 1.00
## depeffort_mostresp_av12x21[4] 0.11 0.07 0.01 0.29 1.00
## depeffort_mostresp_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostresp_av12x21[6] 0.13 0.08 0.02 0.32 1.00
## depeffort_mostresp_av12x21[7] 0.13 0.08 0.02 0.33 1.00
## depeffort_mostresp_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## deplonely_mostresp_devx21[1] 0.25 0.14 0.04 0.58 1.00
## deplonely_mostresp_devx21[2] 0.23 0.14 0.03 0.55 1.00
## deplonely_mostresp_devx21[3] 0.25 0.14 0.04 0.55 1.00
## deplonely_mostresp_devx21[4] 0.27 0.15 0.04 0.62 1.00
## deplonely_mostresp_av12x21[1] 0.13 0.09 0.02 0.35 1.00
## deplonely_mostresp_av12x21[2] 0.14 0.09 0.02 0.35 1.00
## deplonely_mostresp_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## deplonely_mostresp_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## deplonely_mostresp_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## deplonely_mostresp_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## deplonely_mostresp_av12x21[7] 0.12 0.08 0.02 0.32 1.00
## deplonely_mostresp_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## depblues_mostresp_devx21[1] 0.27 0.15 0.04 0.59 1.00
## depblues_mostresp_devx21[2] 0.24 0.14 0.04 0.56 1.00
## depblues_mostresp_devx21[3] 0.23 0.14 0.03 0.55 1.00
## depblues_mostresp_devx21[4] 0.26 0.15 0.04 0.57 1.00
## depblues_mostresp_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## depblues_mostresp_av12x21[2] 0.12 0.07 0.02 0.30 1.00
## depblues_mostresp_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## depblues_mostresp_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## depblues_mostresp_av12x21[5] 0.12 0.08 0.02 0.32 1.00
## depblues_mostresp_av12x21[6] 0.14 0.09 0.02 0.35 1.00
## depblues_mostresp_av12x21[7] 0.14 0.09 0.02 0.35 1.00
## depblues_mostresp_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## depunfair_mostresp_devx21[1] 0.17 0.10 0.02 0.42 1.00
## depunfair_mostresp_devx21[2] 0.27 0.12 0.07 0.54 1.00
## depunfair_mostresp_devx21[3] 0.31 0.13 0.08 0.59 1.00
## depunfair_mostresp_devx21[4] 0.25 0.13 0.04 0.53 1.00
## depunfair_mostresp_av12x21[1] 0.11 0.07 0.01 0.28 1.00
## depunfair_mostresp_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## depunfair_mostresp_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## depunfair_mostresp_av12x21[4] 0.11 0.07 0.02 0.29 1.00
## depunfair_mostresp_av12x21[5] 0.12 0.08 0.02 0.30 1.00
## depunfair_mostresp_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## depunfair_mostresp_av12x21[7] 0.15 0.09 0.02 0.36 1.00
## depunfair_mostresp_av12x21[8] 0.15 0.09 0.02 0.37 1.00
## depmistrt_mostresp_devx21[1] 0.24 0.15 0.03 0.59 1.00
## depmistrt_mostresp_devx21[2] 0.25 0.14 0.04 0.57 1.00
## depmistrt_mostresp_devx21[3] 0.22 0.13 0.03 0.54 1.00
## depmistrt_mostresp_devx21[4] 0.29 0.16 0.05 0.63 1.00
## depmistrt_mostresp_av12x21[1] 0.09 0.06 0.01 0.23 1.00
## depmistrt_mostresp_av12x21[2] 0.11 0.07 0.02 0.27 1.00
## depmistrt_mostresp_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## depmistrt_mostresp_av12x21[4] 0.14 0.08 0.02 0.33 1.00
## depmistrt_mostresp_av12x21[5] 0.12 0.07 0.02 0.30 1.00
## depmistrt_mostresp_av12x21[6] 0.14 0.08 0.02 0.33 1.00
## depmistrt_mostresp_av12x21[7] 0.13 0.08 0.02 0.32 1.01
## depmistrt_mostresp_av12x21[8] 0.13 0.08 0.02 0.31 1.00
## depbetray_mostresp_devx21[1] 0.25 0.14 0.04 0.58 1.00
## depbetray_mostresp_devx21[2] 0.23 0.13 0.04 0.55 1.00
## depbetray_mostresp_devx21[3] 0.24 0.14 0.04 0.57 1.00
## depbetray_mostresp_devx21[4] 0.27 0.15 0.04 0.59 1.00
## depbetray_mostresp_av12x21[1] 0.10 0.07 0.01 0.27 1.00
## depbetray_mostresp_av12x21[2] 0.12 0.08 0.02 0.30 1.00
## depbetray_mostresp_av12x21[3] 0.13 0.08 0.02 0.31 1.00
## depbetray_mostresp_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## depbetray_mostresp_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## depbetray_mostresp_av12x21[6] 0.15 0.09 0.02 0.36 1.00
## depbetray_mostresp_av12x21[7] 0.13 0.08 0.02 0.33 1.00
## depbetray_mostresp_av12x21[8] 0.11 0.07 0.02 0.26 1.00
## Bulk_ESS Tail_ESS
## depcantgo_mostresp_devx21[1] 7244 2643
## depcantgo_mostresp_devx21[2] 5592 2978
## depcantgo_mostresp_devx21[3] 5629 2740
## depcantgo_mostresp_devx21[4] 9155 2984
## depcantgo_mostresp_av12x21[1] 7386 2524
## depcantgo_mostresp_av12x21[2] 8162 2744
## depcantgo_mostresp_av12x21[3] 8931 2320
## depcantgo_mostresp_av12x21[4] 8424 2156
## depcantgo_mostresp_av12x21[5] 6934 2731
## depcantgo_mostresp_av12x21[6] 7535 2657
## depcantgo_mostresp_av12x21[7] 6897 2768
## depcantgo_mostresp_av12x21[8] 6650 2881
## depeffort_mostresp_devx21[1] 8200 3270
## depeffort_mostresp_devx21[2] 7830 2409
## depeffort_mostresp_devx21[3] 6230 3143
## depeffort_mostresp_devx21[4] 7949 2874
## depeffort_mostresp_av12x21[1] 7837 2338
## depeffort_mostresp_av12x21[2] 7429 2944
## depeffort_mostresp_av12x21[3] 7276 2464
## depeffort_mostresp_av12x21[4] 8220 2763
## depeffort_mostresp_av12x21[5] 7848 2198
## depeffort_mostresp_av12x21[6] 8076 2512
## depeffort_mostresp_av12x21[7] 7286 2588
## depeffort_mostresp_av12x21[8] 8570 2493
## deplonely_mostresp_devx21[1] 7492 2622
## deplonely_mostresp_devx21[2] 8151 2692
## deplonely_mostresp_devx21[3] 6436 2842
## deplonely_mostresp_devx21[4] 7924 2680
## deplonely_mostresp_av12x21[1] 8344 2547
## deplonely_mostresp_av12x21[2] 7419 2594
## deplonely_mostresp_av12x21[3] 8285 2746
## deplonely_mostresp_av12x21[4] 6968 2519
## deplonely_mostresp_av12x21[5] 7972 2811
## deplonely_mostresp_av12x21[6] 7909 2795
## deplonely_mostresp_av12x21[7] 7051 2780
## deplonely_mostresp_av12x21[8] 7729 3349
## depblues_mostresp_devx21[1] 8573 2946
## depblues_mostresp_devx21[2] 7187 2340
## depblues_mostresp_devx21[3] 6958 1959
## depblues_mostresp_devx21[4] 8648 2592
## depblues_mostresp_av12x21[1] 7989 2796
## depblues_mostresp_av12x21[2] 6649 2605
## depblues_mostresp_av12x21[3] 8207 2507
## depblues_mostresp_av12x21[4] 6368 2925
## depblues_mostresp_av12x21[5] 10582 2699
## depblues_mostresp_av12x21[6] 6858 2926
## depblues_mostresp_av12x21[7] 6275 2897
## depblues_mostresp_av12x21[8] 7314 2983
## depunfair_mostresp_devx21[1] 7074 3007
## depunfair_mostresp_devx21[2] 7495 2584
## depunfair_mostresp_devx21[3] 5020 2648
## depunfair_mostresp_devx21[4] 8394 2944
## depunfair_mostresp_av12x21[1] 7372 2500
## depunfair_mostresp_av12x21[2] 7693 2351
## depunfair_mostresp_av12x21[3] 8154 2494
## depunfair_mostresp_av12x21[4] 7421 2376
## depunfair_mostresp_av12x21[5] 8076 2521
## depunfair_mostresp_av12x21[6] 8363 2796
## depunfair_mostresp_av12x21[7] 8753 2803
## depunfair_mostresp_av12x21[8] 7990 3020
## depmistrt_mostresp_devx21[1] 7088 2834
## depmistrt_mostresp_devx21[2] 9020 3096
## depmistrt_mostresp_devx21[3] 8419 3182
## depmistrt_mostresp_devx21[4] 7419 3123
## depmistrt_mostresp_av12x21[1] 6271 2672
## depmistrt_mostresp_av12x21[2] 6543 2007
## depmistrt_mostresp_av12x21[3] 7032 2476
## depmistrt_mostresp_av12x21[4] 6772 2758
## depmistrt_mostresp_av12x21[5] 7443 2502
## depmistrt_mostresp_av12x21[6] 8117 2574
## depmistrt_mostresp_av12x21[7] 7909 2709
## depmistrt_mostresp_av12x21[8] 8319 3221
## depbetray_mostresp_devx21[1] 6831 2583
## depbetray_mostresp_devx21[2] 6864 2814
## depbetray_mostresp_devx21[3] 7073 2616
## depbetray_mostresp_devx21[4] 7881 3052
## depbetray_mostresp_av12x21[1] 7162 2694
## depbetray_mostresp_av12x21[2] 8199 2808
## depbetray_mostresp_av12x21[3] 7159 2572
## depbetray_mostresp_av12x21[4] 7518 2802
## depbetray_mostresp_av12x21[5] 7210 2500
## depbetray_mostresp_av12x21[6] 7466 2491
## depbetray_mostresp_av12x21[7] 7510 2769
## depbetray_mostresp_av12x21[8] 7212 3005
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.3.3.5 Prior summary
out.chg.alldepress.stresp.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b depbetray
## normal(0, 0.125) b mostresp_av12x2 depbetray
## normal(0, 0.25) b mostresp_devx2 depbetray
## (flat) b depblues
## normal(0, 0.125) b mostresp_av12x2 depblues
## normal(0, 0.25) b mostresp_devx2 depblues
## (flat) b depcantgo
## normal(0, 0.125) b mostresp_av12x2 depcantgo
## normal(0, 0.25) b mostresp_devx2 depcantgo
## (flat) b depeffort
## normal(0, 0.125) b mostresp_av12x2 depeffort
## normal(0, 0.25) b mostresp_devx2 depeffort
## (flat) b deplonely
## normal(0, 0.125) b mostresp_av12x2 deplonely
## normal(0, 0.25) b mostresp_devx2 deplonely
## (flat) b depmistrt
## normal(0, 0.125) b mostresp_av12x2 depmistrt
## normal(0, 0.25) b mostresp_devx2 depmistrt
## (flat) b depunfair
## normal(0, 0.125) b mostresp_av12x2 depunfair
## normal(0, 0.25) b mostresp_devx2 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 depbetray
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 depbetray
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 depblues
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 depblues
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 depcantgo
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 depcantgo
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 depeffort
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 depeffort
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 deplonely
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 deplonely
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 depmistrt
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 depmistrt
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21 depunfair
## dirichlet(2, 2, 2, 2) simo mostresp_devx21 depunfair
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.3.4 B/W Change Corr: stfair/neg emotions
#Bivariate Change: negative emotions items ~ mo(stfair)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostfair_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostfair_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostfair_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostfair_av12x21',
resp = depdv_names)
)
chg.alldepress.stfair.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stfair_fit",
file_refit = "on_change"
)
out.chg.alldepress.stfair.fit <- ppchecks(chg.alldepress.stfair.fit)6.2.3.4.1 Coefficient plot (intervals)
out.chg.alldepress.stfair.fit[[11]]
6.2.3.4.2 Coefficient plot (distributions)
out.chg.alldepress.stfair.fit[[10]]
6.2.3.4.3 PPcheck (density)
p1 <- out.chg.alldepress.stfair.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.chg.alldepress.stfair.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.chg.alldepress.stfair.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.chg.alldepress.stfair.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.chg.alldepress.stfair.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.chg.alldepress.stfair.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.chg.alldepress.stfair.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
6.2.3.4.4 Fit summary
out.chg.alldepress.stfair.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## depeffort ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## deplonely ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## depblues ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## depunfair ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## depmistrt ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## depbetray ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.29 0.18 0.01 0.68 1.00 695
## sd(depeffort_Intercept) 0.45 0.26 0.03 0.98 1.00 721
## sd(deplonely_Intercept) 0.42 0.24 0.02 0.88 1.01 608
## sd(depblues_Intercept) 0.63 0.33 0.04 1.26 1.00 554
## sd(depunfair_Intercept) 0.25 0.17 0.01 0.64 1.01 956
## sd(depmistrt_Intercept) 0.29 0.21 0.01 0.75 1.00 1006
## sd(depbetray_Intercept) 0.45 0.26 0.02 0.99 1.01 534
## Tail_ESS
## sd(depcantgo_Intercept) 1732
## sd(depeffort_Intercept) 1726
## sd(deplonely_Intercept) 1695
## sd(depblues_Intercept) 1081
## sd(depunfair_Intercept) 1821
## sd(depmistrt_Intercept) 1627
## sd(depbetray_Intercept) 1081
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_Intercept -0.65 0.27 -1.23 -0.17 1.00 5275
## depeffort_Intercept -2.45 0.36 -3.20 -1.79 1.00 3995
## deplonely_Intercept -1.27 0.30 -1.95 -0.74 1.00 3939
## depblues_Intercept -2.18 0.36 -2.94 -1.50 1.00 2707
## depunfair_Intercept -1.80 0.27 -2.38 -1.30 1.00 5264
## depmistrt_Intercept -2.54 0.33 -3.24 -1.96 1.00 4419
## depbetray_Intercept -2.68 0.39 -3.53 -1.98 1.00 3318
## depcantgo_mostfair_devx2 0.18 0.11 -0.02 0.42 1.00 5477
## depcantgo_mostfair_av12x2 -0.01 0.02 -0.06 0.03 1.00 9028
## depeffort_mostfair_devx2 0.24 0.12 0.02 0.50 1.00 4958
## depeffort_mostfair_av12x2 0.05 0.03 -0.02 0.11 1.00 8344
## deplonely_mostfair_devx2 0.10 0.11 -0.11 0.35 1.00 4495
## deplonely_mostfair_av12x2 0.01 0.03 -0.04 0.07 1.00 8350
## depblues_mostfair_devx2 0.03 0.13 -0.24 0.28 1.00 5721
## depblues_mostfair_av12x2 0.03 0.04 -0.04 0.11 1.00 6796
## depunfair_mostfair_devx2 0.27 0.10 0.09 0.49 1.00 6317
## depunfair_mostfair_av12x2 0.05 0.03 -0.00 0.10 1.00 8140
## depmistrt_mostfair_devx2 0.15 0.11 -0.08 0.38 1.00 5272
## depmistrt_mostfair_av12x2 0.14 0.03 0.07 0.20 1.00 5338
## depbetray_mostfair_devx2 0.17 0.13 -0.07 0.43 1.00 4654
## depbetray_mostfair_av12x2 0.13 0.03 0.07 0.20 1.00 7185
## Tail_ESS
## depcantgo_Intercept 3262
## depeffort_Intercept 2935
## deplonely_Intercept 2637
## depblues_Intercept 2844
## depunfair_Intercept 2959
## depmistrt_Intercept 3005
## depbetray_Intercept 2864
## depcantgo_mostfair_devx2 3327
## depcantgo_mostfair_av12x2 2943
## depeffort_mostfair_devx2 3039
## depeffort_mostfair_av12x2 3042
## deplonely_mostfair_devx2 2999
## deplonely_mostfair_av12x2 3207
## depblues_mostfair_devx2 3059
## depblues_mostfair_av12x2 3012
## depunfair_mostfair_devx2 3018
## depunfair_mostfair_av12x2 3048
## depmistrt_mostfair_devx2 2807
## depmistrt_mostfair_av12x2 2816
## depbetray_mostfair_devx2 3231
## depbetray_mostfair_av12x2 3518
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_mostfair_devx21[1] 0.27 0.15 0.04 0.59 1.00
## depcantgo_mostfair_devx21[2] 0.26 0.14 0.05 0.56 1.00
## depcantgo_mostfair_devx21[3] 0.19 0.12 0.03 0.49 1.00
## depcantgo_mostfair_devx21[4] 0.28 0.15 0.04 0.62 1.00
## depcantgo_mostfair_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## depcantgo_mostfair_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## depcantgo_mostfair_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mostfair_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## depcantgo_mostfair_av12x21[5] 0.12 0.08 0.02 0.32 1.00
## depcantgo_mostfair_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## depcantgo_mostfair_av12x21[7] 0.13 0.08 0.02 0.31 1.00
## depcantgo_mostfair_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostfair_devx21[1] 0.22 0.13 0.03 0.52 1.00
## depeffort_mostfair_devx21[2] 0.34 0.14 0.09 0.64 1.00
## depeffort_mostfair_devx21[3] 0.18 0.11 0.02 0.46 1.00
## depeffort_mostfair_devx21[4] 0.26 0.14 0.04 0.56 1.00
## depeffort_mostfair_av12x21[1] 0.13 0.08 0.02 0.32 1.00
## depeffort_mostfair_av12x21[2] 0.13 0.09 0.02 0.34 1.00
## depeffort_mostfair_av12x21[3] 0.13 0.08 0.02 0.32 1.00
## depeffort_mostfair_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## depeffort_mostfair_av12x21[5] 0.11 0.07 0.02 0.30 1.00
## depeffort_mostfair_av12x21[6] 0.12 0.08 0.02 0.30 1.00
## depeffort_mostfair_av12x21[7] 0.12 0.08 0.02 0.32 1.00
## depeffort_mostfair_av12x21[8] 0.14 0.09 0.02 0.35 1.00
## deplonely_mostfair_devx21[1] 0.28 0.15 0.04 0.62 1.00
## deplonely_mostfair_devx21[2] 0.21 0.13 0.03 0.50 1.00
## deplonely_mostfair_devx21[3] 0.24 0.14 0.04 0.56 1.00
## deplonely_mostfair_devx21[4] 0.28 0.15 0.05 0.60 1.00
## deplonely_mostfair_av12x21[1] 0.13 0.08 0.02 0.34 1.00
## deplonely_mostfair_av12x21[2] 0.13 0.08 0.02 0.33 1.00
## deplonely_mostfair_av12x21[3] 0.12 0.08 0.02 0.32 1.00
## deplonely_mostfair_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## deplonely_mostfair_av12x21[5] 0.12 0.08 0.01 0.31 1.00
## deplonely_mostfair_av12x21[6] 0.13 0.08 0.02 0.31 1.00
## deplonely_mostfair_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## deplonely_mostfair_av12x21[8] 0.12 0.08 0.01 0.32 1.00
## depblues_mostfair_devx21[1] 0.26 0.15 0.04 0.59 1.00
## depblues_mostfair_devx21[2] 0.25 0.14 0.03 0.57 1.00
## depblues_mostfair_devx21[3] 0.23 0.13 0.03 0.54 1.00
## depblues_mostfair_devx21[4] 0.26 0.15 0.04 0.59 1.00
## depblues_mostfair_av12x21[1] 0.12 0.08 0.01 0.31 1.00
## depblues_mostfair_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## depblues_mostfair_av12x21[3] 0.12 0.08 0.01 0.30 1.00
## depblues_mostfair_av12x21[4] 0.11 0.08 0.01 0.30 1.00
## depblues_mostfair_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depblues_mostfair_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## depblues_mostfair_av12x21[7] 0.14 0.09 0.02 0.35 1.00
## depblues_mostfair_av12x21[8] 0.14 0.09 0.02 0.36 1.00
## depunfair_mostfair_devx21[1] 0.18 0.11 0.03 0.44 1.00
## depunfair_mostfair_devx21[2] 0.33 0.13 0.09 0.60 1.00
## depunfair_mostfair_devx21[3] 0.24 0.12 0.05 0.50 1.00
## depunfair_mostfair_devx21[4] 0.24 0.13 0.04 0.54 1.00
## depunfair_mostfair_av12x21[1] 0.11 0.07 0.01 0.28 1.00
## depunfair_mostfair_av12x21[2] 0.12 0.08 0.02 0.30 1.00
## depunfair_mostfair_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## depunfair_mostfair_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## depunfair_mostfair_av12x21[5] 0.11 0.07 0.01 0.30 1.00
## depunfair_mostfair_av12x21[6] 0.13 0.08 0.02 0.31 1.00
## depunfair_mostfair_av12x21[7] 0.15 0.10 0.02 0.38 1.00
## depunfair_mostfair_av12x21[8] 0.14 0.08 0.02 0.34 1.00
## depmistrt_mostfair_devx21[1] 0.25 0.13 0.04 0.55 1.00
## depmistrt_mostfair_devx21[2] 0.23 0.13 0.04 0.54 1.00
## depmistrt_mostfair_devx21[3] 0.28 0.14 0.05 0.58 1.00
## depmistrt_mostfair_devx21[4] 0.24 0.14 0.03 0.55 1.00
## depmistrt_mostfair_av12x21[1] 0.09 0.06 0.01 0.23 1.00
## depmistrt_mostfair_av12x21[2] 0.12 0.07 0.02 0.29 1.00
## depmistrt_mostfair_av12x21[3] 0.12 0.07 0.02 0.29 1.00
## depmistrt_mostfair_av12x21[4] 0.14 0.09 0.02 0.35 1.00
## depmistrt_mostfair_av12x21[5] 0.12 0.08 0.02 0.31 1.00
## depmistrt_mostfair_av12x21[6] 0.12 0.07 0.02 0.29 1.00
## depmistrt_mostfair_av12x21[7] 0.19 0.10 0.03 0.41 1.00
## depmistrt_mostfair_av12x21[8] 0.10 0.07 0.01 0.26 1.00
## depbetray_mostfair_devx21[1] 0.26 0.14 0.04 0.58 1.00
## depbetray_mostfair_devx21[2] 0.29 0.14 0.05 0.60 1.00
## depbetray_mostfair_devx21[3] 0.19 0.12 0.02 0.49 1.00
## depbetray_mostfair_devx21[4] 0.26 0.15 0.03 0.59 1.00
## depbetray_mostfair_av12x21[1] 0.11 0.07 0.02 0.28 1.00
## depbetray_mostfair_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## depbetray_mostfair_av12x21[3] 0.14 0.08 0.02 0.33 1.00
## depbetray_mostfair_av12x21[4] 0.15 0.09 0.02 0.35 1.00
## depbetray_mostfair_av12x21[5] 0.10 0.06 0.01 0.25 1.00
## depbetray_mostfair_av12x21[6] 0.10 0.07 0.01 0.28 1.00
## depbetray_mostfair_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## depbetray_mostfair_av12x21[8] 0.14 0.08 0.02 0.33 1.00
## Bulk_ESS Tail_ESS
## depcantgo_mostfair_devx21[1] 7214 2075
## depcantgo_mostfair_devx21[2] 8034 2196
## depcantgo_mostfair_devx21[3] 5477 3222
## depcantgo_mostfair_devx21[4] 9063 3259
## depcantgo_mostfair_av12x21[1] 7801 2357
## depcantgo_mostfair_av12x21[2] 8461 2654
## depcantgo_mostfair_av12x21[3] 8158 2634
## depcantgo_mostfair_av12x21[4] 6754 2558
## depcantgo_mostfair_av12x21[5] 7105 2079
## depcantgo_mostfair_av12x21[6] 8146 2727
## depcantgo_mostfair_av12x21[7] 7946 2963
## depcantgo_mostfair_av12x21[8] 7345 2874
## depeffort_mostfair_devx21[1] 7562 2973
## depeffort_mostfair_devx21[2] 6329 2780
## depeffort_mostfair_devx21[3] 5489 2606
## depeffort_mostfair_devx21[4] 9224 3011
## depeffort_mostfair_av12x21[1] 8330 2351
## depeffort_mostfair_av12x21[2] 9287 2646
## depeffort_mostfair_av12x21[3] 7130 2237
## depeffort_mostfair_av12x21[4] 7572 3082
## depeffort_mostfair_av12x21[5] 7859 2851
## depeffort_mostfair_av12x21[6] 7693 2849
## depeffort_mostfair_av12x21[7] 8948 3150
## depeffort_mostfair_av12x21[8] 8542 2398
## deplonely_mostfair_devx21[1] 7711 3012
## deplonely_mostfair_devx21[2] 6443 2968
## deplonely_mostfair_devx21[3] 7093 2763
## deplonely_mostfair_devx21[4] 7899 2623
## deplonely_mostfair_av12x21[1] 7496 2420
## deplonely_mostfair_av12x21[2] 8649 2429
## deplonely_mostfair_av12x21[3] 8045 2365
## deplonely_mostfair_av12x21[4] 7453 2710
## deplonely_mostfair_av12x21[5] 8268 2127
## deplonely_mostfair_av12x21[6] 8636 2620
## deplonely_mostfair_av12x21[7] 7891 3127
## deplonely_mostfair_av12x21[8] 8155 2382
## depblues_mostfair_devx21[1] 9049 2982
## depblues_mostfair_devx21[2] 7608 2846
## depblues_mostfair_devx21[3] 8016 2766
## depblues_mostfair_devx21[4] 8704 3303
## depblues_mostfair_av12x21[1] 6621 2500
## depblues_mostfair_av12x21[2] 10114 2818
## depblues_mostfair_av12x21[3] 8223 2292
## depblues_mostfair_av12x21[4] 8237 2188
## depblues_mostfair_av12x21[5] 7384 2623
## depblues_mostfair_av12x21[6] 8089 2581
## depblues_mostfair_av12x21[7] 8349 2608
## depblues_mostfair_av12x21[8] 6403 2612
## depunfair_mostfair_devx21[1] 7154 3070
## depunfair_mostfair_devx21[2] 6314 2302
## depunfair_mostfair_devx21[3] 5865 3308
## depunfair_mostfair_devx21[4] 7608 2740
## depunfair_mostfair_av12x21[1] 8039 2412
## depunfair_mostfair_av12x21[2] 8320 2469
## depunfair_mostfair_av12x21[3] 7278 2515
## depunfair_mostfair_av12x21[4] 7348 2677
## depunfair_mostfair_av12x21[5] 7885 2520
## depunfair_mostfair_av12x21[6] 7495 2312
## depunfair_mostfair_av12x21[7] 7139 2583
## depunfair_mostfair_av12x21[8] 7504 2795
## depmistrt_mostfair_devx21[1] 6745 2881
## depmistrt_mostfair_devx21[2] 8660 2703
## depmistrt_mostfair_devx21[3] 6651 2796
## depmistrt_mostfair_devx21[4] 7639 2883
## depmistrt_mostfair_av12x21[1] 8946 2799
## depmistrt_mostfair_av12x21[2] 7194 2176
## depmistrt_mostfair_av12x21[3] 7659 2294
## depmistrt_mostfair_av12x21[4] 9241 2383
## depmistrt_mostfair_av12x21[5] 8777 2430
## depmistrt_mostfair_av12x21[6] 6963 2727
## depmistrt_mostfair_av12x21[7] 7489 2714
## depmistrt_mostfair_av12x21[8] 7354 2617
## depbetray_mostfair_devx21[1] 7891 2705
## depbetray_mostfair_devx21[2] 7871 2824
## depbetray_mostfair_devx21[3] 5890 2673
## depbetray_mostfair_devx21[4] 7721 2278
## depbetray_mostfair_av12x21[1] 7693 2774
## depbetray_mostfair_av12x21[2] 7915 2682
## depbetray_mostfair_av12x21[3] 8015 2920
## depbetray_mostfair_av12x21[4] 9346 2969
## depbetray_mostfair_av12x21[5] 8879 2991
## depbetray_mostfair_av12x21[6] 7312 2443
## depbetray_mostfair_av12x21[7] 7329 2473
## depbetray_mostfair_av12x21[8] 7441 3040
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.3.4.5 Prior summary
out.chg.alldepress.stfair.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b depbetray
## normal(0, 0.125) b mostfair_av12x2 depbetray
## normal(0, 0.25) b mostfair_devx2 depbetray
## (flat) b depblues
## normal(0, 0.125) b mostfair_av12x2 depblues
## normal(0, 0.25) b mostfair_devx2 depblues
## (flat) b depcantgo
## normal(0, 0.125) b mostfair_av12x2 depcantgo
## normal(0, 0.25) b mostfair_devx2 depcantgo
## (flat) b depeffort
## normal(0, 0.125) b mostfair_av12x2 depeffort
## normal(0, 0.25) b mostfair_devx2 depeffort
## (flat) b deplonely
## normal(0, 0.125) b mostfair_av12x2 deplonely
## normal(0, 0.25) b mostfair_devx2 deplonely
## (flat) b depmistrt
## normal(0, 0.125) b mostfair_av12x2 depmistrt
## normal(0, 0.25) b mostfair_devx2 depmistrt
## (flat) b depunfair
## normal(0, 0.125) b mostfair_av12x2 depunfair
## normal(0, 0.25) b mostfair_devx2 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 depbetray
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 depbetray
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 depblues
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 depblues
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 depcantgo
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 depcantgo
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 depeffort
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 depeffort
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 deplonely
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 deplonely
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 depmistrt
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 depmistrt
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21 depunfair
## dirichlet(2, 2, 2, 2) simo mostfair_devx21 depunfair
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.3.5 B/W Change Corr: stjob/neg emotions
#Bivariate Change: negative emotions items ~ mo(stjob)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostjob_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostjob_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostjob_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostjob_av12x21',
resp = depdv_names)
)
chg.alldepress.stjob.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stjob_fit",
file_refit = "on_change"
)
out.chg.alldepress.stjob.fit <- ppchecks(chg.alldepress.stjob.fit)6.2.3.5.1 Coefficient plot (intervals)
out.chg.alldepress.stjob.fit[[11]]
6.2.3.5.2 Coefficient plot (distributions)
out.chg.alldepress.stjob.fit[[10]]
6.2.3.5.3 PPcheck (density)
p1 <- out.chg.alldepress.stjob.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.chg.alldepress.stjob.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.chg.alldepress.stjob.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.chg.alldepress.stjob.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.chg.alldepress.stjob.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.chg.alldepress.stjob.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.chg.alldepress.stjob.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
6.2.3.5.4 Fit summary
out.chg.alldepress.stjob.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## depeffort ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## deplonely ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## depblues ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## depunfair ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## depmistrt ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## depbetray ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.27 0.18 0.01 0.66 1.00 846
## sd(depeffort_Intercept) 0.42 0.27 0.02 0.97 1.00 742
## sd(deplonely_Intercept) 0.41 0.24 0.02 0.88 1.01 577
## sd(depblues_Intercept) 0.66 0.31 0.06 1.25 1.00 531
## sd(depunfair_Intercept) 0.23 0.16 0.01 0.59 1.00 1164
## sd(depmistrt_Intercept) 0.34 0.22 0.02 0.83 1.00 862
## sd(depbetray_Intercept) 0.46 0.27 0.03 1.02 1.01 651
## Tail_ESS
## sd(depcantgo_Intercept) 1542
## sd(depeffort_Intercept) 1397
## sd(deplonely_Intercept) 1298
## sd(depblues_Intercept) 962
## sd(depunfair_Intercept) 1993
## sd(depmistrt_Intercept) 1868
## sd(depbetray_Intercept) 1424
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_Intercept -0.39 0.22 -0.85 0.02 1.00 5106
## depeffort_Intercept -2.14 0.33 -2.82 -1.55 1.00 3948
## deplonely_Intercept -1.16 0.26 -1.69 -0.64 1.00 3818
## depblues_Intercept -1.97 0.33 -2.62 -1.32 1.00 2754
## depunfair_Intercept -1.83 0.27 -2.41 -1.37 1.00 4616
## depmistrt_Intercept -1.82 0.34 -2.43 -1.02 1.00 3399
## depbetray_Intercept -2.62 0.33 -3.33 -2.03 1.00 2705
## depcantgo_mostjob_devx2 0.18 0.09 -0.01 0.37 1.00 5054
## depcantgo_mostjob_av12x2 -0.06 0.02 -0.11 -0.01 1.00 7519
## depeffort_mostjob_devx2 0.14 0.12 -0.10 0.38 1.00 4827
## depeffort_mostjob_av12x2 0.04 0.03 -0.03 0.10 1.00 6877
## deplonely_mostjob_devx2 0.07 0.10 -0.13 0.26 1.00 4857
## deplonely_mostjob_av12x2 0.01 0.03 -0.05 0.06 1.00 8324
## depblues_mostjob_devx2 -0.07 0.12 -0.30 0.16 1.00 6426
## depblues_mostjob_av12x2 0.03 0.03 -0.03 0.10 1.00 6709
## depunfair_mostjob_devx2 0.22 0.10 0.04 0.42 1.00 5560
## depunfair_mostjob_av12x2 0.09 0.03 0.04 0.14 1.00 7301
## depmistrt_mostjob_devx2 0.00 0.13 -0.30 0.23 1.00 4063
## depmistrt_mostjob_av12x2 0.06 0.03 -0.00 0.12 1.00 7381
## depbetray_mostjob_devx2 0.21 0.11 0.00 0.42 1.00 4260
## depbetray_mostjob_av12x2 0.13 0.04 0.06 0.20 1.00 6094
## Tail_ESS
## depcantgo_Intercept 2673
## depeffort_Intercept 3058
## deplonely_Intercept 2733
## depblues_Intercept 2531
## depunfair_Intercept 2761
## depmistrt_Intercept 2658
## depbetray_Intercept 2202
## depcantgo_mostjob_devx2 2894
## depcantgo_mostjob_av12x2 3178
## depeffort_mostjob_devx2 3077
## depeffort_mostjob_av12x2 2813
## deplonely_mostjob_devx2 2902
## deplonely_mostjob_av12x2 3321
## depblues_mostjob_devx2 3122
## depblues_mostjob_av12x2 2960
## depunfair_mostjob_devx2 2925
## depunfair_mostjob_av12x2 3158
## depmistrt_mostjob_devx2 2421
## depmistrt_mostjob_av12x2 3332
## depbetray_mostjob_devx2 2650
## depbetray_mostjob_av12x2 3043
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_mostjob_devx21[1] 0.21 0.13 0.03 0.50 1.00 6130
## depcantgo_mostjob_devx21[2] 0.19 0.11 0.03 0.46 1.00 7577
## depcantgo_mostjob_devx21[3] 0.35 0.15 0.08 0.66 1.00 5349
## depcantgo_mostjob_devx21[4] 0.25 0.14 0.04 0.54 1.00 7668
## depcantgo_mostjob_av12x21[1] 0.10 0.07 0.01 0.27 1.00 6390
## depcantgo_mostjob_av12x21[2] 0.10 0.07 0.02 0.26 1.00 7469
## depcantgo_mostjob_av12x21[3] 0.11 0.07 0.01 0.27 1.00 8435
## depcantgo_mostjob_av12x21[4] 0.12 0.08 0.02 0.30 1.00 7108
## depcantgo_mostjob_av12x21[5] 0.14 0.09 0.02 0.34 1.00 8620
## depcantgo_mostjob_av12x21[6] 0.16 0.09 0.02 0.37 1.00 6959
## depcantgo_mostjob_av12x21[7] 0.14 0.09 0.02 0.35 1.00 6934
## depcantgo_mostjob_av12x21[8] 0.13 0.08 0.02 0.33 1.00 7734
## depeffort_mostjob_devx21[1] 0.24 0.14 0.04 0.57 1.00 7332
## depeffort_mostjob_devx21[2] 0.28 0.15 0.05 0.60 1.00 7249
## depeffort_mostjob_devx21[3] 0.22 0.13 0.03 0.52 1.00 6986
## depeffort_mostjob_devx21[4] 0.25 0.14 0.04 0.56 1.00 7228
## depeffort_mostjob_av12x21[1] 0.12 0.08 0.02 0.31 1.00 9578
## depeffort_mostjob_av12x21[2] 0.13 0.08 0.02 0.32 1.00 8860
## depeffort_mostjob_av12x21[3] 0.12 0.08 0.01 0.31 1.00 8468
## depeffort_mostjob_av12x21[4] 0.12 0.07 0.02 0.30 1.00 7184
## depeffort_mostjob_av12x21[5] 0.12 0.08 0.01 0.31 1.00 6693
## depeffort_mostjob_av12x21[6] 0.12 0.08 0.02 0.32 1.00 7457
## depeffort_mostjob_av12x21[7] 0.14 0.09 0.02 0.34 1.00 7041
## depeffort_mostjob_av12x21[8] 0.14 0.09 0.02 0.36 1.00 7045
## deplonely_mostjob_devx21[1] 0.25 0.15 0.03 0.59 1.00 6294
## deplonely_mostjob_devx21[2] 0.25 0.14 0.04 0.57 1.00 7929
## deplonely_mostjob_devx21[3] 0.24 0.14 0.04 0.55 1.00 6523
## deplonely_mostjob_devx21[4] 0.27 0.15 0.04 0.60 1.00 8071
## deplonely_mostjob_av12x21[1] 0.13 0.08 0.02 0.32 1.00 7300
## deplonely_mostjob_av12x21[2] 0.13 0.08 0.02 0.31 1.00 7164
## deplonely_mostjob_av12x21[3] 0.12 0.08 0.02 0.32 1.00 7857
## deplonely_mostjob_av12x21[4] 0.12 0.08 0.02 0.31 1.00 6985
## deplonely_mostjob_av12x21[5] 0.13 0.08 0.02 0.32 1.00 8468
## deplonely_mostjob_av12x21[6] 0.12 0.08 0.02 0.32 1.00 7824
## deplonely_mostjob_av12x21[7] 0.12 0.08 0.02 0.32 1.00 7385
## deplonely_mostjob_av12x21[8] 0.13 0.08 0.02 0.32 1.00 7401
## depblues_mostjob_devx21[1] 0.25 0.14 0.04 0.57 1.00 8140
## depblues_mostjob_devx21[2] 0.25 0.14 0.04 0.56 1.00 7537
## depblues_mostjob_devx21[3] 0.24 0.14 0.04 0.57 1.00 7298
## depblues_mostjob_devx21[4] 0.26 0.15 0.04 0.60 1.00 7935
## depblues_mostjob_av12x21[1] 0.12 0.08 0.02 0.31 1.00 8169
## depblues_mostjob_av12x21[2] 0.12 0.07 0.02 0.30 1.00 8547
## depblues_mostjob_av12x21[3] 0.12 0.08 0.02 0.31 1.00 6197
## depblues_mostjob_av12x21[4] 0.13 0.08 0.02 0.32 1.00 6916
## depblues_mostjob_av12x21[5] 0.13 0.08 0.02 0.31 1.00 9005
## depblues_mostjob_av12x21[6] 0.13 0.08 0.02 0.34 1.00 7858
## depblues_mostjob_av12x21[7] 0.13 0.08 0.02 0.34 1.00 8068
## depblues_mostjob_av12x21[8] 0.12 0.08 0.02 0.31 1.00 7465
## depunfair_mostjob_devx21[1] 0.22 0.13 0.04 0.52 1.00 6828
## depunfair_mostjob_devx21[2] 0.22 0.12 0.03 0.51 1.00 7841
## depunfair_mostjob_devx21[3] 0.33 0.14 0.07 0.63 1.00 6097
## depunfair_mostjob_devx21[4] 0.23 0.13 0.03 0.52 1.00 7444
## depunfair_mostjob_av12x21[1] 0.10 0.07 0.01 0.26 1.00 6402
## depunfair_mostjob_av12x21[2] 0.11 0.07 0.01 0.27 1.00 7362
## depunfair_mostjob_av12x21[3] 0.12 0.07 0.02 0.30 1.00 6247
## depunfair_mostjob_av12x21[4] 0.12 0.08 0.02 0.31 1.00 7486
## depunfair_mostjob_av12x21[5] 0.12 0.08 0.01 0.31 1.00 7459
## depunfair_mostjob_av12x21[6] 0.11 0.07 0.01 0.28 1.00 6877
## depunfair_mostjob_av12x21[7] 0.19 0.10 0.03 0.41 1.00 7614
## depunfair_mostjob_av12x21[8] 0.14 0.09 0.02 0.34 1.00 7109
## depmistrt_mostjob_devx21[1] 0.27 0.16 0.04 0.63 1.00 5379
## depmistrt_mostjob_devx21[2] 0.23 0.14 0.03 0.54 1.00 6153
## depmistrt_mostjob_devx21[3] 0.24 0.15 0.03 0.59 1.00 4724
## depmistrt_mostjob_devx21[4] 0.26 0.15 0.04 0.59 1.00 6910
## depmistrt_mostjob_av12x21[1] 0.14 0.08 0.02 0.34 1.00 7078
## depmistrt_mostjob_av12x21[2] 0.13 0.08 0.02 0.33 1.00 6949
## depmistrt_mostjob_av12x21[3] 0.12 0.08 0.02 0.33 1.00 7299
## depmistrt_mostjob_av12x21[4] 0.12 0.08 0.02 0.33 1.00 8183
## depmistrt_mostjob_av12x21[5] 0.12 0.08 0.01 0.30 1.00 7833
## depmistrt_mostjob_av12x21[6] 0.11 0.07 0.02 0.29 1.00 8332
## depmistrt_mostjob_av12x21[7] 0.12 0.08 0.02 0.32 1.00 7583
## depmistrt_mostjob_av12x21[8] 0.13 0.08 0.02 0.31 1.00 7467
## depbetray_mostjob_devx21[1] 0.21 0.12 0.03 0.50 1.00 8422
## depbetray_mostjob_devx21[2] 0.28 0.14 0.05 0.57 1.00 8102
## depbetray_mostjob_devx21[3] 0.28 0.14 0.05 0.59 1.00 6257
## depbetray_mostjob_devx21[4] 0.23 0.13 0.03 0.53 1.00 7343
## depbetray_mostjob_av12x21[1] 0.13 0.08 0.02 0.33 1.00 7978
## depbetray_mostjob_av12x21[2] 0.11 0.08 0.01 0.30 1.00 6621
## depbetray_mostjob_av12x21[3] 0.11 0.07 0.01 0.27 1.00 7170
## depbetray_mostjob_av12x21[4] 0.10 0.07 0.01 0.26 1.00 8239
## depbetray_mostjob_av12x21[5] 0.10 0.07 0.01 0.26 1.00 7347
## depbetray_mostjob_av12x21[6] 0.10 0.07 0.01 0.27 1.00 7680
## depbetray_mostjob_av12x21[7] 0.14 0.08 0.02 0.34 1.00 7266
## depbetray_mostjob_av12x21[8] 0.20 0.11 0.03 0.44 1.00 7156
## Tail_ESS
## depcantgo_mostjob_devx21[1] 2763
## depcantgo_mostjob_devx21[2] 2664
## depcantgo_mostjob_devx21[3] 2734
## depcantgo_mostjob_devx21[4] 2791
## depcantgo_mostjob_av12x21[1] 2387
## depcantgo_mostjob_av12x21[2] 2494
## depcantgo_mostjob_av12x21[3] 2625
## depcantgo_mostjob_av12x21[4] 2320
## depcantgo_mostjob_av12x21[5] 2305
## depcantgo_mostjob_av12x21[6] 2375
## depcantgo_mostjob_av12x21[7] 3062
## depcantgo_mostjob_av12x21[8] 2625
## depeffort_mostjob_devx21[1] 2336
## depeffort_mostjob_devx21[2] 2872
## depeffort_mostjob_devx21[3] 2785
## depeffort_mostjob_devx21[4] 2964
## depeffort_mostjob_av12x21[1] 2426
## depeffort_mostjob_av12x21[2] 3037
## depeffort_mostjob_av12x21[3] 2730
## depeffort_mostjob_av12x21[4] 2411
## depeffort_mostjob_av12x21[5] 2398
## depeffort_mostjob_av12x21[6] 2460
## depeffort_mostjob_av12x21[7] 2548
## depeffort_mostjob_av12x21[8] 2524
## deplonely_mostjob_devx21[1] 2389
## deplonely_mostjob_devx21[2] 2596
## deplonely_mostjob_devx21[3] 3089
## deplonely_mostjob_devx21[4] 2568
## deplonely_mostjob_av12x21[1] 2623
## deplonely_mostjob_av12x21[2] 2340
## deplonely_mostjob_av12x21[3] 2774
## deplonely_mostjob_av12x21[4] 2171
## deplonely_mostjob_av12x21[5] 2874
## deplonely_mostjob_av12x21[6] 2707
## deplonely_mostjob_av12x21[7] 3058
## deplonely_mostjob_av12x21[8] 2736
## depblues_mostjob_devx21[1] 2751
## depblues_mostjob_devx21[2] 2882
## depblues_mostjob_devx21[3] 2756
## depblues_mostjob_devx21[4] 2593
## depblues_mostjob_av12x21[1] 2991
## depblues_mostjob_av12x21[2] 2942
## depblues_mostjob_av12x21[3] 1517
## depblues_mostjob_av12x21[4] 2709
## depblues_mostjob_av12x21[5] 2832
## depblues_mostjob_av12x21[6] 2702
## depblues_mostjob_av12x21[7] 2676
## depblues_mostjob_av12x21[8] 2573
## depunfair_mostjob_devx21[1] 2707
## depunfair_mostjob_devx21[2] 2610
## depunfair_mostjob_devx21[3] 2787
## depunfair_mostjob_devx21[4] 2953
## depunfair_mostjob_av12x21[1] 2264
## depunfair_mostjob_av12x21[2] 2355
## depunfair_mostjob_av12x21[3] 2783
## depunfair_mostjob_av12x21[4] 2481
## depunfair_mostjob_av12x21[5] 2573
## depunfair_mostjob_av12x21[6] 2674
## depunfair_mostjob_av12x21[7] 2687
## depunfair_mostjob_av12x21[8] 2507
## depmistrt_mostjob_devx21[1] 2994
## depmistrt_mostjob_devx21[2] 2331
## depmistrt_mostjob_devx21[3] 2907
## depmistrt_mostjob_devx21[4] 2818
## depmistrt_mostjob_av12x21[1] 2805
## depmistrt_mostjob_av12x21[2] 2383
## depmistrt_mostjob_av12x21[3] 1955
## depmistrt_mostjob_av12x21[4] 2663
## depmistrt_mostjob_av12x21[5] 2424
## depmistrt_mostjob_av12x21[6] 2602
## depmistrt_mostjob_av12x21[7] 2943
## depmistrt_mostjob_av12x21[8] 2734
## depbetray_mostjob_devx21[1] 2895
## depbetray_mostjob_devx21[2] 2914
## depbetray_mostjob_devx21[3] 2859
## depbetray_mostjob_devx21[4] 2738
## depbetray_mostjob_av12x21[1] 2689
## depbetray_mostjob_av12x21[2] 1962
## depbetray_mostjob_av12x21[3] 2257
## depbetray_mostjob_av12x21[4] 2648
## depbetray_mostjob_av12x21[5] 2668
## depbetray_mostjob_av12x21[6] 2437
## depbetray_mostjob_av12x21[7] 3110
## depbetray_mostjob_av12x21[8] 2918
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.3.5.5 Prior summary
out.chg.alldepress.stjob.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b depbetray
## normal(0, 0.125) b mostjob_av12x2 depbetray
## normal(0, 0.25) b mostjob_devx2 depbetray
## (flat) b depblues
## normal(0, 0.125) b mostjob_av12x2 depblues
## normal(0, 0.25) b mostjob_devx2 depblues
## (flat) b depcantgo
## normal(0, 0.125) b mostjob_av12x2 depcantgo
## normal(0, 0.25) b mostjob_devx2 depcantgo
## (flat) b depeffort
## normal(0, 0.125) b mostjob_av12x2 depeffort
## normal(0, 0.25) b mostjob_devx2 depeffort
## (flat) b deplonely
## normal(0, 0.125) b mostjob_av12x2 deplonely
## normal(0, 0.25) b mostjob_devx2 deplonely
## (flat) b depmistrt
## normal(0, 0.125) b mostjob_av12x2 depmistrt
## normal(0, 0.25) b mostjob_devx2 depmistrt
## (flat) b depunfair
## normal(0, 0.125) b mostjob_av12x2 depunfair
## normal(0, 0.25) b mostjob_devx2 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 depbetray
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 depbetray
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 depblues
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 depblues
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 depcantgo
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 depcantgo
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 depeffort
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 depeffort
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 deplonely
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 deplonely
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 depmistrt
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 depmistrt
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21 depunfair
## dirichlet(2, 2, 2, 2) simo mostjob_devx21 depunfair
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.3.6 B/W Change Corr: stthft/neg emotions
#Bivariate Change: negative emotions items ~ mo(stthft)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostthft_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostthft_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostthft_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostthft_av12x21',
resp = depdv_names)
)
chg.alldepress.stthft.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stthft_fit",
file_refit = "on_change"
)
out.chg.alldepress.stthft.fit <- ppchecks(chg.alldepress.stthft.fit)6.2.3.6.1 Coefficient plot (intervals)
out.chg.alldepress.stthft.fit[[11]]
6.2.3.6.2 Coefficient plot (distributions)
out.chg.alldepress.stthft.fit[[10]]
6.2.3.6.3 PPcheck (density)
p1 <- out.chg.alldepress.stthft.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.chg.alldepress.stthft.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.chg.alldepress.stthft.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.chg.alldepress.stthft.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.chg.alldepress.stthft.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.chg.alldepress.stthft.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.chg.alldepress.stthft.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
6.2.3.6.4 Fit summary
out.chg.alldepress.stthft.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## depeffort ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## deplonely ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## depblues ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## depunfair ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## depmistrt ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## depbetray ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.28 0.19 0.01 0.69 1.00 673
## sd(depeffort_Intercept) 0.43 0.27 0.02 0.98 1.00 706
## sd(deplonely_Intercept) 0.42 0.24 0.02 0.90 1.01 392
## sd(depblues_Intercept) 0.63 0.32 0.05 1.24 1.01 544
## sd(depunfair_Intercept) 0.25 0.18 0.01 0.64 1.00 968
## sd(depmistrt_Intercept) 0.33 0.22 0.01 0.80 1.01 716
## sd(depbetray_Intercept) 0.53 0.29 0.04 1.10 1.00 497
## Tail_ESS
## sd(depcantgo_Intercept) 1747
## sd(depeffort_Intercept) 1419
## sd(deplonely_Intercept) 973
## sd(depblues_Intercept) 1208
## sd(depunfair_Intercept) 1814
## sd(depmistrt_Intercept) 1605
## sd(depbetray_Intercept) 1422
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_Intercept -0.79 0.24 -1.31 -0.37 1.00 3812
## depeffort_Intercept -1.87 0.30 -2.53 -1.30 1.00 2661
## deplonely_Intercept -1.53 0.27 -2.08 -1.04 1.00 3445
## depblues_Intercept -2.24 0.37 -3.02 -1.58 1.00 2434
## depunfair_Intercept -1.66 0.28 -2.28 -1.16 1.00 3776
## depmistrt_Intercept -1.92 0.27 -2.45 -1.37 1.00 3514
## depbetray_Intercept -2.40 0.32 -3.07 -1.86 1.00 2468
## depcantgo_mostthft_devx2 0.15 0.10 -0.03 0.35 1.00 4134
## depcantgo_mostthft_av12x2 0.05 0.03 -0.01 0.11 1.00 7857
## depeffort_mostthft_devx2 0.10 0.12 -0.14 0.36 1.00 4917
## depeffort_mostthft_av12x2 -0.00 0.04 -0.08 0.07 1.00 6509
## deplonely_mostthft_devx2 0.23 0.11 0.02 0.45 1.00 4177
## deplonely_mostthft_av12x2 0.03 0.03 -0.03 0.09 1.00 6699
## depblues_mostthft_devx2 0.20 0.14 -0.06 0.51 1.00 4846
## depblues_mostthft_av12x2 -0.04 0.04 -0.13 0.05 1.00 6297
## depunfair_mostthft_devx2 0.25 0.11 0.05 0.47 1.00 5000
## depunfair_mostthft_av12x2 0.04 0.03 -0.02 0.11 1.00 6426
## depmistrt_mostthft_devx2 0.13 0.12 -0.12 0.34 1.00 4026
## depmistrt_mostthft_av12x2 0.06 0.04 -0.02 0.13 1.00 6798
## depbetray_mostthft_devx2 0.26 0.11 0.05 0.49 1.00 5065
## depbetray_mostthft_av12x2 0.09 0.04 0.01 0.18 1.00 6078
## Tail_ESS
## depcantgo_Intercept 2608
## depeffort_Intercept 2679
## deplonely_Intercept 2768
## depblues_Intercept 2505
## depunfair_Intercept 2662
## depmistrt_Intercept 2425
## depbetray_Intercept 2703
## depcantgo_mostthft_devx2 2938
## depcantgo_mostthft_av12x2 3052
## depeffort_mostthft_devx2 3143
## depeffort_mostthft_av12x2 2992
## deplonely_mostthft_devx2 3097
## deplonely_mostthft_av12x2 3094
## depblues_mostthft_devx2 2994
## depblues_mostthft_av12x2 2986
## depunfair_mostthft_devx2 2972
## depunfair_mostthft_av12x2 3372
## depmistrt_mostthft_devx2 2379
## depmistrt_mostthft_av12x2 2827
## depbetray_mostthft_devx2 3169
## depbetray_mostthft_av12x2 3266
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_mostthft_devx21[1] 0.23 0.13 0.03 0.52 1.00
## depcantgo_mostthft_devx21[2] 0.28 0.14 0.05 0.59 1.00
## depcantgo_mostthft_devx21[3] 0.24 0.13 0.04 0.54 1.00
## depcantgo_mostthft_devx21[4] 0.25 0.14 0.04 0.56 1.00
## depcantgo_mostthft_av12x21[1] 0.13 0.08 0.02 0.31 1.00
## depcantgo_mostthft_av12x21[2] 0.12 0.07 0.02 0.30 1.00
## depcantgo_mostthft_av12x21[3] 0.12 0.07 0.02 0.30 1.00
## depcantgo_mostthft_av12x21[4] 0.14 0.09 0.02 0.34 1.00
## depcantgo_mostthft_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## depcantgo_mostthft_av12x21[6] 0.13 0.08 0.02 0.34 1.00
## depcantgo_mostthft_av12x21[7] 0.12 0.08 0.02 0.30 1.00
## depcantgo_mostthft_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostthft_devx21[1] 0.26 0.15 0.03 0.59 1.00
## depeffort_mostthft_devx21[2] 0.23 0.14 0.04 0.56 1.00
## depeffort_mostthft_devx21[3] 0.23 0.14 0.03 0.54 1.00
## depeffort_mostthft_devx21[4] 0.27 0.15 0.04 0.61 1.00
## depeffort_mostthft_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## depeffort_mostthft_av12x21[2] 0.12 0.08 0.02 0.32 1.00
## depeffort_mostthft_av12x21[3] 0.12 0.08 0.02 0.30 1.00
## depeffort_mostthft_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## depeffort_mostthft_av12x21[5] 0.12 0.08 0.01 0.31 1.00
## depeffort_mostthft_av12x21[6] 0.13 0.08 0.02 0.32 1.00
## depeffort_mostthft_av12x21[7] 0.13 0.08 0.02 0.33 1.00
## depeffort_mostthft_av12x21[8] 0.13 0.08 0.02 0.31 1.00
## deplonely_mostthft_devx21[1] 0.17 0.11 0.02 0.45 1.00
## deplonely_mostthft_devx21[2] 0.35 0.15 0.09 0.65 1.00
## deplonely_mostthft_devx21[3] 0.21 0.12 0.04 0.49 1.00
## deplonely_mostthft_devx21[4] 0.26 0.14 0.04 0.57 1.00
## deplonely_mostthft_av12x21[1] 0.11 0.08 0.02 0.30 1.00
## deplonely_mostthft_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## deplonely_mostthft_av12x21[3] 0.13 0.08 0.01 0.32 1.00
## deplonely_mostthft_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## deplonely_mostthft_av12x21[5] 0.13 0.08 0.02 0.31 1.00
## deplonely_mostthft_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## deplonely_mostthft_av12x21[7] 0.13 0.08 0.02 0.34 1.00
## deplonely_mostthft_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## depblues_mostthft_devx21[1] 0.25 0.14 0.04 0.57 1.00
## depblues_mostthft_devx21[2] 0.25 0.14 0.04 0.56 1.00
## depblues_mostthft_devx21[3] 0.21 0.12 0.03 0.50 1.00
## depblues_mostthft_devx21[4] 0.29 0.16 0.04 0.62 1.00
## depblues_mostthft_av12x21[1] 0.14 0.09 0.02 0.34 1.00
## depblues_mostthft_av12x21[2] 0.13 0.09 0.02 0.34 1.00
## depblues_mostthft_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## depblues_mostthft_av12x21[4] 0.12 0.08 0.02 0.31 1.00
## depblues_mostthft_av12x21[5] 0.11 0.08 0.01 0.30 1.00
## depblues_mostthft_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## depblues_mostthft_av12x21[7] 0.12 0.08 0.01 0.32 1.00
## depblues_mostthft_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## depunfair_mostthft_devx21[1] 0.22 0.13 0.03 0.50 1.00
## depunfair_mostthft_devx21[2] 0.34 0.14 0.09 0.64 1.00
## depunfair_mostthft_devx21[3] 0.20 0.11 0.03 0.47 1.00
## depunfair_mostthft_devx21[4] 0.23 0.13 0.03 0.53 1.00
## depunfair_mostthft_av12x21[1] 0.11 0.08 0.02 0.30 1.00
## depunfair_mostthft_av12x21[2] 0.11 0.07 0.01 0.28 1.00
## depunfair_mostthft_av12x21[3] 0.11 0.07 0.01 0.28 1.00
## depunfair_mostthft_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## depunfair_mostthft_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## depunfair_mostthft_av12x21[6] 0.14 0.09 0.02 0.35 1.00
## depunfair_mostthft_av12x21[7] 0.14 0.09 0.02 0.36 1.00
## depunfair_mostthft_av12x21[8] 0.14 0.09 0.02 0.36 1.00
## depmistrt_mostthft_devx21[1] 0.23 0.14 0.04 0.55 1.00
## depmistrt_mostthft_devx21[2] 0.21 0.12 0.03 0.50 1.00
## depmistrt_mostthft_devx21[3] 0.33 0.16 0.05 0.65 1.00
## depmistrt_mostthft_devx21[4] 0.24 0.14 0.03 0.56 1.00
## depmistrt_mostthft_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## depmistrt_mostthft_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## depmistrt_mostthft_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## depmistrt_mostthft_av12x21[4] 0.11 0.07 0.01 0.29 1.00
## depmistrt_mostthft_av12x21[5] 0.13 0.08 0.02 0.32 1.00
## depmistrt_mostthft_av12x21[6] 0.13 0.08 0.02 0.32 1.00
## depmistrt_mostthft_av12x21[7] 0.14 0.09 0.02 0.35 1.00
## depmistrt_mostthft_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## depbetray_mostthft_devx21[1] 0.18 0.11 0.02 0.44 1.00
## depbetray_mostthft_devx21[2] 0.32 0.14 0.08 0.60 1.00
## depbetray_mostthft_devx21[3] 0.29 0.13 0.06 0.59 1.00
## depbetray_mostthft_devx21[4] 0.21 0.12 0.03 0.49 1.00
## depbetray_mostthft_av12x21[1] 0.10 0.07 0.01 0.27 1.00
## depbetray_mostthft_av12x21[2] 0.11 0.07 0.01 0.28 1.00
## depbetray_mostthft_av12x21[3] 0.10 0.07 0.01 0.27 1.00
## depbetray_mostthft_av12x21[4] 0.11 0.07 0.01 0.28 1.00
## depbetray_mostthft_av12x21[5] 0.13 0.09 0.02 0.34 1.00
## depbetray_mostthft_av12x21[6] 0.14 0.09 0.02 0.36 1.00
## depbetray_mostthft_av12x21[7] 0.17 0.10 0.02 0.39 1.00
## depbetray_mostthft_av12x21[8] 0.13 0.08 0.02 0.32 1.00
## Bulk_ESS Tail_ESS
## depcantgo_mostthft_devx21[1] 7135 2859
## depcantgo_mostthft_devx21[2] 7977 2529
## depcantgo_mostthft_devx21[3] 6079 2523
## depcantgo_mostthft_devx21[4] 6180 2983
## depcantgo_mostthft_av12x21[1] 7832 2593
## depcantgo_mostthft_av12x21[2] 6625 2525
## depcantgo_mostthft_av12x21[3] 7434 2744
## depcantgo_mostthft_av12x21[4] 8300 2557
## depcantgo_mostthft_av12x21[5] 7426 2915
## depcantgo_mostthft_av12x21[6] 6845 2553
## depcantgo_mostthft_av12x21[7] 5906 2623
## depcantgo_mostthft_av12x21[8] 6675 2704
## depeffort_mostthft_devx21[1] 6423 2655
## depeffort_mostthft_devx21[2] 7495 2216
## depeffort_mostthft_devx21[3] 6693 2372
## depeffort_mostthft_devx21[4] 6885 2597
## depeffort_mostthft_av12x21[1] 7017 2556
## depeffort_mostthft_av12x21[2] 6795 2170
## depeffort_mostthft_av12x21[3] 5738 2250
## depeffort_mostthft_av12x21[4] 7267 1928
## depeffort_mostthft_av12x21[5] 6276 2063
## depeffort_mostthft_av12x21[6] 6548 2836
## depeffort_mostthft_av12x21[7] 8262 2922
## depeffort_mostthft_av12x21[8] 5880 2987
## deplonely_mostthft_devx21[1] 6756 2995
## deplonely_mostthft_devx21[2] 5979 2744
## deplonely_mostthft_devx21[3] 5625 2926
## deplonely_mostthft_devx21[4] 6089 2505
## deplonely_mostthft_av12x21[1] 7355 2422
## deplonely_mostthft_av12x21[2] 6827 2514
## deplonely_mostthft_av12x21[3] 6466 2763
## deplonely_mostthft_av12x21[4] 7754 2687
## deplonely_mostthft_av12x21[5] 7768 2654
## deplonely_mostthft_av12x21[6] 7475 2454
## deplonely_mostthft_av12x21[7] 6728 2808
## deplonely_mostthft_av12x21[8] 5914 3100
## depblues_mostthft_devx21[1] 6251 3005
## depblues_mostthft_devx21[2] 7258 2561
## depblues_mostthft_devx21[3] 6747 3100
## depblues_mostthft_devx21[4] 7643 2778
## depblues_mostthft_av12x21[1] 7236 2467
## depblues_mostthft_av12x21[2] 6220 2671
## depblues_mostthft_av12x21[3] 6838 2657
## depblues_mostthft_av12x21[4] 7240 2798
## depblues_mostthft_av12x21[5] 6807 2844
## depblues_mostthft_av12x21[6] 6795 2751
## depblues_mostthft_av12x21[7] 7444 2769
## depblues_mostthft_av12x21[8] 6840 3039
## depunfair_mostthft_devx21[1] 7231 3273
## depunfair_mostthft_devx21[2] 6845 2403
## depunfair_mostthft_devx21[3] 5718 3076
## depunfair_mostthft_devx21[4] 7338 3052
## depunfair_mostthft_av12x21[1] 8993 2392
## depunfair_mostthft_av12x21[2] 6920 2858
## depunfair_mostthft_av12x21[3] 6594 2521
## depunfair_mostthft_av12x21[4] 7336 2829
## depunfair_mostthft_av12x21[5] 8340 2386
## depunfair_mostthft_av12x21[6] 6424 2385
## depunfair_mostthft_av12x21[7] 6245 2728
## depunfair_mostthft_av12x21[8] 6511 3096
## depmistrt_mostthft_devx21[1] 6200 2820
## depmistrt_mostthft_devx21[2] 7954 2808
## depmistrt_mostthft_devx21[3] 4723 2943
## depmistrt_mostthft_devx21[4] 7403 2914
## depmistrt_mostthft_av12x21[1] 7377 2469
## depmistrt_mostthft_av12x21[2] 7330 2798
## depmistrt_mostthft_av12x21[3] 6761 2516
## depmistrt_mostthft_av12x21[4] 7071 2993
## depmistrt_mostthft_av12x21[5] 7099 2618
## depmistrt_mostthft_av12x21[6] 7720 2372
## depmistrt_mostthft_av12x21[7] 6920 2987
## depmistrt_mostthft_av12x21[8] 6627 2454
## depbetray_mostthft_devx21[1] 6282 2483
## depbetray_mostthft_devx21[2] 6989 2383
## depbetray_mostthft_devx21[3] 5914 2617
## depbetray_mostthft_devx21[4] 7233 2756
## depbetray_mostthft_av12x21[1] 5683 1982
## depbetray_mostthft_av12x21[2] 6370 2308
## depbetray_mostthft_av12x21[3] 7248 2569
## depbetray_mostthft_av12x21[4] 6328 2620
## depbetray_mostthft_av12x21[5] 9901 2560
## depbetray_mostthft_av12x21[6] 7822 2767
## depbetray_mostthft_av12x21[7] 5497 3001
## depbetray_mostthft_av12x21[8] 7671 2450
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.3.6.5 Prior summary
out.chg.alldepress.stthft.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b depbetray
## normal(0, 0.125) b mostthft_av12x2 depbetray
## normal(0, 0.25) b mostthft_devx2 depbetray
## (flat) b depblues
## normal(0, 0.125) b mostthft_av12x2 depblues
## normal(0, 0.25) b mostthft_devx2 depblues
## (flat) b depcantgo
## normal(0, 0.125) b mostthft_av12x2 depcantgo
## normal(0, 0.25) b mostthft_devx2 depcantgo
## (flat) b depeffort
## normal(0, 0.125) b mostthft_av12x2 depeffort
## normal(0, 0.25) b mostthft_devx2 depeffort
## (flat) b deplonely
## normal(0, 0.125) b mostthft_av12x2 deplonely
## normal(0, 0.25) b mostthft_devx2 deplonely
## (flat) b depmistrt
## normal(0, 0.125) b mostthft_av12x2 depmistrt
## normal(0, 0.25) b mostthft_devx2 depmistrt
## (flat) b depunfair
## normal(0, 0.125) b mostthft_av12x2 depunfair
## normal(0, 0.25) b mostthft_devx2 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 depbetray
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 depbetray
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 depblues
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 depblues
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 depcantgo
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 depcantgo
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 depeffort
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 depeffort
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 deplonely
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 deplonely
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 depmistrt
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 depmistrt
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21 depunfair
## dirichlet(2, 2, 2, 2) simo mostthft_devx21 depunfair
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user6.2.3.7 B/W Change Corr: stmug/neg emotions
#Bivariate Change: negative emotions items ~ mo(stmug)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
# set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmug_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmug_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmug_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmug_av12x21',
resp = depdv_names)
)
chg.alldepress.stmug.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stmug_fit",
file_refit = "on_change"
)
out.chg.alldepress.stmug.fit <- ppchecks(chg.alldepress.stmug.fit)6.2.3.7.1 Coefficient plot (intervals)
out.chg.alldepress.stmug.fit[[11]]
6.2.3.7.2 Coefficient plot (distributions)
out.chg.alldepress.stmug.fit[[10]]
6.2.3.7.3 PPcheck (density)
p1 <- out.chg.alldepress.stmug.fit[[3]] + labs(title = "Can't Get Going (T1)")
p2 <- out.chg.alldepress.stmug.fit[[4]] + labs(title = "Everything Effort (T1)")
p3 <- out.chg.alldepress.stmug.fit[[5]] + labs(title = "Lonely (T1)")
p4 <- out.chg.alldepress.stmug.fit[[6]] + labs(title = "Can't Shake Blues (T1)")
p5 <- out.chg.alldepress.stmug.fit[[7]] + labs(title = "Felt Life Unfair (T1)")
p6 <- out.chg.alldepress.stmug.fit[[8]] + labs(title = "Felt Mistreated (T1)")
p7 <- out.chg.alldepress.stmug.fit[[9]] + labs(title = "Felt Betrayed (T1)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
6.2.3.7.4 Fit summary
out.chg.alldepress.stmug.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## depeffort ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## deplonely ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## depblues ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## depunfair ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## depmistrt ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## depbetray ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.28 0.18 0.01 0.67 1.00 567
## sd(depeffort_Intercept) 0.43 0.25 0.02 0.94 1.01 620
## sd(deplonely_Intercept) 0.42 0.23 0.03 0.85 1.01 599
## sd(depblues_Intercept) 0.67 0.32 0.05 1.24 1.01 350
## sd(depunfair_Intercept) 0.23 0.16 0.01 0.60 1.00 754
## sd(depmistrt_Intercept) 0.32 0.21 0.02 0.80 1.01 580
## sd(depbetray_Intercept) 0.47 0.27 0.03 1.01 1.01 439
## Tail_ESS
## sd(depcantgo_Intercept) 1683
## sd(depeffort_Intercept) 1336
## sd(deplonely_Intercept) 1259
## sd(depblues_Intercept) 880
## sd(depunfair_Intercept) 1414
## sd(depmistrt_Intercept) 1589
## sd(depbetray_Intercept) 1205
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_Intercept -0.68 0.26 -1.23 -0.21 1.00 2764
## depeffort_Intercept -1.81 0.30 -2.41 -1.21 1.00 2820
## deplonely_Intercept -1.53 0.25 -2.09 -1.10 1.00 2378
## depblues_Intercept -1.79 0.33 -2.45 -1.12 1.00 1541
## depunfair_Intercept -1.22 0.26 -1.76 -0.70 1.00 3126
## depmistrt_Intercept -1.77 0.31 -2.35 -1.11 1.00 2555
## depbetray_Intercept -2.25 0.32 -2.92 -1.68 1.00 2457
## depcantgo_mostmug_devx2 0.13 0.10 -0.08 0.33 1.00 3011
## depcantgo_mostmug_av12x2 0.02 0.03 -0.04 0.09 1.00 5625
## depeffort_mostmug_devx2 0.05 0.13 -0.22 0.30 1.00 3780
## depeffort_mostmug_av12x2 0.01 0.04 -0.07 0.10 1.00 4886
## deplonely_mostmug_devx2 0.21 0.12 -0.01 0.46 1.00 3006
## deplonely_mostmug_av12x2 0.08 0.04 0.01 0.16 1.00 4375
## depblues_mostmug_devx2 -0.10 0.14 -0.39 0.16 1.00 3285
## depblues_mostmug_av12x2 0.02 0.05 -0.08 0.11 1.00 4869
## depunfair_mostmug_devx2 0.03 0.12 -0.19 0.27 1.00 3533
## depunfair_mostmug_av12x2 0.08 0.04 0.00 0.16 1.00 4944
## depmistrt_mostmug_devx2 -0.02 0.13 -0.28 0.22 1.00 3472
## depmistrt_mostmug_av12x2 0.13 0.04 0.05 0.22 1.00 4648
## depbetray_mostmug_devx2 0.17 0.12 -0.08 0.40 1.00 3978
## depbetray_mostmug_av12x2 0.15 0.05 0.07 0.25 1.00 6104
## Tail_ESS
## depcantgo_Intercept 2715
## depeffort_Intercept 2062
## deplonely_Intercept 1941
## depblues_Intercept 2165
## depunfair_Intercept 2529
## depmistrt_Intercept 2195
## depbetray_Intercept 2616
## depcantgo_mostmug_devx2 2487
## depcantgo_mostmug_av12x2 3040
## depeffort_mostmug_devx2 2308
## depeffort_mostmug_av12x2 2345
## deplonely_mostmug_devx2 2374
## deplonely_mostmug_av12x2 2921
## depblues_mostmug_devx2 2630
## depblues_mostmug_av12x2 3208
## depunfair_mostmug_devx2 2474
## depunfair_mostmug_av12x2 2958
## depmistrt_mostmug_devx2 2335
## depmistrt_mostmug_av12x2 2512
## depbetray_mostmug_devx2 2533
## depbetray_mostmug_av12x2 3380
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## depcantgo_mostmug_devx21[1] 0.25 0.14 0.03 0.57 1.00 4885
## depcantgo_mostmug_devx21[2] 0.30 0.15 0.05 0.61 1.00 5426
## depcantgo_mostmug_devx21[3] 0.22 0.13 0.03 0.52 1.00 5739
## depcantgo_mostmug_devx21[4] 0.23 0.14 0.03 0.56 1.00 5725
## depcantgo_mostmug_av12x21[1] 0.12 0.07 0.01 0.30 1.00 5919
## depcantgo_mostmug_av12x21[2] 0.12 0.08 0.02 0.32 1.00 5692
## depcantgo_mostmug_av12x21[3] 0.13 0.08 0.01 0.33 1.00 4788
## depcantgo_mostmug_av12x21[4] 0.12 0.08 0.02 0.31 1.00 6696
## depcantgo_mostmug_av12x21[5] 0.12 0.08 0.02 0.30 1.00 6154
## depcantgo_mostmug_av12x21[6] 0.13 0.08 0.02 0.32 1.00 6040
## depcantgo_mostmug_av12x21[7] 0.13 0.08 0.02 0.33 1.00 5738
## depcantgo_mostmug_av12x21[8] 0.13 0.08 0.02 0.33 1.00 6062
## depeffort_mostmug_devx21[1] 0.25 0.14 0.04 0.57 1.00 6070
## depeffort_mostmug_devx21[2] 0.23 0.14 0.03 0.55 1.00 6161
## depeffort_mostmug_devx21[3] 0.25 0.14 0.03 0.58 1.00 4779
## depeffort_mostmug_devx21[4] 0.26 0.14 0.04 0.58 1.00 5461
## depeffort_mostmug_av12x21[1] 0.12 0.08 0.02 0.32 1.00 5316
## depeffort_mostmug_av12x21[2] 0.12 0.08 0.02 0.32 1.00 6958
## depeffort_mostmug_av12x21[3] 0.12 0.08 0.02 0.31 1.00 6932
## depeffort_mostmug_av12x21[4] 0.12 0.08 0.01 0.31 1.00 6013
## depeffort_mostmug_av12x21[5] 0.13 0.08 0.02 0.33 1.00 6292
## depeffort_mostmug_av12x21[6] 0.13 0.08 0.02 0.32 1.00 6208
## depeffort_mostmug_av12x21[7] 0.13 0.08 0.02 0.34 1.00 6513
## depeffort_mostmug_av12x21[8] 0.13 0.08 0.02 0.34 1.00 5841
## deplonely_mostmug_devx21[1] 0.21 0.12 0.03 0.50 1.00 4530
## deplonely_mostmug_devx21[2] 0.21 0.12 0.03 0.49 1.00 6057
## deplonely_mostmug_devx21[3] 0.30 0.15 0.06 0.61 1.00 4779
## deplonely_mostmug_devx21[4] 0.28 0.15 0.05 0.61 1.00 5953
## deplonely_mostmug_av12x21[1] 0.12 0.07 0.02 0.30 1.00 4269
## deplonely_mostmug_av12x21[2] 0.12 0.08 0.02 0.31 1.00 5951
## deplonely_mostmug_av12x21[3] 0.13 0.08 0.02 0.31 1.00 5662
## deplonely_mostmug_av12x21[4] 0.12 0.08 0.01 0.31 1.00 6018
## deplonely_mostmug_av12x21[5] 0.13 0.08 0.02 0.32 1.00 6060
## deplonely_mostmug_av12x21[6] 0.14 0.09 0.02 0.34 1.00 5463
## deplonely_mostmug_av12x21[7] 0.13 0.08 0.02 0.32 1.00 6205
## deplonely_mostmug_av12x21[8] 0.11 0.07 0.02 0.30 1.00 6006
## depblues_mostmug_devx21[1] 0.25 0.15 0.04 0.58 1.00 6010
## depblues_mostmug_devx21[2] 0.23 0.13 0.03 0.54 1.00 5842
## depblues_mostmug_devx21[3] 0.24 0.14 0.04 0.56 1.00 6247
## depblues_mostmug_devx21[4] 0.28 0.15 0.04 0.61 1.00 5857
## depblues_mostmug_av12x21[1] 0.12 0.08 0.02 0.31 1.00 7401
## depblues_mostmug_av12x21[2] 0.12 0.08 0.02 0.30 1.00 5923
## depblues_mostmug_av12x21[3] 0.12 0.08 0.01 0.31 1.00 5345
## depblues_mostmug_av12x21[4] 0.12 0.08 0.02 0.31 1.00 6252
## depblues_mostmug_av12x21[5] 0.13 0.08 0.02 0.33 1.00 6283
## depblues_mostmug_av12x21[6] 0.13 0.08 0.02 0.33 1.00 6226
## depblues_mostmug_av12x21[7] 0.13 0.08 0.02 0.33 1.00 6227
## depblues_mostmug_av12x21[8] 0.13 0.08 0.02 0.33 1.00 5392
## depunfair_mostmug_devx21[1] 0.27 0.15 0.04 0.59 1.00 5659
## depunfair_mostmug_devx21[2] 0.23 0.13 0.03 0.53 1.00 5316
## depunfair_mostmug_devx21[3] 0.24 0.13 0.04 0.54 1.00 5061
## depunfair_mostmug_devx21[4] 0.27 0.15 0.04 0.60 1.00 5389
## depunfair_mostmug_av12x21[1] 0.10 0.07 0.01 0.27 1.00 6521
## depunfair_mostmug_av12x21[2] 0.09 0.06 0.01 0.25 1.00 6101
## depunfair_mostmug_av12x21[3] 0.12 0.08 0.02 0.30 1.00 5176
## depunfair_mostmug_av12x21[4] 0.13 0.08 0.02 0.33 1.00 5241
## depunfair_mostmug_av12x21[5] 0.14 0.09 0.02 0.34 1.00 5380
## depunfair_mostmug_av12x21[6] 0.14 0.09 0.02 0.37 1.00 5556
## depunfair_mostmug_av12x21[7] 0.15 0.09 0.02 0.36 1.00 5471
## depunfair_mostmug_av12x21[8] 0.13 0.08 0.02 0.33 1.00 6180
## depmistrt_mostmug_devx21[1] 0.27 0.15 0.04 0.62 1.00 4573
## depmistrt_mostmug_devx21[2] 0.24 0.14 0.03 0.56 1.00 6308
## depmistrt_mostmug_devx21[3] 0.23 0.14 0.03 0.55 1.00 4789
## depmistrt_mostmug_devx21[4] 0.26 0.15 0.03 0.60 1.00 6377
## depmistrt_mostmug_av12x21[1] 0.15 0.08 0.02 0.34 1.00 6613
## depmistrt_mostmug_av12x21[2] 0.12 0.08 0.02 0.30 1.00 5336
## depmistrt_mostmug_av12x21[3] 0.10 0.07 0.01 0.26 1.00 4712
## depmistrt_mostmug_av12x21[4] 0.11 0.08 0.02 0.29 1.00 6706
## depmistrt_mostmug_av12x21[5] 0.12 0.07 0.02 0.29 1.00 7896
## depmistrt_mostmug_av12x21[6] 0.13 0.08 0.02 0.31 1.00 6301
## depmistrt_mostmug_av12x21[7] 0.15 0.09 0.02 0.36 1.00 5224
## depmistrt_mostmug_av12x21[8] 0.12 0.08 0.02 0.31 1.00 5871
## depbetray_mostmug_devx21[1] 0.22 0.13 0.03 0.53 1.00 5908
## depbetray_mostmug_devx21[2] 0.29 0.15 0.05 0.60 1.00 5027
## depbetray_mostmug_devx21[3] 0.25 0.14 0.04 0.56 1.00 6013
## depbetray_mostmug_devx21[4] 0.23 0.13 0.03 0.53 1.00 5826
## depbetray_mostmug_av12x21[1] 0.10 0.06 0.01 0.25 1.00 6159
## depbetray_mostmug_av12x21[2] 0.10 0.06 0.01 0.25 1.00 6368
## depbetray_mostmug_av12x21[3] 0.08 0.06 0.01 0.22 1.00 4808
## depbetray_mostmug_av12x21[4] 0.11 0.07 0.01 0.28 1.00 4591
## depbetray_mostmug_av12x21[5] 0.17 0.10 0.02 0.39 1.00 6198
## depbetray_mostmug_av12x21[6] 0.19 0.11 0.03 0.43 1.00 5175
## depbetray_mostmug_av12x21[7] 0.15 0.09 0.02 0.35 1.00 5115
## depbetray_mostmug_av12x21[8] 0.12 0.07 0.01 0.30 1.00 6885
## Tail_ESS
## depcantgo_mostmug_devx21[1] 2154
## depcantgo_mostmug_devx21[2] 2582
## depcantgo_mostmug_devx21[3] 2735
## depcantgo_mostmug_devx21[4] 2708
## depcantgo_mostmug_av12x21[1] 2529
## depcantgo_mostmug_av12x21[2] 2394
## depcantgo_mostmug_av12x21[3] 2314
## depcantgo_mostmug_av12x21[4] 2540
## depcantgo_mostmug_av12x21[5] 2615
## depcantgo_mostmug_av12x21[6] 2289
## depcantgo_mostmug_av12x21[7] 2645
## depcantgo_mostmug_av12x21[8] 2708
## depeffort_mostmug_devx21[1] 2972
## depeffort_mostmug_devx21[2] 2680
## depeffort_mostmug_devx21[3] 3029
## depeffort_mostmug_devx21[4] 2836
## depeffort_mostmug_av12x21[1] 2188
## depeffort_mostmug_av12x21[2] 2373
## depeffort_mostmug_av12x21[3] 2296
## depeffort_mostmug_av12x21[4] 2390
## depeffort_mostmug_av12x21[5] 2675
## depeffort_mostmug_av12x21[6] 2738
## depeffort_mostmug_av12x21[7] 2527
## depeffort_mostmug_av12x21[8] 2771
## deplonely_mostmug_devx21[1] 2708
## deplonely_mostmug_devx21[2] 2653
## deplonely_mostmug_devx21[3] 2791
## deplonely_mostmug_devx21[4] 2988
## deplonely_mostmug_av12x21[1] 2134
## deplonely_mostmug_av12x21[2] 2228
## deplonely_mostmug_av12x21[3] 2805
## deplonely_mostmug_av12x21[4] 2613
## deplonely_mostmug_av12x21[5] 2433
## deplonely_mostmug_av12x21[6] 2682
## deplonely_mostmug_av12x21[7] 2728
## deplonely_mostmug_av12x21[8] 2870
## depblues_mostmug_devx21[1] 2406
## depblues_mostmug_devx21[2] 2947
## depblues_mostmug_devx21[3] 2703
## depblues_mostmug_devx21[4] 2710
## depblues_mostmug_av12x21[1] 2189
## depblues_mostmug_av12x21[2] 2477
## depblues_mostmug_av12x21[3] 2259
## depblues_mostmug_av12x21[4] 2716
## depblues_mostmug_av12x21[5] 2380
## depblues_mostmug_av12x21[6] 2575
## depblues_mostmug_av12x21[7] 2883
## depblues_mostmug_av12x21[8] 2734
## depunfair_mostmug_devx21[1] 2444
## depunfair_mostmug_devx21[2] 3033
## depunfair_mostmug_devx21[3] 2565
## depunfair_mostmug_devx21[4] 3024
## depunfair_mostmug_av12x21[1] 2218
## depunfair_mostmug_av12x21[2] 2127
## depunfair_mostmug_av12x21[3] 2324
## depunfair_mostmug_av12x21[4] 2042
## depunfair_mostmug_av12x21[5] 2442
## depunfair_mostmug_av12x21[6] 2713
## depunfair_mostmug_av12x21[7] 2851
## depunfair_mostmug_av12x21[8] 2948
## depmistrt_mostmug_devx21[1] 2401
## depmistrt_mostmug_devx21[2] 2244
## depmistrt_mostmug_devx21[3] 2938
## depmistrt_mostmug_devx21[4] 2476
## depmistrt_mostmug_av12x21[1] 2899
## depmistrt_mostmug_av12x21[2] 2070
## depmistrt_mostmug_av12x21[3] 1833
## depmistrt_mostmug_av12x21[4] 2726
## depmistrt_mostmug_av12x21[5] 3347
## depmistrt_mostmug_av12x21[6] 2166
## depmistrt_mostmug_av12x21[7] 2938
## depmistrt_mostmug_av12x21[8] 3003
## depbetray_mostmug_devx21[1] 2225
## depbetray_mostmug_devx21[2] 2842
## depbetray_mostmug_devx21[3] 3092
## depbetray_mostmug_devx21[4] 2707
## depbetray_mostmug_av12x21[1] 2217
## depbetray_mostmug_av12x21[2] 2294
## depbetray_mostmug_av12x21[3] 2169
## depbetray_mostmug_av12x21[4] 2399
## depbetray_mostmug_av12x21[5] 2313
## depbetray_mostmug_av12x21[6] 2673
## depbetray_mostmug_av12x21[7] 2429
## depbetray_mostmug_av12x21[8] 2775
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).6.2.3.7.5 Prior summary
out.chg.alldepress.stmug.fit[[2]]## prior class coef group resp
## (flat) b
## (flat) b depbetray
## normal(0, 0.125) b mostmug_av12x2 depbetray
## normal(0, 0.25) b mostmug_devx2 depbetray
## (flat) b depblues
## normal(0, 0.125) b mostmug_av12x2 depblues
## normal(0, 0.25) b mostmug_devx2 depblues
## (flat) b depcantgo
## normal(0, 0.125) b mostmug_av12x2 depcantgo
## normal(0, 0.25) b mostmug_devx2 depcantgo
## (flat) b depeffort
## normal(0, 0.125) b mostmug_av12x2 depeffort
## normal(0, 0.25) b mostmug_devx2 depeffort
## (flat) b deplonely
## normal(0, 0.125) b mostmug_av12x2 deplonely
## normal(0, 0.25) b mostmug_devx2 deplonely
## (flat) b depmistrt
## normal(0, 0.125) b mostmug_av12x2 depmistrt
## normal(0, 0.25) b mostmug_devx2 depmistrt
## (flat) b depunfair
## normal(0, 0.125) b mostmug_av12x2 depunfair
## normal(0, 0.25) b mostmug_devx2 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 depbetray
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 depbetray
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 depblues
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 depblues
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 depcantgo
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 depcantgo
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 depeffort
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 depeffort
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 deplonely
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 deplonely
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 depmistrt
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 depmistrt
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21 depunfair
## dirichlet(2, 2, 2, 2) simo mostmug_devx21 depunfair
## dpar nlpar lb ub source
## default
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## user
## userNow that we have model fits for criminal intent and negative emotions change models, let’s generate PLME contrasts and visualize them.
save(stress.wide4, file = here("1_Data_Files/Datasets/stress_wide4.Rdata"))
save(stress.long, file = here("1_Data_Files/Datasets/stress_long.Rdata"))7 Figure 3: T1 vs Change (T2-T1) Correlations
(RMD FILE: BDK_2023_Stress_6_Chgcorr_viz)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE7.1 Visualizing “bivariate” change associations
7.1.1 Predictive Marginal Contrasts (Crim Intent)
#load stress.wide (from Fig1 Rmd)
load(here("1_Data_Files/Datasets/stress_wide.Rdata"))
stress.wide <- zap_labels(stress.wide)
stress.wide <- zap_label(stress.wide)
load(here("1_Data_Files/Datasets/stress_wide2.Rdata"))
load(here("1_Data_Files/Datasets/stress_wide3.Rdata"))
load(here("1_Data_Files/Datasets/stress_wide4.Rdata"))
load(here("1_Data_Files/Datasets/stress_long.Rdata"))
#load change brms model fits
# criminal intent
chg.prjcrime.stmony.fit <- readRDS(here("Models/chg_prjcrime_stmony_fit.rds"))
chg.prjcrime.sttran.fit <- readRDS(here("Models/chg_prjcrime_sttran_fit.rds"))
chg.prjcrime.stresp.fit <- readRDS(here("Models/chg_prjcrime_stresp_fit.rds"))
chg.prjcrime.stfair.fit <- readRDS(here("Models/chg_prjcrime_stfair_fit.rds"))
chg.prjcrime.stjob.fit <- readRDS(here("Models/chg_prjcrime_stjob_fit.rds"))
chg.prjcrime.stthft.fit <- readRDS(here("Models/chg_prjcrime_stthft_fit.rds"))
chg.prjcrime.stmug.fit <- readRDS(here("Models/chg_prjcrime_stmug_fit.rds"))
chg.anyprjcrime.stmony.fit <- readRDS(here("Models/chg_anyprjcrime_stmony_fit.rds"))
chg.anyprjcrime.sttran.fit <- readRDS(here("Models/chg_anyprjcrime_sttran_fit.rds"))
chg.anyprjcrime.stresp.fit <- readRDS(here("Models/chg_anyprjcrime_stresp_fit.rds"))
chg.anyprjcrime.stfair.fit <- readRDS(here("Models/chg_anyprjcrime_stfair_fit.rds"))
chg.anyprjcrime.stjob.fit <- readRDS(here("Models/chg_anyprjcrime_stjob_fit.rds"))
chg.anyprjcrime.stthft.fit <- readRDS(here("Models/chg_anyprjcrime_stthft_fit.rds"))
chg.anyprjcrime.stmug.fit <- readRDS(here("Models/chg_anyprjcrime_stmug_fit.rds"))
# negative emotions
chg.alldepress.stmony.fit <- readRDS(here("Models/chg_alldepress_stmony_fit.rds"))
chg.alldepress.sttran.fit <- readRDS(here("Models/chg_alldepress_sttran_fit.rds"))
chg.alldepress.stresp.fit <- readRDS(here("Models/chg_alldepress_stresp_fit.rds"))
chg.alldepress.stfair.fit <- readRDS(here("Models/chg_alldepress_stfair_fit.rds"))
chg.alldepress.stjob.fit <- readRDS(here("Models/chg_alldepress_stjob_fit.rds"))
chg.alldepress.stthft.fit <- readRDS(here("Models/chg_alldepress_stthft_fit.rds"))
chg.alldepress.stmug.fit <- readRDS(here("Models/chg_alldepress_stmug_fit.rds"))
#T1 posterior PLME contrasts
load(here("1_Data_Files/Datasets/twodif_combineddvs3.Rdata"))
load(here("1_Data_Files/Datasets/twodif_combinedprj3.Rdata"))
load(here("1_Data_Files/Datasets/twodif_combineddep3.Rdata"))# mfx <- marginaleffects(chg.prjcrime.stmony.fit)
# summary(mfx)
# https://vincentarelbundock.github.io/marginaleffects/articles/brms.html
# ordinal models not (yet) supported. see:
# https://stackoverflow.com/questions/71333702/calculate-marginal-effects-for-ordinalclmm
# emfx.prjstmony <- chg.prjcrime.stmony.fit %>%
# emmeans(~ stmony_devx2 | rep.meas,
# at = list(stmony_av12x2 = c(2,3,4,5,6,7,8,9,10)),
# epred = TRUE) %>%
# contrast(method = "pairwise")
# emfx.prjstmony.draws <- emfx.prjstmony %>% gather_emmeans_draws()
# emmeans also does not appear to be working as desired w/ or w/out contrasts
# Solution from combining following two sources:
# https://htmlpreview.github.io/?https://github.com/mjskay/uncertainty-examples/blob/master/marginal-effects_categorical-predictor.html#differences-in-ames
# https://www.andrewheiss.com/blog/2021/11/10/ame-bayes-re-guide/#average-marginal-effects
#Manually generate predictive margins data w/following generic structure:
# create new data grid for epred values
# newdata <- mylongdata %>%
# data_grid(xdev, xbar, year)
#
# # generate epred draws (assuming two waves here)
# predmarg_data = epred_draws(mymodelfit,
# newdata = newdata,
# re_formula = NA) %>%
# filter(xdev == -2 & year == 1 |
# xdev == 2 & year == 2) %>%
# group_by(.category, xdev, .draw) %>%
# summarise(`E[y|xdev]` = mean(`.epred`))
# generating ave marginal effects ignoring person-level REs (global grand mean)
# see: https://www.andrewheiss.com/blog/2021/11/10/ame-bayes-re-guide/
# # calculate marginal effect contrasts
# margeff_contrast = predmarg_data %>%
# compare_levels(`E[y|xdev]`, by = xdev) %>% # pairwise diffs in `E[y|x]`, by levels of x
# rename(`difference in E[y|change in x (T2-T1)]` = `E[y|xdev]`) # more accurate column name
# create function to generate new data grid for epred values
gen_newdata <- function(mylongdata, xdev, xave){
mylongdata %>%
data_grid({{xdev}}, {{xave}}, year)
}
# create function to generate epred draws
# keeping only 2-unit increases from year 1 to year 2 (contrast 2_t2 - -2_t1)
gen_predmarg_data <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
filter({{xdev}} == -2 & year == 1 |
{{xdev}} == 2 & year == 2) %>%
group_by(.category, {{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`))
}
# generate epred draws
newdata <- gen_newdata(stress.long, stmony_devx2, stmony_av12x2)
predmarg_stmony_prjcrim_chg = gen_predmarg_data(chg.prjcrime.stmony.fit, stmony_devx2)
newdata <- gen_newdata(stress.long, sttran_devx2, sttran_av12x2)
predmarg_sttran_prjcrim_chg = gen_predmarg_data(chg.prjcrime.sttran.fit, sttran_devx2)
newdata <- gen_newdata(stress.long, stresp_devx2, stresp_av12x2)
predmarg_stresp_prjcrim_chg = gen_predmarg_data(chg.prjcrime.stresp.fit, stresp_devx2)
newdata <- gen_newdata(stress.long, stfair_devx2, stfair_av12x2)
predmarg_stfair_prjcrim_chg = gen_predmarg_data(chg.prjcrime.stfair.fit, stfair_devx2)
newdata <- gen_newdata(stress.long, stjob_devx2, stjob_av12x2)
predmarg_stjob_prjcrim_chg = gen_predmarg_data(chg.prjcrime.stjob.fit, stjob_devx2)
newdata <- gen_newdata(stress.long, stthft_devx2, stthft_av12x2)
predmarg_stthft_prjcrim_chg = gen_predmarg_data(chg.prjcrime.stthft.fit, stthft_devx2)
newdata <- gen_newdata(stress.long, stmug_devx2, stmug_av12x2)
predmarg_stmug_prjcrim_chg = gen_predmarg_data(chg.prjcrime.stmug.fit, stmug_devx2)
# create function to generate epred draws from "any crim intent" models
# keeping only 2-unit increases from year 1 to year 2 (contrast 2_t2 - -2_t1)
gen_predmarg_data <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
filter({{xdev}} == -2 & year == 1 |
{{xdev}} == 2 & year == 2) %>%
group_by({{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
}
# generate "any" epred data & merge with individual outcome item epred data
newdata <- gen_newdata(stress.long, stmony_devx2, stmony_av12x2)
predmarg_stmony_anyprjcrim_chg = gen_predmarg_data(chg.anyprjcrime.stmony.fit, stmony_devx2)
predmarg_stmony_prjcrim_chg <- bind_rows(predmarg_stmony_prjcrim_chg,
predmarg_stmony_anyprjcrim_chg)
rm(predmarg_stmony_anyprjcrim_chg) #clean environment
newdata <- gen_newdata(stress.long, sttran_devx2, sttran_av12x2)
predmarg_sttran_anyprjcrim_chg = gen_predmarg_data(chg.anyprjcrime.sttran.fit, sttran_devx2)
predmarg_sttran_prjcrim_chg <- bind_rows(predmarg_sttran_prjcrim_chg,
predmarg_sttran_anyprjcrim_chg)
rm(predmarg_sttran_anyprjcrim_chg) #clean environment
newdata <- gen_newdata(stress.long, stresp_devx2, stresp_av12x2)
predmarg_stresp_anyprjcrim_chg = gen_predmarg_data(chg.anyprjcrime.stresp.fit, stresp_devx2)
predmarg_stresp_prjcrim_chg <- bind_rows(predmarg_stresp_prjcrim_chg,
predmarg_stresp_anyprjcrim_chg)
rm(predmarg_stresp_anyprjcrim_chg) #clean environment
newdata <- gen_newdata(stress.long, stfair_devx2, stfair_av12x2)
predmarg_stfair_anyprjcrim_chg = gen_predmarg_data(chg.anyprjcrime.stfair.fit, stfair_devx2)
predmarg_stfair_prjcrim_chg <- bind_rows(predmarg_stfair_prjcrim_chg,
predmarg_stfair_anyprjcrim_chg)
rm(predmarg_stfair_anyprjcrim_chg) #clean environment
newdata <- gen_newdata(stress.long, stjob_devx2, stjob_av12x2)
predmarg_stjob_anyprjcrim_chg = gen_predmarg_data(chg.anyprjcrime.stjob.fit, stjob_devx2)
predmarg_stjob_prjcrim_chg <- bind_rows(predmarg_stjob_prjcrim_chg,
predmarg_stjob_anyprjcrim_chg)
rm(predmarg_stjob_anyprjcrim_chg) #clean environment
newdata <- gen_newdata(stress.long, stthft_devx2, stthft_av12x2)
predmarg_stthft_anyprjcrim_chg = gen_predmarg_data(chg.anyprjcrime.stthft.fit, stthft_devx2)
predmarg_stthft_prjcrim_chg <- bind_rows(predmarg_stthft_prjcrim_chg,
predmarg_stthft_anyprjcrim_chg)
rm(predmarg_stthft_anyprjcrim_chg) #clean environment
newdata <- gen_newdata(stress.long, stmug_devx2, stmug_av12x2)
predmarg_stmug_anyprjcrim_chg = gen_predmarg_data(chg.anyprjcrime.stmug.fit, stmug_devx2)
predmarg_stmug_prjcrim_chg <- bind_rows(predmarg_stmug_prjcrim_chg,
predmarg_stmug_anyprjcrim_chg)
rm(predmarg_stmug_anyprjcrim_chg) #clean environment
#function to calculate mean difference contrasts (marginal effect contrast)
calc_ME_chg <- function(predmarg_data, xdev) {
predmarg_data %>%
compare_levels(`E[y|xdev]`, by = xdev) %>% # pairwise diffs in `E[y|x]`, by levels of x
rename(`PLME2chg` = `E[y|xdev]`) # easy colname reflecting ME 2-unit chg contrast
}
#outputs predicted difference in E[y] associated with 2-Likert category increase in stress (T2-T1)
#marg effect contrasts are marginalized over all person-level avg values of stress (AME)
#generate ME contrasts
PLME2_stmony_prjcrim_chg = xfun::cache_rds({calc_ME_chg(predmarg_stmony_prjcrim_chg, "stmony_devx2") %>%
mutate(stress_var = "Stress:\nMoney",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmony_devx2)}, file="cache_6_1")
PLME2_sttran_prjcrim_chg = xfun::cache_rds({calc_ME_chg(predmarg_sttran_prjcrim_chg, "sttran_devx2") %>%
mutate(stress_var = "Stress:\nTransport",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = sttran_devx2)}, file="cache_6_2")
PLME2_stresp_prjcrim_chg = xfun::cache_rds({calc_ME_chg(predmarg_stresp_prjcrim_chg, "stresp_devx2") %>%
mutate(stress_var = "Stress:\nRespect",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stresp_devx2)}, file="cache_6_3")
PLME2_stfair_prjcrim_chg = xfun::cache_rds({calc_ME_chg(predmarg_stfair_prjcrim_chg, "stfair_devx2") %>%
mutate(stress_var = "Stress:\nFair Trtmt",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stfair_devx2)}, file="cache_6_4")
PLME2_stjob_prjcrim_chg = xfun::cache_rds({calc_ME_chg(predmarg_stjob_prjcrim_chg, "stjob_devx2") %>%
mutate(stress_var = "Stress:\nJob",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stjob_devx2)}, file="cache_6_5")
PLME2_stthft_prjcrim_chg = xfun::cache_rds({calc_ME_chg(predmarg_stthft_prjcrim_chg, "stthft_devx2") %>%
mutate(stress_var = "Stress:\nTheft Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stthft_devx2)}, file="cache_6_6")
PLME2_stmug_prjcrim_chg = xfun::cache_rds({calc_ME_chg(predmarg_stmug_prjcrim_chg, "stmug_devx2") %>%
mutate(stress_var = "Stress:\nAssault Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmug_devx2)}, file="cache_6_7")
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
PLME2_combinedprjchg <- bind_rows(
PLME2_stmony_prjcrim_chg,
PLME2_sttran_prjcrim_chg,
PLME2_stresp_prjcrim_chg,
PLME2_stfair_prjcrim_chg,
PLME2_stjob_prjcrim_chg,
PLME2_stthft_prjcrim_chg,
PLME2_stmug_prjcrim_chg) %>%
mutate (stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney")
)
)7.2 Supp. Figure 3A: PLME Stress Chg/Prj Crime
#function to find & drop leading zeroes (used for x-axis label)
dropLeadingZero <- function(l){
str_replace(l, '0(?=.)', '')
}
#calculate row & col averages of P(diff in E[y|stress change]) > 0
#Rows: P(mean_twodif | Stress item) > 0
#aka P(PLME>0|stress)
# View marginal probabilities
# PLME2_combinedprjchg %>%
# group_by(stress_varf) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(`PLME2chg`>0),
# p_gt0 = n_gt0 / n_ests)
# Stress:\nMoney 0.62
# Stress:\nTransport 0.30
# Stress:\nRespect 0.09
# Stress:\nFair Trtmt 0.34
# Stress:\nJob 0.49
# Stress:\nTheft Vctm 0.20
# Stress:\nAssault Vctm 0.42
#Cols: P(mean_twodif | Crime item) > 0
#aka P(PLME>0|crime)
#View marginal probabilities
# PLME2_combinedprjchg %>%
# group_by(.category) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(`PLME2chg`>0),
# p_gt0 = n_gt0 / n_ests)
# prjthflt5 0.45
# prjthfgt5 0.57
# prjthreat 0.18
# prjharm 0.19
# prjusedrg 0.24
# prjhack 0.30
# prjany 0.55
#Add these posterior probabilities into variable name
prjcrimelabsPchg <- c(
"prjthflt5"="Theft <5BAM\nP(ME>0|col)\n=.45",
"prjthfgt5"="Theft >5BAM\nP(ME>0|col)\n=.57",
"prjthreat"="Threaten\nP(ME>0|col)\n=.18",
"prjharm"="Phys. harm\nP(ME>0|col)\n=.19",
"prjusedrg"="Use drugs\nP(ME>0|col)\n=.24",
"prjhack"="Hack info.\nP(ME>0|col)\n=.30",
"prjany"="Any crime\nP(ME>0|col)\n=.55")
stress_varlabsPchg <- c(
"Stress:\nMoney" = "Stress: Money\nP(ME>0|row)\n=.62",
"Stress:\nTransport" = "Stress: Transport\nP(ME>0|row)\n=.30",
"Stress:\nRespect" = "Stress: Respect\nP(ME>0|row)=.09",
"Stress:\nFair Trtmt" = "Stress: Fair Trtmt\nP(ME>0|row)\n=.34",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.49",
"Stress:\nTheft Vctm" = "Stress: Theft\nP(ME>0|row)\n=0.20",
"Stress:\nAssault Vctm" = "Stress: Assault\nP(ME>0|row)\n=0.43")
#First, add p_gt0 variable used above to each stress/item combo in data (i.e. to each plot)
#Create alpha_scale variable =1 if p_gt0 < 0 (i.e., to be fully opaque) & =p_gt0 if p_gt0 > 1
PLME2_combinedprjchg <- PLME2_combinedprjchg %>%
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(`PLME2chg`>0),
p_gt0 = n_gt0 / n_ests,
gt0 = ifelse(`PLME2chg`>0, 1, 0),
alpha_scale = ifelse(`PLME2chg` <=0, 1, p_gt0),
rural.ses.med = as.factor("0")) %>% #add to combine figures 3 & 4 later
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
#Using sequential color palette (scio = lajolla)
#Also using ggnewscale to add multiple fill palettes
SuppFigure3A <- ggplot(data = PLME2_combinedprjchg, mapping = aes(x = `PLME2chg`, y = stress_varf)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(prjcrimelabsPchg))) +
stat_slab(mapping = aes(fill = p_gt0), .width = .95) +
scale_fill_scico(palette = "lajolla", begin = .8, end = .3, #tells it where to start and end palette
name = "P(PLME > 0)", breaks = c(.1, .5, .9), labels = dropLeadingZero) +
new_scale_fill() + #from ggnewscale
stat_halfeye(mapping = aes(fill = stat(x > 0)), .width = .95, show.legend=FALSE) +
scale_fill_manual(values = c("grey80", "NA")) +
geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.5, .5) +
coord_cartesian(xlim=c(-.03,.03)) +
scale_x_continuous(breaks=c(-.02,0,.02), labels = dropLeadingZero) +
xlab("Posterior Predicted Difference in E[y|stress increase]") +
scale_y_discrete(labels=stress_varlabsPchg) +
labs(
title = 'SUPPLEMENTAL FIGURE 3A\nMarginal Effect of 2-Category Increase in Stress on Criminal Intent from T1 to T2, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. "PLME" refers to "Practically Large Marginal Effect" of 2-category increase in stress on outcome. Darker shaded regions indicate larger\nproportion of posterior effect estimates are greater than zero (positive effect is more probable); grey shading indicates portion of posterior distribution of effect estimates equal to or less than zero.') +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
strip.background = element_blank(),
axis.text.y = element_text(size=10),
strip.text.x = element_text(size=10),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
plot.title.position = "plot",
plot.caption.position = "plot")
SuppFigure3A
#Export to image
ggsave("SuppFigure3A.jpeg", width=9, height=6.5, path=here("Output"))Compared to the T1 marginal contrasts for criminal intent, which were quite small themselves, the marginal contrasts comparing estimated changes in predicted outcome probabilities associated with a two-unit increase in stress from T1 to T2 are extremely small, with most point estimates to the left of zero. Let’s turn to computing and visualizing the negative emotions change contrasts.
7.2.1 Predictive Marginal Contrasts (Negative Emotions)
# create function to generate new data grid for epred values
gen_newdata <- function(mylongdata, xdev, xave){
mylongdata %>%
data_grid({{xdev}}, {{xave}}, year)
}
# create function to generate epred draws
# keeping only 2-unit increases from year 1 to year 2 (contrast 2_t2 - -2_t1)
gen_predmarg_data <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
filter({{xdev}} == -2 & year == 1 |
{{xdev}} == 2 & year == 2) %>%
group_by(.category, {{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`))
}
# generate epred draws using gen_newdata & gen_predmarg_data function
newdata <- gen_newdata(stress.long, stmony_devx2, stmony_av12x2)
predmarg_stmony_dep_chg = gen_predmarg_data(chg.alldepress.stmony.fit, stmony_devx2)
newdata <- gen_newdata(stress.long, sttran_devx2, sttran_av12x2)
predmarg_sttran_dep_chg = gen_predmarg_data(chg.alldepress.sttran.fit, sttran_devx2)
newdata <- gen_newdata(stress.long, stresp_devx2, stresp_av12x2)
predmarg_stresp_dep_chg = gen_predmarg_data(chg.alldepress.stresp.fit, stresp_devx2)
newdata <- gen_newdata(stress.long, stfair_devx2, stfair_av12x2)
predmarg_stfair_dep_chg = gen_predmarg_data(chg.alldepress.stfair.fit, stfair_devx2)
newdata <- gen_newdata(stress.long, stjob_devx2, stjob_av12x2)
predmarg_stjob_dep_chg = gen_predmarg_data(chg.alldepress.stjob.fit, stjob_devx2)
newdata <- gen_newdata(stress.long, stthft_devx2, stthft_av12x2)
predmarg_stthft_dep_chg = gen_predmarg_data(chg.alldepress.stthft.fit, stthft_devx2)
newdata <- gen_newdata(stress.long, stmug_devx2, stmug_av12x2)
predmarg_stmug_dep_chg = gen_predmarg_data(chg.alldepress.stmug.fit, stmug_devx2)
#outputs predicted difference in E[y] associated with 2-Likert category increase in stress (T2-T1)
#marg effect contrasts are marginalized over all person-level avg values of stress (AME)
#generate ME contrasts using calc_ME_chg function
PLME2_stmony_dep_chg = xfun::cache_rds({calc_ME_chg(predmarg_stmony_dep_chg, "stmony_devx2") %>%
mutate(stress_var = "Stress:\nMoney",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmony_devx2)}, file="cache_6_8")
PLME2_sttran_dep_chg = xfun::cache_rds({calc_ME_chg(predmarg_sttran_dep_chg, "sttran_devx2") %>%
mutate(stress_var = "Stress:\nTransport",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = sttran_devx2)}, file="cache_6_9")
PLME2_stresp_dep_chg = xfun::cache_rds({calc_ME_chg(predmarg_stresp_dep_chg, "stresp_devx2") %>%
mutate(stress_var = "Stress:\nRespect",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stresp_devx2)}, file="cache_6_10")
PLME2_stfair_dep_chg = xfun::cache_rds({calc_ME_chg(predmarg_stfair_dep_chg, "stfair_devx2") %>%
mutate(stress_var = "Stress:\nFair Trtmt",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stfair_devx2)}, file="cache_6_11")
PLME2_stjob_dep_chg = xfun::cache_rds({calc_ME_chg(predmarg_stjob_dep_chg, "stjob_devx2") %>%
mutate(stress_var = "Stress:\nJob",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stjob_devx2)}, file="cache_6_12")
PLME2_stthft_dep_chg = xfun::cache_rds({calc_ME_chg(predmarg_stthft_dep_chg, "stthft_devx2") %>%
mutate(stress_var = "Stress:\nTheft Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stthft_devx2)}, file="cache_6_13")
PLME2_stmug_dep_chg = xfun::cache_rds({calc_ME_chg(predmarg_stmug_dep_chg, "stmug_devx2") %>%
mutate(stress_var = "Stress:\nAssault Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmug_devx2)}, file="cache_6_14")
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
PLME2_combineddepchg <- bind_rows(
PLME2_stmony_dep_chg,
PLME2_sttran_dep_chg,
PLME2_stresp_dep_chg,
PLME2_stfair_dep_chg,
PLME2_stjob_dep_chg,
PLME2_stthft_dep_chg,
PLME2_stmug_dep_chg) %>%
mutate (stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney")
)
)7.3 Supp. Figure 3B: PLME Stress Chg/Neg Emotions
#calculate row & col averages of P(diff in E[y|stress change]) > 0
#Rows: P(mean_twodif | Stress item) > 0
#aka P(PLME>0|stress)
#View marginal probabilities
# PLME2_combineddepchg %>%
# group_by(stress_varf) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(`PLME2chg`>0),
# p_gt0 = n_gt0 / n_ests)
# Stress:\nMoney 0.78
# Stress:\nTransport 0.69
# Stress:\nRespect 0.76
# Stress:\nFair Trtmt 0.89
# Stress:\nJob 0.77
# Stress:\nTheft Vctm 0.93
# Stress:\nAssault Vctm 0.68
#Cols: P(mean_twodif | Crime item) > 0
#aka P(PLME>0|crime)
#View marginal probabilities
# PLME2_combineddepchg %>%
# group_by(.category) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(`PLME2chg`>0),
# p_gt0 = n_gt0 / n_ests)
# depcantgo 0.94
# depeffort 0.81
# deplonely 0.88
# depblues 0.37
# depunfair 0.94
# depmistrt 0.74
# depbetray 0.81
#Add these posterior probabilities into variable name
depresslabsPchg <- c(
"depcantgo"="Can't go\nP(ME>0|col)\n=.94",
"depeffort"="Effort\nP(ME>0|col)\n=.81",
"deplonely"="Lonely\nP(ME>0|col)\n=.88",
"depblues"="Blues\nP(ME>0|col)\n=.37",
"depunfair"="Unfair\nP(ME>0|col)\n=.94",
"depmistrt"="Mistreated\nP(ME>0|col)\n=.74",
"depbetray"="Betrayed\nP(ME>0|col)\n=.81")
stress_varlabsPchg <- c(
"Stress:\nMoney" = "Stress: Money\nP(ME>0|row)\n=.78",
"Stress:\nTransport" = "Stress: Transport\nP(ME>0|row)\n=.69",
"Stress:\nRespect" = "Stress: Respect\nP(ME>0|row)=.76",
"Stress:\nFair Trtmt" = "Stress: Fair Trtmt\nP(ME>0|row)\n=.89",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.77",
"Stress:\nTheft Vctm" = "Stress: Theft\nP(ME>0|row)\n=0.93",
"Stress:\nAssault Vctm" = "Stress: Assault\nP(ME>0|row)\n=0.68")
#First, add p_gt0 variable used above to each stress/item combo in data (i.e. to each plot)
#Create alpha_scale variable =1 if p_gt0 < 0 (i.e., to be fully opaque) & =p_gt0 if p_gt0 > 1
PLME2_combineddepchg <- PLME2_combineddepchg %>%
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(`PLME2chg`>0),
p_gt0 = n_gt0 / n_ests,
gt0 = ifelse(`PLME2chg`>0, 1, 0),
alpha_scale = ifelse(`PLME2chg` <=0, 1, p_gt0),
rural.ses.med = as.factor("0")) #add to combine figures 3 & 4 later
#Using sequential color palette (scio = lajolla)
#Also using ggnewscale to add multiple fill palettes
SuppFigure3B <- ggplot(data = PLME2_combineddepchg, mapping = aes(x = `PLME2chg`, y = stress_varf)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category= as_labeller(depresslabsPchg))) +
stat_slab(mapping = aes(fill = p_gt0), .width = .95) +
scale_fill_scico(palette = "lajolla", begin = .8, end = .3, #tells it where to start and end palette
name = "P(PLME > 0)", breaks = c(.1, .5, .9), labels = dropLeadingZero) +
new_scale_fill() + #from ggnewscale
stat_halfeye(mapping = aes(fill = stat(x > 0)), .width = .95, show.legend=FALSE) +
scale_fill_manual(values = c("grey80", "NA")) +
geom_vline(xintercept = 0, linetype = "dashed") +
# xlim(-.5, .5) +
coord_cartesian(xlim=c(-.4,.4)) +
scale_x_continuous(breaks=c(-.3,0,.3)) +
xlab("Posterior Predicted Difference in E[y|stress increase]") +
scale_y_discrete(labels=stress_varlabsPchg) +
labs(
title = 'SUPPLEMENTAL FIGURE 3B\nMarginal Effect of 2-Category Increase in Stress on Negative Emotions from T1 to T2, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. "PLME" refers to "Practically Large Marginal Effect" of 2-category increase in stress on outcome. Darker shaded regions indicate larger\nproportion of posterior effect estimates are greater than zero (positive effect is more probable); grey shading indicates portion of posterior distribution of effect estimates equal to or less than zero.') +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
axis.text.y = element_text(size=10),
strip.text.x = element_text(size=10),
strip.background = element_blank(),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
plot.title.position = "plot",
plot.caption.position = "plot")
SuppFigure3B
# Note - had to widen x-axis scale relative to T1 (larger magnitudes, wider dists)
#Export to image
ggsave("SuppFigure3B.jpeg", width=9, height=6.5, path=here("Output"))7.4 Supp. Figure 3X: “Stacked” PLME Stress/Outcomes T2-T1
#collapse outcomes & stress_vars
tempdata <- PLME2_combinedprjchg %>%
dplyr::filter(.category != "prjany")
twodif_combineddvs_chg <-
rbind(tempdata, PLME2_combineddepchg) %>%
mutate(
.category = forcats::fct_collapse(.category,
"prjtheft" = c("prjthflt5", "prjthfgt5"),
"prjviol" = c("prjthreat", "prjharm"),
"prjusedrg" = "prjusedrg",
"prjhack" = "prjhack",
"depsym" = c("depcantgo", "depeffort", "deplonely", "depblues"),
"negemo" = c("depunfair", "depmistrt", "depbetray")),
stress_var = if_else(stress_var %in% c("Stress:\nMoney", "Stress:\nTransport"),
"Stress:\nFinancial", stress_var),
stress_var = if_else(stress_var %in% c("Stress:\nRespect", "Stress:\nFair Trtmt"),
"Stress:\nRelational", stress_var),
stress_var = if_else(stress_var %in% c("Stress:\nTheft Vctm", "Stress:\nAssault Vctm"),
"Stress:\nCrime Vctm", stress_var),
stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nCrime Vctm",
"Stress:\nJob",
"Stress:\nRelational",
"Stress:\nFinancial"))
) %>%
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME2chg>0),
p_gt0 = n_gt0 / n_ests,
gt0 = ifelse(PLME2chg>0, 1, 0),
alpha_scale = ifelse(PLME2chg <=0, 1, p_gt0),
rural.ses.med = as.factor("0"))
#calculate row & col averages of P(mean_twodif) > 0
#Rows: P(mean_twodif | Stress item) > 0
#aka P(PLME>0|stress)
#View marginal probabilities
# twodif_combineddvs_chg %>%
# group_by(stress_varf) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(PLME2chg>0),
# p_gt0 = n_gt0 / n_ests)
# Stress:\nCrime Vctm 0.57
# Stress:\nJob 24000 0.62
# Stress:\nRelational 0.53
# Stress:\nFinancial 0.58
#Cols: P(mean_twodif | Crime item) > 0
#aka P(PLME>0|crime)
#View marginal probabilities
# twodif_combineddvs_chg %>%
# group_by(.category) %>%
# summarise(n_ests = n(),
# n_gt0 = sum(PLME2chg>0),
# p_gt0 = n_gt0 / n_ests)
# prjtheft 0.51
# prjviol 0.18
# prjusedrg 0.24
# prjhack 0.30
# depsym 0.75
# negemo 0.83
#Add these posterior probabilities into variable name
outcomelabsPchg <- c(
"prjtheft"="Theft\nIntent\nP(ME>0|col)\n=.51",
"prjviol"="Violence\nIntent\nP(ME>0|col)\n=.18",
"prjusedrg"="Use drugs\nIntent\nP(ME>0|col)\n=.24",
"prjhack"="Hack info.\nIntent\nP(ME>0|col)\n=.30",
"depsym"="Depressive\nSymptoms\nP(ME>0|col)\n=.75",
"negemo"="Criminogenic\nEmotions\nP(ME>0|col)\n=.83")
stress_varlabsPchg <- c(
"Stress:\nFinancial" = "Stress: Financial\nP(ME>0|row)\n=.58",
"Stress:\nRelational" = "Stress: Personal\nP(ME>0|row)\n=.53",
"Stress:\nJob" = "Stress: Job\nP(ME>0|row)\n=0.62",
"Stress:\nCrime Vctm" = "Stress: Crime Vctm\nP(ME>0|row)\n=0.57")
SuppFigure3X <- ggplot(data = twodif_combineddvs_chg, mapping = aes(x = PLME2chg, y = stress_varf)) +
facet_wrap(~.category, nrow=1, scales="free_x",
labeller = labeller(.category= as_labeller(outcomelabsPchg))) +
stat_slab(color = "#883E3A", fill = NA, normalize="panels", height=.5) +
stat_slab(fill = "#883E3A", alpha=.1, normalize="panels", height=.5) +
stat_pointinterval(width = .95, color = "#883E3A",
position=position_dodge(width=.2, preserve = "single")) +
geom_vline(xintercept = 0, linetype = "dashed") +
xlab("Posterior Predicted Difference in E[y|stress increase]") +
scale_y_discrete(labels=stress_varlabsPchg) +
facetted_pos_scales(
x = list(
.category %in% c("prjtheft", "prjviol", "prjusedrg") ~
scale_x_continuous(breaks=c(-.01,0,.01),
limits=c(-.02,.011), labels = dropLeadingZero),
.category == "prjhack" ~
scale_x_continuous(breaks=c(-.04,0,.04),
limits=c(-.06,.041), labels = dropLeadingZero),
.category %in% c("depsym", "negemo") ~
scale_x_continuous(breaks=c(-.3,0,.3),
limits=c(-.5,.5), labels = dropLeadingZero)
) ) +
labs(
title = 'SUPPLEMENTAL FIGURE 3X\n"Stacked" Marginal Effects of 2-Category Stress Increase on Change in Outcome Probabilities from T1 to T2, Full Sample',
#subtitle = 'Subtitle here',
caption = 'Note: N=489 respondents participating at both survey waves. Estimates derived from 14 multivariate Bayesian logistic regression models, with each of seven \ncombined criminal intent outcomes (using `brms::mvbind()`) and each of seven combined negative emotion outcomes separately specifying a single T1 stress \ntypes as a predictor. Stress items were separated into a cross-time average (Xbar_i) between-person L2 predictor and a within-person change (X_it - Xbar_i) \n"fixed effects" estimator. Both stress predictors were specified as monotonic ordinal predictors with a cumulative probit link function. Models were estimated \nin brms with 4 chains and 4000 total post-warmup posterior draws. Practically Large Marginal Effect (PLME) parameter distributions were first estimated \nfrom the expectation of the posterior predictive distribution for each model as predicted probability difference distributions for 2-category stress increase \ncontrasts from T1 to T2 for each bivariate item pair. The final displayed PLME posterior parameter estimate distributions were then generated by "stacking" \n(with equal model weights) to generate average posterior PLME estimates across models with conceptually similar outcome (e.g., Depressive Symptoms) and \nstress (e.g., Financial) item groupings') +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
strip.background = element_blank(),
axis.text.y = element_text(size=10),
strip.text.x = element_text(size=10),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
plot.title.position = "plot",
plot.caption.position = "plot")
SuppFigure3X 
#Export to image
# ggsave("SuppFigure3X.jpeg", width=9, height=6.5, path=here("Output"))7.5 Supp. Figure 3XX: “Stacked” PLME Combined Stress/Outcomes T1 vs Change
#wrangle PLME T2-T1 change estimates
PLME2chg <- twodif_combineddvs_chg %>%
rename(PLME = PLME2chg) %>%
mutate(
method=as.factor("chg"),
dif_label='diff in E[y|2-cat stress diff/increase]'
) %>%
dplyr::select(-c(n_ests,gt0,n_gt0,p_gt0,alpha_scale, contrast))
#wrangle PLME T1 difference estimates
PLME2T1 <- twodif_combineddvs3 %>%
rename(PLME = mean_twodif) %>%
mutate(
method=as.factor("T1"),
dif_label='diff in E[y|2-cat stress diff/increase]',
.category = forcats::fct_recode(.category,
"prjtheft" = "prjtheftw1",
"prjviol" = "prjviolw1",
"prjusedrg" = "prjusedrgw1f",
"prjhack" = "prjhackw1f",
"depsym" = "depsymw1",
"negemo" = "negemow1"
)
) %>%
dplyr::select(-c(.chain,.iteration,n_ests,gt0,n_gt0,p_gt0,alpha_scale))
#merge change & T1 estimates (method var indicates which data group)
twodif_compare <- bind_rows(PLME2chg, PLME2T1)
outcomelabs <- c(
"prjtheft"="Theft\nIntent",
"prjviol"="Violence\nIntent",
"prjusedrg"="Use drugs\nIntent",
"prjhack"="Hack info.\nIntent",
"depsym"="Depressive\nSymptoms",
"negemo"="Criminogenic\nEmotions")
stress_varlabs <- c(
"Stress:\nFinancial" = "Stress:\nFinancial",
"Stress:\nRelational" = "Stress:\nPersonal",
"Stress:\nJob" = "Stress:\nJob",
"Stress:\nCrime Vctm" = "Stress:\nCrime Vctm")
SuppFigure3XX <- ggplot(data = PLME2chg, mapping = aes(x = PLME, y = stress_varf)) +
facet_wrap(~.category, nrow=1, scales="free_x",
labeller = labeller(.category= as_labeller(outcomelabs))) +
# stat_slab(data=PLME2chg, aes(x=PLME, y=stress_varf), color="#883E3A",
# fill = NA, normalize="panels", height=.5) +
# stat_slab(data=PLME2chg, aes(x=PLME, y=stress_varf), color=NA,
# fill = "#883E3A", alpha=.1, normalize="panels", height=.5) +
# stat_slab(data=PLME2T1, aes(x=PLME, y=stress_varf), color="#E99D53",
# fill = NA, normalize="panels", height=.5) +
# stat_slab(data=PLME2T1, aes(x=PLME, y=stress_varf), color=NA,
# fill = "#E99D53", alpha=.1, normalize="panels", height=.5) +
stat_pointinterval(data=PLME2chg, aes(x=PLME, y=stress_varf), color="#883E3A",
width = .95, position = position_nudge(y = -.1)) +
stat_pointinterval(data=PLME2T1, aes(x=PLME, y=stress_varf), color="#E99D53",
width = .95, position = position_nudge(y = .1)) +
geom_vline(xintercept = 0, linetype = "dashed") +
xlab("Posterior Predicted Difference in E[y|stress increase]") +
scale_y_discrete(labels=stress_varlabs) +
# facetted_pos_scales(
# x = list(
# .category %in% c("prjtheft", "prjviol", "prjusedrg") ~
# scale_x_continuous(breaks=c(-.01,0,.01),
# limits=c(-.02,.011), labels = dropLeadingZero),
# .category == "prjhack" ~
# scale_x_continuous(breaks=c(-.04,0,.04),
# limits=c(-.06,.041), labels = dropLeadingZero),
# .category %in% c("depsym", "negemo") ~
# scale_x_continuous(breaks=c(-.3,0,.3),
# limits=c(-.5,.5), labels = dropLeadingZero)
# ) ) +
labs(
title = 'SUPPLEMENTAL FIGURE 3XX\n"Stacked" Marginal Effects of 2-Category Stress Increase on Change in Outcome Probabilities from T1 to T2, Full Sample') +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
strip.background = element_blank(),
axis.text.y = element_text(size=10),
strip.text.x = element_text(size=10),
plot.title = element_text(size=12, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
plot.title.position = "plot",
plot.caption.position = "plot")
SuppFigure3XX 
#Export to image
# ggsave("SuppFigure3XX.jpeg", width=9, height=6.5, path=here("Output"))Let’s move on from these exploratory stacked posterior plots and return to item-specific estimates. Rather than collapsing categories for the sake of plots that are easier to read, we will opt for more densely packed plots that contain far more information and that present more precise estimates of stress-outcome correlations in these data.
7.6 FIGURE 3: PLME Stress/Outcomes T1 vs Change
#wrangle PLME prj crime T1 difference estimates
PLMEprjT1 <- twodif_combinedprj3 %>%
rename(PLME = mean_twodif) %>%
mutate(
method=as.factor("T1"),
dif_label='diff in E[y|2-cat stress diff/increase]',
.category = forcats::fct_recode(.category,
"prjthflt5" = "prjthflt5w1f",
"prjthfgt5" = "prjthfgt5w1f",
"prjthreat" = "prjthreatw1f",
"prjharm" = "prjharmw1f",
"prjusedrg" = "prjusedrgw1f",
"prjhack" = "prjhackw1f",
"prjany" = "prjanyw1f"
),
p80_gt0 = as.factor(if_else(p_gt0 >= .80, 1, 0))
) %>%
dplyr::select(-c(.chain,.iteration))
#wrangle PLME neg emo T1 difference estimates
PLMEdepT1 <- twodif_combineddep3 %>%
rename(PLME = mean_twodif) %>%
mutate(
method=as.factor("T1"),
dif_label='diff in E[y|2-cat stress diff/increase]',
.category = forcats::fct_recode(.category,
"depcantgo"="depcantgow1f",
"depeffort"="depeffortw1f",
"deplonely"="deplonelyw1f",
"depblues"="depbluesw1f",
"depunfair"="depunfairw1f",
"depmistrt"="depmistrtw1f",
"depbetray"="depbetrayw1f"
),
p80_gt0 = as.factor(if_else(p_gt0 >= .80, 1, 0))
) %>%
dplyr::select(-c(.chain,.iteration))
#wrangle PLME prj crime T2-T1 change estimates
PLMEprjchg <- PLME2_combinedprjchg %>%
rename(PLME = PLME2chg) %>%
mutate(
method=as.factor("chg"),
dif_label='diff in E[y|2-cat stress diff/increase]',
p80_gt0 = as.factor(if_else(p_gt0 >= .80, 1, 0))
) %>%
dplyr::select(-c(contrast))
#wrangle PLME neg emo T2-T1 change estimates
PLMEdepchg <- PLME2_combineddepchg %>%
rename(PLME = PLME2chg) %>%
mutate(
method=as.factor("chg"),
dif_label='diff in E[y|2-cat stress diff/increase]',
p80_gt0 = as.factor(if_else(p_gt0 >= .80, 1, 0))
) %>%
dplyr::select(-c(contrast))
prjlabs <- c(
"prjthflt5"="Theft <5BAM",
"prjthfgt5"="Theft >5BAM",
"prjthreat"="Threaten",
"prjharm"="Phys. harm",
"prjusedrg"="Use drugs",
"prjhack"="Hack info",
"prjany"="Any crime")
deplabs <- c(
"depcantgo"="Can't go",
"depeffort"="Effort",
"deplonely"="Lonely",
"depblues"="Blues",
"depunfair"="Unfair",
"depmistrt"="Mistreated",
"depbetray"="Betrayed")
methodlabs <- c(
"T1"="Between-Person Difference (T1) Estimator",
"chg"="Within-Person Change (T2-T1) \"Fixed Effects\" Estimator")
stress_varlabsalt <- c(
"Stress:\nMoney" = "Money",
"Stress:\nTransport" = "Transport",
"Stress:\nRespect" = "Respect",
"Stress:\nFair Trtmt" = "Fair Trtmt",
"Stress:\nJob" = "Job",
"Stress:\nTheft Vctm" = "Theft",
"Stress:\nAssault Vctm" = "Assault")
prjplotalt <- ggplot(data = PLMEprjT1,
mapping = aes(x = PLME,
y = stress_varf,
color = method)) +
facet_wrap(~.category, nrow=1, scales="free_x",
labeller = labeller(.category= as_labeller(prjlabs))) +
geom_vline(xintercept = 0, linetype = "dashed", linewidth=.5, alpha=.4) +
stat_pointinterval(data=PLMEprjT1, aes(x=PLME, y=stress_varf,
alpha=p80_gt0),
# color="#E99D53",
.width = .95, size = .7,
position = position_nudge(y = .15)) +
stat_pointinterval(data=PLMEprjchg, aes(x=PLME, y=stress_varf,
alpha=p80_gt0),
# color="#883E3A",
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
scale_color_manual(values=c("#883E3A","#E99D53"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
# coord_cartesian(xlim=c(-.03,.11)) +
# scale_x_continuous(breaks=c(0,.05,.10), labels = dropLeadingZero) +
xlab(element_blank()) +
ylab("Stress Item:\n ") +
# labs(subtitle=~underline("Criminal Intent")) +
labs(subtitle="Criminal Intent") +
scale_y_discrete(labels=stress_varlabsalt) +
facetted_pos_scales( # different x-axis scale for "any crime"
x = list(
.category == "prjany" ~
scale_x_continuous(breaks=c(-.05,0,.05,.10,.15),
limits = c(-.08,.16), labels = dropLeadingZero),
.category != "prjany" ~
scale_x_continuous(breaks=c(0,.05,.10),
limits = c(-.03,.11), labels = dropLeadingZero)
) ) +
theme(axis.title.y = element_text(size=10, face="italic"),
legend.position = "bottom",
strip.background = element_blank(),
strip.text.x = element_text(size=10),
axis.text.y = element_text(size=10),
plot.subtitle=element_text(size=10, hjust=0.5, face="italic"),
legend.text = element_text(size = 10)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)),
color = guide_legend(nrow=1, reverse = TRUE))
depplotalt <- ggplot(data = PLMEdepT1,
mapping = aes(x = PLME,
y = stress_varf,
color = method)) +
facet_wrap(~.category, nrow=1, scales="free_x",
labeller = labeller(.category= as_labeller(deplabs))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLMEdepT1, aes(x=PLME, y=stress_varf,
alpha=p80_gt0),
# color="#E99D53",
.width = .95, size = .7,
position = position_nudge(y = .15)) +
stat_pointinterval(data=PLMEdepchg, aes(x=PLME, y=stress_varf,
alpha=p80_gt0),
# color="#883E3A", .
width = .95, size = .7,
position = position_nudge(y = -.15)) +
scale_color_manual(values=c("#883E3A","#E99D53"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.16,.36)) +
scale_x_continuous(breaks=c(0,.1,.2,.3), labels = dropLeadingZero) +
xlab(element_blank()) +
ylab("Stress Item:\n ") +
# labs(subtitle=~underline("Negative Emotions")) +
labs(subtitle="Negative Emotions") +
scale_y_discrete(labels=stress_varlabsalt) +
theme(axis.title.y = element_text(size=10, face="italic"),
legend.position = "bottom",
strip.background = element_blank(),
strip.text.x = element_text(size=10),
axis.text.y = element_text(size=10),
plot.subtitle=element_text(size=12, hjust=0.5, face="italic"),
legend.text = element_text(size = 10)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)),
color = guide_legend(nrow=1, reverse = TRUE))
design <- "
111
222
333
"
Figure3 <- prjplotalt + depplotalt + guide_area() +
plot_layout(design=design, guides = 'collect', heights = c(20,20,1)) +
plot_annotation(
title = 'FIGURE 3\nMarginal Effects of 2-Category Increase in Stress Item on Outcome Probabilities, by Estimator (T1; Fixed Effects)',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Each of the 196 intervals displayed represents the estimated marginal effect of a "practically large" 2-category increase in stress on an outcome probability derived from 196 distinct Bayesian logistic regression models. Of these, 182 estimates are from multivariate models simultaneously regressing (using `brms:: mvbind()`) the six specific criminal intent outcomes or the seven negative emotions outcomes on each of the seven stress types (T1: 13*7=91 models; multilevel "between/within" or B/W: 13*7=91 models). The other 14 estimates are from seven T1 or seven B/W models separately regressing "any criminal intent" on each stress item. In B/W models, stress items were separated into a L2 cross-time average (Xbar_i) between-person predictor and a L1 within-person change (X_it - Xbar_i) "fixed effects" estimator. In all models, stress predictors (L1 & L2) were specified as monotonic ordinal predictors with a cumulative probit link function. Models were estimated in brms with 4 chains and 4000 total post-warmup posterior draws per outcome. Marginal effect contrast distributions were estimated from the expectation of the posterior predictive distribution for each model as predicted probability difference distributions, either averaged over all 2-category stress differences (T1), or for 2-category stress increases (T1 to T2 change) averaged over all between-person stress levels. Median posterior density estimates with 95% intervals displayed. Bold point-intervals indicate at least 80% of posterior contrast estimates are greater than zero.', width=195)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'bottom',
legend.key.height = unit(.5, 'cm'),
legend.margin = margin(0,0,0,0),
legend.spacing.y = unit(0, "mm"),
plot.title.position = "plot",
plot.caption.position = "plot")
Figure3
ggsave("Figure3.jpeg", width=9, height=6.5, path=here("Output"))Figure highlights (with bold point-intervals) all marginal effect contrasts that have at least an 80% posterior probability of being greater than zero (i.e., at least 80% of the posterior estimates for a contrast are greater than zero). These “plausibly positive” estimates are in line with theoretical predictions.
Comparison of between-person and within-person marginal contrasts in Figure 3 reveals that all 17 (out of 42) of the stress-crime correlations observed as plausibly positive (i.e., 80% posterior probability of being greater than zero) at T1 were estimated as null associations by the within-person fixed effects estimator. In fact, of the 42 possible associations examined, the fixed effects estimator shows only two stress-crime change correlations that were positive with at least 80% plausibility. Specifically, a two-unit response category increase in stress about money from T1 to T2 is associated with a very small increase in the predicted probability of reporting any future intent to engage in theft (less than and greater than 5BAM). The other 40 marginal contrasts plotted in Figure 3 suggest that increases in stress from T1 to T2 are likely unrelated to changes in the probability of criminal intent across waves.
Recall that the fixed effects estimator differences out all time-stable effects of (measured or unmeasured) between-person differences. Overall, these patterns suggest that between-person (T1) estimators of stress-criminal intent associations are likely biased by various sources of unmeasured confounding. For example, crime-involved individuals may have more difficulties in interpersonal and work relationships, and their lifestyles or activity routines might expose them to greater victimization risks. In such cases, spurious between-person estimates of stress “effects” on crime may emerge as artifacts of selection or reporting biases.
With respect to negative emotions, 29 out of 49 marginal effect contrasts at T1 were positive with 80% plausibility, whereas the within-person fixed effects estimator generated 32 (of 49) plausibly positive stress-emotions change correlations. Moreover, in most cases, the fixed effects estimator generated larger marginal effect contrasts. Taken together, these results are consistent with theoretical arguments that stress causes depression and other negative emotions.
Let’s pull all these plotted estimates into tables.
7.6.1 Estimates plotted in Fig3
We will specifically report some of the bold or “plausibly positive” (i.e., >=80% posterior probability of PLME > 0) in the text. To make those easier to find, we can filter to keep only those bold estimates. We also provide all estimates plotted in Fig 3 in separate tables.
7.6.1.1 Bold crim intent (T1, >=80% ppd positive)
PLMEprjT1 %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests) %>%
dplyr::filter(p_gt0 >= .80) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (T1) Unadj. PLME Estimates**"),
subtitle = md("Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper)")
)| Criminal Intent (T1) Unadj. PLME Estimates | ||||
| Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| prjthflt5 | ||||
| Stress: Fair Trtmt | 0.040 | 0.0147 | 0.0105 | 0.070 |
| Stress: Job | 0.056 | 0.0185 | 0.0202 | 0.092 |
| Stress: Respect | 0.029 | 0.0168 | -0.0038 | 0.062 |
| Stress: Theft Vctm | 0.039 | 0.0264 | -0.0067 | 0.097 |
| prjthfgt5 | ||||
| Stress: Fair Trtmt | 0.028 | 0.0148 | -0.0011 | 0.057 |
| Stress: Job | 0.052 | 0.0163 | 0.0203 | 0.084 |
| Stress: Respect | 0.018 | 0.0160 | -0.0146 | 0.049 |
| Stress: Theft Vctm | 0.055 | 0.0282 | 0.0086 | 0.117 |
| prjthreat | ||||
| Stress: Fair Trtmt | 0.026 | 0.0110 | 0.0050 | 0.049 |
| Stress: Job | 0.035 | 0.0125 | 0.0120 | 0.061 |
| Stress: Respect | 0.021 | 0.0120 | -0.0034 | 0.044 |
| Stress: Theft Vctm | 0.019 | 0.0183 | -0.0115 | 0.062 |
| prjusedrg | ||||
| Stress: Assault Vctm | 0.012 | 0.0168 | -0.0112 | 0.055 |
| Stress: Job | 0.017 | 0.0100 | -0.0023 | 0.038 |
| Stress: Respect | 0.010 | 0.0099 | -0.0090 | 0.030 |
| prjhack | ||||
| Stress: Fair Trtmt | 0.010 | 0.0086 | -0.0063 | 0.028 |
| Stress: Job | 0.024 | 0.0087 | 0.0094 | 0.044 |
| prjany | ||||
| Stress: Fair Trtmt | 0.056 | 0.0193 | 0.0182 | 0.095 |
| Stress: Job | 0.084 | 0.0223 | 0.0401 | 0.129 |
| Stress: Respect | 0.051 | 0.0202 | 0.0097 | 0.090 |
| Stress: Theft Vctm | 0.033 | 0.0318 | -0.0205 | 0.103 |
7.6.1.2 Bold neg emotions (T1, >=80% ppd positive)
PLMEdepT1 %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests) %>%
dplyr::filter(p_gt0 >= .80) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (T1) Unadj. PLME Estimates**"),
subtitle = md("Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper)")
)| Negative Emotions (T1) Unadj. PLME Estimates | ||||
| Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| depcantgo | ||||
| Stress: Fair Trtmt | 0.028 | 0.029 | -0.02735 | 0.084 |
| Stress: Money | 0.102 | 0.036 | 0.03052 | 0.169 |
| Stress: Theft Vctm | 0.036 | 0.037 | -0.03363 | 0.111 |
| Stress: Transport | 0.087 | 0.037 | 0.01435 | 0.161 |
| depeffort | ||||
| Stress: Assault Vctm | 0.044 | 0.034 | -0.01057 | 0.126 |
| Stress: Fair Trtmt | 0.036 | 0.020 | -0.00317 | 0.075 |
| Stress: Job | 0.025 | 0.020 | -0.01589 | 0.063 |
| Stress: Money | 0.021 | 0.025 | -0.02957 | 0.069 |
| Stress: Respect | 0.038 | 0.018 | 0.00360 | 0.072 |
| deplonely | ||||
| Stress: Assault Vctm | 0.130 | 0.048 | 0.04687 | 0.232 |
| Stress: Fair Trtmt | 0.066 | 0.024 | 0.01909 | 0.113 |
| Stress: Respect | 0.065 | 0.023 | 0.01881 | 0.110 |
| Stress: Theft Vctm | 0.117 | 0.039 | 0.04421 | 0.197 |
| depblues | ||||
| Stress: Fair Trtmt | 0.025 | 0.022 | -0.02186 | 0.067 |
| Stress: Money | 0.054 | 0.028 | -0.00211 | 0.104 |
| Stress: Transport | 0.064 | 0.029 | 0.00634 | 0.120 |
| depunfair | ||||
| Stress: Money | 0.107 | 0.028 | 0.05388 | 0.162 |
| Stress: Transport | 0.102 | 0.028 | 0.04733 | 0.156 |
| depmistrt | ||||
| Stress: Assault Vctm | 0.159 | 0.040 | 0.08860 | 0.245 |
| Stress: Fair Trtmt | 0.094 | 0.019 | 0.05711 | 0.131 |
| Stress: Money | 0.055 | 0.026 | 0.00212 | 0.103 |
| Stress: Respect | 0.094 | 0.020 | 0.05626 | 0.135 |
| Stress: Theft Vctm | 0.051 | 0.028 | 0.00011 | 0.110 |
| depbetray | ||||
| Stress: Assault Vctm | 0.140 | 0.039 | 0.07243 | 0.225 |
| Stress: Fair Trtmt | 0.090 | 0.019 | 0.05118 | 0.129 |
| Stress: Job | 0.033 | 0.020 | -0.00825 | 0.072 |
| Stress: Money | 0.059 | 0.024 | 0.01175 | 0.106 |
| Stress: Respect | 0.078 | 0.020 | 0.03865 | 0.117 |
| Stress: Theft Vctm | 0.075 | 0.029 | 0.02189 | 0.138 |
7.6.1.3 Bold crim intent (chg, >=80% ppd positive)
PLMEprjchg %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests) %>%
dplyr::filter(p_gt0 >= .80) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (Chg) Unadj. PLME Estimates**"),
subtitle = md("Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper)")
)| Criminal Intent (Chg) Unadj. PLME Estimates | ||||
| Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| prjthflt5 | ||||
| Stress: Money | 0.0028 | 0.0052 | -0.0027 | 0.018 |
| prjthfgt5 | ||||
| Stress: Money | 0.0060 | 0.0072 | -0.0017 | 0.026 |
| prjany | ||||
| Stress: Money | 0.0260 | 0.0240 | -0.0039 | 0.089 |
7.6.1.4 Bold neg emotions (chg, >=80% ppd positive)
PLMEdepchg %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests) %>%
dplyr::filter(p_gt0 >= .80) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (Chg) Unadj. PLME Estimates**"),
subtitle = md("Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper)")
)| Negative Emotions (Chg) Unadj. PLME Estimates | ||||
| Plausibly positive bold only, Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| depcantgo | ||||
| Stress: Assault Vctm | 0.122 | 0.096 | -0.07470 | 0.31 |
| Stress: Fair Trtmt | 0.163 | 0.101 | -0.01468 | 0.39 |
| Stress: Job | 0.168 | 0.087 | -0.00614 | 0.35 |
| Stress: Money | 0.195 | 0.100 | 0.02534 | 0.42 |
| Stress: Respect | 0.085 | 0.089 | -0.08764 | 0.26 |
| Stress: Theft Vctm | 0.146 | 0.090 | -0.02714 | 0.33 |
| Stress: Transport | 0.177 | 0.081 | 0.02942 | 0.35 |
| depeffort | ||||
| Stress: Fair Trtmt | 0.117 | 0.064 | 0.00869 | 0.27 |
| Stress: Job | 0.068 | 0.062 | -0.04879 | 0.20 |
| Stress: Money | 0.067 | 0.064 | -0.05622 | 0.20 |
| Stress: Respect | 0.066 | 0.060 | -0.03916 | 0.20 |
| deplonely | ||||
| Stress: Assault Vctm | 0.171 | 0.098 | -0.00690 | 0.39 |
| Stress: Fair Trtmt | 0.073 | 0.089 | -0.08836 | 0.27 |
| Stress: Money | 0.104 | 0.079 | -0.06594 | 0.25 |
| Stress: Respect | 0.062 | 0.075 | -0.07676 | 0.22 |
| Stress: Theft Vctm | 0.176 | 0.087 | 0.01633 | 0.37 |
| Stress: Transport | 0.106 | 0.083 | -0.04889 | 0.28 |
| depblues | ||||
| Stress: Theft Vctm | 0.083 | 0.068 | -0.02288 | 0.24 |
| depunfair | ||||
| Stress: Fair Trtmt | 0.197 | 0.077 | 0.06834 | 0.37 |
| Stress: Job | 0.161 | 0.073 | 0.03007 | 0.31 |
| Stress: Money | 0.206 | 0.068 | 0.08339 | 0.36 |
| Stress: Respect | 0.194 | 0.070 | 0.07269 | 0.35 |
| Stress: Theft Vctm | 0.186 | 0.081 | 0.04006 | 0.36 |
| Stress: Transport | 0.230 | 0.065 | 0.12253 | 0.37 |
| depmistrt | ||||
| Stress: Fair Trtmt | 0.078 | 0.061 | -0.03941 | 0.20 |
| Stress: Money | 0.080 | 0.069 | -0.04490 | 0.23 |
| Stress: Theft Vctm | 0.084 | 0.075 | -0.07812 | 0.22 |
| depbetray | ||||
| Stress: Assault Vctm | 0.112 | 0.080 | -0.05336 | 0.27 |
| Stress: Fair Trtmt | 0.083 | 0.064 | -0.03771 | 0.22 |
| Stress: Job | 0.106 | 0.057 | 0.00044 | 0.23 |
| Stress: Theft Vctm | 0.150 | 0.066 | 0.03186 | 0.29 |
7.6.1.5 All crim intent (T1)
PLMEprjT1 %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (T1) Unadj. PLME Estimates**"),
subtitle = md("Plotted in FIG3 (FIG4 in paper)")
)| Criminal Intent (T1) Unadj. PLME Estimates | ||||
| Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| prjthflt5 | ||||
| Stress: Assault Vctm | 0.00830 | 0.0252 | -0.0383 | 0.0613 |
| Stress: Fair Trtmt | 0.03996 | 0.0147 | 0.0105 | 0.0696 |
| Stress: Job | 0.05612 | 0.0185 | 0.0202 | 0.0919 |
| Stress: Money | 0.00192 | 0.0223 | -0.0427 | 0.0452 |
| Stress: Respect | 0.02920 | 0.0168 | -0.0038 | 0.0619 |
| Stress: Theft Vctm | 0.03876 | 0.0264 | -0.0067 | 0.0966 |
| Stress: Transport | -0.01207 | 0.0232 | -0.0595 | 0.0311 |
| prjthfgt5 | ||||
| Stress: Assault Vctm | 0.01656 | 0.0240 | -0.0259 | 0.0700 |
| Stress: Fair Trtmt | 0.02795 | 0.0148 | -0.0011 | 0.0572 |
| Stress: Job | 0.05194 | 0.0163 | 0.0203 | 0.0842 |
| Stress: Money | 0.00486 | 0.0206 | -0.0336 | 0.0466 |
| Stress: Respect | 0.01774 | 0.0160 | -0.0146 | 0.0486 |
| Stress: Theft Vctm | 0.05515 | 0.0282 | 0.0086 | 0.1175 |
| Stress: Transport | 0.00951 | 0.0210 | -0.0327 | 0.0484 |
| prjthreat | ||||
| Stress: Assault Vctm | 0.00565 | 0.0184 | -0.0232 | 0.0472 |
| Stress: Fair Trtmt | 0.02637 | 0.0110 | 0.0050 | 0.0486 |
| Stress: Job | 0.03487 | 0.0125 | 0.0120 | 0.0614 |
| Stress: Money | -0.01476 | 0.0173 | -0.0512 | 0.0159 |
| Stress: Respect | 0.02068 | 0.0120 | -0.0034 | 0.0440 |
| Stress: Theft Vctm | 0.01905 | 0.0183 | -0.0115 | 0.0617 |
| Stress: Transport | -0.00483 | 0.0164 | -0.0391 | 0.0258 |
| prjharm | ||||
| Stress: Assault Vctm | -0.00033 | 0.0150 | -0.0233 | 0.0346 |
| Stress: Fair Trtmt | 0.00562 | 0.0107 | -0.0162 | 0.0261 |
| Stress: Job | 0.00924 | 0.0112 | -0.0127 | 0.0318 |
| Stress: Money | -0.01821 | 0.0148 | -0.0499 | 0.0084 |
| Stress: Respect | 0.00590 | 0.0110 | -0.0172 | 0.0270 |
| Stress: Theft Vctm | 0.00595 | 0.0146 | -0.0176 | 0.0397 |
| Stress: Transport | -0.01948 | 0.0162 | -0.0563 | 0.0080 |
| prjusedrg | ||||
| Stress: Assault Vctm | 0.01247 | 0.0168 | -0.0112 | 0.0552 |
| Stress: Fair Trtmt | 0.00644 | 0.0098 | -0.0127 | 0.0259 |
| Stress: Job | 0.01679 | 0.0100 | -0.0023 | 0.0376 |
| Stress: Money | -0.01395 | 0.0145 | -0.0460 | 0.0123 |
| Stress: Respect | 0.01041 | 0.0099 | -0.0090 | 0.0304 |
| Stress: Theft Vctm | -0.00126 | 0.0123 | -0.0205 | 0.0282 |
| Stress: Transport | -0.00489 | 0.0131 | -0.0339 | 0.0188 |
| prjhack | ||||
| Stress: Assault Vctm | 0.00146 | 0.0127 | -0.0162 | 0.0342 |
| Stress: Fair Trtmt | 0.01023 | 0.0086 | -0.0063 | 0.0280 |
| Stress: Job | 0.02423 | 0.0087 | 0.0094 | 0.0437 |
| Stress: Money | -0.00586 | 0.0123 | -0.0327 | 0.0154 |
| Stress: Respect | 0.00493 | 0.0087 | -0.0119 | 0.0230 |
| Stress: Theft Vctm | -0.00092 | 0.0107 | -0.0170 | 0.0256 |
| Stress: Transport | -0.00229 | 0.0122 | -0.0292 | 0.0200 |
| prjany | ||||
| Stress: Assault Vctm | 0.01521 | 0.0316 | -0.0371 | 0.0860 |
| Stress: Fair Trtmt | 0.05650 | 0.0193 | 0.0182 | 0.0949 |
| Stress: Job | 0.08447 | 0.0223 | 0.0401 | 0.1288 |
| Stress: Money | -0.03412 | 0.0310 | -0.0967 | 0.0240 |
| Stress: Respect | 0.05082 | 0.0202 | 0.0097 | 0.0898 |
| Stress: Theft Vctm | 0.03313 | 0.0318 | -0.0205 | 0.1032 |
| Stress: Transport | -0.01737 | 0.0361 | -0.0866 | 0.0525 |
7.6.1.6 All neg emotions (T1)
PLMEdepT1 %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (T1) Unadj. PLME Estimates**"),
subtitle = md("Plotted in FIG3 (FIG4 in paper)")
)| Negative Emotions (T1) Unadj. PLME Estimates | ||||
| Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| depcantgo | ||||
| Stress: Assault Vctm | 0.02967 | 0.041 | -0.05466 | 0.1110 |
| Stress: Fair Trtmt | 0.02752 | 0.029 | -0.02735 | 0.0840 |
| Stress: Job | -0.06216 | 0.034 | -0.12992 | 0.0045 |
| Stress: Money | 0.10163 | 0.036 | 0.03052 | 0.1693 |
| Stress: Respect | 0.01254 | 0.030 | -0.04723 | 0.0735 |
| Stress: Theft Vctm | 0.03552 | 0.037 | -0.03363 | 0.1109 |
| Stress: Transport | 0.08736 | 0.037 | 0.01435 | 0.1614 |
| depeffort | ||||
| Stress: Assault Vctm | 0.04431 | 0.034 | -0.01057 | 0.1260 |
| Stress: Fair Trtmt | 0.03621 | 0.020 | -0.00317 | 0.0752 |
| Stress: Job | 0.02515 | 0.020 | -0.01589 | 0.0630 |
| Stress: Money | 0.02089 | 0.025 | -0.02957 | 0.0690 |
| Stress: Respect | 0.03805 | 0.018 | 0.00360 | 0.0719 |
| Stress: Theft Vctm | 0.01631 | 0.027 | -0.02936 | 0.0745 |
| Stress: Transport | 0.00850 | 0.025 | -0.04171 | 0.0553 |
| deplonely | ||||
| Stress: Assault Vctm | 0.12963 | 0.048 | 0.04687 | 0.2317 |
| Stress: Fair Trtmt | 0.06591 | 0.024 | 0.01909 | 0.1129 |
| Stress: Job | -0.00550 | 0.026 | -0.06013 | 0.0432 |
| Stress: Money | 0.02308 | 0.031 | -0.03896 | 0.0829 |
| Stress: Respect | 0.06471 | 0.023 | 0.01881 | 0.1103 |
| Stress: Theft Vctm | 0.11670 | 0.039 | 0.04421 | 0.1966 |
| Stress: Transport | -0.00078 | 0.032 | -0.06678 | 0.0616 |
| depblues | ||||
| Stress: Assault Vctm | 0.01265 | 0.037 | -0.04418 | 0.1002 |
| Stress: Fair Trtmt | 0.02517 | 0.022 | -0.02186 | 0.0667 |
| Stress: Job | 0.01215 | 0.023 | -0.03392 | 0.0565 |
| Stress: Money | 0.05365 | 0.028 | -0.00211 | 0.1045 |
| Stress: Respect | 0.01128 | 0.025 | -0.03722 | 0.0569 |
| Stress: Theft Vctm | 0.01357 | 0.034 | -0.04296 | 0.0900 |
| Stress: Transport | 0.06421 | 0.029 | 0.00634 | 0.1205 |
| depunfair | ||||
| Stress: Assault Vctm | 0.02649 | 0.040 | -0.03454 | 0.1189 |
| Stress: Fair Trtmt | 0.01702 | 0.023 | -0.03090 | 0.0613 |
| Stress: Job | 0.01543 | 0.024 | -0.03234 | 0.0592 |
| Stress: Money | 0.10670 | 0.028 | 0.05388 | 0.1617 |
| Stress: Respect | 0.01403 | 0.027 | -0.04097 | 0.0624 |
| Stress: Theft Vctm | 0.01216 | 0.033 | -0.04261 | 0.0869 |
| Stress: Transport | 0.10163 | 0.028 | 0.04733 | 0.1562 |
| depmistrt | ||||
| Stress: Assault Vctm | 0.15927 | 0.040 | 0.08860 | 0.2451 |
| Stress: Fair Trtmt | 0.09353 | 0.019 | 0.05711 | 0.1310 |
| Stress: Job | 0.00953 | 0.022 | -0.03649 | 0.0488 |
| Stress: Money | 0.05495 | 0.026 | 0.00212 | 0.1033 |
| Stress: Respect | 0.09414 | 0.020 | 0.05626 | 0.1348 |
| Stress: Theft Vctm | 0.05145 | 0.028 | 0.00011 | 0.1096 |
| Stress: Transport | 0.00332 | 0.029 | -0.05603 | 0.0580 |
| depbetray | ||||
| Stress: Assault Vctm | 0.13990 | 0.039 | 0.07243 | 0.2254 |
| Stress: Fair Trtmt | 0.08968 | 0.019 | 0.05118 | 0.1285 |
| Stress: Job | 0.03302 | 0.020 | -0.00825 | 0.0715 |
| Stress: Money | 0.05943 | 0.024 | 0.01175 | 0.1059 |
| Stress: Respect | 0.07793 | 0.020 | 0.03865 | 0.1167 |
| Stress: Theft Vctm | 0.07543 | 0.029 | 0.02189 | 0.1383 |
| Stress: Transport | 0.01202 | 0.027 | -0.04361 | 0.0656 |
7.6.1.7 All crim intent (chg)
PLMEprjchg %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (Chg) Unadj. PLME Estimates**"),
subtitle = md("Plotted in FIG3 (FIG4 in paper)")
)| Criminal Intent (Chg) Unadj. PLME Estimates | ||||
| Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| prjthflt5 | ||||
| Stress: Assault Vctm | -0.000672 | 0.0050 | -0.0122 | 0.007312 |
| Stress: Fair Trtmt | 0.000649 | 0.0053 | -0.0090 | 0.012876 |
| Stress: Job | -0.000460 | 0.0050 | -0.0112 | 0.009539 |
| Stress: Money | 0.002795 | 0.0052 | -0.0027 | 0.017624 |
| Stress: Respect | -0.003280 | 0.0059 | -0.0199 | 0.002158 |
| Stress: Theft Vctm | -0.001878 | 0.0066 | -0.0190 | 0.006708 |
| Stress: Transport | 0.000086 | 0.0044 | -0.0079 | 0.009795 |
| prjthfgt5 | ||||
| Stress: Assault Vctm | 0.000745 | 0.0070 | -0.0126 | 0.016305 |
| Stress: Fair Trtmt | 0.002285 | 0.0065 | -0.0082 | 0.018537 |
| Stress: Job | 0.001492 | 0.0072 | -0.0107 | 0.018537 |
| Stress: Money | 0.006011 | 0.0072 | -0.0017 | 0.026363 |
| Stress: Respect | -0.003715 | 0.0072 | -0.0242 | 0.003775 |
| Stress: Theft Vctm | -0.001825 | 0.0116 | -0.0286 | 0.018030 |
| Stress: Transport | 0.000462 | 0.0062 | -0.0110 | 0.014733 |
| prjthreat | ||||
| Stress: Assault Vctm | -0.001266 | 0.0038 | -0.0112 | 0.001903 |
| Stress: Fair Trtmt | -0.001889 | 0.0036 | -0.0122 | 0.000985 |
| Stress: Job | -0.000499 | 0.0037 | -0.0088 | 0.005971 |
| Stress: Money | -0.000375 | 0.0026 | -0.0065 | 0.003946 |
| Stress: Respect | -0.002778 | 0.0048 | -0.0162 | 0.000543 |
| Stress: Theft Vctm | -0.002601 | 0.0059 | -0.0209 | 0.001632 |
| Stress: Transport | -0.001257 | 0.0032 | -0.0093 | 0.001331 |
| prjharm | ||||
| Stress: Assault Vctm | -0.000661 | 0.0044 | -0.0109 | 0.006305 |
| Stress: Fair Trtmt | -0.002054 | 0.0042 | -0.0141 | 0.001528 |
| Stress: Job | -0.000764 | 0.0041 | -0.0103 | 0.005779 |
| Stress: Money | -0.000570 | 0.0038 | -0.0098 | 0.005814 |
| Stress: Respect | -0.002447 | 0.0045 | -0.0149 | 0.001728 |
| Stress: Theft Vctm | -0.004166 | 0.0072 | -0.0258 | 0.000117 |
| Stress: Transport | -0.003450 | 0.0057 | -0.0199 | 0.000430 |
| prjusedrg | ||||
| Stress: Assault Vctm | 0.000177 | 0.0043 | -0.0071 | 0.008470 |
| Stress: Fair Trtmt | -0.001541 | 0.0035 | -0.0111 | 0.001933 |
| Stress: Job | -0.000501 | 0.0041 | -0.0099 | 0.006461 |
| Stress: Money | -0.000867 | 0.0029 | -0.0083 | 0.002727 |
| Stress: Respect | -0.002950 | 0.0041 | -0.0150 | 0.000015 |
| Stress: Theft Vctm | -0.000695 | 0.0030 | -0.0082 | 0.003177 |
| Stress: Transport | -0.001557 | 0.0036 | -0.0111 | 0.001376 |
| prjhack | ||||
| Stress: Assault Vctm | -0.000819 | 0.0187 | -0.0375 | 0.037038 |
| Stress: Fair Trtmt | -0.012488 | 0.0157 | -0.0534 | 0.009395 |
| Stress: Job | -0.000017 | 0.0145 | -0.0299 | 0.030009 |
| Stress: Money | 0.003731 | 0.0154 | -0.0247 | 0.038322 |
| Stress: Respect | -0.014460 | 0.0186 | -0.0646 | 0.008575 |
| Stress: Theft Vctm | -0.011463 | 0.0203 | -0.0666 | 0.015074 |
| Stress: Transport | -0.014022 | 0.0187 | -0.0663 | 0.008119 |
| prjany | ||||
| Stress: Assault Vctm | 0.000793 | 0.0216 | -0.0421 | 0.045480 |
| Stress: Fair Trtmt | 0.005569 | 0.0205 | -0.0323 | 0.051726 |
| Stress: Job | 0.010845 | 0.0219 | -0.0291 | 0.060448 |
| Stress: Money | 0.025973 | 0.0240 | -0.0039 | 0.088965 |
| Stress: Respect | -0.018389 | 0.0217 | -0.0742 | 0.012077 |
| Stress: Theft Vctm | -0.014331 | 0.0271 | -0.0816 | 0.028117 |
| Stress: Transport | 0.006167 | 0.0204 | -0.0292 | 0.051096 |
7.6.1.8 All neg emotions (chg)
PLMEdepchg %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (Chg) Unadj. PLME Estimates**"),
subtitle = md("Plotted in FIG3 (FIG4 in paper)")
)| Negative Emotions (Chg) Unadj. PLME Estimates | ||||
| Plotted in FIG3 (FIG4 in paper) | ||||
| stress_var | m | s | ll | ul |
|---|---|---|---|---|
| depcantgo | ||||
| Stress: Assault Vctm | 0.1224 | 0.096 | -0.07470 | 0.306 |
| Stress: Fair Trtmt | 0.1628 | 0.101 | -0.01468 | 0.387 |
| Stress: Job | 0.1682 | 0.087 | -0.00614 | 0.346 |
| Stress: Money | 0.1945 | 0.100 | 0.02534 | 0.416 |
| Stress: Respect | 0.0848 | 0.089 | -0.08764 | 0.263 |
| Stress: Theft Vctm | 0.1463 | 0.090 | -0.02714 | 0.327 |
| Stress: Transport | 0.1771 | 0.081 | 0.02942 | 0.349 |
| depeffort | ||||
| Stress: Assault Vctm | 0.0300 | 0.072 | -0.11694 | 0.168 |
| Stress: Fair Trtmt | 0.1173 | 0.064 | 0.00869 | 0.272 |
| Stress: Job | 0.0678 | 0.062 | -0.04879 | 0.200 |
| Stress: Money | 0.0667 | 0.064 | -0.05622 | 0.199 |
| Stress: Respect | 0.0664 | 0.060 | -0.03916 | 0.199 |
| Stress: Theft Vctm | 0.0497 | 0.067 | -0.07495 | 0.197 |
| Stress: Transport | 0.0164 | 0.065 | -0.11370 | 0.150 |
| deplonely | ||||
| Stress: Assault Vctm | 0.1710 | 0.098 | -0.00690 | 0.389 |
| Stress: Fair Trtmt | 0.0729 | 0.089 | -0.08836 | 0.275 |
| Stress: Job | 0.0533 | 0.079 | -0.09742 | 0.211 |
| Stress: Money | 0.1040 | 0.079 | -0.06594 | 0.246 |
| Stress: Respect | 0.0620 | 0.075 | -0.07676 | 0.224 |
| Stress: Theft Vctm | 0.1765 | 0.087 | 0.01633 | 0.367 |
| Stress: Transport | 0.1056 | 0.083 | -0.04889 | 0.280 |
| depblues | ||||
| Stress: Assault Vctm | -0.0434 | 0.062 | -0.17499 | 0.071 |
| Stress: Fair Trtmt | 0.0145 | 0.056 | -0.10285 | 0.122 |
| Stress: Job | -0.0320 | 0.051 | -0.13296 | 0.070 |
| Stress: Money | -0.0500 | 0.061 | -0.19278 | 0.051 |
| Stress: Respect | -0.0293 | 0.049 | -0.13479 | 0.064 |
| Stress: Theft Vctm | 0.0826 | 0.068 | -0.02288 | 0.245 |
| Stress: Transport | -0.0500 | 0.056 | -0.16196 | 0.061 |
| depunfair | ||||
| Stress: Assault Vctm | 0.0263 | 0.095 | -0.15321 | 0.222 |
| Stress: Fair Trtmt | 0.1971 | 0.077 | 0.06834 | 0.374 |
| Stress: Job | 0.1611 | 0.073 | 0.03007 | 0.314 |
| Stress: Money | 0.2060 | 0.068 | 0.08339 | 0.355 |
| Stress: Respect | 0.1936 | 0.070 | 0.07269 | 0.349 |
| Stress: Theft Vctm | 0.1856 | 0.081 | 0.04006 | 0.362 |
| Stress: Transport | 0.2299 | 0.065 | 0.12253 | 0.372 |
| depmistrt | ||||
| Stress: Assault Vctm | -0.0133 | 0.086 | -0.19494 | 0.151 |
| Stress: Fair Trtmt | 0.0777 | 0.061 | -0.03941 | 0.202 |
| Stress: Job | 0.0091 | 0.077 | -0.18283 | 0.132 |
| Stress: Money | 0.0798 | 0.069 | -0.04490 | 0.231 |
| Stress: Respect | 0.0426 | 0.063 | -0.07361 | 0.174 |
| Stress: Theft Vctm | 0.0839 | 0.075 | -0.07812 | 0.219 |
| Stress: Transport | 0.0370 | 0.077 | -0.14106 | 0.170 |
| depbetray | ||||
| Stress: Assault Vctm | 0.1122 | 0.080 | -0.05336 | 0.266 |
| Stress: Fair Trtmt | 0.0826 | 0.064 | -0.03771 | 0.217 |
| Stress: Job | 0.1063 | 0.057 | 0.00044 | 0.227 |
| Stress: Money | 0.0228 | 0.062 | -0.09990 | 0.151 |
| Stress: Respect | 0.0402 | 0.059 | -0.06941 | 0.168 |
| Stress: Theft Vctm | 0.1499 | 0.066 | 0.03186 | 0.292 |
| Stress: Transport | -0.0071 | 0.066 | -0.13997 | 0.123 |
7.6.2 Median unweighted (meta-analytic avg) estimates from Fig 4
We can also summarize the median unweighted PLME estimates and 80% posterior interval ranges for these model estimates.
7.6.2.1 Crim intent (T1)
PLMEprjT1 %>%
ungroup() %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .10),
ul = quantile(PLME, probs = .90)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (T1) Median Unadj. PLME Estimate**")
)| Criminal Intent (T1) Median Unadj. PLME Estimate | |||
| m | s | ll | ul |
|---|---|---|---|
| 0.011 | 0.03 | -0.019 | 0.053 |
7.6.2.2 Neg emotions (T1)
PLMEdepT1 %>%
# summarize PLME estimates across the MCMC draws
ungroup() %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .10),
ul = quantile(PLME, probs = .90)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (T1) Median Unadj. PLME Estimate**"),
)| Negative Emotions (T1) Median Unadj. PLME Estimate | |||
| m | s | ll | ul |
|---|---|---|---|
| 0.04 | 0.052 | -0.013 | 0.11 |
7.6.2.3 Crim intent (chg)
PLMEprjchg %>%
ungroup() %>%
# summarize PLME estimates across the MCMC draws
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .10),
ul = quantile(PLME, probs = .90)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (Chg) Median Unadj. PLME Estimate**"),
)| Criminal Intent (Chg) Median Unadj. PLME Estimate | |||
| m | s | ll | ul |
|---|---|---|---|
| -0.0011 | 0.014 | -0.014 | 0.0088 |
7.6.2.4 Neg emotions (chg)
PLMEdepchg %>%
ungroup() %>%
# summarize PLME estimates across the MCMC draws
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .10),
ul = quantile(PLME, probs = .90)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (Chg) Median Unadj. PLME Estimate**"),
)| Negative Emotions (Chg) Median Unadj. PLME Estimate | |||
| m | s | ll | ul |
|---|---|---|---|
| 0.08 | 0.11 | -0.046 | 0.23 |
Let’s now turn to examining community variations.
8 Modeling change correlations by community
(RMD FILE: BDK_2023_Stress_7_Chgcorr_comm_mods)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE8.1 RQ3. Stress/Past Crime (T2-T1) x Community Bivariate Correlations
RQ3 (Stress amplification): Are within-person change (fixed effects) correlations between subjective stress and posited outcomes - self-reported criminal intent and negative emotions - positive and strongest in low-SES urban communities?
Recall that the overall (full sample) change associations displayed in Figure 3 may mask important and theoretically expected community variations in the stress-crime relationship. In particular, stress process theories imply that such associations should be strongest where stress is most chronic - that is, in urban areas and particularly low-SES urban communities.
So, let’s model these potential community variations in item pair associations and then see if we can add them to the plot without overloading it.
8.1.1 Criminal Intent & Stress x Community Correlation Models
8.1.1.1 Corr X Community: stmony & prj crime
load(here("1_Data_Files/Datasets/stress_long.Rdata"))#Community Change: criminal intent items ~ mo(stmony)
#Vectorize priors:
#list of colnames for projected crime DVs
prjdv_names <- noquote(c("prjthflt5", "prjthfgt5", "prjthreat", "prjharm",
"prjusedrg", "prjhack"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmony_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmony_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmony_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmony_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.stmony.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stmony_devx2) + mo(stmony_av12x2) +
rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stmony_comm_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="prjthflt5")
ppcheckdv2 <-pp_check(modelfit, resp="prjthfgt5")
ppcheckdv3 <-pp_check(modelfit, resp="prjthreat")
ppcheckdv4 <-pp_check(modelfit, resp="prjharm")
ppcheckdv5 <-pp_check(modelfit, resp="prjusedrg")
ppcheckdv6 <-pp_check(modelfit, resp="prjhack")
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6,
plotcoefs, plotcoefs2)
return(allchecks)
}
out.chg.prjcrime.stmony.comm.fit <- ppchecks(chg.prjcrime.stmony.comm.fit)8.1.1.1.1 Coefficient plot (intervals)
out.chg.prjcrime.stmony.comm.fit[[10]]
8.1.1.1.2 PPcheck (density)
p1 <- out.chg.prjcrime.stmony.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stmony.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stmony.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stmony.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stmony.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stmony.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
8.1.1.1.3 Fit summary
out.chg.prjcrime.stmony.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## prjthreat ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## prjharm ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## prjusedrg ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## prjhack ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 4.15 0.60 3.10 5.43 1.00 1430
## sd(prjthfgt5_Intercept) 3.50 0.51 2.58 4.57 1.00 1225
## sd(prjthreat_Intercept) 3.27 0.57 2.21 4.48 1.00 1676
## sd(prjharm_Intercept) 3.02 0.56 2.04 4.21 1.00 1747
## sd(prjusedrg_Intercept) 2.94 0.54 1.96 4.10 1.00 1760
## sd(prjhack_Intercept) 0.87 0.56 0.05 2.04 1.00 847
## Tail_ESS
## sd(prjthflt5_Intercept) 2572
## sd(prjthfgt5_Intercept) 1779
## sd(prjthreat_Intercept) 2830
## sd(prjharm_Intercept) 2564
## sd(prjusedrg_Intercept) 2593
## sd(prjhack_Intercept) 1710
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_Intercept -6.15 0.91 -8.04 -4.38
## prjthfgt5_Intercept -5.86 0.85 -7.60 -4.25
## prjthreat_Intercept -6.09 0.98 -8.11 -4.26
## prjharm_Intercept -5.72 0.97 -7.69 -3.93
## prjusedrg_Intercept -5.80 0.96 -7.74 -3.99
## prjhack_Intercept -4.31 0.78 -5.91 -2.87
## prjthflt5_rural.ses.med2 -0.23 0.87 -1.89 1.49
## prjthflt5_rural.ses.med3 0.64 0.82 -1.02 2.17
## prjthflt5_rural.ses.med4 1.03 0.88 -0.75 2.66
## prjthfgt5_rural.ses.med2 -0.40 0.88 -2.13 1.38
## prjthfgt5_rural.ses.med3 0.61 0.79 -1.01 2.07
## prjthfgt5_rural.ses.med4 0.80 0.85 -0.87 2.40
## prjthreat_rural.ses.med2 -0.25 0.90 -2.04 1.56
## prjthreat_rural.ses.med3 0.52 0.83 -1.10 2.12
## prjthreat_rural.ses.med4 1.21 0.88 -0.65 2.82
## prjharm_rural.ses.med2 -0.10 0.81 -1.63 1.54
## prjharm_rural.ses.med3 0.39 0.80 -1.26 1.96
## prjharm_rural.ses.med4 0.73 0.77 -0.84 2.21
## prjusedrg_rural.ses.med2 -0.39 0.87 -2.08 1.39
## prjusedrg_rural.ses.med3 -0.12 0.87 -1.84 1.59
## prjusedrg_rural.ses.med4 1.34 0.85 -0.44 2.87
## prjhack_rural.ses.med2 -0.50 0.88 -2.20 1.23
## prjhack_rural.ses.med3 0.33 0.79 -1.31 1.84
## prjhack_rural.ses.med4 0.57 0.76 -1.01 1.99
## prjthflt5_mostmony_devx2 0.01 0.22 -0.43 0.41
## prjthflt5_mostmony_av12x2 -0.05 0.09 -0.24 0.13
## prjthflt5_mostmony_devx2:rural.ses.med2 -0.95 0.59 -2.21 0.10
## prjthflt5_mostmony_devx2:rural.ses.med3 0.59 0.38 -0.13 1.40
## prjthflt5_mostmony_devx2:rural.ses.med4 0.78 0.40 0.07 1.67
## prjthfgt5_mostmony_devx2 0.06 0.21 -0.38 0.44
## prjthfgt5_mostmony_av12x2 -0.03 0.09 -0.21 0.14
## prjthfgt5_mostmony_devx2:rural.ses.med2 -0.85 0.60 -2.13 0.23
## prjthfgt5_mostmony_devx2:rural.ses.med3 0.52 0.36 -0.19 1.25
## prjthfgt5_mostmony_devx2:rural.ses.med4 0.75 0.38 0.03 1.57
## prjthreat_mostmony_devx2 -0.11 0.21 -0.52 0.28
## prjthreat_mostmony_av12x2 -0.10 0.09 -0.29 0.09
## prjthreat_mostmony_devx2:rural.ses.med2 -0.72 0.63 -2.07 0.42
## prjthreat_mostmony_devx2:rural.ses.med3 0.09 0.47 -0.89 0.97
## prjthreat_mostmony_devx2:rural.ses.med4 0.37 0.49 -0.59 1.39
## prjharm_mostmony_devx2 -0.09 0.21 -0.50 0.32
## prjharm_mostmony_av12x2 -0.12 0.09 -0.30 0.06
## prjharm_mostmony_devx2:rural.ses.med2 -0.33 0.50 -1.40 0.61
## prjharm_mostmony_devx2:rural.ses.med3 -0.06 0.51 -1.17 0.88
## prjharm_mostmony_devx2:rural.ses.med4 0.24 0.43 -0.62 1.10
## prjusedrg_mostmony_devx2 -0.16 0.21 -0.56 0.25
## prjusedrg_mostmony_av12x2 -0.05 0.09 -0.24 0.12
## prjusedrg_mostmony_devx2:rural.ses.med2 -0.22 0.56 -1.39 0.81
## prjusedrg_mostmony_devx2:rural.ses.med3 -0.37 0.55 -1.51 0.63
## prjusedrg_mostmony_devx2:rural.ses.med4 0.26 0.47 -0.63 1.22
## prjhack_mostmony_devx2 0.04 0.21 -0.36 0.45
## prjhack_mostmony_av12x2 -0.05 0.08 -0.20 0.11
## prjhack_mostmony_devx2:rural.ses.med2 -0.39 0.56 -1.54 0.67
## prjhack_mostmony_devx2:rural.ses.med3 -0.19 0.56 -1.44 0.81
## prjhack_mostmony_devx2:rural.ses.med4 0.20 0.38 -0.59 0.94
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1.00 2819 2903
## prjthfgt5_Intercept 1.00 2419 2950
## prjthreat_Intercept 1.00 2543 2994
## prjharm_Intercept 1.00 2349 3037
## prjusedrg_Intercept 1.00 3041 3268
## prjhack_Intercept 1.00 2012 2940
## prjthflt5_rural.ses.med2 1.00 5780 3353
## prjthflt5_rural.ses.med3 1.00 5107 3428
## prjthflt5_rural.ses.med4 1.00 4430 2845
## prjthfgt5_rural.ses.med2 1.00 5804 3213
## prjthfgt5_rural.ses.med3 1.00 4017 2968
## prjthfgt5_rural.ses.med4 1.00 4606 3304
## prjthreat_rural.ses.med2 1.00 5347 2731
## prjthreat_rural.ses.med3 1.00 6110 3617
## prjthreat_rural.ses.med4 1.00 4274 3103
## prjharm_rural.ses.med2 1.00 5393 3192
## prjharm_rural.ses.med3 1.00 5272 3625
## prjharm_rural.ses.med4 1.00 5002 3353
## prjusedrg_rural.ses.med2 1.00 5771 3296
## prjusedrg_rural.ses.med3 1.00 5105 2946
## prjusedrg_rural.ses.med4 1.00 4400 3613
## prjhack_rural.ses.med2 1.00 5272 3184
## prjhack_rural.ses.med3 1.00 4350 2981
## prjhack_rural.ses.med4 1.00 4338 3057
## prjthflt5_mostmony_devx2 1.00 5188 3418
## prjthflt5_mostmony_av12x2 1.00 3282 3145
## prjthflt5_mostmony_devx2:rural.ses.med2 1.00 4480 3255
## prjthflt5_mostmony_devx2:rural.ses.med3 1.00 4073 3158
## prjthflt5_mostmony_devx2:rural.ses.med4 1.00 3405 2824
## prjthfgt5_mostmony_devx2 1.00 3792 3195
## prjthfgt5_mostmony_av12x2 1.00 3720 3380
## prjthfgt5_mostmony_devx2:rural.ses.med2 1.00 4422 2665
## prjthfgt5_mostmony_devx2:rural.ses.med3 1.00 3210 2714
## prjthfgt5_mostmony_devx2:rural.ses.med4 1.00 3216 3029
## prjthreat_mostmony_devx2 1.00 5205 3136
## prjthreat_mostmony_av12x2 1.00 4003 3326
## prjthreat_mostmony_devx2:rural.ses.med2 1.00 4136 2684
## prjthreat_mostmony_devx2:rural.ses.med3 1.00 3381 2773
## prjthreat_mostmony_devx2:rural.ses.med4 1.00 2708 3156
## prjharm_mostmony_devx2 1.00 5320 3161
## prjharm_mostmony_av12x2 1.00 4964 3388
## prjharm_mostmony_devx2:rural.ses.med2 1.00 4288 3289
## prjharm_mostmony_devx2:rural.ses.med3 1.00 4082 3111
## prjharm_mostmony_devx2:rural.ses.med4 1.00 4031 2961
## prjusedrg_mostmony_devx2 1.00 5988 3160
## prjusedrg_mostmony_av12x2 1.00 3939 2827
## prjusedrg_mostmony_devx2:rural.ses.med2 1.00 4180 3204
## prjusedrg_mostmony_devx2:rural.ses.med3 1.00 3979 2888
## prjusedrg_mostmony_devx2:rural.ses.med4 1.00 3061 3140
## prjhack_mostmony_devx2 1.00 5650 3223
## prjhack_mostmony_av12x2 1.00 6767 3411
## prjhack_mostmony_devx2:rural.ses.med2 1.00 4256 3423
## prjhack_mostmony_devx2:rural.ses.med3 1.00 3623 3578
## prjhack_mostmony_devx2:rural.ses.med4 1.00 3593 2671
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## prjthflt5_mostmony_devx21[1] 0.25 0.15 0.03
## prjthflt5_mostmony_devx21[2] 0.24 0.14 0.03
## prjthflt5_mostmony_devx21[3] 0.24 0.14 0.03
## prjthflt5_mostmony_devx21[4] 0.26 0.15 0.04
## prjthflt5_mostmony_av12x21[1] 0.13 0.08 0.02
## prjthflt5_mostmony_av12x21[2] 0.13 0.08 0.02
## prjthflt5_mostmony_av12x21[3] 0.13 0.08 0.02
## prjthflt5_mostmony_av12x21[4] 0.13 0.08 0.01
## prjthflt5_mostmony_av12x21[5] 0.12 0.08 0.02
## prjthflt5_mostmony_av12x21[6] 0.12 0.08 0.02
## prjthflt5_mostmony_av12x21[7] 0.12 0.08 0.02
## prjthflt5_mostmony_av12x21[8] 0.12 0.08 0.02
## prjthflt5_mostmony_devx2:rural.ses.med21[1] 0.27 0.18 0.01
## prjthflt5_mostmony_devx2:rural.ses.med21[2] 0.21 0.17 0.01
## prjthflt5_mostmony_devx2:rural.ses.med21[3] 0.29 0.20 0.02
## prjthflt5_mostmony_devx2:rural.ses.med21[4] 0.23 0.18 0.01
## prjthflt5_mostmony_devx2:rural.ses.med31[1] 0.25 0.18 0.01
## prjthflt5_mostmony_devx2:rural.ses.med31[2] 0.31 0.20 0.02
## prjthflt5_mostmony_devx2:rural.ses.med31[3] 0.20 0.16 0.01
## prjthflt5_mostmony_devx2:rural.ses.med31[4] 0.24 0.18 0.01
## prjthflt5_mostmony_devx2:rural.ses.med41[1] 0.33 0.20 0.02
## prjthflt5_mostmony_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjthflt5_mostmony_devx2:rural.ses.med41[3] 0.21 0.16 0.01
## prjthflt5_mostmony_devx2:rural.ses.med41[4] 0.23 0.17 0.01
## prjthfgt5_mostmony_devx21[1] 0.25 0.14 0.04
## prjthfgt5_mostmony_devx21[2] 0.25 0.15 0.04
## prjthfgt5_mostmony_devx21[3] 0.24 0.14 0.03
## prjthfgt5_mostmony_devx21[4] 0.25 0.15 0.04
## prjthfgt5_mostmony_av12x21[1] 0.13 0.08 0.02
## prjthfgt5_mostmony_av12x21[2] 0.13 0.08 0.02
## prjthfgt5_mostmony_av12x21[3] 0.12 0.08 0.02
## prjthfgt5_mostmony_av12x21[4] 0.12 0.08 0.02
## prjthfgt5_mostmony_av12x21[5] 0.12 0.08 0.02
## prjthfgt5_mostmony_av12x21[6] 0.12 0.08 0.02
## prjthfgt5_mostmony_av12x21[7] 0.13 0.08 0.02
## prjthfgt5_mostmony_av12x21[8] 0.13 0.08 0.02
## prjthfgt5_mostmony_devx2:rural.ses.med21[1] 0.30 0.20 0.02
## prjthfgt5_mostmony_devx2:rural.ses.med21[2] 0.19 0.16 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med21[3] 0.27 0.19 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med21[4] 0.24 0.18 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med31[2] 0.32 0.21 0.02
## prjthfgt5_mostmony_devx2:rural.ses.med31[3] 0.19 0.16 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med31[4] 0.24 0.19 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med41[1] 0.29 0.19 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med41[2] 0.31 0.20 0.02
## prjthfgt5_mostmony_devx2:rural.ses.med41[3] 0.17 0.14 0.01
## prjthfgt5_mostmony_devx2:rural.ses.med41[4] 0.23 0.17 0.01
## prjthreat_mostmony_devx21[1] 0.26 0.15 0.04
## prjthreat_mostmony_devx21[2] 0.24 0.15 0.03
## prjthreat_mostmony_devx21[3] 0.26 0.15 0.04
## prjthreat_mostmony_devx21[4] 0.24 0.15 0.03
## prjthreat_mostmony_av12x21[1] 0.13 0.08 0.02
## prjthreat_mostmony_av12x21[2] 0.12 0.08 0.01
## prjthreat_mostmony_av12x21[3] 0.13 0.08 0.02
## prjthreat_mostmony_av12x21[4] 0.13 0.08 0.02
## prjthreat_mostmony_av12x21[5] 0.14 0.09 0.02
## prjthreat_mostmony_av12x21[6] 0.13 0.08 0.01
## prjthreat_mostmony_av12x21[7] 0.12 0.08 0.02
## prjthreat_mostmony_av12x21[8] 0.12 0.08 0.01
## prjthreat_mostmony_devx2:rural.ses.med21[1] 0.28 0.20 0.01
## prjthreat_mostmony_devx2:rural.ses.med21[2] 0.20 0.17 0.01
## prjthreat_mostmony_devx2:rural.ses.med21[3] 0.26 0.20 0.01
## prjthreat_mostmony_devx2:rural.ses.med21[4] 0.25 0.19 0.01
## prjthreat_mostmony_devx2:rural.ses.med31[1] 0.24 0.19 0.01
## prjthreat_mostmony_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## prjthreat_mostmony_devx2:rural.ses.med31[3] 0.23 0.18 0.01
## prjthreat_mostmony_devx2:rural.ses.med31[4] 0.30 0.21 0.01
## prjthreat_mostmony_devx2:rural.ses.med41[1] 0.30 0.21 0.01
## prjthreat_mostmony_devx2:rural.ses.med41[2] 0.24 0.18 0.01
## prjthreat_mostmony_devx2:rural.ses.med41[3] 0.18 0.17 0.01
## prjthreat_mostmony_devx2:rural.ses.med41[4] 0.28 0.20 0.01
## prjharm_mostmony_devx21[1] 0.26 0.15 0.04
## prjharm_mostmony_devx21[2] 0.24 0.14 0.03
## prjharm_mostmony_devx21[3] 0.24 0.15 0.03
## prjharm_mostmony_devx21[4] 0.25 0.15 0.03
## prjharm_mostmony_av12x21[1] 0.12 0.08 0.01
## prjharm_mostmony_av12x21[2] 0.12 0.08 0.02
## prjharm_mostmony_av12x21[3] 0.13 0.08 0.02
## prjharm_mostmony_av12x21[4] 0.13 0.08 0.02
## prjharm_mostmony_av12x21[5] 0.13 0.08 0.02
## prjharm_mostmony_av12x21[6] 0.14 0.09 0.02
## prjharm_mostmony_av12x21[7] 0.12 0.08 0.02
## prjharm_mostmony_av12x21[8] 0.12 0.08 0.02
## prjharm_mostmony_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## prjharm_mostmony_devx2:rural.ses.med21[2] 0.23 0.18 0.01
## prjharm_mostmony_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## prjharm_mostmony_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## prjharm_mostmony_devx2:rural.ses.med31[1] 0.23 0.18 0.01
## prjharm_mostmony_devx2:rural.ses.med31[2] 0.22 0.19 0.01
## prjharm_mostmony_devx2:rural.ses.med31[3] 0.25 0.19 0.01
## prjharm_mostmony_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjharm_mostmony_devx2:rural.ses.med41[1] 0.26 0.20 0.01
## prjharm_mostmony_devx2:rural.ses.med41[2] 0.23 0.19 0.01
## prjharm_mostmony_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## prjharm_mostmony_devx2:rural.ses.med41[4] 0.28 0.21 0.01
## prjusedrg_mostmony_devx21[1] 0.27 0.15 0.04
## prjusedrg_mostmony_devx21[2] 0.25 0.15 0.03
## prjusedrg_mostmony_devx21[3] 0.23 0.14 0.03
## prjusedrg_mostmony_devx21[4] 0.25 0.15 0.03
## prjusedrg_mostmony_av12x21[1] 0.13 0.08 0.02
## prjusedrg_mostmony_av12x21[2] 0.12 0.08 0.01
## prjusedrg_mostmony_av12x21[3] 0.13 0.08 0.02
## prjusedrg_mostmony_av12x21[4] 0.13 0.08 0.02
## prjusedrg_mostmony_av12x21[5] 0.12 0.08 0.02
## prjusedrg_mostmony_av12x21[6] 0.12 0.08 0.01
## prjusedrg_mostmony_av12x21[7] 0.12 0.08 0.02
## prjusedrg_mostmony_av12x21[8] 0.12 0.08 0.02
## prjusedrg_mostmony_devx2:rural.ses.med21[1] 0.27 0.20 0.01
## prjusedrg_mostmony_devx2:rural.ses.med21[2] 0.21 0.18 0.01
## prjusedrg_mostmony_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## prjusedrg_mostmony_devx2:rural.ses.med21[4] 0.29 0.20 0.01
## prjusedrg_mostmony_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## prjusedrg_mostmony_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjusedrg_mostmony_devx2:rural.ses.med31[3] 0.23 0.18 0.01
## prjusedrg_mostmony_devx2:rural.ses.med31[4] 0.28 0.20 0.01
## prjusedrg_mostmony_devx2:rural.ses.med41[1] 0.31 0.22 0.01
## prjusedrg_mostmony_devx2:rural.ses.med41[2] 0.21 0.18 0.01
## prjusedrg_mostmony_devx2:rural.ses.med41[3] 0.20 0.18 0.01
## prjusedrg_mostmony_devx2:rural.ses.med41[4] 0.28 0.21 0.01
## prjhack_mostmony_devx21[1] 0.25 0.15 0.03
## prjhack_mostmony_devx21[2] 0.24 0.14 0.04
## prjhack_mostmony_devx21[3] 0.25 0.14 0.04
## prjhack_mostmony_devx21[4] 0.25 0.14 0.04
## prjhack_mostmony_av12x21[1] 0.13 0.08 0.02
## prjhack_mostmony_av12x21[2] 0.13 0.08 0.02
## prjhack_mostmony_av12x21[3] 0.12 0.08 0.02
## prjhack_mostmony_av12x21[4] 0.12 0.08 0.02
## prjhack_mostmony_av12x21[5] 0.12 0.08 0.01
## prjhack_mostmony_av12x21[6] 0.12 0.08 0.01
## prjhack_mostmony_av12x21[7] 0.13 0.08 0.02
## prjhack_mostmony_av12x21[8] 0.13 0.08 0.02
## prjhack_mostmony_devx2:rural.ses.med21[1] 0.28 0.20 0.01
## prjhack_mostmony_devx2:rural.ses.med21[2] 0.24 0.18 0.01
## prjhack_mostmony_devx2:rural.ses.med21[3] 0.21 0.18 0.01
## prjhack_mostmony_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## prjhack_mostmony_devx2:rural.ses.med31[1] 0.22 0.18 0.01
## prjhack_mostmony_devx2:rural.ses.med31[2] 0.21 0.19 0.01
## prjhack_mostmony_devx2:rural.ses.med31[3] 0.29 0.22 0.01
## prjhack_mostmony_devx2:rural.ses.med31[4] 0.28 0.21 0.01
## prjhack_mostmony_devx2:rural.ses.med41[1] 0.26 0.20 0.01
## prjhack_mostmony_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjhack_mostmony_devx2:rural.ses.med41[3] 0.23 0.18 0.01
## prjhack_mostmony_devx2:rural.ses.med41[4] 0.29 0.21 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## prjthflt5_mostmony_devx21[1] 0.60 1.00 6631 2750
## prjthflt5_mostmony_devx21[2] 0.56 1.00 7623 2864
## prjthflt5_mostmony_devx21[3] 0.57 1.00 6822 2546
## prjthflt5_mostmony_devx21[4] 0.60 1.00 8112 3040
## prjthflt5_mostmony_av12x21[1] 0.32 1.00 7605 2494
## prjthflt5_mostmony_av12x21[2] 0.33 1.00 6219 2690
## prjthflt5_mostmony_av12x21[3] 0.32 1.00 6135 2379
## prjthflt5_mostmony_av12x21[4] 0.33 1.00 7239 1995
## prjthflt5_mostmony_av12x21[5] 0.31 1.00 6785 2706
## prjthflt5_mostmony_av12x21[6] 0.32 1.00 6490 2502
## prjthflt5_mostmony_av12x21[7] 0.31 1.00 7077 3059
## prjthflt5_mostmony_av12x21[8] 0.31 1.00 5440 3026
## prjthflt5_mostmony_devx2:rural.ses.med21[1] 0.69 1.00 4736 2438
## prjthflt5_mostmony_devx2:rural.ses.med21[2] 0.64 1.00 5943 2940
## prjthflt5_mostmony_devx2:rural.ses.med21[3] 0.72 1.00 5758 3015
## prjthflt5_mostmony_devx2:rural.ses.med21[4] 0.64 1.00 6069 2295
## prjthflt5_mostmony_devx2:rural.ses.med31[1] 0.67 1.00 5519 2681
## prjthflt5_mostmony_devx2:rural.ses.med31[2] 0.76 1.00 3971 2497
## prjthflt5_mostmony_devx2:rural.ses.med31[3] 0.61 1.00 4733 3050
## prjthflt5_mostmony_devx2:rural.ses.med31[4] 0.67 1.00 5517 2947
## prjthflt5_mostmony_devx2:rural.ses.med41[1] 0.76 1.00 4060 2298
## prjthflt5_mostmony_devx2:rural.ses.med41[2] 0.65 1.00 4223 2851
## prjthflt5_mostmony_devx2:rural.ses.med41[3] 0.60 1.00 5255 3215
## prjthflt5_mostmony_devx2:rural.ses.med41[4] 0.64 1.00 5328 2815
## prjthfgt5_mostmony_devx21[1] 0.59 1.00 6837 2457
## prjthfgt5_mostmony_devx21[2] 0.59 1.00 6400 2985
## prjthfgt5_mostmony_devx21[3] 0.57 1.00 6851 2782
## prjthfgt5_mostmony_devx21[4] 0.58 1.00 6585 3065
## prjthfgt5_mostmony_av12x21[1] 0.31 1.00 6564 2285
## prjthfgt5_mostmony_av12x21[2] 0.32 1.00 7263 2727
## prjthfgt5_mostmony_av12x21[3] 0.31 1.00 8022 2627
## prjthfgt5_mostmony_av12x21[4] 0.32 1.00 7335 2583
## prjthfgt5_mostmony_av12x21[5] 0.32 1.00 5797 2765
## prjthfgt5_mostmony_av12x21[6] 0.31 1.00 6894 2968
## prjthfgt5_mostmony_av12x21[7] 0.31 1.00 6682 3016
## prjthfgt5_mostmony_av12x21[8] 0.32 1.00 7564 3180
## prjthfgt5_mostmony_devx2:rural.ses.med21[1] 0.72 1.00 6014 2966
## prjthfgt5_mostmony_devx2:rural.ses.med21[2] 0.60 1.00 6739 2543
## prjthfgt5_mostmony_devx2:rural.ses.med21[3] 0.70 1.00 6351 3065
## prjthfgt5_mostmony_devx2:rural.ses.med21[4] 0.66 1.00 6547 3003
## prjthfgt5_mostmony_devx2:rural.ses.med31[1] 0.69 1.00 5840 2966
## prjthfgt5_mostmony_devx2:rural.ses.med31[2] 0.76 1.00 4646 2883
## prjthfgt5_mostmony_devx2:rural.ses.med31[3] 0.58 1.00 4262 2982
## prjthfgt5_mostmony_devx2:rural.ses.med31[4] 0.68 1.00 5393 2838
## prjthfgt5_mostmony_devx2:rural.ses.med41[1] 0.70 1.00 6013 2656
## prjthfgt5_mostmony_devx2:rural.ses.med41[2] 0.74 1.00 4794 2090
## prjthfgt5_mostmony_devx2:rural.ses.med41[3] 0.53 1.00 3928 3003
## prjthfgt5_mostmony_devx2:rural.ses.med41[4] 0.62 1.00 5908 2921
## prjthreat_mostmony_devx21[1] 0.59 1.00 7384 2730
## prjthreat_mostmony_devx21[2] 0.57 1.00 6565 2339
## prjthreat_mostmony_devx21[3] 0.60 1.00 6923 2779
## prjthreat_mostmony_devx21[4] 0.57 1.00 5506 2405
## prjthreat_mostmony_av12x21[1] 0.32 1.00 6135 2697
## prjthreat_mostmony_av12x21[2] 0.33 1.00 8581 2596
## prjthreat_mostmony_av12x21[3] 0.32 1.00 6897 2463
## prjthreat_mostmony_av12x21[4] 0.34 1.00 6604 2617
## prjthreat_mostmony_av12x21[5] 0.35 1.00 7445 2532
## prjthreat_mostmony_av12x21[6] 0.32 1.00 6063 2501
## prjthreat_mostmony_av12x21[7] 0.31 1.00 6979 2814
## prjthreat_mostmony_av12x21[8] 0.30 1.00 7052 2333
## prjthreat_mostmony_devx2:rural.ses.med21[1] 0.72 1.00 5652 2350
## prjthreat_mostmony_devx2:rural.ses.med21[2] 0.63 1.00 6066 2512
## prjthreat_mostmony_devx2:rural.ses.med21[3] 0.72 1.00 7074 2877
## prjthreat_mostmony_devx2:rural.ses.med21[4] 0.71 1.00 5831 2568
## prjthreat_mostmony_devx2:rural.ses.med31[1] 0.69 1.00 6345 2330
## prjthreat_mostmony_devx2:rural.ses.med31[2] 0.67 1.00 5207 2581
## prjthreat_mostmony_devx2:rural.ses.med31[3] 0.68 1.00 5976 2696
## prjthreat_mostmony_devx2:rural.ses.med31[4] 0.75 1.00 5729 2774
## prjthreat_mostmony_devx2:rural.ses.med41[1] 0.76 1.00 4440 2637
## prjthreat_mostmony_devx2:rural.ses.med41[2] 0.68 1.00 5461 3137
## prjthreat_mostmony_devx2:rural.ses.med41[3] 0.61 1.00 4150 3073
## prjthreat_mostmony_devx2:rural.ses.med41[4] 0.73 1.00 6589 2621
## prjharm_mostmony_devx21[1] 0.60 1.00 7494 2609
## prjharm_mostmony_devx21[2] 0.56 1.00 6707 2881
## prjharm_mostmony_devx21[3] 0.60 1.01 7965 3020
## prjharm_mostmony_devx21[4] 0.59 1.00 6320 2988
## prjharm_mostmony_av12x21[1] 0.31 1.00 7237 2312
## prjharm_mostmony_av12x21[2] 0.31 1.00 6752 1991
## prjharm_mostmony_av12x21[3] 0.32 1.00 7789 2146
## prjharm_mostmony_av12x21[4] 0.31 1.00 6472 2599
## prjharm_mostmony_av12x21[5] 0.33 1.00 7805 2526
## prjharm_mostmony_av12x21[6] 0.34 1.00 6883 2728
## prjharm_mostmony_av12x21[7] 0.31 1.00 5911 2866
## prjharm_mostmony_av12x21[8] 0.30 1.00 6440 3021
## prjharm_mostmony_devx2:rural.ses.med21[1] 0.70 1.00 6010 2317
## prjharm_mostmony_devx2:rural.ses.med21[2] 0.66 1.00 6124 2848
## prjharm_mostmony_devx2:rural.ses.med21[3] 0.66 1.00 6102 2237
## prjharm_mostmony_devx2:rural.ses.med21[4] 0.74 1.00 5120 3067
## prjharm_mostmony_devx2:rural.ses.med31[1] 0.67 1.00 5738 2481
## prjharm_mostmony_devx2:rural.ses.med31[2] 0.69 1.00 5037 2411
## prjharm_mostmony_devx2:rural.ses.med31[3] 0.72 1.00 5255 2838
## prjharm_mostmony_devx2:rural.ses.med31[4] 0.74 1.00 6830 3086
## prjharm_mostmony_devx2:rural.ses.med41[1] 0.72 1.00 7511 2586
## prjharm_mostmony_devx2:rural.ses.med41[2] 0.68 1.00 5931 2497
## prjharm_mostmony_devx2:rural.ses.med41[3] 0.65 1.00 6309 2909
## prjharm_mostmony_devx2:rural.ses.med41[4] 0.75 1.00 6044 2832
## prjusedrg_mostmony_devx21[1] 0.60 1.00 8297 2551
## prjusedrg_mostmony_devx21[2] 0.59 1.00 8930 2341
## prjusedrg_mostmony_devx21[3] 0.55 1.00 7748 2906
## prjusedrg_mostmony_devx21[4] 0.59 1.00 7785 2816
## prjusedrg_mostmony_av12x21[1] 0.32 1.00 6441 2563
## prjusedrg_mostmony_av12x21[2] 0.32 1.00 5849 2057
## prjusedrg_mostmony_av12x21[3] 0.31 1.00 6871 2510
## prjusedrg_mostmony_av12x21[4] 0.32 1.00 6328 3064
## prjusedrg_mostmony_av12x21[5] 0.32 1.00 7700 2779
## prjusedrg_mostmony_av12x21[6] 0.33 1.00 6995 2444
## prjusedrg_mostmony_av12x21[7] 0.30 1.00 7330 3328
## prjusedrg_mostmony_av12x21[8] 0.31 1.00 6803 2978
## prjusedrg_mostmony_devx2:rural.ses.med21[1] 0.74 1.00 5474 2234
## prjusedrg_mostmony_devx2:rural.ses.med21[2] 0.65 1.00 6157 2521
## prjusedrg_mostmony_devx2:rural.ses.med21[3] 0.67 1.00 5676 2677
## prjusedrg_mostmony_devx2:rural.ses.med21[4] 0.73 1.00 4928 2769
## prjusedrg_mostmony_devx2:rural.ses.med31[1] 0.71 1.00 5342 2427
## prjusedrg_mostmony_devx2:rural.ses.med31[2] 0.66 1.00 6359 2424
## prjusedrg_mostmony_devx2:rural.ses.med31[3] 0.69 1.00 5882 2324
## prjusedrg_mostmony_devx2:rural.ses.med31[4] 0.73 1.00 4886 2800
## prjusedrg_mostmony_devx2:rural.ses.med41[1] 0.78 1.00 4089 2539
## prjusedrg_mostmony_devx2:rural.ses.med41[2] 0.67 1.00 5784 2539
## prjusedrg_mostmony_devx2:rural.ses.med41[3] 0.66 1.00 4281 2965
## prjusedrg_mostmony_devx2:rural.ses.med41[4] 0.75 1.00 5929 3103
## prjhack_mostmony_devx21[1] 0.59 1.00 8527 2597
## prjhack_mostmony_devx21[2] 0.56 1.00 6115 2227
## prjhack_mostmony_devx21[3] 0.58 1.00 6545 3252
## prjhack_mostmony_devx21[4] 0.57 1.00 6254 2954
## prjhack_mostmony_av12x21[1] 0.32 1.00 7320 2266
## prjhack_mostmony_av12x21[2] 0.34 1.00 7132 2650
## prjhack_mostmony_av12x21[3] 0.32 1.00 5903 2153
## prjhack_mostmony_av12x21[4] 0.31 1.00 7110 2667
## prjhack_mostmony_av12x21[5] 0.30 1.00 6136 2413
## prjhack_mostmony_av12x21[6] 0.31 1.00 8614 2678
## prjhack_mostmony_av12x21[7] 0.32 1.00 7226 2986
## prjhack_mostmony_av12x21[8] 0.32 1.00 6712 2919
## prjhack_mostmony_devx2:rural.ses.med21[1] 0.74 1.00 4753 2246
## prjhack_mostmony_devx2:rural.ses.med21[2] 0.68 1.00 6622 2576
## prjhack_mostmony_devx2:rural.ses.med21[3] 0.66 1.00 5539 2637
## prjhack_mostmony_devx2:rural.ses.med21[4] 0.73 1.00 6067 2984
## prjhack_mostmony_devx2:rural.ses.med31[1] 0.66 1.00 5357 2546
## prjhack_mostmony_devx2:rural.ses.med31[2] 0.69 1.00 4550 3215
## prjhack_mostmony_devx2:rural.ses.med31[3] 0.78 1.00 4474 3461
## prjhack_mostmony_devx2:rural.ses.med31[4] 0.74 1.00 5711 2650
## prjhack_mostmony_devx2:rural.ses.med41[1] 0.72 1.00 6390 2576
## prjhack_mostmony_devx2:rural.ses.med41[2] 0.67 1.00 6446 2618
## prjhack_mostmony_devx2:rural.ses.med41[3] 0.65 1.00 5715 2929
## prjhack_mostmony_devx2:rural.ses.med41[4] 0.75 1.00 5735 2495
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.1.1.4 Prior summary
out.chg.prjcrime.stmony.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## group resp dpar nlpar lb ub source
## default
## prjhack user
## prjhack user
## prjhack user
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjharm user
## prjharm user
## prjharm user
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthflt5 user
## prjthflt5 user
## prjthflt5 user
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthreat user
## prjthreat user
## prjthreat user
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjusedrg user
## prjusedrg user
## prjusedrg user
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## default
## prjhack user
## prjharm user
## prjthfgt5 user
## prjthflt5 user
## prjthreat user
## prjusedrg user
## prjhack 0 default
## prjharm 0 default
## prjthfgt5 0 default
## prjthflt5 0 default
## prjthreat 0 default
## prjusedrg 0 default
## id prjhack 0 (vectorized)
## id prjhack 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjusedrg 0 (vectorized)
## id prjusedrg 0 (vectorized)
## prjhack user
## prjhack default
## prjhack default
## prjhack default
## prjhack user
## prjharm user
## prjharm default
## prjharm default
## prjharm default
## prjharm user
## prjthfgt5 user
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 user
## prjthflt5 user
## prjthflt5 default
## prjthflt5 default
## prjthflt5 default
## prjthflt5 user
## prjthreat user
## prjthreat default
## prjthreat default
## prjthreat default
## prjthreat user
## prjusedrg user
## prjusedrg default
## prjusedrg default
## prjusedrg default
## prjusedrg user8.1.1.2 Corr X Community: sttran & prj crime
#Community Change: criminal intent items ~ mo(sttran)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mosttran_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mosttran_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mosttran_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mosttran_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.sttran.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(sttran_devx2) + mo(sttran_av12x2) +
rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_sttran_comm_fit",
file_refit = "on_change"
)
out.chg.prjcrime.sttran.comm.fit <- ppchecks(chg.prjcrime.sttran.comm.fit)8.1.1.2.1 Coefficient plot (intervals)
out.chg.prjcrime.sttran.comm.fit[[10]]
8.1.1.2.2 PPcheck (density)
p1 <- out.chg.prjcrime.sttran.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.sttran.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.sttran.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.sttran.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.sttran.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.sttran.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
8.1.1.2.3 Fit summary
out.chg.prjcrime.sttran.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## prjthreat ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## prjharm ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## prjusedrg ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## prjhack ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.93 0.57 2.87 5.15 1.00 1593
## sd(prjthfgt5_Intercept) 3.36 0.50 2.45 4.42 1.00 1522
## sd(prjthreat_Intercept) 3.31 0.58 2.32 4.57 1.00 1713
## sd(prjharm_Intercept) 3.13 0.59 2.11 4.40 1.00 1763
## sd(prjusedrg_Intercept) 3.01 0.56 2.02 4.20 1.00 1889
## sd(prjhack_Intercept) 0.93 0.58 0.04 2.13 1.00 747
## Tail_ESS
## sd(prjthflt5_Intercept) 2409
## sd(prjthfgt5_Intercept) 2437
## sd(prjthreat_Intercept) 2794
## sd(prjharm_Intercept) 2669
## sd(prjusedrg_Intercept) 2376
## sd(prjhack_Intercept) 1667
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_Intercept -5.31 0.88 -7.12 -3.67
## prjthfgt5_Intercept -5.03 0.82 -6.69 -3.47
## prjthreat_Intercept -5.98 0.95 -7.95 -4.17
## prjharm_Intercept -5.13 0.97 -7.17 -3.30
## prjusedrg_Intercept -5.55 0.95 -7.55 -3.80
## prjhack_Intercept -3.84 0.78 -5.49 -2.39
## prjthflt5_rural.ses.med2 -0.48 0.88 -2.15 1.28
## prjthflt5_rural.ses.med3 1.14 0.80 -0.53 2.63
## prjthflt5_rural.ses.med4 1.26 0.85 -0.48 2.81
## prjthfgt5_rural.ses.med2 -0.51 0.89 -2.21 1.30
## prjthfgt5_rural.ses.med3 1.18 0.78 -0.45 2.62
## prjthfgt5_rural.ses.med4 1.11 0.79 -0.49 2.63
## prjthreat_rural.ses.med2 -0.31 0.91 -2.02 1.48
## prjthreat_rural.ses.med3 0.54 0.79 -1.09 2.07
## prjthreat_rural.ses.med4 1.55 0.87 -0.29 3.20
## prjharm_rural.ses.med2 0.20 0.88 -1.52 1.93
## prjharm_rural.ses.med3 -0.04 0.80 -1.68 1.47
## prjharm_rural.ses.med4 1.36 0.76 -0.24 2.75
## prjusedrg_rural.ses.med2 -0.27 0.85 -1.90 1.44
## prjusedrg_rural.ses.med3 -0.04 0.83 -1.64 1.58
## prjusedrg_rural.ses.med4 1.67 0.80 -0.04 3.15
## prjhack_rural.ses.med2 -0.38 0.84 -2.06 1.33
## prjhack_rural.ses.med3 0.10 0.74 -1.39 1.60
## prjhack_rural.ses.med4 1.25 0.68 -0.12 2.58
## prjthflt5_mosttran_devx2 -0.09 0.19 -0.46 0.29
## prjthflt5_mosttran_av12x2 -0.12 0.10 -0.31 0.07
## prjthflt5_mosttran_devx2:rural.ses.med2 -0.64 0.54 -1.77 0.36
## prjthflt5_mosttran_devx2:rural.ses.med3 0.23 0.42 -0.55 1.12
## prjthflt5_mosttran_devx2:rural.ses.med4 0.65 0.45 -0.14 1.62
## prjthfgt5_mosttran_devx2 -0.06 0.20 -0.48 0.32
## prjthfgt5_mosttran_av12x2 -0.07 0.09 -0.25 0.11
## prjthfgt5_mosttran_devx2:rural.ses.med2 -0.76 0.51 -1.81 0.20
## prjthfgt5_mosttran_devx2:rural.ses.med3 0.06 0.42 -0.77 0.95
## prjthfgt5_mosttran_devx2:rural.ses.med4 0.53 0.36 -0.14 1.29
## prjthreat_mosttran_devx2 -0.18 0.21 -0.59 0.22
## prjthreat_mosttran_av12x2 -0.06 0.09 -0.25 0.12
## prjthreat_mosttran_devx2:rural.ses.med2 -0.76 0.63 -2.08 0.39
## prjthreat_mosttran_devx2:rural.ses.med3 -0.03 0.47 -0.97 0.89
## prjthreat_mosttran_devx2:rural.ses.med4 0.08 0.48 -0.83 1.09
## prjharm_mosttran_devx2 -0.28 0.21 -0.70 0.14
## prjharm_mosttran_av12x2 -0.17 0.10 -0.35 0.03
## prjharm_mosttran_devx2:rural.ses.med2 -0.79 0.58 -2.06 0.25
## prjharm_mosttran_devx2:rural.ses.med3 0.19 0.48 -0.81 1.08
## prjharm_mosttran_devx2:rural.ses.med4 -0.41 0.54 -1.53 0.64
## prjusedrg_mosttran_devx2 -0.18 0.21 -0.60 0.23
## prjusedrg_mosttran_av12x2 -0.07 0.10 -0.27 0.12
## prjusedrg_mosttran_devx2:rural.ses.med2 -0.56 0.60 -1.87 0.51
## prjusedrg_mosttran_devx2:rural.ses.med3 -0.69 0.58 -1.90 0.37
## prjusedrg_mosttran_devx2:rural.ses.med4 -0.08 0.50 -1.10 0.88
## prjhack_mosttran_devx2 -0.19 0.21 -0.59 0.21
## prjhack_mosttran_av12x2 -0.03 0.09 -0.20 0.14
## prjhack_mosttran_devx2:rural.ses.med2 -0.69 0.61 -1.96 0.44
## prjhack_mosttran_devx2:rural.ses.med3 -0.03 0.47 -1.05 0.82
## prjhack_mosttran_devx2:rural.ses.med4 -0.31 0.42 -1.27 0.45
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1.00 2917 3185
## prjthfgt5_Intercept 1.00 2650 2720
## prjthreat_Intercept 1.00 3296 3132
## prjharm_Intercept 1.00 2898 2519
## prjusedrg_Intercept 1.00 3364 3442
## prjhack_Intercept 1.00 2231 2351
## prjthflt5_rural.ses.med2 1.00 5716 2906
## prjthflt5_rural.ses.med3 1.00 4594 2902
## prjthflt5_rural.ses.med4 1.00 4894 3034
## prjthfgt5_rural.ses.med2 1.00 5379 3445
## prjthfgt5_rural.ses.med3 1.00 4437 3212
## prjthfgt5_rural.ses.med4 1.00 4526 2910
## prjthreat_rural.ses.med2 1.00 5490 3044
## prjthreat_rural.ses.med3 1.00 5482 2993
## prjthreat_rural.ses.med4 1.00 4426 3092
## prjharm_rural.ses.med2 1.00 5878 3011
## prjharm_rural.ses.med3 1.00 5500 3224
## prjharm_rural.ses.med4 1.00 4373 3006
## prjusedrg_rural.ses.med2 1.00 5636 3208
## prjusedrg_rural.ses.med3 1.00 5614 3378
## prjusedrg_rural.ses.med4 1.00 3972 3287
## prjhack_rural.ses.med2 1.00 5977 3163
## prjhack_rural.ses.med3 1.00 4919 3133
## prjhack_rural.ses.med4 1.00 4583 2899
## prjthflt5_mosttran_devx2 1.00 4943 3307
## prjthflt5_mosttran_av12x2 1.00 3510 3163
## prjthflt5_mosttran_devx2:rural.ses.med2 1.00 3879 2862
## prjthflt5_mosttran_devx2:rural.ses.med3 1.00 3397 2733
## prjthflt5_mosttran_devx2:rural.ses.med4 1.00 2885 2401
## prjthfgt5_mosttran_devx2 1.00 4672 2964
## prjthfgt5_mosttran_av12x2 1.00 4224 2741
## prjthfgt5_mosttran_devx2:rural.ses.med2 1.00 3176 2681
## prjthfgt5_mosttran_devx2:rural.ses.med3 1.00 2624 2716
## prjthfgt5_mosttran_devx2:rural.ses.med4 1.00 3827 3188
## prjthreat_mosttran_devx2 1.00 5729 3202
## prjthreat_mosttran_av12x2 1.00 5051 3057
## prjthreat_mosttran_devx2:rural.ses.med2 1.00 4504 3241
## prjthreat_mosttran_devx2:rural.ses.med3 1.00 3266 2971
## prjthreat_mosttran_devx2:rural.ses.med4 1.00 2975 2809
## prjharm_mosttran_devx2 1.00 5372 3446
## prjharm_mosttran_av12x2 1.00 5103 3402
## prjharm_mosttran_devx2:rural.ses.med2 1.00 4361 2946
## prjharm_mosttran_devx2:rural.ses.med3 1.00 3924 2785
## prjharm_mosttran_devx2:rural.ses.med4 1.00 3247 2921
## prjusedrg_mosttran_devx2 1.00 5729 2824
## prjusedrg_mosttran_av12x2 1.00 4798 2857
## prjusedrg_mosttran_devx2:rural.ses.med2 1.00 4579 2943
## prjusedrg_mosttran_devx2:rural.ses.med3 1.00 4508 3037
## prjusedrg_mosttran_devx2:rural.ses.med4 1.00 2440 2934
## prjhack_mosttran_devx2 1.00 5769 3020
## prjhack_mosttran_av12x2 1.00 5700 3037
## prjhack_mosttran_devx2:rural.ses.med2 1.00 4518 3211
## prjhack_mosttran_devx2:rural.ses.med3 1.00 4022 3061
## prjhack_mosttran_devx2:rural.ses.med4 1.00 4000 2768
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## prjthflt5_mosttran_devx21[1] 0.27 0.15 0.04
## prjthflt5_mosttran_devx21[2] 0.24 0.14 0.03
## prjthflt5_mosttran_devx21[3] 0.24 0.14 0.04
## prjthflt5_mosttran_devx21[4] 0.25 0.15 0.04
## prjthflt5_mosttran_av12x21[1] 0.13 0.08 0.02
## prjthflt5_mosttran_av12x21[2] 0.14 0.09 0.02
## prjthflt5_mosttran_av12x21[3] 0.12 0.08 0.01
## prjthflt5_mosttran_av12x21[4] 0.13 0.08 0.02
## prjthflt5_mosttran_av12x21[5] 0.12 0.08 0.02
## prjthflt5_mosttran_av12x21[6] 0.12 0.07 0.02
## prjthflt5_mosttran_av12x21[7] 0.12 0.08 0.02
## prjthflt5_mosttran_av12x21[8] 0.12 0.08 0.02
## prjthflt5_mosttran_devx2:rural.ses.med21[1] 0.31 0.21 0.01
## prjthflt5_mosttran_devx2:rural.ses.med21[2] 0.22 0.17 0.01
## prjthflt5_mosttran_devx2:rural.ses.med21[3] 0.21 0.17 0.01
## prjthflt5_mosttran_devx2:rural.ses.med21[4] 0.26 0.20 0.01
## prjthflt5_mosttran_devx2:rural.ses.med31[1] 0.29 0.21 0.01
## prjthflt5_mosttran_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjthflt5_mosttran_devx2:rural.ses.med31[3] 0.20 0.17 0.01
## prjthflt5_mosttran_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjthflt5_mosttran_devx2:rural.ses.med41[1] 0.36 0.22 0.02
## prjthflt5_mosttran_devx2:rural.ses.med41[2] 0.16 0.15 0.00
## prjthflt5_mosttran_devx2:rural.ses.med41[3] 0.16 0.15 0.00
## prjthflt5_mosttran_devx2:rural.ses.med41[4] 0.32 0.22 0.01
## prjthfgt5_mosttran_devx21[1] 0.27 0.15 0.03
## prjthfgt5_mosttran_devx21[2] 0.24 0.14 0.04
## prjthfgt5_mosttran_devx21[3] 0.23 0.14 0.03
## prjthfgt5_mosttran_devx21[4] 0.26 0.15 0.04
## prjthfgt5_mosttran_av12x21[1] 0.13 0.09 0.02
## prjthfgt5_mosttran_av12x21[2] 0.13 0.09 0.02
## prjthfgt5_mosttran_av12x21[3] 0.12 0.08 0.02
## prjthfgt5_mosttran_av12x21[4] 0.13 0.08 0.02
## prjthfgt5_mosttran_av12x21[5] 0.12 0.08 0.02
## prjthfgt5_mosttran_av12x21[6] 0.11 0.08 0.02
## prjthfgt5_mosttran_av12x21[7] 0.12 0.08 0.01
## prjthfgt5_mosttran_av12x21[8] 0.13 0.08 0.02
## prjthfgt5_mosttran_devx2:rural.ses.med21[1] 0.33 0.21 0.02
## prjthfgt5_mosttran_devx2:rural.ses.med21[2] 0.24 0.18 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med21[3] 0.19 0.16 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med21[4] 0.25 0.19 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med31[2] 0.21 0.18 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med31[3] 0.22 0.18 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med31[4] 0.30 0.22 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med41[1] 0.29 0.20 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med41[3] 0.22 0.17 0.01
## prjthfgt5_mosttran_devx2:rural.ses.med41[4] 0.25 0.19 0.01
## prjthreat_mosttran_devx21[1] 0.27 0.15 0.04
## prjthreat_mosttran_devx21[2] 0.25 0.14 0.04
## prjthreat_mosttran_devx21[3] 0.24 0.14 0.03
## prjthreat_mosttran_devx21[4] 0.25 0.15 0.03
## prjthreat_mosttran_av12x21[1] 0.13 0.08 0.02
## prjthreat_mosttran_av12x21[2] 0.13 0.08 0.02
## prjthreat_mosttran_av12x21[3] 0.13 0.08 0.02
## prjthreat_mosttran_av12x21[4] 0.13 0.08 0.02
## prjthreat_mosttran_av12x21[5] 0.12 0.08 0.02
## prjthreat_mosttran_av12x21[6] 0.12 0.08 0.02
## prjthreat_mosttran_av12x21[7] 0.12 0.08 0.01
## prjthreat_mosttran_av12x21[8] 0.13 0.08 0.02
## prjthreat_mosttran_devx2:rural.ses.med21[1] 0.29 0.20 0.01
## prjthreat_mosttran_devx2:rural.ses.med21[2] 0.21 0.17 0.00
## prjthreat_mosttran_devx2:rural.ses.med21[3] 0.26 0.19 0.01
## prjthreat_mosttran_devx2:rural.ses.med21[4] 0.25 0.19 0.01
## prjthreat_mosttran_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjthreat_mosttran_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjthreat_mosttran_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## prjthreat_mosttran_devx2:rural.ses.med31[4] 0.30 0.22 0.01
## prjthreat_mosttran_devx2:rural.ses.med41[1] 0.30 0.23 0.01
## prjthreat_mosttran_devx2:rural.ses.med41[2] 0.21 0.18 0.01
## prjthreat_mosttran_devx2:rural.ses.med41[3] 0.21 0.18 0.01
## prjthreat_mosttran_devx2:rural.ses.med41[4] 0.28 0.21 0.01
## prjharm_mosttran_devx21[1] 0.27 0.15 0.04
## prjharm_mosttran_devx21[2] 0.24 0.14 0.04
## prjharm_mosttran_devx21[3] 0.27 0.15 0.05
## prjharm_mosttran_devx21[4] 0.23 0.14 0.03
## prjharm_mosttran_av12x21[1] 0.12 0.07 0.02
## prjharm_mosttran_av12x21[2] 0.12 0.08 0.02
## prjharm_mosttran_av12x21[3] 0.14 0.09 0.02
## prjharm_mosttran_av12x21[4] 0.14 0.09 0.02
## prjharm_mosttran_av12x21[5] 0.14 0.08 0.02
## prjharm_mosttran_av12x21[6] 0.12 0.07 0.02
## prjharm_mosttran_av12x21[7] 0.11 0.07 0.02
## prjharm_mosttran_av12x21[8] 0.11 0.07 0.01
## prjharm_mosttran_devx2:rural.ses.med21[1] 0.29 0.20 0.01
## prjharm_mosttran_devx2:rural.ses.med21[2] 0.19 0.16 0.01
## prjharm_mosttran_devx2:rural.ses.med21[3] 0.27 0.20 0.01
## prjharm_mosttran_devx2:rural.ses.med21[4] 0.25 0.19 0.01
## prjharm_mosttran_devx2:rural.ses.med31[1] 0.23 0.18 0.01
## prjharm_mosttran_devx2:rural.ses.med31[2] 0.26 0.19 0.01
## prjharm_mosttran_devx2:rural.ses.med31[3] 0.23 0.18 0.01
## prjharm_mosttran_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjharm_mosttran_devx2:rural.ses.med41[1] 0.19 0.18 0.01
## prjharm_mosttran_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## prjharm_mosttran_devx2:rural.ses.med41[3] 0.32 0.22 0.01
## prjharm_mosttran_devx2:rural.ses.med41[4] 0.28 0.20 0.01
## prjusedrg_mosttran_devx21[1] 0.26 0.14 0.04
## prjusedrg_mosttran_devx21[2] 0.24 0.14 0.04
## prjusedrg_mosttran_devx21[3] 0.25 0.14 0.04
## prjusedrg_mosttran_devx21[4] 0.25 0.14 0.04
## prjusedrg_mosttran_av12x21[1] 0.13 0.08 0.02
## prjusedrg_mosttran_av12x21[2] 0.12 0.08 0.02
## prjusedrg_mosttran_av12x21[3] 0.13 0.08 0.02
## prjusedrg_mosttran_av12x21[4] 0.13 0.09 0.02
## prjusedrg_mosttran_av12x21[5] 0.12 0.08 0.01
## prjusedrg_mosttran_av12x21[6] 0.12 0.08 0.02
## prjusedrg_mosttran_av12x21[7] 0.12 0.08 0.01
## prjusedrg_mosttran_av12x21[8] 0.12 0.08 0.01
## prjusedrg_mosttran_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## prjusedrg_mosttran_devx2:rural.ses.med21[2] 0.20 0.17 0.00
## prjusedrg_mosttran_devx2:rural.ses.med21[3] 0.27 0.20 0.01
## prjusedrg_mosttran_devx2:rural.ses.med21[4] 0.27 0.20 0.01
## prjusedrg_mosttran_devx2:rural.ses.med31[1] 0.25 0.18 0.01
## prjusedrg_mosttran_devx2:rural.ses.med31[2] 0.20 0.17 0.01
## prjusedrg_mosttran_devx2:rural.ses.med31[3] 0.29 0.21 0.01
## prjusedrg_mosttran_devx2:rural.ses.med31[4] 0.26 0.20 0.01
## prjusedrg_mosttran_devx2:rural.ses.med41[1] 0.26 0.22 0.01
## prjusedrg_mosttran_devx2:rural.ses.med41[2] 0.21 0.18 0.01
## prjusedrg_mosttran_devx2:rural.ses.med41[3] 0.24 0.19 0.01
## prjusedrg_mosttran_devx2:rural.ses.med41[4] 0.29 0.21 0.01
## prjhack_mosttran_devx21[1] 0.27 0.15 0.04
## prjhack_mosttran_devx21[2] 0.25 0.14 0.04
## prjhack_mosttran_devx21[3] 0.23 0.14 0.03
## prjhack_mosttran_devx21[4] 0.25 0.15 0.03
## prjhack_mosttran_av12x21[1] 0.13 0.08 0.02
## prjhack_mosttran_av12x21[2] 0.13 0.08 0.02
## prjhack_mosttran_av12x21[3] 0.13 0.08 0.02
## prjhack_mosttran_av12x21[4] 0.13 0.08 0.02
## prjhack_mosttran_av12x21[5] 0.12 0.08 0.02
## prjhack_mosttran_av12x21[6] 0.12 0.08 0.01
## prjhack_mosttran_av12x21[7] 0.12 0.08 0.01
## prjhack_mosttran_av12x21[8] 0.13 0.08 0.02
## prjhack_mosttran_devx2:rural.ses.med21[1] 0.27 0.20 0.01
## prjhack_mosttran_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## prjhack_mosttran_devx2:rural.ses.med21[3] 0.26 0.19 0.01
## prjhack_mosttran_devx2:rural.ses.med21[4] 0.26 0.19 0.01
## prjhack_mosttran_devx2:rural.ses.med31[1] 0.24 0.19 0.01
## prjhack_mosttran_devx2:rural.ses.med31[2] 0.23 0.19 0.01
## prjhack_mosttran_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## prjhack_mosttran_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjhack_mosttran_devx2:rural.ses.med41[1] 0.21 0.17 0.01
## prjhack_mosttran_devx2:rural.ses.med41[2] 0.27 0.20 0.01
## prjhack_mosttran_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## prjhack_mosttran_devx2:rural.ses.med41[4] 0.30 0.21 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## prjthflt5_mosttran_devx21[1] 0.61 1.00 6897 2670
## prjthflt5_mosttran_devx21[2] 0.57 1.00 6472 2443
## prjthflt5_mosttran_devx21[3] 0.57 1.00 6513 2754
## prjthflt5_mosttran_devx21[4] 0.58 1.00 6905 2782
## prjthflt5_mosttran_av12x21[1] 0.32 1.00 7035 2713
## prjthflt5_mosttran_av12x21[2] 0.35 1.00 7739 2471
## prjthflt5_mosttran_av12x21[3] 0.31 1.00 7310 2362
## prjthflt5_mosttran_av12x21[4] 0.32 1.00 6722 2808
## prjthflt5_mosttran_av12x21[5] 0.31 1.00 6588 2680
## prjthflt5_mosttran_av12x21[6] 0.30 1.00 6848 2475
## prjthflt5_mosttran_av12x21[7] 0.31 1.00 6128 2856
## prjthflt5_mosttran_av12x21[8] 0.31 1.00 6025 3115
## prjthflt5_mosttran_devx2:rural.ses.med21[1] 0.75 1.00 4670 2912
## prjthflt5_mosttran_devx2:rural.ses.med21[2] 0.64 1.00 5189 2395
## prjthflt5_mosttran_devx2:rural.ses.med21[3] 0.63 1.00 5890 2341
## prjthflt5_mosttran_devx2:rural.ses.med21[4] 0.71 1.00 6132 2212
## prjthflt5_mosttran_devx2:rural.ses.med31[1] 0.74 1.00 4387 2425
## prjthflt5_mosttran_devx2:rural.ses.med31[2] 0.66 1.00 5015 2354
## prjthflt5_mosttran_devx2:rural.ses.med31[3] 0.63 1.00 4394 2770
## prjthflt5_mosttran_devx2:rural.ses.med31[4] 0.76 1.01 6082 2583
## prjthflt5_mosttran_devx2:rural.ses.med41[1] 0.79 1.00 5159 2827
## prjthflt5_mosttran_devx2:rural.ses.med41[2] 0.57 1.00 4732 2723
## prjthflt5_mosttran_devx2:rural.ses.med41[3] 0.55 1.00 4893 2769
## prjthflt5_mosttran_devx2:rural.ses.med41[4] 0.77 1.00 5302 2901
## prjthfgt5_mosttran_devx21[1] 0.62 1.00 6747 2195
## prjthfgt5_mosttran_devx21[2] 0.57 1.00 6935 2449
## prjthfgt5_mosttran_devx21[3] 0.56 1.00 6782 3011
## prjthfgt5_mosttran_devx21[4] 0.60 1.00 6202 3018
## prjthfgt5_mosttran_av12x21[1] 0.33 1.00 7789 2689
## prjthfgt5_mosttran_av12x21[2] 0.35 1.00 6433 2459
## prjthfgt5_mosttran_av12x21[3] 0.31 1.00 6334 2503
## prjthfgt5_mosttran_av12x21[4] 0.33 1.00 6623 2388
## prjthfgt5_mosttran_av12x21[5] 0.30 1.00 6891 3006
## prjthfgt5_mosttran_av12x21[6] 0.31 1.00 7112 2634
## prjthfgt5_mosttran_av12x21[7] 0.32 1.00 5407 2528
## prjthfgt5_mosttran_av12x21[8] 0.32 1.00 6899 2711
## prjthfgt5_mosttran_devx2:rural.ses.med21[1] 0.75 1.00 4795 2649
## prjthfgt5_mosttran_devx2:rural.ses.med21[2] 0.67 1.00 5909 2597
## prjthfgt5_mosttran_devx2:rural.ses.med21[3] 0.59 1.00 5000 2535
## prjthfgt5_mosttran_devx2:rural.ses.med21[4] 0.69 1.00 5503 2293
## prjthfgt5_mosttran_devx2:rural.ses.med31[1] 0.74 1.00 4707 2871
## prjthfgt5_mosttran_devx2:rural.ses.med31[2] 0.66 1.00 6782 2840
## prjthfgt5_mosttran_devx2:rural.ses.med31[3] 0.68 1.00 4432 2391
## prjthfgt5_mosttran_devx2:rural.ses.med31[4] 0.76 1.00 5250 2751
## prjthfgt5_mosttran_devx2:rural.ses.med41[1] 0.72 1.00 5663 2547
## prjthfgt5_mosttran_devx2:rural.ses.med41[2] 0.65 1.00 5037 2720
## prjthfgt5_mosttran_devx2:rural.ses.med41[3] 0.66 1.00 4161 2560
## prjthfgt5_mosttran_devx2:rural.ses.med41[4] 0.67 1.00 5372 2532
## prjthreat_mosttran_devx21[1] 0.59 1.00 7663 2472
## prjthreat_mosttran_devx21[2] 0.56 1.00 5778 2594
## prjthreat_mosttran_devx21[3] 0.55 1.00 6808 2441
## prjthreat_mosttran_devx21[4] 0.59 1.00 7471 2154
## prjthreat_mosttran_av12x21[1] 0.33 1.00 7604 2649
## prjthreat_mosttran_av12x21[2] 0.33 1.00 6432 1989
## prjthreat_mosttran_av12x21[3] 0.32 1.00 6637 2592
## prjthreat_mosttran_av12x21[4] 0.32 1.00 6645 2511
## prjthreat_mosttran_av12x21[5] 0.32 1.00 7518 2719
## prjthreat_mosttran_av12x21[6] 0.31 1.00 6239 2645
## prjthreat_mosttran_av12x21[7] 0.32 1.00 7212 2448
## prjthreat_mosttran_av12x21[8] 0.33 1.00 5961 2915
## prjthreat_mosttran_devx2:rural.ses.med21[1] 0.74 1.00 5881 2767
## prjthreat_mosttran_devx2:rural.ses.med21[2] 0.64 1.00 5992 2390
## prjthreat_mosttran_devx2:rural.ses.med21[3] 0.69 1.00 5297 2289
## prjthreat_mosttran_devx2:rural.ses.med21[4] 0.69 1.00 6384 2840
## prjthreat_mosttran_devx2:rural.ses.med31[1] 0.69 1.00 6092 2377
## prjthreat_mosttran_devx2:rural.ses.med31[2] 0.68 1.00 6500 2693
## prjthreat_mosttran_devx2:rural.ses.med31[3] 0.69 1.00 5403 2276
## prjthreat_mosttran_devx2:rural.ses.med31[4] 0.78 1.00 5126 2216
## prjthreat_mosttran_devx2:rural.ses.med41[1] 0.79 1.00 3532 2724
## prjthreat_mosttran_devx2:rural.ses.med41[2] 0.65 1.00 4701 2962
## prjthreat_mosttran_devx2:rural.ses.med41[3] 0.66 1.00 4553 2649
## prjthreat_mosttran_devx2:rural.ses.med41[4] 0.75 1.00 5348 2676
## prjharm_mosttran_devx21[1] 0.61 1.00 7092 2186
## prjharm_mosttran_devx21[2] 0.55 1.00 7624 2668
## prjharm_mosttran_devx21[3] 0.59 1.00 7145 2841
## prjharm_mosttran_devx21[4] 0.54 1.00 7331 2554
## prjharm_mosttran_av12x21[1] 0.30 1.00 6550 2632
## prjharm_mosttran_av12x21[2] 0.31 1.00 7190 2574
## prjharm_mosttran_av12x21[3] 0.35 1.00 7829 2197
## prjharm_mosttran_av12x21[4] 0.36 1.00 7124 2631
## prjharm_mosttran_av12x21[5] 0.34 1.00 6773 2766
## prjharm_mosttran_av12x21[6] 0.29 1.00 6492 2186
## prjharm_mosttran_av12x21[7] 0.29 1.00 6290 2489
## prjharm_mosttran_av12x21[8] 0.28 1.00 5969 2961
## prjharm_mosttran_devx2:rural.ses.med21[1] 0.73 1.00 5356 2294
## prjharm_mosttran_devx2:rural.ses.med21[2] 0.61 1.00 5633 2349
## prjharm_mosttran_devx2:rural.ses.med21[3] 0.73 1.00 6163 2497
## prjharm_mosttran_devx2:rural.ses.med21[4] 0.69 1.00 5727 2710
## prjharm_mosttran_devx2:rural.ses.med31[1] 0.66 1.00 6146 2476
## prjharm_mosttran_devx2:rural.ses.med31[2] 0.70 1.00 4197 2756
## prjharm_mosttran_devx2:rural.ses.med31[3] 0.67 1.00 6052 3027
## prjharm_mosttran_devx2:rural.ses.med31[4] 0.74 1.00 4924 2559
## prjharm_mosttran_devx2:rural.ses.med41[1] 0.67 1.00 4114 2715
## prjharm_mosttran_devx2:rural.ses.med41[2] 0.65 1.00 4905 2745
## prjharm_mosttran_devx2:rural.ses.med41[3] 0.79 1.00 4652 2657
## prjharm_mosttran_devx2:rural.ses.med41[4] 0.72 1.00 5764 2801
## prjusedrg_mosttran_devx21[1] 0.58 1.00 6774 2692
## prjusedrg_mosttran_devx21[2] 0.56 1.00 6140 2533
## prjusedrg_mosttran_devx21[3] 0.57 1.00 5773 2611
## prjusedrg_mosttran_devx21[4] 0.57 1.00 6870 2648
## prjusedrg_mosttran_av12x21[1] 0.31 1.00 6352 2789
## prjusedrg_mosttran_av12x21[2] 0.31 1.00 7360 1869
## prjusedrg_mosttran_av12x21[3] 0.33 1.00 6850 2407
## prjusedrg_mosttran_av12x21[4] 0.34 1.00 5618 2263
## prjusedrg_mosttran_av12x21[5] 0.32 1.00 6472 2615
## prjusedrg_mosttran_av12x21[6] 0.31 1.00 7321 2739
## prjusedrg_mosttran_av12x21[7] 0.32 1.00 6552 2229
## prjusedrg_mosttran_av12x21[8] 0.32 1.00 6141 3133
## prjusedrg_mosttran_devx2:rural.ses.med21[1] 0.71 1.00 6008 2267
## prjusedrg_mosttran_devx2:rural.ses.med21[2] 0.64 1.00 5851 2368
## prjusedrg_mosttran_devx2:rural.ses.med21[3] 0.74 1.00 5926 2892
## prjusedrg_mosttran_devx2:rural.ses.med21[4] 0.73 1.00 5726 2746
## prjusedrg_mosttran_devx2:rural.ses.med31[1] 0.67 1.00 5690 2269
## prjusedrg_mosttran_devx2:rural.ses.med31[2] 0.62 1.00 5608 2858
## prjusedrg_mosttran_devx2:rural.ses.med31[3] 0.75 1.00 5087 2988
## prjusedrg_mosttran_devx2:rural.ses.med31[4] 0.70 1.00 6741 2743
## prjusedrg_mosttran_devx2:rural.ses.med41[1] 0.77 1.00 3751 2912
## prjusedrg_mosttran_devx2:rural.ses.med41[2] 0.66 1.00 5608 2654
## prjusedrg_mosttran_devx2:rural.ses.med41[3] 0.71 1.00 3979 2653
## prjusedrg_mosttran_devx2:rural.ses.med41[4] 0.76 1.00 5848 2978
## prjhack_mosttran_devx21[1] 0.61 1.00 7799 2744
## prjhack_mosttran_devx21[2] 0.57 1.00 7448 2782
## prjhack_mosttran_devx21[3] 0.55 1.00 6448 2687
## prjhack_mosttran_devx21[4] 0.59 1.00 8608 2861
## prjhack_mosttran_av12x21[1] 0.32 1.00 7578 2907
## prjhack_mosttran_av12x21[2] 0.33 1.00 7823 2425
## prjhack_mosttran_av12x21[3] 0.33 1.00 8070 2235
## prjhack_mosttran_av12x21[4] 0.32 1.00 6190 2452
## prjhack_mosttran_av12x21[5] 0.31 1.00 7537 2678
## prjhack_mosttran_av12x21[6] 0.31 1.00 6562 2259
## prjhack_mosttran_av12x21[7] 0.31 1.00 6252 2986
## prjhack_mosttran_av12x21[8] 0.32 1.00 7039 2502
## prjhack_mosttran_devx2:rural.ses.med21[1] 0.71 1.00 5509 2510
## prjhack_mosttran_devx2:rural.ses.med21[2] 0.66 1.00 5405 2157
## prjhack_mosttran_devx2:rural.ses.med21[3] 0.70 1.00 5505 2372
## prjhack_mosttran_devx2:rural.ses.med21[4] 0.71 1.00 5670 2265
## prjhack_mosttran_devx2:rural.ses.med31[1] 0.68 1.00 6341 2423
## prjhack_mosttran_devx2:rural.ses.med31[2] 0.70 1.00 5643 2733
## prjhack_mosttran_devx2:rural.ses.med31[3] 0.70 1.00 5447 2189
## prjhack_mosttran_devx2:rural.ses.med31[4] 0.76 1.00 5177 3010
## prjhack_mosttran_devx2:rural.ses.med41[1] 0.65 1.00 4783 2406
## prjhack_mosttran_devx2:rural.ses.med41[2] 0.73 1.00 5256 2567
## prjhack_mosttran_devx2:rural.ses.med41[3] 0.65 1.00 5491 2823
## prjhack_mosttran_devx2:rural.ses.med41[4] 0.77 1.00 4651 2256
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.1.2.4 Prior summary
out.chg.prjcrime.sttran.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## group resp dpar nlpar lb ub source
## default
## prjhack user
## prjhack user
## prjhack user
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjharm user
## prjharm user
## prjharm user
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthflt5 user
## prjthflt5 user
## prjthflt5 user
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthreat user
## prjthreat user
## prjthreat user
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjusedrg user
## prjusedrg user
## prjusedrg user
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## default
## prjhack user
## prjharm user
## prjthfgt5 user
## prjthflt5 user
## prjthreat user
## prjusedrg user
## prjhack 0 default
## prjharm 0 default
## prjthfgt5 0 default
## prjthflt5 0 default
## prjthreat 0 default
## prjusedrg 0 default
## id prjhack 0 (vectorized)
## id prjhack 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjusedrg 0 (vectorized)
## id prjusedrg 0 (vectorized)
## prjhack user
## prjhack default
## prjhack default
## prjhack default
## prjhack user
## prjharm user
## prjharm default
## prjharm default
## prjharm default
## prjharm user
## prjthfgt5 user
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 user
## prjthflt5 user
## prjthflt5 default
## prjthflt5 default
## prjthflt5 default
## prjthflt5 user
## prjthreat user
## prjthreat default
## prjthreat default
## prjthreat default
## prjthreat user
## prjusedrg user
## prjusedrg default
## prjusedrg default
## prjusedrg default
## prjusedrg user8.1.1.3 Corr X Community: stresp & prj crime
#Community Change: criminal intent items ~ mo(stresp)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostresp_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostresp_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostresp_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostresp_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.stresp.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stresp_devx2) + mo(stresp_av12x2) +
rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stresp_comm_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stresp.comm.fit <- ppchecks(chg.prjcrime.stresp.comm.fit)8.1.1.3.1 Coefficient plot (intervals)
out.chg.prjcrime.stresp.comm.fit[[10]]
8.1.1.3.2 PPcheck (density)
p1 <- out.chg.prjcrime.stresp.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stresp.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stresp.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stresp.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stresp.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stresp.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
8.1.1.3.3 Fit summary
out.chg.prjcrime.stresp.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## prjthreat ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## prjharm ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## prjusedrg ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## prjhack ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.86 0.56 2.86 5.05 1.00 1696
## sd(prjthfgt5_Intercept) 3.30 0.50 2.43 4.36 1.00 1587
## sd(prjthreat_Intercept) 3.04 0.55 2.07 4.24 1.00 1533
## sd(prjharm_Intercept) 3.00 0.55 2.03 4.18 1.01 1375
## sd(prjusedrg_Intercept) 2.80 0.54 1.83 3.96 1.00 1411
## sd(prjhack_Intercept) 0.85 0.55 0.04 2.03 1.00 961
## Tail_ESS
## sd(prjthflt5_Intercept) 2646
## sd(prjthfgt5_Intercept) 2256
## sd(prjthreat_Intercept) 2559
## sd(prjharm_Intercept) 2560
## sd(prjusedrg_Intercept) 1799
## sd(prjhack_Intercept) 1791
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_Intercept -5.76 0.86 -7.49 -4.11
## prjthfgt5_Intercept -5.23 0.82 -6.94 -3.72
## prjthreat_Intercept -6.38 0.92 -8.22 -4.66
## prjharm_Intercept -6.06 0.95 -7.97 -4.32
## prjusedrg_Intercept -6.00 0.93 -7.92 -4.33
## prjhack_Intercept -4.31 0.76 -5.93 -2.93
## prjthflt5_rural.ses.med2 -0.54 0.85 -2.13 1.17
## prjthflt5_rural.ses.med3 1.27 0.70 -0.14 2.64
## prjthflt5_rural.ses.med4 1.31 0.84 -0.43 2.86
## prjthfgt5_rural.ses.med2 -0.70 0.89 -2.39 1.10
## prjthfgt5_rural.ses.med3 1.20 0.69 -0.11 2.54
## prjthfgt5_rural.ses.med4 1.27 0.80 -0.39 2.75
## prjthreat_rural.ses.med2 -0.56 0.87 -2.20 1.16
## prjthreat_rural.ses.med3 0.73 0.74 -0.71 2.20
## prjthreat_rural.ses.med4 1.24 0.79 -0.36 2.72
## prjharm_rural.ses.med2 -0.24 0.81 -1.79 1.35
## prjharm_rural.ses.med3 0.03 0.76 -1.49 1.49
## prjharm_rural.ses.med4 1.35 0.76 -0.16 2.85
## prjusedrg_rural.ses.med2 -0.33 0.83 -1.94 1.30
## prjusedrg_rural.ses.med3 -0.18 0.80 -1.71 1.42
## prjusedrg_rural.ses.med4 1.30 0.78 -0.38 2.76
## prjhack_rural.ses.med2 -0.62 0.86 -2.26 1.12
## prjhack_rural.ses.med3 0.53 0.70 -0.84 1.90
## prjhack_rural.ses.med4 0.80 0.69 -0.56 2.17
## prjthflt5_mostresp_devx2 -0.24 0.21 -0.66 0.16
## prjthflt5_mostresp_av12x2 0.10 0.08 -0.06 0.26
## prjthflt5_mostresp_devx2:rural.ses.med2 -0.68 0.50 -1.73 0.24
## prjthflt5_mostresp_devx2:rural.ses.med3 0.01 0.37 -0.84 0.67
## prjthflt5_mostresp_devx2:rural.ses.med4 0.41 0.38 -0.33 1.17
## prjthfgt5_mostresp_devx2 -0.21 0.20 -0.62 0.18
## prjthfgt5_mostresp_av12x2 0.07 0.08 -0.09 0.22
## prjthfgt5_mostresp_devx2:rural.ses.med2 -0.67 0.53 -1.79 0.32
## prjthfgt5_mostresp_devx2:rural.ses.med3 -0.02 0.39 -0.91 0.60
## prjthfgt5_mostresp_devx2:rural.ses.med4 0.34 0.36 -0.38 1.06
## prjthreat_mostresp_devx2 -0.33 0.21 -0.75 0.07
## prjthreat_mostresp_av12x2 0.15 0.08 -0.02 0.30
## prjthreat_mostresp_devx2:rural.ses.med2 -0.32 0.59 -1.53 0.75
## prjthreat_mostresp_devx2:rural.ses.med3 -0.10 0.40 -0.95 0.63
## prjthreat_mostresp_devx2:rural.ses.med4 0.33 0.39 -0.41 1.14
## prjharm_mostresp_devx2 -0.20 0.22 -0.63 0.22
## prjharm_mostresp_av12x2 0.05 0.09 -0.13 0.21
## prjharm_mostresp_devx2:rural.ses.med2 -0.30 0.51 -1.41 0.69
## prjharm_mostresp_devx2:rural.ses.med3 0.13 0.42 -0.76 0.91
## prjharm_mostresp_devx2:rural.ses.med4 -0.34 0.44 -1.28 0.47
## prjusedrg_mostresp_devx2 -0.35 0.21 -0.74 0.06
## prjusedrg_mostresp_av12x2 0.12 0.09 -0.06 0.29
## prjusedrg_mostresp_devx2:rural.ses.med2 -0.29 0.54 -1.39 0.74
## prjusedrg_mostresp_devx2:rural.ses.med3 -0.41 0.51 -1.50 0.56
## prjusedrg_mostresp_devx2:rural.ses.med4 0.23 0.40 -0.53 1.04
## prjhack_mostresp_devx2 -0.21 0.21 -0.63 0.20
## prjhack_mostresp_av12x2 0.09 0.07 -0.05 0.23
## prjhack_mostresp_devx2:rural.ses.med2 -0.38 0.57 -1.56 0.75
## prjhack_mostresp_devx2:rural.ses.med3 -0.35 0.43 -1.29 0.42
## prjhack_mostresp_devx2:rural.ses.med4 -0.04 0.39 -0.93 0.64
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1.00 4102 3505
## prjthfgt5_Intercept 1.00 3391 3202
## prjthreat_Intercept 1.00 3972 2833
## prjharm_Intercept 1.00 3041 3308
## prjusedrg_Intercept 1.00 3311 2933
## prjhack_Intercept 1.00 3019 2777
## prjthflt5_rural.ses.med2 1.00 6762 3191
## prjthflt5_rural.ses.med3 1.00 5289 3651
## prjthflt5_rural.ses.med4 1.00 5803 3138
## prjthfgt5_rural.ses.med2 1.00 6182 3085
## prjthfgt5_rural.ses.med3 1.00 4846 3277
## prjthfgt5_rural.ses.med4 1.00 4665 2962
## prjthreat_rural.ses.med2 1.00 7348 3178
## prjthreat_rural.ses.med3 1.00 5921 3117
## prjthreat_rural.ses.med4 1.00 5281 3018
## prjharm_rural.ses.med2 1.00 6829 2852
## prjharm_rural.ses.med3 1.00 6551 3218
## prjharm_rural.ses.med4 1.00 6081 3172
## prjusedrg_rural.ses.med2 1.00 6440 3359
## prjusedrg_rural.ses.med3 1.00 6861 2997
## prjusedrg_rural.ses.med4 1.00 5189 2991
## prjhack_rural.ses.med2 1.00 6492 3124
## prjhack_rural.ses.med3 1.00 6585 3161
## prjhack_rural.ses.med4 1.00 6793 3321
## prjthflt5_mostresp_devx2 1.00 5075 3604
## prjthflt5_mostresp_av12x2 1.00 3707 3121
## prjthflt5_mostresp_devx2:rural.ses.med2 1.00 4583 2947
## prjthflt5_mostresp_devx2:rural.ses.med3 1.00 3391 2431
## prjthflt5_mostresp_devx2:rural.ses.med4 1.00 3714 3072
## prjthfgt5_mostresp_devx2 1.00 4346 3332
## prjthfgt5_mostresp_av12x2 1.00 4220 2905
## prjthfgt5_mostresp_devx2:rural.ses.med2 1.00 4126 3322
## prjthfgt5_mostresp_devx2:rural.ses.med3 1.00 3296 2240
## prjthfgt5_mostresp_devx2:rural.ses.med4 1.00 3964 3211
## prjthreat_mostresp_devx2 1.00 5506 3376
## prjthreat_mostresp_av12x2 1.00 4326 3381
## prjthreat_mostresp_devx2:rural.ses.med2 1.00 4796 3615
## prjthreat_mostresp_devx2:rural.ses.med3 1.00 4995 2902
## prjthreat_mostresp_devx2:rural.ses.med4 1.00 4969 3107
## prjharm_mostresp_devx2 1.00 6330 3064
## prjharm_mostresp_av12x2 1.00 5739 2985
## prjharm_mostresp_devx2:rural.ses.med2 1.00 5170 3230
## prjharm_mostresp_devx2:rural.ses.med3 1.00 4952 2588
## prjharm_mostresp_devx2:rural.ses.med4 1.00 4876 2713
## prjusedrg_mostresp_devx2 1.00 6695 3251
## prjusedrg_mostresp_av12x2 1.00 4841 3191
## prjusedrg_mostresp_devx2:rural.ses.med2 1.00 4576 3025
## prjusedrg_mostresp_devx2:rural.ses.med3 1.00 4965 3337
## prjusedrg_mostresp_devx2:rural.ses.med4 1.00 4387 3363
## prjhack_mostresp_devx2 1.00 6532 2758
## prjhack_mostresp_av12x2 1.00 7325 2726
## prjhack_mostresp_devx2:rural.ses.med2 1.00 4916 2981
## prjhack_mostresp_devx2:rural.ses.med3 1.00 5207 2921
## prjhack_mostresp_devx2:rural.ses.med4 1.00 4905 2592
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## prjthflt5_mostresp_devx21[1] 0.29 0.16 0.04
## prjthflt5_mostresp_devx21[2] 0.25 0.14 0.04
## prjthflt5_mostresp_devx21[3] 0.21 0.13 0.03
## prjthflt5_mostresp_devx21[4] 0.25 0.15 0.04
## prjthflt5_mostresp_av12x21[1] 0.12 0.08 0.02
## prjthflt5_mostresp_av12x21[2] 0.12 0.08 0.02
## prjthflt5_mostresp_av12x21[3] 0.12 0.08 0.02
## prjthflt5_mostresp_av12x21[4] 0.12 0.08 0.02
## prjthflt5_mostresp_av12x21[5] 0.12 0.08 0.02
## prjthflt5_mostresp_av12x21[6] 0.13 0.08 0.02
## prjthflt5_mostresp_av12x21[7] 0.13 0.08 0.02
## prjthflt5_mostresp_av12x21[8] 0.12 0.08 0.01
## prjthflt5_mostresp_devx2:rural.ses.med21[1] 0.31 0.20 0.02
## prjthflt5_mostresp_devx2:rural.ses.med21[2] 0.26 0.19 0.01
## prjthflt5_mostresp_devx2:rural.ses.med21[3] 0.17 0.16 0.01
## prjthflt5_mostresp_devx2:rural.ses.med21[4] 0.26 0.19 0.01
## prjthflt5_mostresp_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjthflt5_mostresp_devx2:rural.ses.med31[2] 0.23 0.19 0.01
## prjthflt5_mostresp_devx2:rural.ses.med31[3] 0.22 0.18 0.01
## prjthflt5_mostresp_devx2:rural.ses.med31[4] 0.30 0.22 0.01
## prjthflt5_mostresp_devx2:rural.ses.med41[1] 0.34 0.22 0.02
## prjthflt5_mostresp_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## prjthflt5_mostresp_devx2:rural.ses.med41[3] 0.20 0.17 0.01
## prjthflt5_mostresp_devx2:rural.ses.med41[4] 0.25 0.19 0.01
## prjthfgt5_mostresp_devx21[1] 0.28 0.15 0.04
## prjthfgt5_mostresp_devx21[2] 0.25 0.14 0.04
## prjthfgt5_mostresp_devx21[3] 0.22 0.13 0.03
## prjthfgt5_mostresp_devx21[4] 0.26 0.15 0.04
## prjthfgt5_mostresp_av12x21[1] 0.13 0.08 0.02
## prjthfgt5_mostresp_av12x21[2] 0.13 0.08 0.02
## prjthfgt5_mostresp_av12x21[3] 0.12 0.08 0.02
## prjthfgt5_mostresp_av12x21[4] 0.12 0.08 0.02
## prjthfgt5_mostresp_av12x21[5] 0.12 0.08 0.02
## prjthfgt5_mostresp_av12x21[6] 0.13 0.08 0.01
## prjthfgt5_mostresp_av12x21[7] 0.13 0.08 0.02
## prjthfgt5_mostresp_av12x21[8] 0.12 0.08 0.02
## prjthfgt5_mostresp_devx2:rural.ses.med21[1] 0.33 0.21 0.02
## prjthfgt5_mostresp_devx2:rural.ses.med21[2] 0.25 0.19 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med21[3] 0.16 0.15 0.00
## prjthfgt5_mostresp_devx2:rural.ses.med21[4] 0.26 0.19 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med31[3] 0.22 0.18 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med31[4] 0.30 0.22 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med41[1] 0.32 0.22 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med41[3] 0.21 0.17 0.01
## prjthfgt5_mostresp_devx2:rural.ses.med41[4] 0.26 0.19 0.01
## prjthreat_mostresp_devx21[1] 0.27 0.15 0.04
## prjthreat_mostresp_devx21[2] 0.31 0.16 0.05
## prjthreat_mostresp_devx21[3] 0.20 0.13 0.03
## prjthreat_mostresp_devx21[4] 0.22 0.13 0.03
## prjthreat_mostresp_av12x21[1] 0.11 0.07 0.01
## prjthreat_mostresp_av12x21[2] 0.11 0.08 0.01
## prjthreat_mostresp_av12x21[3] 0.12 0.08 0.02
## prjthreat_mostresp_av12x21[4] 0.13 0.08 0.02
## prjthreat_mostresp_av12x21[5] 0.13 0.08 0.02
## prjthreat_mostresp_av12x21[6] 0.13 0.08 0.02
## prjthreat_mostresp_av12x21[7] 0.14 0.09 0.02
## prjthreat_mostresp_av12x21[8] 0.12 0.08 0.02
## prjthreat_mostresp_devx2:rural.ses.med21[1] 0.29 0.21 0.01
## prjthreat_mostresp_devx2:rural.ses.med21[2] 0.21 0.18 0.01
## prjthreat_mostresp_devx2:rural.ses.med21[3] 0.22 0.18 0.01
## prjthreat_mostresp_devx2:rural.ses.med21[4] 0.28 0.21 0.01
## prjthreat_mostresp_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjthreat_mostresp_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjthreat_mostresp_devx2:rural.ses.med31[3] 0.23 0.18 0.01
## prjthreat_mostresp_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjthreat_mostresp_devx2:rural.ses.med41[1] 0.30 0.21 0.01
## prjthreat_mostresp_devx2:rural.ses.med41[2] 0.19 0.17 0.00
## prjthreat_mostresp_devx2:rural.ses.med41[3] 0.24 0.19 0.01
## prjthreat_mostresp_devx2:rural.ses.med41[4] 0.27 0.20 0.01
## prjharm_mostresp_devx21[1] 0.27 0.15 0.04
## prjharm_mostresp_devx21[2] 0.27 0.15 0.04
## prjharm_mostresp_devx21[3] 0.22 0.14 0.03
## prjharm_mostresp_devx21[4] 0.25 0.14 0.04
## prjharm_mostresp_av12x21[1] 0.13 0.08 0.02
## prjharm_mostresp_av12x21[2] 0.12 0.08 0.02
## prjharm_mostresp_av12x21[3] 0.12 0.08 0.02
## prjharm_mostresp_av12x21[4] 0.12 0.08 0.01
## prjharm_mostresp_av12x21[5] 0.13 0.08 0.02
## prjharm_mostresp_av12x21[6] 0.13 0.08 0.02
## prjharm_mostresp_av12x21[7] 0.12 0.08 0.02
## prjharm_mostresp_av12x21[8] 0.12 0.08 0.02
## prjharm_mostresp_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## prjharm_mostresp_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## prjharm_mostresp_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## prjharm_mostresp_devx2:rural.ses.med21[4] 0.29 0.21 0.01
## prjharm_mostresp_devx2:rural.ses.med31[1] 0.24 0.18 0.01
## prjharm_mostresp_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## prjharm_mostresp_devx2:rural.ses.med31[3] 0.26 0.20 0.01
## prjharm_mostresp_devx2:rural.ses.med31[4] 0.28 0.21 0.01
## prjharm_mostresp_devx2:rural.ses.med41[1] 0.21 0.18 0.01
## prjharm_mostresp_devx2:rural.ses.med41[2] 0.26 0.19 0.01
## prjharm_mostresp_devx2:rural.ses.med41[3] 0.24 0.19 0.01
## prjharm_mostresp_devx2:rural.ses.med41[4] 0.30 0.21 0.01
## prjusedrg_mostresp_devx21[1] 0.24 0.14 0.03
## prjusedrg_mostresp_devx21[2] 0.30 0.16 0.05
## prjusedrg_mostresp_devx21[3] 0.25 0.14 0.04
## prjusedrg_mostresp_devx21[4] 0.21 0.13 0.03
## prjusedrg_mostresp_av12x21[1] 0.11 0.07 0.02
## prjusedrg_mostresp_av12x21[2] 0.11 0.07 0.02
## prjusedrg_mostresp_av12x21[3] 0.12 0.08 0.02
## prjusedrg_mostresp_av12x21[4] 0.13 0.08 0.02
## prjusedrg_mostresp_av12x21[5] 0.13 0.08 0.02
## prjusedrg_mostresp_av12x21[6] 0.13 0.08 0.02
## prjusedrg_mostresp_av12x21[7] 0.13 0.08 0.02
## prjusedrg_mostresp_av12x21[8] 0.13 0.08 0.02
## prjusedrg_mostresp_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## prjusedrg_mostresp_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## prjusedrg_mostresp_devx2:rural.ses.med21[3] 0.23 0.19 0.01
## prjusedrg_mostresp_devx2:rural.ses.med21[4] 0.29 0.21 0.01
## prjusedrg_mostresp_devx2:rural.ses.med31[1] 0.26 0.19 0.01
## prjusedrg_mostresp_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## prjusedrg_mostresp_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## prjusedrg_mostresp_devx2:rural.ses.med31[4] 0.27 0.20 0.01
## prjusedrg_mostresp_devx2:rural.ses.med41[1] 0.29 0.21 0.01
## prjusedrg_mostresp_devx2:rural.ses.med41[2] 0.22 0.18 0.01
## prjusedrg_mostresp_devx2:rural.ses.med41[3] 0.21 0.18 0.01
## prjusedrg_mostresp_devx2:rural.ses.med41[4] 0.28 0.20 0.01
## prjhack_mostresp_devx21[1] 0.28 0.15 0.05
## prjhack_mostresp_devx21[2] 0.25 0.14 0.04
## prjhack_mostresp_devx21[3] 0.22 0.13 0.03
## prjhack_mostresp_devx21[4] 0.25 0.14 0.04
## prjhack_mostresp_av12x21[1] 0.12 0.08 0.01
## prjhack_mostresp_av12x21[2] 0.12 0.08 0.02
## prjhack_mostresp_av12x21[3] 0.13 0.08 0.02
## prjhack_mostresp_av12x21[4] 0.13 0.08 0.02
## prjhack_mostresp_av12x21[5] 0.13 0.08 0.02
## prjhack_mostresp_av12x21[6] 0.13 0.08 0.02
## prjhack_mostresp_av12x21[7] 0.12 0.08 0.02
## prjhack_mostresp_av12x21[8] 0.12 0.08 0.02
## prjhack_mostresp_devx2:rural.ses.med21[1] 0.28 0.21 0.01
## prjhack_mostresp_devx2:rural.ses.med21[2] 0.23 0.18 0.01
## prjhack_mostresp_devx2:rural.ses.med21[3] 0.21 0.18 0.01
## prjhack_mostresp_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## prjhack_mostresp_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjhack_mostresp_devx2:rural.ses.med31[2] 0.22 0.17 0.01
## prjhack_mostresp_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## prjhack_mostresp_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjhack_mostresp_devx2:rural.ses.med41[1] 0.24 0.18 0.01
## prjhack_mostresp_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjhack_mostresp_devx2:rural.ses.med41[3] 0.24 0.19 0.01
## prjhack_mostresp_devx2:rural.ses.med41[4] 0.29 0.22 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## prjthflt5_mostresp_devx21[1] 0.63 1.00 9154 2478
## prjthflt5_mostresp_devx21[2] 0.57 1.00 7334 2726
## prjthflt5_mostresp_devx21[3] 0.52 1.00 6066 2701
## prjthflt5_mostresp_devx21[4] 0.58 1.00 6865 2863
## prjthflt5_mostresp_av12x21[1] 0.32 1.00 7475 2530
## prjthflt5_mostresp_av12x21[2] 0.32 1.00 7053 2672
## prjthflt5_mostresp_av12x21[3] 0.32 1.00 7789 2197
## prjthflt5_mostresp_av12x21[4] 0.31 1.00 6980 2287
## prjthflt5_mostresp_av12x21[5] 0.31 1.00 6792 2839
## prjthflt5_mostresp_av12x21[6] 0.33 1.00 7107 2525
## prjthflt5_mostresp_av12x21[7] 0.34 1.00 6866 2588
## prjthflt5_mostresp_av12x21[8] 0.30 1.00 6664 2637
## prjthflt5_mostresp_devx2:rural.ses.med21[1] 0.75 1.00 5198 2686
## prjthflt5_mostresp_devx2:rural.ses.med21[2] 0.70 1.00 5956 3323
## prjthflt5_mostresp_devx2:rural.ses.med21[3] 0.58 1.00 5994 2958
## prjthflt5_mostresp_devx2:rural.ses.med21[4] 0.70 1.00 6580 2919
## prjthflt5_mostresp_devx2:rural.ses.med31[1] 0.71 1.00 7050 2959
## prjthflt5_mostresp_devx2:rural.ses.med31[2] 0.67 1.00 4606 2687
## prjthflt5_mostresp_devx2:rural.ses.med31[3] 0.66 1.00 5466 3039
## prjthflt5_mostresp_devx2:rural.ses.med31[4] 0.79 1.00 5243 3067
## prjthflt5_mostresp_devx2:rural.ses.med41[1] 0.79 1.00 4941 2971
## prjthflt5_mostresp_devx2:rural.ses.med41[2] 0.64 1.00 5178 2566
## prjthflt5_mostresp_devx2:rural.ses.med41[3] 0.64 1.00 5863 3221
## prjthflt5_mostresp_devx2:rural.ses.med41[4] 0.69 1.00 6350 3193
## prjthfgt5_mostresp_devx21[1] 0.62 1.00 7978 3064
## prjthfgt5_mostresp_devx21[2] 0.57 1.00 7508 2887
## prjthfgt5_mostresp_devx21[3] 0.54 1.00 6437 2691
## prjthfgt5_mostresp_devx21[4] 0.58 1.00 7618 2885
## prjthfgt5_mostresp_av12x21[1] 0.32 1.00 7136 2544
## prjthfgt5_mostresp_av12x21[2] 0.32 1.00 7187 2240
## prjthfgt5_mostresp_av12x21[3] 0.32 1.00 7399 2657
## prjthfgt5_mostresp_av12x21[4] 0.32 1.00 8048 2155
## prjthfgt5_mostresp_av12x21[5] 0.31 1.00 6843 2229
## prjthfgt5_mostresp_av12x21[6] 0.32 1.00 7562 2642
## prjthfgt5_mostresp_av12x21[7] 0.33 1.00 7990 3154
## prjthfgt5_mostresp_av12x21[8] 0.31 1.00 8315 3202
## prjthfgt5_mostresp_devx2:rural.ses.med21[1] 0.77 1.00 6171 2031
## prjthfgt5_mostresp_devx2:rural.ses.med21[2] 0.69 1.00 6493 2717
## prjthfgt5_mostresp_devx2:rural.ses.med21[3] 0.58 1.00 5000 3071
## prjthfgt5_mostresp_devx2:rural.ses.med21[4] 0.69 1.00 5744 2610
## prjthfgt5_mostresp_devx2:rural.ses.med31[1] 0.69 1.00 5858 2880
## prjthfgt5_mostresp_devx2:rural.ses.med31[2] 0.66 1.00 5596 3149
## prjthfgt5_mostresp_devx2:rural.ses.med31[3] 0.68 1.00 5517 2941
## prjthfgt5_mostresp_devx2:rural.ses.med31[4] 0.81 1.00 4338 2976
## prjthfgt5_mostresp_devx2:rural.ses.med41[1] 0.78 1.00 4775 2515
## prjthfgt5_mostresp_devx2:rural.ses.med41[2] 0.63 1.00 5194 2657
## prjthfgt5_mostresp_devx2:rural.ses.med41[3] 0.65 1.00 5166 2544
## prjthfgt5_mostresp_devx2:rural.ses.med41[4] 0.70 1.00 6128 2480
## prjthreat_mostresp_devx21[1] 0.60 1.00 7540 3161
## prjthreat_mostresp_devx21[2] 0.64 1.00 6386 3092
## prjthreat_mostresp_devx21[3] 0.51 1.00 6788 2890
## prjthreat_mostresp_devx21[4] 0.52 1.00 6974 2966
## prjthreat_mostresp_av12x21[1] 0.29 1.00 6125 2202
## prjthreat_mostresp_av12x21[2] 0.30 1.00 7083 2620
## prjthreat_mostresp_av12x21[3] 0.31 1.00 7375 2500
## prjthreat_mostresp_av12x21[4] 0.31 1.00 8346 2334
## prjthreat_mostresp_av12x21[5] 0.32 1.00 6351 2406
## prjthreat_mostresp_av12x21[6] 0.33 1.00 7491 2612
## prjthreat_mostresp_av12x21[7] 0.35 1.00 8362 3018
## prjthreat_mostresp_av12x21[8] 0.32 1.00 8222 2946
## prjthreat_mostresp_devx2:rural.ses.med21[1] 0.75 1.00 5809 2858
## prjthreat_mostresp_devx2:rural.ses.med21[2] 0.66 1.00 5739 2640
## prjthreat_mostresp_devx2:rural.ses.med21[3] 0.68 1.00 5887 3014
## prjthreat_mostresp_devx2:rural.ses.med21[4] 0.74 1.00 6841 3177
## prjthreat_mostresp_devx2:rural.ses.med31[1] 0.71 1.00 6779 2257
## prjthreat_mostresp_devx2:rural.ses.med31[2] 0.68 1.00 6508 2729
## prjthreat_mostresp_devx2:rural.ses.med31[3] 0.67 1.00 6086 2826
## prjthreat_mostresp_devx2:rural.ses.med31[4] 0.75 1.00 5862 2814
## prjthreat_mostresp_devx2:rural.ses.med41[1] 0.75 1.00 5608 2892
## prjthreat_mostresp_devx2:rural.ses.med41[2] 0.63 1.00 5244 2601
## prjthreat_mostresp_devx2:rural.ses.med41[3] 0.69 1.00 5915 2789
## prjthreat_mostresp_devx2:rural.ses.med41[4] 0.74 1.00 6154 2721
## prjharm_mostresp_devx21[1] 0.61 1.00 9660 2911
## prjharm_mostresp_devx21[2] 0.59 1.00 6998 2850
## prjharm_mostresp_devx21[3] 0.53 1.00 6081 3074
## prjharm_mostresp_devx21[4] 0.58 1.00 8571 2594
## prjharm_mostresp_av12x21[1] 0.33 1.00 6975 2494
## prjharm_mostresp_av12x21[2] 0.31 1.00 7097 2829
## prjharm_mostresp_av12x21[3] 0.31 1.00 6280 2605
## prjharm_mostresp_av12x21[4] 0.32 1.00 6753 2063
## prjharm_mostresp_av12x21[5] 0.32 1.00 7475 2435
## prjharm_mostresp_av12x21[6] 0.33 1.00 6764 2299
## prjharm_mostresp_av12x21[7] 0.32 1.00 6989 2697
## prjharm_mostresp_av12x21[8] 0.32 1.00 6676 2919
## prjharm_mostresp_devx2:rural.ses.med21[1] 0.70 1.00 6710 2517
## prjharm_mostresp_devx2:rural.ses.med21[2] 0.65 1.00 7791 2842
## prjharm_mostresp_devx2:rural.ses.med21[3] 0.67 1.00 5417 2653
## prjharm_mostresp_devx2:rural.ses.med21[4] 0.74 1.00 6808 2917
## prjharm_mostresp_devx2:rural.ses.med31[1] 0.68 1.00 6020 2032
## prjharm_mostresp_devx2:rural.ses.med31[2] 0.67 1.00 6985 1966
## prjharm_mostresp_devx2:rural.ses.med31[3] 0.73 1.00 5591 2058
## prjharm_mostresp_devx2:rural.ses.med31[4] 0.75 1.00 5183 2573
## prjharm_mostresp_devx2:rural.ses.med41[1] 0.66 1.00 5940 2587
## prjharm_mostresp_devx2:rural.ses.med41[2] 0.69 1.00 5440 3241
## prjharm_mostresp_devx2:rural.ses.med41[3] 0.69 1.00 5900 3299
## prjharm_mostresp_devx2:rural.ses.med41[4] 0.75 1.00 5886 2777
## prjusedrg_mostresp_devx21[1] 0.56 1.00 8715 2788
## prjusedrg_mostresp_devx21[2] 0.64 1.00 6918 3073
## prjusedrg_mostresp_devx21[3] 0.58 1.00 6591 2796
## prjusedrg_mostresp_devx21[4] 0.50 1.00 5760 3295
## prjusedrg_mostresp_av12x21[1] 0.29 1.00 6423 2941
## prjusedrg_mostresp_av12x21[2] 0.30 1.00 7150 2488
## prjusedrg_mostresp_av12x21[3] 0.33 1.00 7809 2572
## prjusedrg_mostresp_av12x21[4] 0.32 1.00 7945 2167
## prjusedrg_mostresp_av12x21[5] 0.33 1.00 8770 2708
## prjusedrg_mostresp_av12x21[6] 0.33 1.00 7442 2807
## prjusedrg_mostresp_av12x21[7] 0.33 1.00 6867 2908
## prjusedrg_mostresp_av12x21[8] 0.33 1.00 6774 2965
## prjusedrg_mostresp_devx2:rural.ses.med21[1] 0.71 1.00 5559 2837
## prjusedrg_mostresp_devx2:rural.ses.med21[2] 0.67 1.00 6754 2625
## prjusedrg_mostresp_devx2:rural.ses.med21[3] 0.68 1.00 5754 2849
## prjusedrg_mostresp_devx2:rural.ses.med21[4] 0.75 1.00 5706 2702
## prjusedrg_mostresp_devx2:rural.ses.med31[1] 0.69 1.00 5958 2306
## prjusedrg_mostresp_devx2:rural.ses.med31[2] 0.67 1.00 6353 2564
## prjusedrg_mostresp_devx2:rural.ses.med31[3] 0.70 1.00 7529 2723
## prjusedrg_mostresp_devx2:rural.ses.med31[4] 0.73 1.00 5067 2423
## prjusedrg_mostresp_devx2:rural.ses.med41[1] 0.75 1.00 5288 3064
## prjusedrg_mostresp_devx2:rural.ses.med41[2] 0.66 1.00 5702 2185
## prjusedrg_mostresp_devx2:rural.ses.med41[3] 0.66 1.00 5463 2930
## prjusedrg_mostresp_devx2:rural.ses.med41[4] 0.74 1.00 7234 2423
## prjhack_mostresp_devx21[1] 0.61 1.00 7886 2596
## prjhack_mostresp_devx21[2] 0.58 1.00 8771 2616
## prjhack_mostresp_devx21[3] 0.53 1.00 6280 2638
## prjhack_mostresp_devx21[4] 0.57 1.00 9261 2612
## prjhack_mostresp_av12x21[1] 0.33 1.00 8418 2529
## prjhack_mostresp_av12x21[2] 0.31 1.00 7856 2613
## prjhack_mostresp_av12x21[3] 0.32 1.00 7598 2270
## prjhack_mostresp_av12x21[4] 0.32 1.00 6664 2475
## prjhack_mostresp_av12x21[5] 0.34 1.00 7253 2916
## prjhack_mostresp_av12x21[6] 0.32 1.00 8170 3056
## prjhack_mostresp_av12x21[7] 0.31 1.00 6846 2685
## prjhack_mostresp_av12x21[8] 0.32 1.00 7851 2850
## prjhack_mostresp_devx2:rural.ses.med21[1] 0.75 1.00 6781 3025
## prjhack_mostresp_devx2:rural.ses.med21[2] 0.67 1.00 6129 2772
## prjhack_mostresp_devx2:rural.ses.med21[3] 0.67 1.00 4784 2526
## prjhack_mostresp_devx2:rural.ses.med21[4] 0.73 1.00 5866 2797
## prjhack_mostresp_devx2:rural.ses.med31[1] 0.69 1.00 5999 2527
## prjhack_mostresp_devx2:rural.ses.med31[2] 0.65 1.00 6514 2874
## prjhack_mostresp_devx2:rural.ses.med31[3] 0.69 1.00 6554 2318
## prjhack_mostresp_devx2:rural.ses.med31[4] 0.76 1.00 6088 2658
## prjhack_mostresp_devx2:rural.ses.med41[1] 0.68 1.00 8089 2794
## prjhack_mostresp_devx2:rural.ses.med41[2] 0.67 1.00 6180 2247
## prjhack_mostresp_devx2:rural.ses.med41[3] 0.68 1.00 6258 2752
## prjhack_mostresp_devx2:rural.ses.med41[4] 0.77 1.00 4775 2725
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.1.3.4 Prior summary
out.chg.prjcrime.stresp.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## group resp dpar nlpar lb ub source
## default
## prjhack user
## prjhack user
## prjhack user
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjharm user
## prjharm user
## prjharm user
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthflt5 user
## prjthflt5 user
## prjthflt5 user
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthreat user
## prjthreat user
## prjthreat user
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjusedrg user
## prjusedrg user
## prjusedrg user
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## default
## prjhack user
## prjharm user
## prjthfgt5 user
## prjthflt5 user
## prjthreat user
## prjusedrg user
## prjhack 0 default
## prjharm 0 default
## prjthfgt5 0 default
## prjthflt5 0 default
## prjthreat 0 default
## prjusedrg 0 default
## id prjhack 0 (vectorized)
## id prjhack 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjusedrg 0 (vectorized)
## id prjusedrg 0 (vectorized)
## prjhack user
## prjhack default
## prjhack default
## prjhack default
## prjhack user
## prjharm user
## prjharm default
## prjharm default
## prjharm default
## prjharm user
## prjthfgt5 user
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 user
## prjthflt5 user
## prjthflt5 default
## prjthflt5 default
## prjthflt5 default
## prjthflt5 user
## prjthreat user
## prjthreat default
## prjthreat default
## prjthreat default
## prjthreat user
## prjusedrg user
## prjusedrg default
## prjusedrg default
## prjusedrg default
## prjusedrg user8.1.1.4 Corr X Community: stfair & prj crime
#Community Change: criminal intent items ~ mo(stfair)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostfair_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostfair_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostfair_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostfair_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.stfair.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stfair_devx2) + mo(stfair_av12x2) +
rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stfair_comm_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stfair.comm.fit <- ppchecks(chg.prjcrime.stfair.comm.fit)8.1.1.4.1 Coefficient plot (intervals)
out.chg.prjcrime.stfair.comm.fit[[10]]
8.1.1.4.2 PPcheck (density)
p1 <- out.chg.prjcrime.stfair.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stfair.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stfair.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stfair.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stfair.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stfair.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
8.1.1.4.3 Fit summary
out.chg.prjcrime.stfair.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## prjthreat ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## prjharm ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## prjusedrg ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## prjhack ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.80 0.55 2.83 4.99 1.00 1670
## sd(prjthfgt5_Intercept) 3.33 0.49 2.46 4.39 1.00 1580
## sd(prjthreat_Intercept) 3.15 0.55 2.15 4.31 1.00 1914
## sd(prjharm_Intercept) 3.05 0.57 2.03 4.26 1.00 1727
## sd(prjusedrg_Intercept) 2.90 0.54 1.93 4.05 1.00 2099
## sd(prjhack_Intercept) 0.81 0.53 0.03 1.99 1.00 1016
## Tail_ESS
## sd(prjthflt5_Intercept) 2291
## sd(prjthfgt5_Intercept) 2400
## sd(prjthreat_Intercept) 2922
## sd(prjharm_Intercept) 2424
## sd(prjusedrg_Intercept) 2746
## sd(prjhack_Intercept) 2101
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_Intercept -6.36 0.87 -8.13 -4.72
## prjthfgt5_Intercept -5.83 0.83 -7.58 -4.32
## prjthreat_Intercept -6.46 0.96 -8.46 -4.69
## prjharm_Intercept -5.91 0.96 -7.82 -4.12
## prjusedrg_Intercept -6.04 0.96 -8.04 -4.24
## prjhack_Intercept -4.41 0.75 -5.96 -3.02
## prjthflt5_rural.ses.med2 -0.47 0.89 -2.22 1.31
## prjthflt5_rural.ses.med3 0.61 0.78 -1.00 2.05
## prjthflt5_rural.ses.med4 1.51 0.90 -0.37 3.12
## prjthfgt5_rural.ses.med2 -0.50 0.92 -2.23 1.31
## prjthfgt5_rural.ses.med3 0.85 0.77 -0.75 2.29
## prjthfgt5_rural.ses.med4 1.00 0.84 -0.77 2.53
## prjthreat_rural.ses.med2 -0.29 0.89 -2.01 1.54
## prjthreat_rural.ses.med3 0.35 0.77 -1.16 1.87
## prjthreat_rural.ses.med4 1.25 0.84 -0.45 2.81
## prjharm_rural.ses.med2 -0.20 0.82 -1.81 1.47
## prjharm_rural.ses.med3 0.02 0.76 -1.48 1.52
## prjharm_rural.ses.med4 1.21 0.79 -0.44 2.72
## prjusedrg_rural.ses.med2 0.01 0.85 -1.62 1.69
## prjusedrg_rural.ses.med3 -0.23 0.81 -1.84 1.40
## prjusedrg_rural.ses.med4 1.00 0.83 -0.72 2.55
## prjhack_rural.ses.med2 -0.58 0.87 -2.27 1.17
## prjhack_rural.ses.med3 0.23 0.73 -1.21 1.70
## prjhack_rural.ses.med4 0.93 0.72 -0.54 2.29
## prjthflt5_mostfair_devx2 -0.06 0.20 -0.47 0.33
## prjthflt5_mostfair_av12x2 0.13 0.08 -0.03 0.29
## prjthflt5_mostfair_devx2:rural.ses.med2 -0.65 0.53 -1.72 0.34
## prjthflt5_mostfair_devx2:rural.ses.med3 0.47 0.34 -0.18 1.16
## prjthflt5_mostfair_devx2:rural.ses.med4 0.29 0.48 -0.66 1.26
## prjthfgt5_mostfair_devx2 -0.00 0.20 -0.41 0.38
## prjthfgt5_mostfair_av12x2 0.07 0.08 -0.09 0.22
## prjthfgt5_mostfair_devx2:rural.ses.med2 -0.74 0.53 -1.85 0.26
## prjthfgt5_mostfair_devx2:rural.ses.med3 0.27 0.35 -0.41 0.98
## prjthfgt5_mostfair_devx2:rural.ses.med4 0.52 0.39 -0.21 1.31
## prjthreat_mostfair_devx2 -0.29 0.21 -0.71 0.12
## prjthreat_mostfair_av12x2 0.10 0.08 -0.06 0.27
## prjthreat_mostfair_devx2:rural.ses.med2 -0.65 0.63 -1.99 0.52
## prjthreat_mostfair_devx2:rural.ses.med3 0.17 0.42 -0.68 1.00
## prjthreat_mostfair_devx2:rural.ses.med4 0.29 0.45 -0.59 1.21
## prjharm_mostfair_devx2 -0.24 0.21 -0.65 0.17
## prjharm_mostfair_av12x2 0.01 0.08 -0.16 0.17
## prjharm_mostfair_devx2:rural.ses.med2 -0.31 0.54 -1.42 0.70
## prjharm_mostfair_devx2:rural.ses.med3 0.18 0.42 -0.67 1.00
## prjharm_mostfair_devx2:rural.ses.med4 -0.34 0.63 -1.71 0.77
## prjusedrg_mostfair_devx2 -0.26 0.21 -0.67 0.14
## prjusedrg_mostfair_av12x2 0.06 0.08 -0.11 0.22
## prjusedrg_mostfair_devx2:rural.ses.med2 -0.67 0.59 -1.93 0.42
## prjusedrg_mostfair_devx2:rural.ses.med3 -0.36 0.53 -1.50 0.63
## prjusedrg_mostfair_devx2:rural.ses.med4 0.49 0.41 -0.28 1.30
## prjhack_mostfair_devx2 -0.19 0.21 -0.60 0.21
## prjhack_mostfair_av12x2 0.11 0.07 -0.04 0.25
## prjhack_mostfair_devx2:rural.ses.med2 -0.38 0.59 -1.58 0.75
## prjhack_mostfair_devx2:rural.ses.med3 -0.15 0.45 -1.13 0.66
## prjhack_mostfair_devx2:rural.ses.med4 -0.17 0.46 -1.17 0.68
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1.00 3999 3296
## prjthfgt5_Intercept 1.00 3179 3054
## prjthreat_Intercept 1.00 4240 3337
## prjharm_Intercept 1.00 3620 3303
## prjusedrg_Intercept 1.00 3637 2958
## prjhack_Intercept 1.00 3703 3205
## prjthflt5_rural.ses.med2 1.00 6674 3385
## prjthflt5_rural.ses.med3 1.00 5542 3268
## prjthflt5_rural.ses.med4 1.00 4498 3439
## prjthfgt5_rural.ses.med2 1.00 6482 2984
## prjthfgt5_rural.ses.med3 1.00 5460 3016
## prjthfgt5_rural.ses.med4 1.00 5572 3420
## prjthreat_rural.ses.med2 1.00 7523 2432
## prjthreat_rural.ses.med3 1.00 5458 2983
## prjthreat_rural.ses.med4 1.00 5738 3621
## prjharm_rural.ses.med2 1.00 5872 2851
## prjharm_rural.ses.med3 1.00 6422 2882
## prjharm_rural.ses.med4 1.00 6392 3187
## prjusedrg_rural.ses.med2 1.00 6317 2670
## prjusedrg_rural.ses.med3 1.00 6743 3032
## prjusedrg_rural.ses.med4 1.00 5964 3103
## prjhack_rural.ses.med2 1.00 6020 3564
## prjhack_rural.ses.med3 1.00 6498 3023
## prjhack_rural.ses.med4 1.00 5644 3265
## prjthflt5_mostfair_devx2 1.00 5487 3055
## prjthflt5_mostfair_av12x2 1.00 4493 3082
## prjthflt5_mostfair_devx2:rural.ses.med2 1.00 4497 3343
## prjthflt5_mostfair_devx2:rural.ses.med3 1.00 3888 2921
## prjthflt5_mostfair_devx2:rural.ses.med4 1.00 3104 3074
## prjthfgt5_mostfair_devx2 1.00 5593 3450
## prjthfgt5_mostfair_av12x2 1.00 4145 3356
## prjthfgt5_mostfair_devx2:rural.ses.med2 1.00 4338 3163
## prjthfgt5_mostfair_devx2:rural.ses.med3 1.00 4218 2838
## prjthfgt5_mostfair_devx2:rural.ses.med4 1.00 3976 2435
## prjthreat_mostfair_devx2 1.00 7202 3484
## prjthreat_mostfair_av12x2 1.00 3828 2520
## prjthreat_mostfair_devx2:rural.ses.med2 1.00 5170 3050
## prjthreat_mostfair_devx2:rural.ses.med3 1.00 5215 3347
## prjthreat_mostfair_devx2:rural.ses.med4 1.00 4788 3503
## prjharm_mostfair_devx2 1.00 5806 3134
## prjharm_mostfair_av12x2 1.00 4734 3657
## prjharm_mostfair_devx2:rural.ses.med2 1.00 4388 2835
## prjharm_mostfair_devx2:rural.ses.med3 1.00 4791 3323
## prjharm_mostfair_devx2:rural.ses.med4 1.00 3691 3618
## prjusedrg_mostfair_devx2 1.00 6133 3450
## prjusedrg_mostfair_av12x2 1.00 5193 3234
## prjusedrg_mostfair_devx2:rural.ses.med2 1.00 5809 3187
## prjusedrg_mostfair_devx2:rural.ses.med3 1.00 4857 3467
## prjusedrg_mostfair_devx2:rural.ses.med4 1.00 4442 2855
## prjhack_mostfair_devx2 1.00 6864 3319
## prjhack_mostfair_av12x2 1.00 8366 3238
## prjhack_mostfair_devx2:rural.ses.med2 1.00 4441 3307
## prjhack_mostfair_devx2:rural.ses.med3 1.00 5376 2876
## prjhack_mostfair_devx2:rural.ses.med4 1.00 4785 2908
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## prjthflt5_mostfair_devx21[1] 0.27 0.15 0.04
## prjthflt5_mostfair_devx21[2] 0.24 0.14 0.04
## prjthflt5_mostfair_devx21[3] 0.24 0.14 0.04
## prjthflt5_mostfair_devx21[4] 0.26 0.14 0.04
## prjthflt5_mostfair_av12x21[1] 0.12 0.08 0.02
## prjthflt5_mostfair_av12x21[2] 0.12 0.08 0.02
## prjthflt5_mostfair_av12x21[3] 0.13 0.08 0.02
## prjthflt5_mostfair_av12x21[4] 0.12 0.08 0.02
## prjthflt5_mostfair_av12x21[5] 0.12 0.08 0.02
## prjthflt5_mostfair_av12x21[6] 0.13 0.08 0.02
## prjthflt5_mostfair_av12x21[7] 0.13 0.09 0.02
## prjthflt5_mostfair_av12x21[8] 0.13 0.08 0.02
## prjthflt5_mostfair_devx2:rural.ses.med21[1] 0.32 0.21 0.01
## prjthflt5_mostfair_devx2:rural.ses.med21[2] 0.21 0.17 0.01
## prjthflt5_mostfair_devx2:rural.ses.med21[3] 0.21 0.17 0.01
## prjthflt5_mostfair_devx2:rural.ses.med21[4] 0.26 0.19 0.01
## prjthflt5_mostfair_devx2:rural.ses.med31[1] 0.26 0.19 0.01
## prjthflt5_mostfair_devx2:rural.ses.med31[2] 0.26 0.19 0.01
## prjthflt5_mostfair_devx2:rural.ses.med31[3] 0.24 0.18 0.01
## prjthflt5_mostfair_devx2:rural.ses.med31[4] 0.24 0.18 0.01
## prjthflt5_mostfair_devx2:rural.ses.med41[1] 0.34 0.23 0.01
## prjthflt5_mostfair_devx2:rural.ses.med41[2] 0.20 0.17 0.00
## prjthflt5_mostfair_devx2:rural.ses.med41[3] 0.19 0.17 0.00
## prjthflt5_mostfair_devx2:rural.ses.med41[4] 0.27 0.20 0.01
## prjthfgt5_mostfair_devx21[1] 0.26 0.15 0.04
## prjthfgt5_mostfair_devx21[2] 0.24 0.14 0.03
## prjthfgt5_mostfair_devx21[3] 0.24 0.14 0.03
## prjthfgt5_mostfair_devx21[4] 0.26 0.15 0.03
## prjthfgt5_mostfair_av12x21[1] 0.13 0.08 0.02
## prjthfgt5_mostfair_av12x21[2] 0.12 0.08 0.02
## prjthfgt5_mostfair_av12x21[3] 0.13 0.08 0.02
## prjthfgt5_mostfair_av12x21[4] 0.12 0.08 0.02
## prjthfgt5_mostfair_av12x21[5] 0.12 0.08 0.01
## prjthfgt5_mostfair_av12x21[6] 0.13 0.08 0.02
## prjthfgt5_mostfair_av12x21[7] 0.13 0.08 0.02
## prjthfgt5_mostfair_av12x21[8] 0.12 0.08 0.02
## prjthfgt5_mostfair_devx2:rural.ses.med21[1] 0.32 0.21 0.02
## prjthfgt5_mostfair_devx2:rural.ses.med21[2] 0.23 0.18 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med21[3] 0.19 0.17 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med21[4] 0.25 0.19 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med31[1] 0.29 0.20 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med31[3] 0.23 0.18 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med31[4] 0.27 0.20 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med41[1] 0.32 0.21 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med41[2] 0.25 0.19 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med41[3] 0.19 0.16 0.01
## prjthfgt5_mostfair_devx2:rural.ses.med41[4] 0.24 0.19 0.01
## prjthreat_mostfair_devx21[1] 0.26 0.14 0.04
## prjthreat_mostfair_devx21[2] 0.26 0.14 0.04
## prjthreat_mostfair_devx21[3] 0.26 0.15 0.04
## prjthreat_mostfair_devx21[4] 0.23 0.14 0.03
## prjthreat_mostfair_av12x21[1] 0.12 0.08 0.01
## prjthreat_mostfair_av12x21[2] 0.12 0.08 0.01
## prjthreat_mostfair_av12x21[3] 0.12 0.08 0.02
## prjthreat_mostfair_av12x21[4] 0.13 0.08 0.02
## prjthreat_mostfair_av12x21[5] 0.13 0.08 0.02
## prjthreat_mostfair_av12x21[6] 0.12 0.08 0.02
## prjthreat_mostfair_av12x21[7] 0.13 0.08 0.02
## prjthreat_mostfair_av12x21[8] 0.13 0.08 0.02
## prjthreat_mostfair_devx2:rural.ses.med21[1] 0.28 0.20 0.01
## prjthreat_mostfair_devx2:rural.ses.med21[2] 0.20 0.17 0.01
## prjthreat_mostfair_devx2:rural.ses.med21[3] 0.26 0.19 0.01
## prjthreat_mostfair_devx2:rural.ses.med21[4] 0.26 0.20 0.01
## prjthreat_mostfair_devx2:rural.ses.med31[1] 0.24 0.19 0.01
## prjthreat_mostfair_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## prjthreat_mostfair_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## prjthreat_mostfair_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjthreat_mostfair_devx2:rural.ses.med41[1] 0.31 0.21 0.01
## prjthreat_mostfair_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## prjthreat_mostfair_devx2:rural.ses.med41[3] 0.20 0.17 0.01
## prjthreat_mostfair_devx2:rural.ses.med41[4] 0.28 0.21 0.01
## prjharm_mostfair_devx21[1] 0.26 0.14 0.04
## prjharm_mostfair_devx21[2] 0.27 0.15 0.04
## prjharm_mostfair_devx21[3] 0.23 0.14 0.04
## prjharm_mostfair_devx21[4] 0.24 0.14 0.03
## prjharm_mostfair_av12x21[1] 0.13 0.08 0.02
## prjharm_mostfair_av12x21[2] 0.13 0.08 0.02
## prjharm_mostfair_av12x21[3] 0.13 0.08 0.02
## prjharm_mostfair_av12x21[4] 0.12 0.08 0.02
## prjharm_mostfair_av12x21[5] 0.13 0.08 0.02
## prjharm_mostfair_av12x21[6] 0.12 0.08 0.02
## prjharm_mostfair_av12x21[7] 0.12 0.08 0.02
## prjharm_mostfair_av12x21[8] 0.13 0.08 0.02
## prjharm_mostfair_devx2:rural.ses.med21[1] 0.26 0.20 0.01
## prjharm_mostfair_devx2:rural.ses.med21[2] 0.23 0.18 0.01
## prjharm_mostfair_devx2:rural.ses.med21[3] 0.23 0.19 0.01
## prjharm_mostfair_devx2:rural.ses.med21[4] 0.28 0.21 0.01
## prjharm_mostfair_devx2:rural.ses.med31[1] 0.23 0.18 0.01
## prjharm_mostfair_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjharm_mostfair_devx2:rural.ses.med31[3] 0.26 0.20 0.01
## prjharm_mostfair_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjharm_mostfair_devx2:rural.ses.med41[1] 0.20 0.19 0.01
## prjharm_mostfair_devx2:rural.ses.med41[2] 0.18 0.17 0.01
## prjharm_mostfair_devx2:rural.ses.med41[3] 0.34 0.24 0.01
## prjharm_mostfair_devx2:rural.ses.med41[4] 0.27 0.20 0.01
## prjusedrg_mostfair_devx21[1] 0.26 0.14 0.04
## prjusedrg_mostfair_devx21[2] 0.27 0.15 0.05
## prjusedrg_mostfair_devx21[3] 0.24 0.14 0.04
## prjusedrg_mostfair_devx21[4] 0.23 0.14 0.04
## prjusedrg_mostfair_av12x21[1] 0.12 0.08 0.01
## prjusedrg_mostfair_av12x21[2] 0.12 0.08 0.02
## prjusedrg_mostfair_av12x21[3] 0.13 0.08 0.02
## prjusedrg_mostfair_av12x21[4] 0.12 0.08 0.02
## prjusedrg_mostfair_av12x21[5] 0.12 0.08 0.02
## prjusedrg_mostfair_av12x21[6] 0.13 0.08 0.02
## prjusedrg_mostfair_av12x21[7] 0.13 0.08 0.02
## prjusedrg_mostfair_av12x21[8] 0.13 0.08 0.02
## prjusedrg_mostfair_devx2:rural.ses.med21[1] 0.24 0.18 0.01
## prjusedrg_mostfair_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## prjusedrg_mostfair_devx2:rural.ses.med21[3] 0.28 0.21 0.01
## prjusedrg_mostfair_devx2:rural.ses.med21[4] 0.25 0.20 0.01
## prjusedrg_mostfair_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## prjusedrg_mostfair_devx2:rural.ses.med31[2] 0.21 0.18 0.01
## prjusedrg_mostfair_devx2:rural.ses.med31[3] 0.24 0.18 0.01
## prjusedrg_mostfair_devx2:rural.ses.med31[4] 0.29 0.20 0.01
## prjusedrg_mostfair_devx2:rural.ses.med41[1] 0.31 0.20 0.01
## prjusedrg_mostfair_devx2:rural.ses.med41[2] 0.19 0.16 0.01
## prjusedrg_mostfair_devx2:rural.ses.med41[3] 0.25 0.18 0.01
## prjusedrg_mostfair_devx2:rural.ses.med41[4] 0.25 0.19 0.01
## prjhack_mostfair_devx21[1] 0.25 0.14 0.04
## prjhack_mostfair_devx21[2] 0.28 0.15 0.05
## prjhack_mostfair_devx21[3] 0.23 0.14 0.03
## prjhack_mostfair_devx21[4] 0.24 0.14 0.04
## prjhack_mostfair_av12x21[1] 0.12 0.07 0.02
## prjhack_mostfair_av12x21[2] 0.11 0.08 0.02
## prjhack_mostfair_av12x21[3] 0.12 0.08 0.02
## prjhack_mostfair_av12x21[4] 0.13 0.08 0.02
## prjhack_mostfair_av12x21[5] 0.13 0.08 0.02
## prjhack_mostfair_av12x21[6] 0.13 0.08 0.02
## prjhack_mostfair_av12x21[7] 0.14 0.08 0.02
## prjhack_mostfair_av12x21[8] 0.12 0.08 0.02
## prjhack_mostfair_devx2:rural.ses.med21[1] 0.28 0.21 0.01
## prjhack_mostfair_devx2:rural.ses.med21[2] 0.23 0.18 0.01
## prjhack_mostfair_devx2:rural.ses.med21[3] 0.21 0.18 0.01
## prjhack_mostfair_devx2:rural.ses.med21[4] 0.28 0.21 0.01
## prjhack_mostfair_devx2:rural.ses.med31[1] 0.24 0.19 0.01
## prjhack_mostfair_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjhack_mostfair_devx2:rural.ses.med31[3] 0.25 0.19 0.01
## prjhack_mostfair_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## prjhack_mostfair_devx2:rural.ses.med41[1] 0.23 0.19 0.01
## prjhack_mostfair_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjhack_mostfair_devx2:rural.ses.med41[3] 0.25 0.19 0.01
## prjhack_mostfair_devx2:rural.ses.med41[4] 0.30 0.21 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## prjthflt5_mostfair_devx21[1] 0.61 1.00 7172 3012
## prjthflt5_mostfair_devx21[2] 0.55 1.00 7996 2392
## prjthflt5_mostfair_devx21[3] 0.55 1.00 5642 3074
## prjthflt5_mostfair_devx21[4] 0.59 1.00 7561 2662
## prjthflt5_mostfair_av12x21[1] 0.30 1.00 7528 2418
## prjthflt5_mostfair_av12x21[2] 0.32 1.00 6980 2608
## prjthflt5_mostfair_av12x21[3] 0.34 1.00 6756 2689
## prjthflt5_mostfair_av12x21[4] 0.32 1.00 8018 2608
## prjthflt5_mostfair_av12x21[5] 0.31 1.00 8666 2408
## prjthflt5_mostfair_av12x21[6] 0.32 1.00 8205 2710
## prjthflt5_mostfair_av12x21[7] 0.34 1.00 7521 2315
## prjthflt5_mostfair_av12x21[8] 0.32 1.00 7822 2799
## prjthflt5_mostfair_devx2:rural.ses.med21[1] 0.77 1.00 5506 2721
## prjthflt5_mostfair_devx2:rural.ses.med21[2] 0.63 1.00 6374 2988
## prjthflt5_mostfair_devx2:rural.ses.med21[3] 0.65 1.00 5709 2665
## prjthflt5_mostfair_devx2:rural.ses.med21[4] 0.70 1.00 5948 2562
## prjthflt5_mostfair_devx2:rural.ses.med31[1] 0.69 1.00 6344 2098
## prjthflt5_mostfair_devx2:rural.ses.med31[2] 0.70 1.00 5772 2904
## prjthflt5_mostfair_devx2:rural.ses.med31[3] 0.67 1.00 5296 3139
## prjthflt5_mostfair_devx2:rural.ses.med31[4] 0.66 1.00 5917 3113
## prjthflt5_mostfair_devx2:rural.ses.med41[1] 0.81 1.00 3988 3139
## prjthflt5_mostfair_devx2:rural.ses.med41[2] 0.64 1.00 6269 2603
## prjthflt5_mostfair_devx2:rural.ses.med41[3] 0.65 1.00 3974 3114
## prjthflt5_mostfair_devx2:rural.ses.med41[4] 0.74 1.00 6863 2789
## prjthfgt5_mostfair_devx21[1] 0.60 1.00 8350 3135
## prjthfgt5_mostfair_devx21[2] 0.57 1.00 8041 2507
## prjthfgt5_mostfair_devx21[3] 0.55 1.00 5874 2993
## prjthfgt5_mostfair_devx21[4] 0.60 1.00 7180 2885
## prjthfgt5_mostfair_av12x21[1] 0.32 1.00 7468 2611
## prjthfgt5_mostfair_av12x21[2] 0.32 1.00 7262 2624
## prjthfgt5_mostfair_av12x21[3] 0.33 1.00 8718 2472
## prjthfgt5_mostfair_av12x21[4] 0.31 1.00 9022 2951
## prjthfgt5_mostfair_av12x21[5] 0.32 1.00 7091 2247
## prjthfgt5_mostfair_av12x21[6] 0.33 1.00 8352 2507
## prjthfgt5_mostfair_av12x21[7] 0.33 1.00 7549 2929
## prjthfgt5_mostfair_av12x21[8] 0.31 1.00 7344 2514
## prjthfgt5_mostfair_devx2:rural.ses.med21[1] 0.76 1.00 5205 2881
## prjthfgt5_mostfair_devx2:rural.ses.med21[2] 0.65 1.00 6197 2681
## prjthfgt5_mostfair_devx2:rural.ses.med21[3] 0.62 1.00 4973 2506
## prjthfgt5_mostfair_devx2:rural.ses.med21[4] 0.69 1.00 6398 3110
## prjthfgt5_mostfair_devx2:rural.ses.med31[1] 0.72 1.00 5339 2238
## prjthfgt5_mostfair_devx2:rural.ses.med31[2] 0.66 1.00 5488 2304
## prjthfgt5_mostfair_devx2:rural.ses.med31[3] 0.66 1.00 4484 2560
## prjthfgt5_mostfair_devx2:rural.ses.med31[4] 0.72 1.00 5946 3265
## prjthfgt5_mostfair_devx2:rural.ses.med41[1] 0.77 1.00 5874 2898
## prjthfgt5_mostfair_devx2:rural.ses.med41[2] 0.70 1.00 5340 2750
## prjthfgt5_mostfair_devx2:rural.ses.med41[3] 0.60 1.00 4572 2863
## prjthfgt5_mostfair_devx2:rural.ses.med41[4] 0.68 1.00 5506 2543
## prjthreat_mostfair_devx21[1] 0.57 1.00 7690 2722
## prjthreat_mostfair_devx21[2] 0.59 1.00 7924 2641
## prjthreat_mostfair_devx21[3] 0.59 1.00 6343 2880
## prjthreat_mostfair_devx21[4] 0.55 1.00 7360 2792
## prjthreat_mostfair_av12x21[1] 0.31 1.00 8133 2153
## prjthreat_mostfair_av12x21[2] 0.31 1.00 7143 2705
## prjthreat_mostfair_av12x21[3] 0.30 1.00 8027 2775
## prjthreat_mostfair_av12x21[4] 0.32 1.00 8298 2857
## prjthreat_mostfair_av12x21[5] 0.32 1.00 7705 2501
## prjthreat_mostfair_av12x21[6] 0.31 1.00 7974 2392
## prjthreat_mostfair_av12x21[7] 0.33 1.00 6863 2786
## prjthreat_mostfair_av12x21[8] 0.33 1.00 7579 2891
## prjthreat_mostfair_devx2:rural.ses.med21[1] 0.72 1.00 6985 2896
## prjthreat_mostfair_devx2:rural.ses.med21[2] 0.63 1.00 5791 2503
## prjthreat_mostfair_devx2:rural.ses.med21[3] 0.72 1.00 5556 2967
## prjthreat_mostfair_devx2:rural.ses.med21[4] 0.72 1.00 7165 2933
## prjthreat_mostfair_devx2:rural.ses.med31[1] 0.70 1.00 7494 2536
## prjthreat_mostfair_devx2:rural.ses.med31[2] 0.66 1.00 6000 2669
## prjthreat_mostfair_devx2:rural.ses.med31[3] 0.68 1.00 6256 2780
## prjthreat_mostfair_devx2:rural.ses.med31[4] 0.75 1.00 5518 2137
## prjthreat_mostfair_devx2:rural.ses.med41[1] 0.76 1.00 5375 3029
## prjthreat_mostfair_devx2:rural.ses.med41[2] 0.64 1.00 7141 2953
## prjthreat_mostfair_devx2:rural.ses.med41[3] 0.64 1.00 5444 2639
## prjthreat_mostfair_devx2:rural.ses.med41[4] 0.75 1.00 6510 3089
## prjharm_mostfair_devx21[1] 0.58 1.00 8373 2572
## prjharm_mostfair_devx21[2] 0.60 1.00 7218 3010
## prjharm_mostfair_devx21[3] 0.55 1.00 7208 2975
## prjharm_mostfair_devx21[4] 0.56 1.00 7543 2522
## prjharm_mostfair_av12x21[1] 0.32 1.00 8117 2266
## prjharm_mostfair_av12x21[2] 0.32 1.00 7382 2454
## prjharm_mostfair_av12x21[3] 0.32 1.00 7898 2294
## prjharm_mostfair_av12x21[4] 0.32 1.00 6672 2623
## prjharm_mostfair_av12x21[5] 0.32 1.00 9053 2197
## prjharm_mostfair_av12x21[6] 0.32 1.00 7779 2554
## prjharm_mostfair_av12x21[7] 0.32 1.00 6979 2735
## prjharm_mostfair_av12x21[8] 0.31 1.00 7635 2743
## prjharm_mostfair_devx2:rural.ses.med21[1] 0.71 1.00 6545 2296
## prjharm_mostfair_devx2:rural.ses.med21[2] 0.67 1.00 6001 2140
## prjharm_mostfair_devx2:rural.ses.med21[3] 0.69 1.00 7221 2556
## prjharm_mostfair_devx2:rural.ses.med21[4] 0.76 1.00 5761 2874
## prjharm_mostfair_devx2:rural.ses.med31[1] 0.67 1.00 5962 2735
## prjharm_mostfair_devx2:rural.ses.med31[2] 0.66 1.00 5505 2730
## prjharm_mostfair_devx2:rural.ses.med31[3] 0.72 1.00 5373 3091
## prjharm_mostfair_devx2:rural.ses.med31[4] 0.75 1.00 6028 2732
## prjharm_mostfair_devx2:rural.ses.med41[1] 0.69 1.00 4291 2855
## prjharm_mostfair_devx2:rural.ses.med41[2] 0.61 1.00 5401 2469
## prjharm_mostfair_devx2:rural.ses.med41[3] 0.82 1.00 4116 3112
## prjharm_mostfair_devx2:rural.ses.med41[4] 0.73 1.00 6247 2474
## prjusedrg_mostfair_devx21[1] 0.58 1.00 8122 2454
## prjusedrg_mostfair_devx21[2] 0.60 1.00 6442 2629
## prjusedrg_mostfair_devx21[3] 0.55 1.00 6832 2964
## prjusedrg_mostfair_devx21[4] 0.56 1.00 9087 2914
## prjusedrg_mostfair_av12x21[1] 0.31 1.00 6645 2585
## prjusedrg_mostfair_av12x21[2] 0.31 1.00 7727 2734
## prjusedrg_mostfair_av12x21[3] 0.31 1.00 7163 2729
## prjusedrg_mostfair_av12x21[4] 0.33 1.00 8291 2492
## prjusedrg_mostfair_av12x21[5] 0.31 1.00 6391 2226
## prjusedrg_mostfair_av12x21[6] 0.32 1.00 8761 1923
## prjusedrg_mostfair_av12x21[7] 0.33 1.00 6031 2718
## prjusedrg_mostfair_av12x21[8] 0.33 1.00 7082 2761
## prjusedrg_mostfair_devx2:rural.ses.med21[1] 0.66 1.00 6192 2684
## prjusedrg_mostfair_devx2:rural.ses.med21[2] 0.66 1.00 5928 2567
## prjusedrg_mostfair_devx2:rural.ses.med21[3] 0.75 1.00 5875 2832
## prjusedrg_mostfair_devx2:rural.ses.med21[4] 0.73 1.00 6310 2444
## prjusedrg_mostfair_devx2:rural.ses.med31[1] 0.73 1.00 5660 2683
## prjusedrg_mostfair_devx2:rural.ses.med31[2] 0.64 1.00 6207 2564
## prjusedrg_mostfair_devx2:rural.ses.med31[3] 0.69 1.00 5770 2951
## prjusedrg_mostfair_devx2:rural.ses.med31[4] 0.74 1.00 5640 2531
## prjusedrg_mostfair_devx2:rural.ses.med41[1] 0.73 1.00 4844 2360
## prjusedrg_mostfair_devx2:rural.ses.med41[2] 0.59 1.00 5941 2926
## prjusedrg_mostfair_devx2:rural.ses.med41[3] 0.67 1.00 5423 2759
## prjusedrg_mostfair_devx2:rural.ses.med41[4] 0.69 1.00 6488 2785
## prjhack_mostfair_devx21[1] 0.57 1.00 8361 2647
## prjhack_mostfair_devx21[2] 0.60 1.00 7196 3402
## prjhack_mostfair_devx21[3] 0.55 1.00 7610 2795
## prjhack_mostfair_devx21[4] 0.58 1.00 8012 2621
## prjhack_mostfair_av12x21[1] 0.29 1.00 7715 2559
## prjhack_mostfair_av12x21[2] 0.30 1.00 7962 2507
## prjhack_mostfair_av12x21[3] 0.31 1.00 7685 2420
## prjhack_mostfair_av12x21[4] 0.32 1.00 7285 2252
## prjhack_mostfair_av12x21[5] 0.33 1.00 7823 2554
## prjhack_mostfair_av12x21[6] 0.34 1.00 6769 2619
## prjhack_mostfair_av12x21[7] 0.33 1.00 8073 3088
## prjhack_mostfair_av12x21[8] 0.31 1.00 7843 2592
## prjhack_mostfair_devx2:rural.ses.med21[1] 0.75 1.00 5748 2982
## prjhack_mostfair_devx2:rural.ses.med21[2] 0.67 1.00 6008 2459
## prjhack_mostfair_devx2:rural.ses.med21[3] 0.65 1.00 5171 2716
## prjhack_mostfair_devx2:rural.ses.med21[4] 0.74 1.00 6470 2834
## prjhack_mostfair_devx2:rural.ses.med31[1] 0.69 1.00 5229 2118
## prjhack_mostfair_devx2:rural.ses.med31[2] 0.65 1.00 6580 2820
## prjhack_mostfair_devx2:rural.ses.med31[3] 0.69 1.00 6846 2994
## prjhack_mostfair_devx2:rural.ses.med31[4] 0.76 1.00 5423 2331
## prjhack_mostfair_devx2:rural.ses.med41[1] 0.69 1.00 5855 2568
## prjhack_mostfair_devx2:rural.ses.med41[2] 0.66 1.00 6786 2762
## prjhack_mostfair_devx2:rural.ses.med41[3] 0.70 1.00 5622 2821
## prjhack_mostfair_devx2:rural.ses.med41[4] 0.77 1.00 5533 3169
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.1.4.4 Prior summary
out.chg.prjcrime.stfair.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## group resp dpar nlpar lb ub source
## default
## prjhack user
## prjhack user
## prjhack user
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjharm user
## prjharm user
## prjharm user
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthflt5 user
## prjthflt5 user
## prjthflt5 user
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthreat user
## prjthreat user
## prjthreat user
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjusedrg user
## prjusedrg user
## prjusedrg user
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## default
## prjhack user
## prjharm user
## prjthfgt5 user
## prjthflt5 user
## prjthreat user
## prjusedrg user
## prjhack 0 default
## prjharm 0 default
## prjthfgt5 0 default
## prjthflt5 0 default
## prjthreat 0 default
## prjusedrg 0 default
## id prjhack 0 (vectorized)
## id prjhack 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjusedrg 0 (vectorized)
## id prjusedrg 0 (vectorized)
## prjhack user
## prjhack default
## prjhack default
## prjhack default
## prjhack user
## prjharm user
## prjharm default
## prjharm default
## prjharm default
## prjharm user
## prjthfgt5 user
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 user
## prjthflt5 user
## prjthflt5 default
## prjthflt5 default
## prjthflt5 default
## prjthflt5 user
## prjthreat user
## prjthreat default
## prjthreat default
## prjthreat default
## prjthreat user
## prjusedrg user
## prjusedrg default
## prjusedrg default
## prjusedrg default
## prjusedrg user8.1.1.5 Corr X Community: stjob & prj crime
#Community Change: criminal intent items ~ mo(stjob)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostjob_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostjob_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostjob_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostjob_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.stjob.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stjob_devx2) + mo(stjob_av12x2) +
rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stjob_comm_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stjob.comm.fit <- ppchecks(chg.prjcrime.stjob.comm.fit)8.1.1.5.1 Coefficient plot (intervals)
out.chg.prjcrime.stjob.comm.fit[[10]]
8.1.1.5.2 PPcheck (density)
p1 <- out.chg.prjcrime.stjob.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stjob.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stjob.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stjob.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stjob.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stjob.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
8.1.1.5.3 Fit summary
out.chg.prjcrime.stjob.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## prjthreat ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## prjharm ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## prjusedrg ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## prjhack ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.92 0.58 2.90 5.16 1.00 1589
## sd(prjthfgt5_Intercept) 3.22 0.49 2.33 4.28 1.00 1296
## sd(prjthreat_Intercept) 2.86 0.50 1.97 3.95 1.00 1743
## sd(prjharm_Intercept) 2.90 0.54 1.97 4.03 1.00 1742
## sd(prjusedrg_Intercept) 2.61 0.51 1.69 3.70 1.00 1668
## sd(prjhack_Intercept) 0.60 0.41 0.03 1.50 1.00 1568
## Tail_ESS
## sd(prjthflt5_Intercept) 2531
## sd(prjthfgt5_Intercept) 2606
## sd(prjthreat_Intercept) 2538
## sd(prjharm_Intercept) 2841
## sd(prjusedrg_Intercept) 2432
## sd(prjhack_Intercept) 1960
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_Intercept -6.22 0.86 -8.00 -4.62
## prjthfgt5_Intercept -5.83 0.81 -7.50 -4.32
## prjthreat_Intercept -6.93 0.92 -8.89 -5.17
## prjharm_Intercept -6.40 0.92 -8.27 -4.72
## prjusedrg_Intercept -6.50 0.91 -8.36 -4.83
## prjhack_Intercept -5.44 0.75 -6.98 -4.04
## prjthflt5_rural.ses.med2 -0.18 0.84 -1.78 1.49
## prjthflt5_rural.ses.med3 0.50 0.80 -1.12 2.01
## prjthflt5_rural.ses.med4 1.27 0.91 -0.63 2.97
## prjthfgt5_rural.ses.med2 -0.48 0.90 -2.25 1.35
## prjthfgt5_rural.ses.med3 0.33 0.78 -1.29 1.81
## prjthfgt5_rural.ses.med4 1.24 0.92 -0.62 2.91
## prjthreat_rural.ses.med2 -0.39 0.85 -2.05 1.30
## prjthreat_rural.ses.med3 0.41 0.76 -1.06 1.90
## prjthreat_rural.ses.med4 1.18 0.87 -0.63 2.81
## prjharm_rural.ses.med2 -0.13 0.83 -1.67 1.55
## prjharm_rural.ses.med3 0.41 0.75 -1.06 1.87
## prjharm_rural.ses.med4 1.05 0.79 -0.61 2.56
## prjusedrg_rural.ses.med2 -0.32 0.84 -1.95 1.34
## prjusedrg_rural.ses.med3 -0.15 0.81 -1.74 1.50
## prjusedrg_rural.ses.med4 1.39 0.84 -0.38 2.90
## prjhack_rural.ses.med2 -0.46 0.87 -2.14 1.32
## prjhack_rural.ses.med3 0.34 0.70 -1.02 1.68
## prjhack_rural.ses.med4 0.76 0.74 -0.72 2.17
## prjthflt5_mostjob_devx2 -0.14 0.20 -0.54 0.26
## prjthflt5_mostjob_av12x2 0.11 0.08 -0.04 0.26
## prjthflt5_mostjob_devx2:rural.ses.med2 -1.06 0.58 -2.30 -0.02
## prjthflt5_mostjob_devx2:rural.ses.med3 0.58 0.36 -0.07 1.35
## prjthflt5_mostjob_devx2:rural.ses.med4 0.47 0.45 -0.48 1.34
## prjthfgt5_mostjob_devx2 -0.09 0.21 -0.50 0.32
## prjthfgt5_mostjob_av12x2 0.12 0.07 -0.03 0.26
## prjthfgt5_mostjob_devx2:rural.ses.med2 -0.80 0.59 -2.06 0.26
## prjthfgt5_mostjob_devx2:rural.ses.med3 0.57 0.33 -0.06 1.27
## prjthfgt5_mostjob_devx2:rural.ses.med4 0.35 0.47 -0.64 1.20
## prjthreat_mostjob_devx2 -0.10 0.22 -0.55 0.33
## prjthreat_mostjob_av12x2 0.18 0.08 0.02 0.34
## prjthreat_mostjob_devx2:rural.ses.med2 -0.32 0.56 -1.48 0.70
## prjthreat_mostjob_devx2:rural.ses.med3 0.17 0.40 -0.66 0.93
## prjthreat_mostjob_devx2:rural.ses.med4 0.29 0.48 -0.79 1.13
## prjharm_mostjob_devx2 -0.02 0.21 -0.45 0.39
## prjharm_mostjob_av12x2 0.09 0.08 -0.07 0.24
## prjharm_mostjob_devx2:rural.ses.med2 -0.37 0.51 -1.48 0.55
## prjharm_mostjob_devx2:rural.ses.med3 -0.22 0.44 -1.16 0.59
## prjharm_mostjob_devx2:rural.ses.med4 -0.19 0.51 -1.26 0.75
## prjusedrg_mostjob_devx2 -0.04 0.22 -0.48 0.37
## prjusedrg_mostjob_av12x2 0.15 0.08 -0.00 0.31
## prjusedrg_mostjob_devx2:rural.ses.med2 -0.23 0.53 -1.37 0.73
## prjusedrg_mostjob_devx2:rural.ses.med3 -0.41 0.50 -1.48 0.46
## prjusedrg_mostjob_devx2:rural.ses.med4 0.13 0.44 -0.77 1.01
## prjhack_mostjob_devx2 0.06 0.21 -0.36 0.46
## prjhack_mostjob_av12x2 0.25 0.08 0.11 0.41
## prjhack_mostjob_devx2:rural.ses.med2 -0.42 0.54 -1.54 0.60
## prjhack_mostjob_devx2:rural.ses.med3 -0.29 0.43 -1.23 0.46
## prjhack_mostjob_devx2:rural.ses.med4 -0.24 0.52 -1.40 0.63
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1.00 3355 3280
## prjthfgt5_Intercept 1.00 3102 3276
## prjthreat_Intercept 1.00 3413 2970
## prjharm_Intercept 1.00 3525 3186
## prjusedrg_Intercept 1.00 3157 2598
## prjhack_Intercept 1.00 5651 3380
## prjthflt5_rural.ses.med2 1.00 7986 3376
## prjthflt5_rural.ses.med3 1.00 5928 3653
## prjthflt5_rural.ses.med4 1.00 7190 3581
## prjthfgt5_rural.ses.med2 1.00 7126 3234
## prjthfgt5_rural.ses.med3 1.00 6418 3238
## prjthfgt5_rural.ses.med4 1.00 4431 3283
## prjthreat_rural.ses.med2 1.00 7163 2936
## prjthreat_rural.ses.med3 1.00 5726 3179
## prjthreat_rural.ses.med4 1.00 5891 3365
## prjharm_rural.ses.med2 1.00 7175 3095
## prjharm_rural.ses.med3 1.00 7362 3010
## prjharm_rural.ses.med4 1.00 6451 2840
## prjusedrg_rural.ses.med2 1.00 8150 3263
## prjusedrg_rural.ses.med3 1.00 8809 2849
## prjusedrg_rural.ses.med4 1.00 5306 3275
## prjhack_rural.ses.med2 1.00 7370 3480
## prjhack_rural.ses.med3 1.00 6609 3487
## prjhack_rural.ses.med4 1.00 6965 3568
## prjthflt5_mostjob_devx2 1.00 6505 2872
## prjthflt5_mostjob_av12x2 1.00 3554 3274
## prjthflt5_mostjob_devx2:rural.ses.med2 1.00 5597 3232
## prjthflt5_mostjob_devx2:rural.ses.med3 1.00 4421 3039
## prjthflt5_mostjob_devx2:rural.ses.med4 1.00 3992 3345
## prjthfgt5_mostjob_devx2 1.00 5293 3214
## prjthfgt5_mostjob_av12x2 1.00 4889 3268
## prjthfgt5_mostjob_devx2:rural.ses.med2 1.00 5691 3185
## prjthfgt5_mostjob_devx2:rural.ses.med3 1.00 4415 3051
## prjthfgt5_mostjob_devx2:rural.ses.med4 1.00 2838 3208
## prjthreat_mostjob_devx2 1.00 6347 3266
## prjthreat_mostjob_av12x2 1.00 4538 2869
## prjthreat_mostjob_devx2:rural.ses.med2 1.00 4482 3386
## prjthreat_mostjob_devx2:rural.ses.med3 1.00 4519 2743
## prjthreat_mostjob_devx2:rural.ses.med4 1.00 3678 3203
## prjharm_mostjob_devx2 1.00 5800 3127
## prjharm_mostjob_av12x2 1.00 5471 3487
## prjharm_mostjob_devx2:rural.ses.med2 1.00 4334 3047
## prjharm_mostjob_devx2:rural.ses.med3 1.00 4962 2989
## prjharm_mostjob_devx2:rural.ses.med4 1.00 4757 3014
## prjusedrg_mostjob_devx2 1.00 6055 3250
## prjusedrg_mostjob_av12x2 1.00 5431 3513
## prjusedrg_mostjob_devx2:rural.ses.med2 1.00 5104 2930
## prjusedrg_mostjob_devx2:rural.ses.med3 1.00 5230 3003
## prjusedrg_mostjob_devx2:rural.ses.med4 1.00 4123 3195
## prjhack_mostjob_devx2 1.00 7167 3182
## prjhack_mostjob_av12x2 1.00 7811 2648
## prjhack_mostjob_devx2:rural.ses.med2 1.00 5452 3191
## prjhack_mostjob_devx2:rural.ses.med3 1.00 5895 3351
## prjhack_mostjob_devx2:rural.ses.med4 1.00 5117 3142
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_mostjob_devx21[1] 0.26 0.15 0.04 0.59
## prjthflt5_mostjob_devx21[2] 0.24 0.14 0.04 0.55
## prjthflt5_mostjob_devx21[3] 0.26 0.14 0.04 0.59
## prjthflt5_mostjob_devx21[4] 0.24 0.14 0.04 0.55
## prjthflt5_mostjob_av12x21[1] 0.11 0.07 0.02 0.30
## prjthflt5_mostjob_av12x21[2] 0.12 0.08 0.02 0.31
## prjthflt5_mostjob_av12x21[3] 0.12 0.08 0.02 0.30
## prjthflt5_mostjob_av12x21[4] 0.12 0.08 0.02 0.32
## prjthflt5_mostjob_av12x21[5] 0.13 0.08 0.02 0.33
## prjthflt5_mostjob_av12x21[6] 0.14 0.09 0.02 0.35
## prjthflt5_mostjob_av12x21[7] 0.13 0.08 0.02 0.32
## prjthflt5_mostjob_av12x21[8] 0.13 0.08 0.02 0.33
## prjthflt5_mostjob_devx2:rural.ses.med21[1] 0.24 0.18 0.01 0.67
## prjthflt5_mostjob_devx2:rural.ses.med21[2] 0.24 0.18 0.01 0.68
## prjthflt5_mostjob_devx2:rural.ses.med21[3] 0.30 0.20 0.02 0.72
## prjthflt5_mostjob_devx2:rural.ses.med21[4] 0.22 0.17 0.01 0.64
## prjthflt5_mostjob_devx2:rural.ses.med31[1] 0.27 0.19 0.01 0.70
## prjthflt5_mostjob_devx2:rural.ses.med31[2] 0.27 0.19 0.01 0.71
## prjthflt5_mostjob_devx2:rural.ses.med31[3] 0.20 0.16 0.01 0.59
## prjthflt5_mostjob_devx2:rural.ses.med31[4] 0.26 0.19 0.01 0.67
## prjthflt5_mostjob_devx2:rural.ses.med41[1] 0.36 0.23 0.02 0.81
## prjthflt5_mostjob_devx2:rural.ses.med41[2] 0.23 0.18 0.01 0.69
## prjthflt5_mostjob_devx2:rural.ses.med41[3] 0.17 0.16 0.01 0.59
## prjthflt5_mostjob_devx2:rural.ses.med41[4] 0.24 0.18 0.01 0.68
## prjthfgt5_mostjob_devx21[1] 0.26 0.14 0.04 0.57
## prjthfgt5_mostjob_devx21[2] 0.23 0.14 0.03 0.55
## prjthfgt5_mostjob_devx21[3] 0.26 0.15 0.04 0.59
## prjthfgt5_mostjob_devx21[4] 0.25 0.15 0.04 0.59
## prjthfgt5_mostjob_av12x21[1] 0.12 0.08 0.01 0.31
## prjthfgt5_mostjob_av12x21[2] 0.12 0.08 0.02 0.31
## prjthfgt5_mostjob_av12x21[3] 0.12 0.08 0.02 0.29
## prjthfgt5_mostjob_av12x21[4] 0.12 0.08 0.02 0.31
## prjthfgt5_mostjob_av12x21[5] 0.13 0.08 0.02 0.32
## prjthfgt5_mostjob_av12x21[6] 0.14 0.09 0.02 0.34
## prjthfgt5_mostjob_av12x21[7] 0.12 0.08 0.02 0.31
## prjthfgt5_mostjob_av12x21[8] 0.13 0.09 0.02 0.34
## prjthfgt5_mostjob_devx2:rural.ses.med21[1] 0.30 0.20 0.01 0.75
## prjthfgt5_mostjob_devx2:rural.ses.med21[2] 0.19 0.16 0.01 0.61
## prjthfgt5_mostjob_devx2:rural.ses.med21[3] 0.27 0.20 0.01 0.72
## prjthfgt5_mostjob_devx2:rural.ses.med21[4] 0.24 0.19 0.01 0.68
## prjthfgt5_mostjob_devx2:rural.ses.med31[1] 0.25 0.18 0.01 0.66
## prjthfgt5_mostjob_devx2:rural.ses.med31[2] 0.30 0.19 0.02 0.73
## prjthfgt5_mostjob_devx2:rural.ses.med31[3] 0.19 0.15 0.01 0.58
## prjthfgt5_mostjob_devx2:rural.ses.med31[4] 0.26 0.18 0.01 0.65
## prjthfgt5_mostjob_devx2:rural.ses.med41[1] 0.31 0.22 0.01 0.78
## prjthfgt5_mostjob_devx2:rural.ses.med41[2] 0.26 0.20 0.01 0.70
## prjthfgt5_mostjob_devx2:rural.ses.med41[3] 0.18 0.17 0.00 0.65
## prjthfgt5_mostjob_devx2:rural.ses.med41[4] 0.25 0.19 0.01 0.69
## prjthreat_mostjob_devx21[1] 0.27 0.15 0.04 0.60
## prjthreat_mostjob_devx21[2] 0.24 0.14 0.04 0.56
## prjthreat_mostjob_devx21[3] 0.24 0.14 0.03 0.57
## prjthreat_mostjob_devx21[4] 0.26 0.14 0.04 0.58
## prjthreat_mostjob_av12x21[1] 0.11 0.07 0.01 0.28
## prjthreat_mostjob_av12x21[2] 0.12 0.07 0.01 0.29
## prjthreat_mostjob_av12x21[3] 0.12 0.08 0.02 0.30
## prjthreat_mostjob_av12x21[4] 0.12 0.08 0.02 0.30
## prjthreat_mostjob_av12x21[5] 0.12 0.08 0.02 0.31
## prjthreat_mostjob_av12x21[6] 0.13 0.09 0.01 0.34
## prjthreat_mostjob_av12x21[7] 0.15 0.09 0.02 0.37
## prjthreat_mostjob_av12x21[8] 0.13 0.08 0.02 0.34
## prjthreat_mostjob_devx2:rural.ses.med21[1] 0.28 0.20 0.01 0.74
## prjthreat_mostjob_devx2:rural.ses.med21[2] 0.21 0.18 0.01 0.67
## prjthreat_mostjob_devx2:rural.ses.med21[3] 0.23 0.19 0.01 0.70
## prjthreat_mostjob_devx2:rural.ses.med21[4] 0.28 0.21 0.01 0.75
## prjthreat_mostjob_devx2:rural.ses.med31[1] 0.24 0.19 0.01 0.69
## prjthreat_mostjob_devx2:rural.ses.med31[2] 0.22 0.18 0.01 0.65
## prjthreat_mostjob_devx2:rural.ses.med31[3] 0.25 0.19 0.01 0.70
## prjthreat_mostjob_devx2:rural.ses.med31[4] 0.28 0.21 0.01 0.75
## prjthreat_mostjob_devx2:rural.ses.med41[1] 0.29 0.21 0.01 0.75
## prjthreat_mostjob_devx2:rural.ses.med41[2] 0.25 0.19 0.01 0.71
## prjthreat_mostjob_devx2:rural.ses.med41[3] 0.20 0.18 0.00 0.69
## prjthreat_mostjob_devx2:rural.ses.med41[4] 0.26 0.19 0.01 0.70
## prjharm_mostjob_devx21[1] 0.26 0.15 0.04 0.60
## prjharm_mostjob_devx21[2] 0.24 0.14 0.04 0.56
## prjharm_mostjob_devx21[3] 0.24 0.14 0.03 0.57
## prjharm_mostjob_devx21[4] 0.26 0.15 0.04 0.59
## prjharm_mostjob_av12x21[1] 0.12 0.08 0.02 0.32
## prjharm_mostjob_av12x21[2] 0.12 0.08 0.02 0.31
## prjharm_mostjob_av12x21[3] 0.12 0.08 0.02 0.31
## prjharm_mostjob_av12x21[4] 0.13 0.08 0.02 0.33
## prjharm_mostjob_av12x21[5] 0.12 0.08 0.02 0.31
## prjharm_mostjob_av12x21[6] 0.13 0.08 0.02 0.34
## prjharm_mostjob_av12x21[7] 0.13 0.08 0.02 0.32
## prjharm_mostjob_av12x21[8] 0.12 0.07 0.02 0.30
## prjharm_mostjob_devx2:rural.ses.med21[1] 0.26 0.19 0.01 0.71
## prjharm_mostjob_devx2:rural.ses.med21[2] 0.23 0.18 0.01 0.67
## prjharm_mostjob_devx2:rural.ses.med21[3] 0.23 0.19 0.01 0.68
## prjharm_mostjob_devx2:rural.ses.med21[4] 0.28 0.20 0.01 0.74
## prjharm_mostjob_devx2:rural.ses.med31[1] 0.24 0.19 0.01 0.68
## prjharm_mostjob_devx2:rural.ses.med31[2] 0.23 0.18 0.01 0.67
## prjharm_mostjob_devx2:rural.ses.med31[3] 0.24 0.18 0.01 0.67
## prjharm_mostjob_devx2:rural.ses.med31[4] 0.29 0.21 0.01 0.75
## prjharm_mostjob_devx2:rural.ses.med41[1] 0.22 0.19 0.01 0.67
## prjharm_mostjob_devx2:rural.ses.med41[2] 0.21 0.18 0.01 0.66
## prjharm_mostjob_devx2:rural.ses.med41[3] 0.28 0.21 0.01 0.76
## prjharm_mostjob_devx2:rural.ses.med41[4] 0.29 0.20 0.01 0.74
## prjusedrg_mostjob_devx21[1] 0.26 0.15 0.04 0.61
## prjusedrg_mostjob_devx21[2] 0.24 0.14 0.04 0.56
## prjusedrg_mostjob_devx21[3] 0.24 0.14 0.03 0.57
## prjusedrg_mostjob_devx21[4] 0.26 0.14 0.04 0.58
## prjusedrg_mostjob_av12x21[1] 0.11 0.07 0.02 0.29
## prjusedrg_mostjob_av12x21[2] 0.12 0.08 0.01 0.31
## prjusedrg_mostjob_av12x21[3] 0.12 0.07 0.02 0.30
## prjusedrg_mostjob_av12x21[4] 0.12 0.08 0.02 0.30
## prjusedrg_mostjob_av12x21[5] 0.13 0.08 0.02 0.33
## prjusedrg_mostjob_av12x21[6] 0.13 0.08 0.02 0.33
## prjusedrg_mostjob_av12x21[7] 0.15 0.09 0.02 0.38
## prjusedrg_mostjob_av12x21[8] 0.12 0.07 0.02 0.29
## prjusedrg_mostjob_devx2:rural.ses.med21[1] 0.26 0.20 0.01 0.71
## prjusedrg_mostjob_devx2:rural.ses.med21[2] 0.22 0.18 0.01 0.67
## prjusedrg_mostjob_devx2:rural.ses.med21[3] 0.24 0.19 0.01 0.68
## prjusedrg_mostjob_devx2:rural.ses.med21[4] 0.29 0.21 0.01 0.75
## prjusedrg_mostjob_devx2:rural.ses.med31[1] 0.26 0.20 0.01 0.72
## prjusedrg_mostjob_devx2:rural.ses.med31[2] 0.22 0.17 0.01 0.65
## prjusedrg_mostjob_devx2:rural.ses.med31[3] 0.24 0.19 0.01 0.68
## prjusedrg_mostjob_devx2:rural.ses.med31[4] 0.28 0.21 0.01 0.74
## prjusedrg_mostjob_devx2:rural.ses.med41[1] 0.29 0.22 0.01 0.78
## prjusedrg_mostjob_devx2:rural.ses.med41[2] 0.21 0.18 0.01 0.64
## prjusedrg_mostjob_devx2:rural.ses.med41[3] 0.22 0.18 0.01 0.65
## prjusedrg_mostjob_devx2:rural.ses.med41[4] 0.28 0.21 0.01 0.75
## prjhack_mostjob_devx21[1] 0.25 0.14 0.04 0.57
## prjhack_mostjob_devx21[2] 0.25 0.14 0.04 0.58
## prjhack_mostjob_devx21[3] 0.26 0.15 0.04 0.59
## prjhack_mostjob_devx21[4] 0.25 0.15 0.04 0.60
## prjhack_mostjob_av12x21[1] 0.10 0.07 0.01 0.26
## prjhack_mostjob_av12x21[2] 0.10 0.07 0.01 0.27
## prjhack_mostjob_av12x21[3] 0.11 0.07 0.01 0.28
## prjhack_mostjob_av12x21[4] 0.10 0.07 0.01 0.27
## prjhack_mostjob_av12x21[5] 0.12 0.08 0.01 0.31
## prjhack_mostjob_av12x21[6] 0.14 0.09 0.02 0.34
## prjhack_mostjob_av12x21[7] 0.19 0.10 0.03 0.41
## prjhack_mostjob_av12x21[8] 0.14 0.08 0.02 0.33
## prjhack_mostjob_devx2:rural.ses.med21[1] 0.28 0.20 0.01 0.74
## prjhack_mostjob_devx2:rural.ses.med21[2] 0.24 0.19 0.01 0.68
## prjhack_mostjob_devx2:rural.ses.med21[3] 0.21 0.18 0.01 0.66
## prjhack_mostjob_devx2:rural.ses.med21[4] 0.28 0.20 0.01 0.73
## prjhack_mostjob_devx2:rural.ses.med31[1] 0.23 0.18 0.01 0.68
## prjhack_mostjob_devx2:rural.ses.med31[2] 0.23 0.18 0.01 0.67
## prjhack_mostjob_devx2:rural.ses.med31[3] 0.25 0.19 0.01 0.69
## prjhack_mostjob_devx2:rural.ses.med31[4] 0.29 0.21 0.01 0.75
## prjhack_mostjob_devx2:rural.ses.med41[1] 0.21 0.18 0.01 0.66
## prjhack_mostjob_devx2:rural.ses.med41[2] 0.20 0.18 0.00 0.67
## prjhack_mostjob_devx2:rural.ses.med41[3] 0.31 0.23 0.01 0.80
## prjhack_mostjob_devx2:rural.ses.med41[4] 0.28 0.21 0.01 0.75
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_mostjob_devx21[1] 1.00 8559 2854
## prjthflt5_mostjob_devx21[2] 1.00 6973 2609
## prjthflt5_mostjob_devx21[3] 1.00 7683 2756
## prjthflt5_mostjob_devx21[4] 1.00 7220 2298
## prjthflt5_mostjob_av12x21[1] 1.00 7731 2923
## prjthflt5_mostjob_av12x21[2] 1.00 8800 2872
## prjthflt5_mostjob_av12x21[3] 1.00 8371 2559
## prjthflt5_mostjob_av12x21[4] 1.00 8226 2963
## prjthflt5_mostjob_av12x21[5] 1.00 8558 2341
## prjthflt5_mostjob_av12x21[6] 1.00 7876 2397
## prjthflt5_mostjob_av12x21[7] 1.00 7008 2522
## prjthflt5_mostjob_av12x21[8] 1.00 7835 2342
## prjthflt5_mostjob_devx2:rural.ses.med21[1] 1.00 7123 2587
## prjthflt5_mostjob_devx2:rural.ses.med21[2] 1.00 6237 2711
## prjthflt5_mostjob_devx2:rural.ses.med21[3] 1.00 5815 3175
## prjthflt5_mostjob_devx2:rural.ses.med21[4] 1.00 6068 2865
## prjthflt5_mostjob_devx2:rural.ses.med31[1] 1.00 7289 2621
## prjthflt5_mostjob_devx2:rural.ses.med31[2] 1.00 7174 2908
## prjthflt5_mostjob_devx2:rural.ses.med31[3] 1.00 5470 2837
## prjthflt5_mostjob_devx2:rural.ses.med31[4] 1.00 6801 3364
## prjthflt5_mostjob_devx2:rural.ses.med41[1] 1.00 5238 3016
## prjthflt5_mostjob_devx2:rural.ses.med41[2] 1.00 6480 2592
## prjthflt5_mostjob_devx2:rural.ses.med41[3] 1.00 4955 2853
## prjthflt5_mostjob_devx2:rural.ses.med41[4] 1.00 6273 2930
## prjthfgt5_mostjob_devx21[1] 1.00 8600 2476
## prjthfgt5_mostjob_devx21[2] 1.00 7617 2858
## prjthfgt5_mostjob_devx21[3] 1.00 7513 2712
## prjthfgt5_mostjob_devx21[4] 1.00 8912 2735
## prjthfgt5_mostjob_av12x21[1] 1.00 7841 2429
## prjthfgt5_mostjob_av12x21[2] 1.00 6940 2628
## prjthfgt5_mostjob_av12x21[3] 1.00 8628 2301
## prjthfgt5_mostjob_av12x21[4] 1.00 7091 2990
## prjthfgt5_mostjob_av12x21[5] 1.00 7248 2623
## prjthfgt5_mostjob_av12x21[6] 1.00 8622 2794
## prjthfgt5_mostjob_av12x21[7] 1.00 6608 2580
## prjthfgt5_mostjob_av12x21[8] 1.00 6773 2401
## prjthfgt5_mostjob_devx2:rural.ses.med21[1] 1.00 6484 2418
## prjthfgt5_mostjob_devx2:rural.ses.med21[2] 1.00 7093 2814
## prjthfgt5_mostjob_devx2:rural.ses.med21[3] 1.00 7376 2806
## prjthfgt5_mostjob_devx2:rural.ses.med21[4] 1.00 6027 3123
## prjthfgt5_mostjob_devx2:rural.ses.med31[1] 1.00 6824 2813
## prjthfgt5_mostjob_devx2:rural.ses.med31[2] 1.00 7635 3048
## prjthfgt5_mostjob_devx2:rural.ses.med31[3] 1.00 5139 3070
## prjthfgt5_mostjob_devx2:rural.ses.med31[4] 1.00 6475 2792
## prjthfgt5_mostjob_devx2:rural.ses.med41[1] 1.00 4985 3192
## prjthfgt5_mostjob_devx2:rural.ses.med41[2] 1.00 4921 2745
## prjthfgt5_mostjob_devx2:rural.ses.med41[3] 1.00 3762 2841
## prjthfgt5_mostjob_devx2:rural.ses.med41[4] 1.00 7078 2606
## prjthreat_mostjob_devx21[1] 1.00 8712 2636
## prjthreat_mostjob_devx21[2] 1.00 8269 2694
## prjthreat_mostjob_devx21[3] 1.00 8161 2728
## prjthreat_mostjob_devx21[4] 1.00 7222 2563
## prjthreat_mostjob_av12x21[1] 1.00 7906 2064
## prjthreat_mostjob_av12x21[2] 1.00 8188 2403
## prjthreat_mostjob_av12x21[3] 1.00 7888 2809
## prjthreat_mostjob_av12x21[4] 1.00 7116 2209
## prjthreat_mostjob_av12x21[5] 1.00 6905 2538
## prjthreat_mostjob_av12x21[6] 1.00 8048 2430
## prjthreat_mostjob_av12x21[7] 1.00 6570 3235
## prjthreat_mostjob_av12x21[8] 1.00 6844 3000
## prjthreat_mostjob_devx2:rural.ses.med21[1] 1.00 6249 2482
## prjthreat_mostjob_devx2:rural.ses.med21[2] 1.00 6578 2706
## prjthreat_mostjob_devx2:rural.ses.med21[3] 1.00 6272 2643
## prjthreat_mostjob_devx2:rural.ses.med21[4] 1.00 5552 2416
## prjthreat_mostjob_devx2:rural.ses.med31[1] 1.00 7186 2658
## prjthreat_mostjob_devx2:rural.ses.med31[2] 1.00 7111 2971
## prjthreat_mostjob_devx2:rural.ses.med31[3] 1.00 5771 2635
## prjthreat_mostjob_devx2:rural.ses.med31[4] 1.00 6052 3163
## prjthreat_mostjob_devx2:rural.ses.med41[1] 1.00 5481 2765
## prjthreat_mostjob_devx2:rural.ses.med41[2] 1.00 5196 3003
## prjthreat_mostjob_devx2:rural.ses.med41[3] 1.00 3947 2626
## prjthreat_mostjob_devx2:rural.ses.med41[4] 1.00 6620 2762
## prjharm_mostjob_devx21[1] 1.00 8827 2701
## prjharm_mostjob_devx21[2] 1.00 6828 2306
## prjharm_mostjob_devx21[3] 1.00 6784 3111
## prjharm_mostjob_devx21[4] 1.00 8310 2823
## prjharm_mostjob_av12x21[1] 1.00 7082 2332
## prjharm_mostjob_av12x21[2] 1.00 7477 2685
## prjharm_mostjob_av12x21[3] 1.00 8056 2483
## prjharm_mostjob_av12x21[4] 1.00 7786 2363
## prjharm_mostjob_av12x21[5] 1.00 8504 2310
## prjharm_mostjob_av12x21[6] 1.00 7737 2673
## prjharm_mostjob_av12x21[7] 1.00 8810 3049
## prjharm_mostjob_av12x21[8] 1.00 6514 2242
## prjharm_mostjob_devx2:rural.ses.med21[1] 1.00 7800 2940
## prjharm_mostjob_devx2:rural.ses.med21[2] 1.00 6508 2640
## prjharm_mostjob_devx2:rural.ses.med21[3] 1.00 6698 2639
## prjharm_mostjob_devx2:rural.ses.med21[4] 1.00 5888 2205
## prjharm_mostjob_devx2:rural.ses.med31[1] 1.00 7902 2764
## prjharm_mostjob_devx2:rural.ses.med31[2] 1.00 6647 2807
## prjharm_mostjob_devx2:rural.ses.med31[3] 1.00 6306 2835
## prjharm_mostjob_devx2:rural.ses.med31[4] 1.00 6214 2582
## prjharm_mostjob_devx2:rural.ses.med41[1] 1.00 6027 2647
## prjharm_mostjob_devx2:rural.ses.med41[2] 1.00 6544 2728
## prjharm_mostjob_devx2:rural.ses.med41[3] 1.00 5673 2900
## prjharm_mostjob_devx2:rural.ses.med41[4] 1.00 5878 2605
## prjusedrg_mostjob_devx21[1] 1.00 8468 2742
## prjusedrg_mostjob_devx21[2] 1.00 7396 2648
## prjusedrg_mostjob_devx21[3] 1.00 6837 3077
## prjusedrg_mostjob_devx21[4] 1.00 7482 2392
## prjusedrg_mostjob_av12x21[1] 1.00 7625 2671
## prjusedrg_mostjob_av12x21[2] 1.00 7089 2167
## prjusedrg_mostjob_av12x21[3] 1.00 7956 2684
## prjusedrg_mostjob_av12x21[4] 1.00 8541 2950
## prjusedrg_mostjob_av12x21[5] 1.00 7843 2520
## prjusedrg_mostjob_av12x21[6] 1.00 8571 2746
## prjusedrg_mostjob_av12x21[7] 1.00 7710 2750
## prjusedrg_mostjob_av12x21[8] 1.00 6909 2593
## prjusedrg_mostjob_devx2:rural.ses.med21[1] 1.00 7302 2281
## prjusedrg_mostjob_devx2:rural.ses.med21[2] 1.00 7715 2630
## prjusedrg_mostjob_devx2:rural.ses.med21[3] 1.00 6498 2469
## prjusedrg_mostjob_devx2:rural.ses.med21[4] 1.00 5910 2863
## prjusedrg_mostjob_devx2:rural.ses.med31[1] 1.00 6144 2614
## prjusedrg_mostjob_devx2:rural.ses.med31[2] 1.00 6061 3074
## prjusedrg_mostjob_devx2:rural.ses.med31[3] 1.00 6970 2524
## prjusedrg_mostjob_devx2:rural.ses.med31[4] 1.00 5184 2800
## prjusedrg_mostjob_devx2:rural.ses.med41[1] 1.00 4945 2763
## prjusedrg_mostjob_devx2:rural.ses.med41[2] 1.00 6291 2593
## prjusedrg_mostjob_devx2:rural.ses.med41[3] 1.00 5086 2765
## prjusedrg_mostjob_devx2:rural.ses.med41[4] 1.00 6534 2940
## prjhack_mostjob_devx21[1] 1.00 7513 2155
## prjhack_mostjob_devx21[2] 1.00 9228 1911
## prjhack_mostjob_devx21[3] 1.00 6327 2781
## prjhack_mostjob_devx21[4] 1.00 8007 2693
## prjhack_mostjob_av12x21[1] 1.00 9040 2687
## prjhack_mostjob_av12x21[2] 1.00 7711 2996
## prjhack_mostjob_av12x21[3] 1.00 9575 2665
## prjhack_mostjob_av12x21[4] 1.00 7762 2761
## prjhack_mostjob_av12x21[5] 1.01 7864 2588
## prjhack_mostjob_av12x21[6] 1.00 7788 2873
## prjhack_mostjob_av12x21[7] 1.00 8388 2394
## prjhack_mostjob_av12x21[8] 1.00 6524 2709
## prjhack_mostjob_devx2:rural.ses.med21[1] 1.00 6021 2471
## prjhack_mostjob_devx2:rural.ses.med21[2] 1.00 5575 2601
## prjhack_mostjob_devx2:rural.ses.med21[3] 1.00 5817 3219
## prjhack_mostjob_devx2:rural.ses.med21[4] 1.00 6985 2882
## prjhack_mostjob_devx2:rural.ses.med31[1] 1.00 7287 2277
## prjhack_mostjob_devx2:rural.ses.med31[2] 1.00 7670 2892
## prjhack_mostjob_devx2:rural.ses.med31[3] 1.00 6133 2589
## prjhack_mostjob_devx2:rural.ses.med31[4] 1.00 6595 2074
## prjhack_mostjob_devx2:rural.ses.med41[1] 1.00 5420 2742
## prjhack_mostjob_devx2:rural.ses.med41[2] 1.00 5518 2920
## prjhack_mostjob_devx2:rural.ses.med41[3] 1.00 5427 2726
## prjhack_mostjob_devx2:rural.ses.med41[4] 1.00 6385 2769
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.1.5.4 Prior summary
out.chg.prjcrime.stjob.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## group resp dpar nlpar lb ub source
## default
## prjhack user
## prjhack user
## prjhack user
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjharm user
## prjharm user
## prjharm user
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthflt5 user
## prjthflt5 user
## prjthflt5 user
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthreat user
## prjthreat user
## prjthreat user
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjusedrg user
## prjusedrg user
## prjusedrg user
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## default
## prjhack user
## prjharm user
## prjthfgt5 user
## prjthflt5 user
## prjthreat user
## prjusedrg user
## prjhack 0 default
## prjharm 0 default
## prjthfgt5 0 default
## prjthflt5 0 default
## prjthreat 0 default
## prjusedrg 0 default
## id prjhack 0 (vectorized)
## id prjhack 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjusedrg 0 (vectorized)
## id prjusedrg 0 (vectorized)
## prjhack user
## prjhack default
## prjhack default
## prjhack default
## prjhack user
## prjharm user
## prjharm default
## prjharm default
## prjharm default
## prjharm user
## prjthfgt5 user
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 user
## prjthflt5 user
## prjthflt5 default
## prjthflt5 default
## prjthflt5 default
## prjthflt5 user
## prjthreat user
## prjthreat default
## prjthreat default
## prjthreat default
## prjthreat user
## prjusedrg user
## prjusedrg default
## prjusedrg default
## prjusedrg default
## prjusedrg user8.1.1.6 Corr X Community: stthft & prj crime
#Community Change: criminal intent items ~ mo(stthft)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostthft_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostthft_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostthft_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostthft_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.stthft.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stthft_devx2) + mo(stthft_av12x2) +
rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stthft_comm_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stthft.comm.fit <- ppchecks(chg.prjcrime.stthft.comm.fit)8.1.1.6.1 Coefficient plot (intervals)
out.chg.prjcrime.stthft.comm.fit[[10]]
8.1.1.6.2 PPcheck (density)
p1 <- out.chg.prjcrime.stthft.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stthft.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stthft.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stthft.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stthft.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stthft.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
8.1.1.6.3 Fit summary
out.chg.prjcrime.stthft.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## prjthreat ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## prjharm ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## prjusedrg ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## prjhack ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.96 0.57 2.94 5.21 1.00 1487
## sd(prjthfgt5_Intercept) 3.30 0.49 2.40 4.34 1.00 1609
## sd(prjthreat_Intercept) 3.14 0.55 2.19 4.29 1.00 1515
## sd(prjharm_Intercept) 3.09 0.57 2.07 4.30 1.00 1550
## sd(prjusedrg_Intercept) 2.87 0.53 1.93 3.97 1.00 1483
## sd(prjhack_Intercept) 0.90 0.57 0.05 2.12 1.00 698
## Tail_ESS
## sd(prjthflt5_Intercept) 2564
## sd(prjthfgt5_Intercept) 1721
## sd(prjthreat_Intercept) 2166
## sd(prjharm_Intercept) 2378
## sd(prjusedrg_Intercept) 2382
## sd(prjhack_Intercept) 1389
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_Intercept -5.60 0.84 -7.34 -4.03
## prjthfgt5_Intercept -5.29 0.78 -6.85 -3.77
## prjthreat_Intercept -6.12 0.92 -8.00 -4.37
## prjharm_Intercept -5.71 0.93 -7.64 -3.99
## prjusedrg_Intercept -5.75 0.91 -7.60 -4.09
## prjhack_Intercept -3.99 0.74 -5.59 -2.64
## prjthflt5_rural.ses.med2 -0.34 0.90 -2.08 1.44
## prjthflt5_rural.ses.med3 0.66 0.80 -1.00 2.15
## prjthflt5_rural.ses.med4 1.67 0.75 0.15 3.08
## prjthfgt5_rural.ses.med2 -0.31 0.92 -2.11 1.44
## prjthfgt5_rural.ses.med3 0.62 0.77 -0.96 2.06
## prjthfgt5_rural.ses.med4 1.35 0.76 -0.21 2.77
## prjthreat_rural.ses.med2 -0.53 0.92 -2.27 1.34
## prjthreat_rural.ses.med3 0.48 0.76 -1.07 1.97
## prjthreat_rural.ses.med4 1.61 0.77 0.04 3.03
## prjharm_rural.ses.med2 -0.02 0.87 -1.63 1.76
## prjharm_rural.ses.med3 0.21 0.77 -1.29 1.75
## prjharm_rural.ses.med4 0.98 0.75 -0.53 2.43
## prjusedrg_rural.ses.med2 -0.41 0.86 -2.09 1.32
## prjusedrg_rural.ses.med3 -0.15 0.82 -1.76 1.48
## prjusedrg_rural.ses.med4 1.54 0.77 -0.02 3.04
## prjhack_rural.ses.med2 -0.31 0.87 -2.02 1.44
## prjhack_rural.ses.med3 -0.06 0.71 -1.48 1.37
## prjhack_rural.ses.med4 0.96 0.70 -0.38 2.32
## prjthflt5_mostthft_devx2 -0.20 0.20 -0.60 0.19
## prjthflt5_mostthft_av12x2 0.06 0.09 -0.12 0.23
## prjthflt5_mostthft_devx2:rural.ses.med2 -0.80 0.53 -1.92 0.19
## prjthflt5_mostthft_devx2:rural.ses.med3 0.45 0.41 -0.27 1.35
## prjthflt5_mostthft_devx2:rural.ses.med4 0.25 0.35 -0.45 0.93
## prjthfgt5_mostthft_devx2 -0.15 0.21 -0.58 0.26
## prjthfgt5_mostthft_av12x2 0.10 0.08 -0.05 0.27
## prjthfgt5_mostthft_devx2:rural.ses.med2 -0.91 0.51 -1.96 0.05
## prjthfgt5_mostthft_devx2:rural.ses.med3 0.36 0.39 -0.31 1.18
## prjthfgt5_mostthft_devx2:rural.ses.med4 0.31 0.33 -0.36 0.95
## prjthreat_mostthft_devx2 -0.25 0.22 -0.69 0.19
## prjthreat_mostthft_av12x2 0.10 0.09 -0.08 0.28
## prjthreat_mostthft_devx2:rural.ses.med2 -0.38 0.63 -1.62 0.90
## prjthreat_mostthft_devx2:rural.ses.med3 -0.05 0.47 -1.02 0.84
## prjthreat_mostthft_devx2:rural.ses.med4 0.02 0.41 -0.86 0.81
## prjharm_mostthft_devx2 -0.33 0.21 -0.75 0.10
## prjharm_mostthft_av12x2 0.05 0.09 -0.14 0.22
## prjharm_mostthft_devx2:rural.ses.med2 -0.50 0.56 -1.72 0.57
## prjharm_mostthft_devx2:rural.ses.med3 -0.08 0.46 -1.06 0.78
## prjharm_mostthft_devx2:rural.ses.med4 -0.07 0.44 -1.03 0.72
## prjusedrg_mostthft_devx2 -0.13 0.22 -0.57 0.28
## prjusedrg_mostthft_av12x2 -0.07 0.09 -0.25 0.11
## prjusedrg_mostthft_devx2:rural.ses.med2 -0.24 0.56 -1.41 0.83
## prjusedrg_mostthft_devx2:rural.ses.med3 -0.48 0.60 -1.78 0.59
## prjusedrg_mostthft_devx2:rural.ses.med4 0.14 0.39 -0.71 0.86
## prjhack_mostthft_devx2 -0.18 0.21 -0.59 0.24
## prjhack_mostthft_av12x2 -0.01 0.08 -0.16 0.14
## prjhack_mostthft_devx2:rural.ses.med2 -0.68 0.63 -2.04 0.49
## prjhack_mostthft_devx2:rural.ses.med3 0.12 0.40 -0.73 0.87
## prjhack_mostthft_devx2:rural.ses.med4 -0.05 0.38 -0.78 0.74
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1.00 2512 2921
## prjthfgt5_Intercept 1.00 2832 2943
## prjthreat_Intercept 1.00 2723 3045
## prjharm_Intercept 1.00 2961 2968
## prjusedrg_Intercept 1.00 2594 2720
## prjhack_Intercept 1.00 1555 2329
## prjthflt5_rural.ses.med2 1.00 5429 3297
## prjthflt5_rural.ses.med3 1.00 4022 3092
## prjthflt5_rural.ses.med4 1.00 4089 3057
## prjthfgt5_rural.ses.med2 1.00 4844 3099
## prjthfgt5_rural.ses.med3 1.00 4013 3099
## prjthfgt5_rural.ses.med4 1.00 3997 2759
## prjthreat_rural.ses.med2 1.00 4757 3012
## prjthreat_rural.ses.med3 1.00 5472 3331
## prjthreat_rural.ses.med4 1.00 4310 3057
## prjharm_rural.ses.med2 1.00 5433 3226
## prjharm_rural.ses.med3 1.00 5457 3001
## prjharm_rural.ses.med4 1.00 4232 3120
## prjusedrg_rural.ses.med2 1.00 5382 3616
## prjusedrg_rural.ses.med3 1.00 5343 2778
## prjusedrg_rural.ses.med4 1.00 5192 3130
## prjhack_rural.ses.med2 1.00 5024 3188
## prjhack_rural.ses.med3 1.00 5076 3005
## prjhack_rural.ses.med4 1.00 4513 3208
## prjthflt5_mostthft_devx2 1.00 4532 2980
## prjthflt5_mostthft_av12x2 1.00 3066 3105
## prjthflt5_mostthft_devx2:rural.ses.med2 1.00 3224 2909
## prjthflt5_mostthft_devx2:rural.ses.med3 1.00 3288 2681
## prjthflt5_mostthft_devx2:rural.ses.med4 1.00 3357 3126
## prjthfgt5_mostthft_devx2 1.00 4312 3362
## prjthfgt5_mostthft_av12x2 1.00 3455 3280
## prjthfgt5_mostthft_devx2:rural.ses.med2 1.00 3875 2778
## prjthfgt5_mostthft_devx2:rural.ses.med3 1.00 3131 2537
## prjthfgt5_mostthft_devx2:rural.ses.med4 1.00 3545 3034
## prjthreat_mostthft_devx2 1.00 4438 3016
## prjthreat_mostthft_av12x2 1.00 4478 3176
## prjthreat_mostthft_devx2:rural.ses.med2 1.00 2936 2727
## prjthreat_mostthft_devx2:rural.ses.med3 1.00 3806 2561
## prjthreat_mostthft_devx2:rural.ses.med4 1.00 3235 2834
## prjharm_mostthft_devx2 1.00 5087 2870
## prjharm_mostthft_av12x2 1.00 4418 3262
## prjharm_mostthft_devx2:rural.ses.med2 1.00 3839 2666
## prjharm_mostthft_devx2:rural.ses.med3 1.00 3526 2479
## prjharm_mostthft_devx2:rural.ses.med4 1.00 3734 2897
## prjusedrg_mostthft_devx2 1.00 5087 3211
## prjusedrg_mostthft_av12x2 1.00 4630 3259
## prjusedrg_mostthft_devx2:rural.ses.med2 1.00 3679 3217
## prjusedrg_mostthft_devx2:rural.ses.med3 1.00 3924 2626
## prjusedrg_mostthft_devx2:rural.ses.med4 1.00 3646 2522
## prjhack_mostthft_devx2 1.00 4700 3243
## prjhack_mostthft_av12x2 1.00 6489 3067
## prjhack_mostthft_devx2:rural.ses.med2 1.00 4270 2866
## prjhack_mostthft_devx2:rural.ses.med3 1.00 4078 2661
## prjhack_mostthft_devx2:rural.ses.med4 1.00 3321 2867
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## prjthflt5_mostthft_devx21[1] 0.28 0.15 0.05
## prjthflt5_mostthft_devx21[2] 0.26 0.14 0.04
## prjthflt5_mostthft_devx21[3] 0.23 0.13 0.03
## prjthflt5_mostthft_devx21[4] 0.24 0.14 0.03
## prjthflt5_mostthft_av12x21[1] 0.13 0.08 0.02
## prjthflt5_mostthft_av12x21[2] 0.12 0.08 0.02
## prjthflt5_mostthft_av12x21[3] 0.12 0.08 0.02
## prjthflt5_mostthft_av12x21[4] 0.12 0.08 0.02
## prjthflt5_mostthft_av12x21[5] 0.13 0.08 0.02
## prjthflt5_mostthft_av12x21[6] 0.13 0.08 0.02
## prjthflt5_mostthft_av12x21[7] 0.12 0.08 0.01
## prjthflt5_mostthft_av12x21[8] 0.13 0.08 0.02
## prjthflt5_mostthft_devx2:rural.ses.med21[1] 0.31 0.20 0.02
## prjthflt5_mostthft_devx2:rural.ses.med21[2] 0.26 0.19 0.01
## prjthflt5_mostthft_devx2:rural.ses.med21[3] 0.19 0.16 0.01
## prjthflt5_mostthft_devx2:rural.ses.med21[4] 0.24 0.18 0.01
## prjthflt5_mostthft_devx2:rural.ses.med31[1] 0.29 0.21 0.01
## prjthflt5_mostthft_devx2:rural.ses.med31[2] 0.21 0.17 0.01
## prjthflt5_mostthft_devx2:rural.ses.med31[3] 0.18 0.16 0.00
## prjthflt5_mostthft_devx2:rural.ses.med31[4] 0.31 0.21 0.01
## prjthflt5_mostthft_devx2:rural.ses.med41[1] 0.29 0.20 0.01
## prjthflt5_mostthft_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjthflt5_mostthft_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## prjthflt5_mostthft_devx2:rural.ses.med41[4] 0.25 0.20 0.01
## prjthfgt5_mostthft_devx21[1] 0.27 0.15 0.04
## prjthfgt5_mostthft_devx21[2] 0.26 0.15 0.04
## prjthfgt5_mostthft_devx21[3] 0.22 0.13 0.03
## prjthfgt5_mostthft_devx21[4] 0.25 0.14 0.04
## prjthfgt5_mostthft_av12x21[1] 0.13 0.08 0.02
## prjthfgt5_mostthft_av12x21[2] 0.12 0.08 0.02
## prjthfgt5_mostthft_av12x21[3] 0.12 0.08 0.02
## prjthfgt5_mostthft_av12x21[4] 0.12 0.08 0.02
## prjthfgt5_mostthft_av12x21[5] 0.13 0.08 0.02
## prjthfgt5_mostthft_av12x21[6] 0.13 0.08 0.02
## prjthfgt5_mostthft_av12x21[7] 0.12 0.07 0.02
## prjthfgt5_mostthft_av12x21[8] 0.12 0.08 0.02
## prjthfgt5_mostthft_devx2:rural.ses.med21[1] 0.30 0.19 0.02
## prjthfgt5_mostthft_devx2:rural.ses.med21[2] 0.30 0.20 0.02
## prjthfgt5_mostthft_devx2:rural.ses.med21[3] 0.17 0.15 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med21[4] 0.23 0.18 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med31[1] 0.29 0.21 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med31[2] 0.21 0.18 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med31[3] 0.20 0.17 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med31[4] 0.31 0.21 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med41[1] 0.26 0.19 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med41[2] 0.24 0.19 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med41[3] 0.26 0.19 0.01
## prjthfgt5_mostthft_devx2:rural.ses.med41[4] 0.24 0.19 0.01
## prjthreat_mostthft_devx21[1] 0.27 0.15 0.04
## prjthreat_mostthft_devx21[2] 0.28 0.15 0.04
## prjthreat_mostthft_devx21[3] 0.21 0.13 0.03
## prjthreat_mostthft_devx21[4] 0.24 0.14 0.03
## prjthreat_mostthft_av12x21[1] 0.13 0.08 0.02
## prjthreat_mostthft_av12x21[2] 0.13 0.08 0.02
## prjthreat_mostthft_av12x21[3] 0.12 0.08 0.02
## prjthreat_mostthft_av12x21[4] 0.13 0.08 0.02
## prjthreat_mostthft_av12x21[5] 0.13 0.08 0.02
## prjthreat_mostthft_av12x21[6] 0.13 0.08 0.02
## prjthreat_mostthft_av12x21[7] 0.13 0.08 0.02
## prjthreat_mostthft_av12x21[8] 0.12 0.08 0.01
## prjthreat_mostthft_devx2:rural.ses.med21[1] 0.30 0.21 0.01
## prjthreat_mostthft_devx2:rural.ses.med21[2] 0.23 0.18 0.01
## prjthreat_mostthft_devx2:rural.ses.med21[3] 0.20 0.18 0.01
## prjthreat_mostthft_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## prjthreat_mostthft_devx2:rural.ses.med31[1] 0.24 0.19 0.01
## prjthreat_mostthft_devx2:rural.ses.med31[2] 0.22 0.19 0.01
## prjthreat_mostthft_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## prjthreat_mostthft_devx2:rural.ses.med31[4] 0.30 0.21 0.01
## prjthreat_mostthft_devx2:rural.ses.med41[1] 0.26 0.20 0.01
## prjthreat_mostthft_devx2:rural.ses.med41[2] 0.22 0.18 0.01
## prjthreat_mostthft_devx2:rural.ses.med41[3] 0.23 0.18 0.01
## prjthreat_mostthft_devx2:rural.ses.med41[4] 0.28 0.21 0.01
## prjharm_mostthft_devx21[1] 0.26 0.14 0.04
## prjharm_mostthft_devx21[2] 0.30 0.15 0.05
## prjharm_mostthft_devx21[3] 0.22 0.13 0.03
## prjharm_mostthft_devx21[4] 0.22 0.13 0.03
## prjharm_mostthft_av12x21[1] 0.13 0.08 0.02
## prjharm_mostthft_av12x21[2] 0.13 0.08 0.02
## prjharm_mostthft_av12x21[3] 0.12 0.08 0.01
## prjharm_mostthft_av12x21[4] 0.12 0.08 0.01
## prjharm_mostthft_av12x21[5] 0.12 0.08 0.02
## prjharm_mostthft_av12x21[6] 0.12 0.08 0.02
## prjharm_mostthft_av12x21[7] 0.13 0.08 0.02
## prjharm_mostthft_av12x21[8] 0.13 0.08 0.02
## prjharm_mostthft_devx2:rural.ses.med21[1] 0.25 0.19 0.01
## prjharm_mostthft_devx2:rural.ses.med21[2] 0.25 0.19 0.01
## prjharm_mostthft_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## prjharm_mostthft_devx2:rural.ses.med21[4] 0.27 0.20 0.01
## prjharm_mostthft_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjharm_mostthft_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## prjharm_mostthft_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## prjharm_mostthft_devx2:rural.ses.med31[4] 0.30 0.21 0.01
## prjharm_mostthft_devx2:rural.ses.med41[1] 0.24 0.19 0.01
## prjharm_mostthft_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjharm_mostthft_devx2:rural.ses.med41[3] 0.24 0.19 0.01
## prjharm_mostthft_devx2:rural.ses.med41[4] 0.29 0.22 0.01
## prjusedrg_mostthft_devx21[1] 0.27 0.15 0.04
## prjusedrg_mostthft_devx21[2] 0.24 0.14 0.04
## prjusedrg_mostthft_devx21[3] 0.23 0.14 0.03
## prjusedrg_mostthft_devx21[4] 0.26 0.14 0.04
## prjusedrg_mostthft_av12x21[1] 0.13 0.08 0.02
## prjusedrg_mostthft_av12x21[2] 0.13 0.08 0.02
## prjusedrg_mostthft_av12x21[3] 0.13 0.08 0.02
## prjusedrg_mostthft_av12x21[4] 0.13 0.08 0.02
## prjusedrg_mostthft_av12x21[5] 0.12 0.08 0.02
## prjusedrg_mostthft_av12x21[6] 0.12 0.08 0.02
## prjusedrg_mostthft_av12x21[7] 0.12 0.08 0.02
## prjusedrg_mostthft_av12x21[8] 0.13 0.08 0.02
## prjusedrg_mostthft_devx2:rural.ses.med21[1] 0.26 0.20 0.01
## prjusedrg_mostthft_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## prjusedrg_mostthft_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## prjusedrg_mostthft_devx2:rural.ses.med21[4] 0.29 0.21 0.01
## prjusedrg_mostthft_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## prjusedrg_mostthft_devx2:rural.ses.med31[2] 0.19 0.17 0.01
## prjusedrg_mostthft_devx2:rural.ses.med31[3] 0.29 0.21 0.01
## prjusedrg_mostthft_devx2:rural.ses.med31[4] 0.27 0.20 0.01
## prjusedrg_mostthft_devx2:rural.ses.med41[1] 0.27 0.20 0.01
## prjusedrg_mostthft_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## prjusedrg_mostthft_devx2:rural.ses.med41[3] 0.24 0.19 0.01
## prjusedrg_mostthft_devx2:rural.ses.med41[4] 0.27 0.21 0.01
## prjhack_mostthft_devx21[1] 0.26 0.14 0.04
## prjhack_mostthft_devx21[2] 0.26 0.14 0.04
## prjhack_mostthft_devx21[3] 0.25 0.14 0.04
## prjhack_mostthft_devx21[4] 0.23 0.14 0.03
## prjhack_mostthft_av12x21[1] 0.13 0.08 0.02
## prjhack_mostthft_av12x21[2] 0.12 0.08 0.02
## prjhack_mostthft_av12x21[3] 0.12 0.08 0.02
## prjhack_mostthft_av12x21[4] 0.12 0.08 0.02
## prjhack_mostthft_av12x21[5] 0.12 0.08 0.02
## prjhack_mostthft_av12x21[6] 0.13 0.08 0.02
## prjhack_mostthft_av12x21[7] 0.13 0.08 0.02
## prjhack_mostthft_av12x21[8] 0.13 0.08 0.02
## prjhack_mostthft_devx2:rural.ses.med21[1] 0.27 0.19 0.01
## prjhack_mostthft_devx2:rural.ses.med21[2] 0.21 0.17 0.01
## prjhack_mostthft_devx2:rural.ses.med21[3] 0.26 0.19 0.01
## prjhack_mostthft_devx2:rural.ses.med21[4] 0.26 0.20 0.01
## prjhack_mostthft_devx2:rural.ses.med31[1] 0.24 0.19 0.01
## prjhack_mostthft_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## prjhack_mostthft_devx2:rural.ses.med31[3] 0.25 0.19 0.01
## prjhack_mostthft_devx2:rural.ses.med31[4] 0.28 0.21 0.01
## prjhack_mostthft_devx2:rural.ses.med41[1] 0.24 0.19 0.01
## prjhack_mostthft_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## prjhack_mostthft_devx2:rural.ses.med41[3] 0.25 0.19 0.01
## prjhack_mostthft_devx2:rural.ses.med41[4] 0.28 0.21 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## prjthflt5_mostthft_devx21[1] 0.62 1.00 7637 3134
## prjthflt5_mostthft_devx21[2] 0.59 1.00 5855 2883
## prjthflt5_mostthft_devx21[3] 0.54 1.00 7238 2885
## prjthflt5_mostthft_devx21[4] 0.55 1.00 7103 2989
## prjthflt5_mostthft_av12x21[1] 0.33 1.00 8089 2580
## prjthflt5_mostthft_av12x21[2] 0.32 1.00 6305 2380
## prjthflt5_mostthft_av12x21[3] 0.31 1.00 7284 2531
## prjthflt5_mostthft_av12x21[4] 0.31 1.00 6858 2270
## prjthflt5_mostthft_av12x21[5] 0.33 1.00 6060 2470
## prjthflt5_mostthft_av12x21[6] 0.32 1.00 6798 2943
## prjthflt5_mostthft_av12x21[7] 0.32 1.00 6862 2752
## prjthflt5_mostthft_av12x21[8] 0.32 1.00 6105 2938
## prjthflt5_mostthft_devx2:rural.ses.med21[1] 0.72 1.00 4706 2844
## prjthflt5_mostthft_devx2:rural.ses.med21[2] 0.70 1.00 4539 2220
## prjthflt5_mostthft_devx2:rural.ses.med21[3] 0.60 1.00 5492 2407
## prjthflt5_mostthft_devx2:rural.ses.med21[4] 0.66 1.00 5187 2804
## prjthflt5_mostthft_devx2:rural.ses.med31[1] 0.75 1.00 4908 2236
## prjthflt5_mostthft_devx2:rural.ses.med31[2] 0.64 1.00 5329 2890
## prjthflt5_mostthft_devx2:rural.ses.med31[3] 0.60 1.00 4491 2988
## prjthflt5_mostthft_devx2:rural.ses.med31[4] 0.77 1.00 5207 2919
## prjthflt5_mostthft_devx2:rural.ses.med41[1] 0.75 1.00 5212 2442
## prjthflt5_mostthft_devx2:rural.ses.med41[2] 0.67 1.00 5303 2491
## prjthflt5_mostthft_devx2:rural.ses.med41[3] 0.64 1.00 5115 2823
## prjthflt5_mostthft_devx2:rural.ses.med41[4] 0.74 1.00 5925 2771
## prjthfgt5_mostthft_devx21[1] 0.60 1.00 6074 2979
## prjthfgt5_mostthft_devx21[2] 0.59 1.00 6882 2580
## prjthfgt5_mostthft_devx21[3] 0.54 1.00 5913 2828
## prjthfgt5_mostthft_devx21[4] 0.56 1.00 7369 2857
## prjthfgt5_mostthft_av12x21[1] 0.31 1.00 7241 2224
## prjthfgt5_mostthft_av12x21[2] 0.32 1.00 7054 2078
## prjthfgt5_mostthft_av12x21[3] 0.31 1.00 7009 2761
## prjthfgt5_mostthft_av12x21[4] 0.32 1.00 7189 2575
## prjthfgt5_mostthft_av12x21[5] 0.33 1.00 6169 2485
## prjthfgt5_mostthft_av12x21[6] 0.33 1.00 6622 2920
## prjthfgt5_mostthft_av12x21[7] 0.30 1.00 6693 2708
## prjthfgt5_mostthft_av12x21[8] 0.31 1.00 7518 2903
## prjthfgt5_mostthft_devx2:rural.ses.med21[1] 0.73 1.00 6098 2813
## prjthfgt5_mostthft_devx2:rural.ses.med21[2] 0.74 1.00 6437 2882
## prjthfgt5_mostthft_devx2:rural.ses.med21[3] 0.56 1.00 6140 2749
## prjthfgt5_mostthft_devx2:rural.ses.med21[4] 0.64 1.00 5533 2816
## prjthfgt5_mostthft_devx2:rural.ses.med31[1] 0.75 1.00 4227 2207
## prjthfgt5_mostthft_devx2:rural.ses.med31[2] 0.66 1.00 4814 2747
## prjthfgt5_mostthft_devx2:rural.ses.med31[3] 0.63 1.00 4348 3114
## prjthfgt5_mostthft_devx2:rural.ses.med31[4] 0.77 1.00 4189 3030
## prjthfgt5_mostthft_devx2:rural.ses.med41[1] 0.71 1.00 5918 2759
## prjthfgt5_mostthft_devx2:rural.ses.med41[2] 0.69 1.00 5746 2851
## prjthfgt5_mostthft_devx2:rural.ses.med41[3] 0.70 1.00 4558 2227
## prjthfgt5_mostthft_devx2:rural.ses.med41[4] 0.70 1.00 5402 2656
## prjthreat_mostthft_devx21[1] 0.61 1.00 6686 2389
## prjthreat_mostthft_devx21[2] 0.61 1.00 6574 2820
## prjthreat_mostthft_devx21[3] 0.52 1.00 6143 3187
## prjthreat_mostthft_devx21[4] 0.56 1.00 7208 3295
## prjthreat_mostthft_av12x21[1] 0.31 1.00 8205 3081
## prjthreat_mostthft_av12x21[2] 0.31 1.00 6954 2289
## prjthreat_mostthft_av12x21[3] 0.31 1.00 7352 2110
## prjthreat_mostthft_av12x21[4] 0.32 1.00 6863 2538
## prjthreat_mostthft_av12x21[5] 0.33 1.00 7088 2585
## prjthreat_mostthft_av12x21[6] 0.32 1.00 6762 2775
## prjthreat_mostthft_av12x21[7] 0.32 1.00 5911 2589
## prjthreat_mostthft_av12x21[8] 0.31 1.00 5710 2796
## prjthreat_mostthft_devx2:rural.ses.med21[1] 0.76 1.00 4941 2328
## prjthreat_mostthft_devx2:rural.ses.med21[2] 0.69 1.00 6364 2504
## prjthreat_mostthft_devx2:rural.ses.med21[3] 0.66 1.00 3758 2566
## prjthreat_mostthft_devx2:rural.ses.med21[4] 0.74 1.00 5787 2953
## prjthreat_mostthft_devx2:rural.ses.med31[1] 0.70 1.00 5210 2282
## prjthreat_mostthft_devx2:rural.ses.med31[2] 0.67 1.00 5324 2291
## prjthreat_mostthft_devx2:rural.ses.med31[3] 0.70 1.00 5549 3370
## prjthreat_mostthft_devx2:rural.ses.med31[4] 0.77 1.00 5697 2880
## prjthreat_mostthft_devx2:rural.ses.med41[1] 0.74 1.00 5091 2930
## prjthreat_mostthft_devx2:rural.ses.med41[2] 0.65 1.00 5303 2475
## prjthreat_mostthft_devx2:rural.ses.med41[3] 0.68 1.00 6128 2877
## prjthreat_mostthft_devx2:rural.ses.med41[4] 0.75 1.00 5447 2966
## prjharm_mostthft_devx21[1] 0.58 1.00 7005 2370
## prjharm_mostthft_devx21[2] 0.63 1.00 5779 2673
## prjharm_mostthft_devx21[3] 0.53 1.00 5972 2964
## prjharm_mostthft_devx21[4] 0.53 1.00 7326 2936
## prjharm_mostthft_av12x21[1] 0.31 1.00 7420 2335
## prjharm_mostthft_av12x21[2] 0.33 1.00 6955 2460
## prjharm_mostthft_av12x21[3] 0.31 1.00 8367 2322
## prjharm_mostthft_av12x21[4] 0.32 1.00 7620 2378
## prjharm_mostthft_av12x21[5] 0.32 1.00 6557 2668
## prjharm_mostthft_av12x21[6] 0.32 1.00 6346 2455
## prjharm_mostthft_av12x21[7] 0.33 1.00 6161 2703
## prjharm_mostthft_av12x21[8] 0.32 1.00 6243 2539
## prjharm_mostthft_devx2:rural.ses.med21[1] 0.70 1.00 4873 2022
## prjharm_mostthft_devx2:rural.ses.med21[2] 0.70 1.00 5631 2738
## prjharm_mostthft_devx2:rural.ses.med21[3] 0.66 1.00 5387 2504
## prjharm_mostthft_devx2:rural.ses.med21[4] 0.74 1.00 5552 2605
## prjharm_mostthft_devx2:rural.ses.med31[1] 0.70 1.00 7447 2541
## prjharm_mostthft_devx2:rural.ses.med31[2] 0.66 1.00 6366 2493
## prjharm_mostthft_devx2:rural.ses.med31[3] 0.68 1.00 5467 2652
## prjharm_mostthft_devx2:rural.ses.med31[4] 0.77 1.00 4822 3187
## prjharm_mostthft_devx2:rural.ses.med41[1] 0.68 1.00 6246 2497
## prjharm_mostthft_devx2:rural.ses.med41[2] 0.66 1.00 6196 2447
## prjharm_mostthft_devx2:rural.ses.med41[3] 0.70 1.00 5585 2685
## prjharm_mostthft_devx2:rural.ses.med41[4] 0.78 1.00 4829 2706
## prjusedrg_mostthft_devx21[1] 0.60 1.00 6245 2542
## prjusedrg_mostthft_devx21[2] 0.57 1.00 8421 2855
## prjusedrg_mostthft_devx21[3] 0.56 1.00 6018 2594
## prjusedrg_mostthft_devx21[4] 0.57 1.00 6477 2578
## prjusedrg_mostthft_av12x21[1] 0.32 1.00 7675 2579
## prjusedrg_mostthft_av12x21[2] 0.31 1.00 6499 2700
## prjusedrg_mostthft_av12x21[3] 0.33 1.00 7381 2306
## prjusedrg_mostthft_av12x21[4] 0.33 1.00 7064 2285
## prjusedrg_mostthft_av12x21[5] 0.31 1.00 7037 2214
## prjusedrg_mostthft_av12x21[6] 0.31 1.00 6143 2535
## prjusedrg_mostthft_av12x21[7] 0.32 1.00 7631 2904
## prjusedrg_mostthft_av12x21[8] 0.32 1.00 7599 2442
## prjusedrg_mostthft_devx2:rural.ses.med21[1] 0.72 1.00 4603 2538
## prjusedrg_mostthft_devx2:rural.ses.med21[2] 0.65 1.00 6325 3034
## prjusedrg_mostthft_devx2:rural.ses.med21[3] 0.68 1.00 5566 2497
## prjusedrg_mostthft_devx2:rural.ses.med21[4] 0.74 1.00 5951 3026
## prjusedrg_mostthft_devx2:rural.ses.med31[1] 0.69 1.00 5144 2209
## prjusedrg_mostthft_devx2:rural.ses.med31[2] 0.62 1.00 3889 2329
## prjusedrg_mostthft_devx2:rural.ses.med31[3] 0.77 1.00 5309 2680
## prjusedrg_mostthft_devx2:rural.ses.med31[4] 0.74 1.00 5913 2561
## prjusedrg_mostthft_devx2:rural.ses.med41[1] 0.73 1.00 5306 2657
## prjusedrg_mostthft_devx2:rural.ses.med41[2] 0.64 1.00 4459 1802
## prjusedrg_mostthft_devx2:rural.ses.med41[3] 0.68 1.00 5368 2223
## prjusedrg_mostthft_devx2:rural.ses.med41[4] 0.76 1.00 4555 2823
## prjhack_mostthft_devx21[1] 0.59 1.00 6955 2197
## prjhack_mostthft_devx21[2] 0.59 1.00 7219 2810
## prjhack_mostthft_devx21[3] 0.57 1.00 6592 2629
## prjhack_mostthft_devx21[4] 0.56 1.00 6893 2685
## prjhack_mostthft_av12x21[1] 0.32 1.00 8880 2326
## prjhack_mostthft_av12x21[2] 0.32 1.00 6458 2297
## prjhack_mostthft_av12x21[3] 0.32 1.00 6483 2222
## prjhack_mostthft_av12x21[4] 0.32 1.00 6842 2618
## prjhack_mostthft_av12x21[5] 0.31 1.00 6307 2030
## prjhack_mostthft_av12x21[6] 0.32 1.00 7308 3049
## prjhack_mostthft_av12x21[7] 0.33 1.00 7624 3031
## prjhack_mostthft_av12x21[8] 0.32 1.00 6667 2713
## prjhack_mostthft_devx2:rural.ses.med21[1] 0.71 1.00 5313 2493
## prjhack_mostthft_devx2:rural.ses.med21[2] 0.63 1.00 5699 2376
## prjhack_mostthft_devx2:rural.ses.med21[3] 0.70 1.00 5876 2641
## prjhack_mostthft_devx2:rural.ses.med21[4] 0.71 1.00 4928 2369
## prjhack_mostthft_devx2:rural.ses.med31[1] 0.70 1.00 5427 2080
## prjhack_mostthft_devx2:rural.ses.med31[2] 0.66 1.00 5350 2291
## prjhack_mostthft_devx2:rural.ses.med31[3] 0.68 1.00 5934 2431
## prjhack_mostthft_devx2:rural.ses.med31[4] 0.75 1.00 5037 2162
## prjhack_mostthft_devx2:rural.ses.med41[1] 0.68 1.00 5532 2265
## prjhack_mostthft_devx2:rural.ses.med41[2] 0.67 1.00 4348 2350
## prjhack_mostthft_devx2:rural.ses.med41[3] 0.71 1.00 5742 2987
## prjhack_mostthft_devx2:rural.ses.med41[4] 0.76 1.00 4867 3025
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.1.6.4 Prior summary
out.chg.prjcrime.stthft.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## group resp dpar nlpar lb ub source
## default
## prjhack user
## prjhack user
## prjhack user
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjharm user
## prjharm user
## prjharm user
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthflt5 user
## prjthflt5 user
## prjthflt5 user
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthreat user
## prjthreat user
## prjthreat user
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjusedrg user
## prjusedrg user
## prjusedrg user
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## default
## prjhack user
## prjharm user
## prjthfgt5 user
## prjthflt5 user
## prjthreat user
## prjusedrg user
## prjhack 0 default
## prjharm 0 default
## prjthfgt5 0 default
## prjthflt5 0 default
## prjthreat 0 default
## prjusedrg 0 default
## id prjhack 0 (vectorized)
## id prjhack 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjusedrg 0 (vectorized)
## id prjusedrg 0 (vectorized)
## prjhack user
## prjhack default
## prjhack default
## prjhack default
## prjhack user
## prjharm user
## prjharm default
## prjharm default
## prjharm default
## prjharm user
## prjthfgt5 user
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 user
## prjthflt5 user
## prjthflt5 default
## prjthflt5 default
## prjthflt5 default
## prjthflt5 user
## prjthreat user
## prjthreat default
## prjthreat default
## prjthreat default
## prjthreat user
## prjusedrg user
## prjusedrg default
## prjusedrg default
## prjusedrg default
## prjusedrg user8.1.1.7 Corr X Community: stmug & prj crime
#Community Change: criminal intent items ~ mo(stmug)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmug_devx2',
resp = prjdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmug_av12x2',
resp = prjdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmug_devx21',
resp = prjdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmug_av12x21',
resp = prjdv_names)
)
# drop year from model to avoid inappropriately partially out systematic stress change differences.
# also, with two waves, can only add random int OR random slope for year
chg.prjcrime.stmug.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
mo(stmug_devx2) + mo(stmug_av12x2) +
rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stmug_comm_fit",
file_refit = "on_change"
)
out.chg.prjcrime.stmug.comm.fit <- ppchecks(chg.prjcrime.stmug.comm.fit)8.1.1.7.1 Coefficient plot (intervals)
out.chg.prjcrime.stmug.comm.fit[[10]]
8.1.1.7.2 PPcheck (density)
p1 <- out.chg.prjcrime.stmug.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stmug.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stmug.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stmug.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stmug.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stmug.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
8.1.1.7.3 Fit summary
out.chg.prjcrime.stmug.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## prjthreat ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## prjharm ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## prjusedrg ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## prjhack ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.93 0.56 2.96 5.13 1.01 1407
## sd(prjthfgt5_Intercept) 3.36 0.50 2.47 4.42 1.00 1583
## sd(prjthreat_Intercept) 3.24 0.56 2.25 4.45 1.00 1675
## sd(prjharm_Intercept) 2.99 0.54 2.03 4.16 1.00 1391
## sd(prjusedrg_Intercept) 2.85 0.52 1.95 3.93 1.00 1738
## sd(prjhack_Intercept) 0.85 0.55 0.04 2.02 1.00 723
## Tail_ESS
## sd(prjthflt5_Intercept) 2333
## sd(prjthfgt5_Intercept) 2468
## sd(prjthreat_Intercept) 2788
## sd(prjharm_Intercept) 2389
## sd(prjusedrg_Intercept) 2526
## sd(prjhack_Intercept) 1707
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_Intercept -5.61 0.84 -7.37 -4.06
## prjthfgt5_Intercept -5.31 0.80 -6.88 -3.78
## prjthreat_Intercept -6.07 0.91 -7.95 -4.39
## prjharm_Intercept -6.09 0.92 -7.96 -4.35
## prjusedrg_Intercept -6.01 0.95 -7.95 -4.21
## prjhack_Intercept -4.33 0.72 -5.87 -2.99
## prjthflt5_rural.ses.med2 -0.57 0.86 -2.26 1.15
## prjthflt5_rural.ses.med3 1.25 0.80 -0.43 2.75
## prjthflt5_rural.ses.med4 1.33 0.85 -0.44 2.86
## prjthfgt5_rural.ses.med2 -0.56 0.89 -2.26 1.22
## prjthfgt5_rural.ses.med3 1.08 0.77 -0.50 2.53
## prjthfgt5_rural.ses.med4 1.14 0.79 -0.48 2.61
## prjthreat_rural.ses.med2 -0.57 0.91 -2.29 1.24
## prjthreat_rural.ses.med3 0.57 0.79 -0.97 2.07
## prjthreat_rural.ses.med4 1.76 0.78 0.13 3.24
## prjharm_rural.ses.med2 -0.63 0.82 -2.26 1.04
## prjharm_rural.ses.med3 0.68 0.79 -0.81 2.26
## prjharm_rural.ses.med4 0.95 0.77 -0.63 2.44
## prjusedrg_rural.ses.med2 -0.37 0.84 -2.01 1.34
## prjusedrg_rural.ses.med3 -0.06 0.83 -1.64 1.61
## prjusedrg_rural.ses.med4 1.20 0.76 -0.31 2.70
## prjhack_rural.ses.med2 -0.52 0.86 -2.19 1.17
## prjhack_rural.ses.med3 0.01 0.76 -1.53 1.52
## prjhack_rural.ses.med4 1.04 0.72 -0.37 2.46
## prjthflt5_mostmug_devx2 -0.11 0.21 -0.51 0.29
## prjthflt5_mostmug_av12x2 -0.03 0.09 -0.23 0.14
## prjthflt5_mostmug_devx2:rural.ses.med2 -0.64 0.53 -1.78 0.37
## prjthflt5_mostmug_devx2:rural.ses.med3 0.11 0.45 -0.75 1.04
## prjthflt5_mostmug_devx2:rural.ses.med4 0.48 0.41 -0.27 1.34
## prjthfgt5_mostmug_devx2 -0.05 0.23 -0.50 0.39
## prjthfgt5_mostmug_av12x2 -0.04 0.09 -0.21 0.14
## prjthfgt5_mostmug_devx2:rural.ses.med2 -0.72 0.52 -1.79 0.25
## prjthfgt5_mostmug_devx2:rural.ses.med3 0.16 0.43 -0.71 1.04
## prjthfgt5_mostmug_devx2:rural.ses.med4 0.51 0.36 -0.17 1.27
## prjthreat_mostmug_devx2 -0.17 0.22 -0.62 0.26
## prjthreat_mostmug_av12x2 -0.03 0.09 -0.22 0.16
## prjthreat_mostmug_devx2:rural.ses.med2 -0.43 0.58 -1.58 0.67
## prjthreat_mostmug_devx2:rural.ses.med3 -0.05 0.49 -1.06 0.94
## prjthreat_mostmug_devx2:rural.ses.med4 -0.08 0.45 -1.09 0.72
## prjharm_mostmug_devx2 -0.08 0.23 -0.54 0.37
## prjharm_mostmug_av12x2 -0.01 0.10 -0.21 0.19
## prjharm_mostmug_devx2:rural.ses.med2 0.17 0.51 -0.80 1.21
## prjharm_mostmug_devx2:rural.ses.med3 -0.32 0.49 -1.37 0.61
## prjharm_mostmug_devx2:rural.ses.med4 0.04 0.42 -0.88 0.80
## prjusedrg_mostmug_devx2 -0.08 0.25 -0.58 0.40
## prjusedrg_mostmug_av12x2 -0.01 0.09 -0.19 0.17
## prjusedrg_mostmug_devx2:rural.ses.med2 -0.23 0.53 -1.36 0.76
## prjusedrg_mostmug_devx2:rural.ses.med3 -0.40 0.53 -1.53 0.59
## prjusedrg_mostmug_devx2:rural.ses.med4 0.46 0.39 -0.33 1.17
## prjhack_mostmug_devx2 -0.00 0.22 -0.46 0.42
## prjhack_mostmug_av12x2 0.01 0.08 -0.15 0.17
## prjhack_mostmug_devx2:rural.ses.med2 -0.40 0.55 -1.50 0.64
## prjhack_mostmug_devx2:rural.ses.med3 0.06 0.46 -0.88 0.91
## prjhack_mostmug_devx2:rural.ses.med4 -0.12 0.40 -0.99 0.60
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1.00 2539 3277
## prjthfgt5_Intercept 1.00 3211 2746
## prjthreat_Intercept 1.00 2898 3266
## prjharm_Intercept 1.00 2431 2803
## prjusedrg_Intercept 1.00 2667 3059
## prjhack_Intercept 1.00 2443 2750
## prjthflt5_rural.ses.med2 1.00 7097 2978
## prjthflt5_rural.ses.med3 1.00 4479 2914
## prjthflt5_rural.ses.med4 1.00 5271 3327
## prjthfgt5_rural.ses.med2 1.00 5456 3261
## prjthfgt5_rural.ses.med3 1.00 5171 3001
## prjthfgt5_rural.ses.med4 1.00 5538 3443
## prjthreat_rural.ses.med2 1.00 5373 3023
## prjthreat_rural.ses.med3 1.00 5602 3494
## prjthreat_rural.ses.med4 1.00 5477 3300
## prjharm_rural.ses.med2 1.00 6183 2838
## prjharm_rural.ses.med3 1.00 5148 3106
## prjharm_rural.ses.med4 1.00 5665 3036
## prjusedrg_rural.ses.med2 1.00 6701 3319
## prjusedrg_rural.ses.med3 1.00 6886 3069
## prjusedrg_rural.ses.med4 1.00 4548 2832
## prjhack_rural.ses.med2 1.00 6311 3195
## prjhack_rural.ses.med3 1.00 4771 2625
## prjhack_rural.ses.med4 1.00 5978 3418
## prjthflt5_mostmug_devx2 1.00 5334 3415
## prjthflt5_mostmug_av12x2 1.00 4515 3219
## prjthflt5_mostmug_devx2:rural.ses.med2 1.00 3939 2621
## prjthflt5_mostmug_devx2:rural.ses.med3 1.00 3244 2770
## prjthflt5_mostmug_devx2:rural.ses.med4 1.00 3493 2926
## prjthfgt5_mostmug_devx2 1.00 4508 3350
## prjthfgt5_mostmug_av12x2 1.00 4010 3522
## prjthfgt5_mostmug_devx2:rural.ses.med2 1.00 4169 3329
## prjthfgt5_mostmug_devx2:rural.ses.med3 1.00 3087 2550
## prjthfgt5_mostmug_devx2:rural.ses.med4 1.00 3929 3233
## prjthreat_mostmug_devx2 1.00 4787 3230
## prjthreat_mostmug_av12x2 1.00 5407 3274
## prjthreat_mostmug_devx2:rural.ses.med2 1.00 4628 3097
## prjthreat_mostmug_devx2:rural.ses.med3 1.00 3828 2955
## prjthreat_mostmug_devx2:rural.ses.med4 1.00 3702 2880
## prjharm_mostmug_devx2 1.00 4737 3089
## prjharm_mostmug_av12x2 1.00 5645 2729
## prjharm_mostmug_devx2:rural.ses.med2 1.00 3651 3071
## prjharm_mostmug_devx2:rural.ses.med3 1.00 4340 2845
## prjharm_mostmug_devx2:rural.ses.med4 1.00 5111 3150
## prjusedrg_mostmug_devx2 1.00 3666 3453
## prjusedrg_mostmug_av12x2 1.00 5037 3276
## prjusedrg_mostmug_devx2:rural.ses.med2 1.00 4916 3254
## prjusedrg_mostmug_devx2:rural.ses.med3 1.00 4719 3320
## prjusedrg_mostmug_devx2:rural.ses.med4 1.00 3223 2668
## prjhack_mostmug_devx2 1.00 5358 3380
## prjhack_mostmug_av12x2 1.00 7504 2997
## prjhack_mostmug_devx2:rural.ses.med2 1.00 4812 3149
## prjhack_mostmug_devx2:rural.ses.med3 1.00 3935 2853
## prjhack_mostmug_devx2:rural.ses.med4 1.00 4628 3081
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI
## prjthflt5_mostmug_devx21[1] 0.27 0.15 0.04 0.60
## prjthflt5_mostmug_devx21[2] 0.26 0.14 0.04 0.58
## prjthflt5_mostmug_devx21[3] 0.23 0.14 0.03 0.57
## prjthflt5_mostmug_devx21[4] 0.25 0.15 0.03 0.59
## prjthflt5_mostmug_av12x21[1] 0.12 0.08 0.02 0.31
## prjthflt5_mostmug_av12x21[2] 0.12 0.08 0.02 0.31
## prjthflt5_mostmug_av12x21[3] 0.12 0.08 0.02 0.31
## prjthflt5_mostmug_av12x21[4] 0.13 0.08 0.02 0.32
## prjthflt5_mostmug_av12x21[5] 0.13 0.08 0.02 0.33
## prjthflt5_mostmug_av12x21[6] 0.13 0.08 0.02 0.33
## prjthflt5_mostmug_av12x21[7] 0.13 0.08 0.02 0.33
## prjthflt5_mostmug_av12x21[8] 0.13 0.08 0.02 0.33
## prjthflt5_mostmug_devx2:rural.ses.med21[1] 0.31 0.21 0.01 0.76
## prjthflt5_mostmug_devx2:rural.ses.med21[2] 0.23 0.18 0.01 0.66
## prjthflt5_mostmug_devx2:rural.ses.med21[3] 0.21 0.17 0.01 0.63
## prjthflt5_mostmug_devx2:rural.ses.med21[4] 0.26 0.19 0.01 0.70
## prjthflt5_mostmug_devx2:rural.ses.med31[1] 0.28 0.21 0.01 0.75
## prjthflt5_mostmug_devx2:rural.ses.med31[2] 0.21 0.18 0.01 0.66
## prjthflt5_mostmug_devx2:rural.ses.med31[3] 0.22 0.18 0.01 0.65
## prjthflt5_mostmug_devx2:rural.ses.med31[4] 0.30 0.22 0.01 0.77
## prjthflt5_mostmug_devx2:rural.ses.med41[1] 0.33 0.21 0.02 0.78
## prjthflt5_mostmug_devx2:rural.ses.med41[2] 0.23 0.18 0.01 0.66
## prjthflt5_mostmug_devx2:rural.ses.med41[3] 0.19 0.16 0.01 0.60
## prjthflt5_mostmug_devx2:rural.ses.med41[4] 0.25 0.19 0.01 0.69
## prjthfgt5_mostmug_devx21[1] 0.26 0.15 0.04 0.61
## prjthfgt5_mostmug_devx21[2] 0.25 0.15 0.03 0.58
## prjthfgt5_mostmug_devx21[3] 0.24 0.15 0.03 0.60
## prjthfgt5_mostmug_devx21[4] 0.25 0.14 0.03 0.58
## prjthfgt5_mostmug_av12x21[1] 0.12 0.08 0.02 0.31
## prjthfgt5_mostmug_av12x21[2] 0.12 0.08 0.02 0.31
## prjthfgt5_mostmug_av12x21[3] 0.12 0.08 0.02 0.31
## prjthfgt5_mostmug_av12x21[4] 0.13 0.08 0.02 0.33
## prjthfgt5_mostmug_av12x21[5] 0.13 0.08 0.02 0.33
## prjthfgt5_mostmug_av12x21[6] 0.13 0.08 0.02 0.33
## prjthfgt5_mostmug_av12x21[7] 0.13 0.09 0.01 0.34
## prjthfgt5_mostmug_av12x21[8] 0.13 0.08 0.02 0.33
## prjthfgt5_mostmug_devx2:rural.ses.med21[1] 0.31 0.21 0.01 0.76
## prjthfgt5_mostmug_devx2:rural.ses.med21[2] 0.25 0.19 0.01 0.70
## prjthfgt5_mostmug_devx2:rural.ses.med21[3] 0.19 0.16 0.00 0.59
## prjthfgt5_mostmug_devx2:rural.ses.med21[4] 0.25 0.19 0.01 0.70
## prjthfgt5_mostmug_devx2:rural.ses.med31[1] 0.28 0.21 0.01 0.75
## prjthfgt5_mostmug_devx2:rural.ses.med31[2] 0.20 0.18 0.01 0.65
## prjthfgt5_mostmug_devx2:rural.ses.med31[3] 0.23 0.19 0.01 0.69
## prjthfgt5_mostmug_devx2:rural.ses.med31[4] 0.29 0.21 0.01 0.76
## prjthfgt5_mostmug_devx2:rural.ses.med41[1] 0.29 0.20 0.01 0.72
## prjthfgt5_mostmug_devx2:rural.ses.med41[2] 0.25 0.19 0.01 0.68
## prjthfgt5_mostmug_devx2:rural.ses.med41[3] 0.22 0.17 0.01 0.66
## prjthfgt5_mostmug_devx2:rural.ses.med41[4] 0.24 0.19 0.01 0.67
## prjthreat_mostmug_devx21[1] 0.28 0.16 0.04 0.61
## prjthreat_mostmug_devx21[2] 0.26 0.14 0.04 0.58
## prjthreat_mostmug_devx21[3] 0.22 0.13 0.03 0.53
## prjthreat_mostmug_devx21[4] 0.25 0.14 0.04 0.57
## prjthreat_mostmug_av12x21[1] 0.12 0.08 0.02 0.31
## prjthreat_mostmug_av12x21[2] 0.12 0.08 0.02 0.31
## prjthreat_mostmug_av12x21[3] 0.13 0.08 0.02 0.32
## prjthreat_mostmug_av12x21[4] 0.12 0.08 0.02 0.31
## prjthreat_mostmug_av12x21[5] 0.13 0.08 0.02 0.33
## prjthreat_mostmug_av12x21[6] 0.13 0.08 0.02 0.32
## prjthreat_mostmug_av12x21[7] 0.13 0.08 0.02 0.32
## prjthreat_mostmug_av12x21[8] 0.13 0.08 0.02 0.33
## prjthreat_mostmug_devx2:rural.ses.med21[1] 0.29 0.21 0.01 0.76
## prjthreat_mostmug_devx2:rural.ses.med21[2] 0.22 0.18 0.01 0.67
## prjthreat_mostmug_devx2:rural.ses.med21[3] 0.21 0.18 0.00 0.66
## prjthreat_mostmug_devx2:rural.ses.med21[4] 0.28 0.21 0.01 0.74
## prjthreat_mostmug_devx2:rural.ses.med31[1] 0.24 0.19 0.01 0.68
## prjthreat_mostmug_devx2:rural.ses.med31[2] 0.23 0.18 0.01 0.68
## prjthreat_mostmug_devx2:rural.ses.med31[3] 0.23 0.18 0.01 0.68
## prjthreat_mostmug_devx2:rural.ses.med31[4] 0.30 0.21 0.01 0.77
## prjthreat_mostmug_devx2:rural.ses.med41[1] 0.24 0.20 0.01 0.71
## prjthreat_mostmug_devx2:rural.ses.med41[2] 0.22 0.18 0.01 0.68
## prjthreat_mostmug_devx2:rural.ses.med41[3] 0.24 0.19 0.01 0.68
## prjthreat_mostmug_devx2:rural.ses.med41[4] 0.30 0.22 0.01 0.77
## prjharm_mostmug_devx21[1] 0.27 0.15 0.04 0.61
## prjharm_mostmug_devx21[2] 0.25 0.14 0.04 0.57
## prjharm_mostmug_devx21[3] 0.23 0.14 0.03 0.55
## prjharm_mostmug_devx21[4] 0.25 0.14 0.04 0.59
## prjharm_mostmug_av12x21[1] 0.12 0.08 0.02 0.32
## prjharm_mostmug_av12x21[2] 0.12 0.08 0.02 0.32
## prjharm_mostmug_av12x21[3] 0.12 0.08 0.02 0.32
## prjharm_mostmug_av12x21[4] 0.12 0.08 0.02 0.31
## prjharm_mostmug_av12x21[5] 0.13 0.08 0.02 0.32
## prjharm_mostmug_av12x21[6] 0.13 0.08 0.02 0.32
## prjharm_mostmug_av12x21[7] 0.13 0.08 0.02 0.34
## prjharm_mostmug_av12x21[8] 0.13 0.08 0.01 0.32
## prjharm_mostmug_devx2:rural.ses.med21[1] 0.22 0.18 0.01 0.68
## prjharm_mostmug_devx2:rural.ses.med21[2] 0.21 0.18 0.01 0.65
## prjharm_mostmug_devx2:rural.ses.med21[3] 0.25 0.19 0.01 0.71
## prjharm_mostmug_devx2:rural.ses.med21[4] 0.31 0.22 0.01 0.79
## prjharm_mostmug_devx2:rural.ses.med31[1] 0.22 0.18 0.01 0.65
## prjharm_mostmug_devx2:rural.ses.med31[2] 0.24 0.18 0.01 0.68
## prjharm_mostmug_devx2:rural.ses.med31[3] 0.24 0.18 0.01 0.67
## prjharm_mostmug_devx2:rural.ses.med31[4] 0.29 0.20 0.02 0.73
## prjharm_mostmug_devx2:rural.ses.med41[1] 0.24 0.19 0.01 0.70
## prjharm_mostmug_devx2:rural.ses.med41[2] 0.23 0.19 0.01 0.69
## prjharm_mostmug_devx2:rural.ses.med41[3] 0.24 0.19 0.01 0.71
## prjharm_mostmug_devx2:rural.ses.med41[4] 0.28 0.21 0.01 0.76
## prjusedrg_mostmug_devx21[1] 0.27 0.16 0.04 0.61
## prjusedrg_mostmug_devx21[2] 0.24 0.14 0.03 0.57
## prjusedrg_mostmug_devx21[3] 0.24 0.16 0.03 0.61
## prjusedrg_mostmug_devx21[4] 0.25 0.15 0.03 0.58
## prjusedrg_mostmug_av12x21[1] 0.12 0.08 0.02 0.32
## prjusedrg_mostmug_av12x21[2] 0.12 0.08 0.02 0.32
## prjusedrg_mostmug_av12x21[3] 0.12 0.08 0.02 0.32
## prjusedrg_mostmug_av12x21[4] 0.12 0.08 0.02 0.31
## prjusedrg_mostmug_av12x21[5] 0.13 0.08 0.02 0.32
## prjusedrg_mostmug_av12x21[6] 0.13 0.08 0.02 0.32
## prjusedrg_mostmug_av12x21[7] 0.13 0.08 0.02 0.32
## prjusedrg_mostmug_av12x21[8] 0.13 0.08 0.02 0.32
## prjusedrg_mostmug_devx2:rural.ses.med21[1] 0.26 0.19 0.01 0.70
## prjusedrg_mostmug_devx2:rural.ses.med21[2] 0.22 0.18 0.01 0.66
## prjusedrg_mostmug_devx2:rural.ses.med21[3] 0.24 0.19 0.01 0.69
## prjusedrg_mostmug_devx2:rural.ses.med21[4] 0.29 0.21 0.01 0.76
## prjusedrg_mostmug_devx2:rural.ses.med31[1] 0.26 0.20 0.01 0.71
## prjusedrg_mostmug_devx2:rural.ses.med31[2] 0.23 0.18 0.01 0.66
## prjusedrg_mostmug_devx2:rural.ses.med31[3] 0.24 0.18 0.01 0.67
## prjusedrg_mostmug_devx2:rural.ses.med31[4] 0.28 0.21 0.01 0.73
## prjusedrg_mostmug_devx2:rural.ses.med41[1] 0.24 0.18 0.01 0.67
## prjusedrg_mostmug_devx2:rural.ses.med41[2] 0.17 0.16 0.00 0.61
## prjusedrg_mostmug_devx2:rural.ses.med41[3] 0.38 0.22 0.02 0.81
## prjusedrg_mostmug_devx2:rural.ses.med41[4] 0.21 0.18 0.01 0.67
## prjhack_mostmug_devx21[1] 0.26 0.15 0.04 0.59
## prjhack_mostmug_devx21[2] 0.24 0.14 0.04 0.56
## prjhack_mostmug_devx21[3] 0.25 0.15 0.04 0.59
## prjhack_mostmug_devx21[4] 0.25 0.14 0.04 0.57
## prjhack_mostmug_av12x21[1] 0.12 0.08 0.02 0.32
## prjhack_mostmug_av12x21[2] 0.12 0.08 0.02 0.32
## prjhack_mostmug_av12x21[3] 0.12 0.08 0.02 0.31
## prjhack_mostmug_av12x21[4] 0.13 0.08 0.02 0.32
## prjhack_mostmug_av12x21[5] 0.12 0.08 0.02 0.31
## prjhack_mostmug_av12x21[6] 0.13 0.08 0.02 0.32
## prjhack_mostmug_av12x21[7] 0.13 0.08 0.02 0.32
## prjhack_mostmug_av12x21[8] 0.13 0.08 0.02 0.32
## prjhack_mostmug_devx2:rural.ses.med21[1] 0.28 0.20 0.01 0.72
## prjhack_mostmug_devx2:rural.ses.med21[2] 0.24 0.19 0.01 0.68
## prjhack_mostmug_devx2:rural.ses.med21[3] 0.21 0.18 0.01 0.66
## prjhack_mostmug_devx2:rural.ses.med21[4] 0.27 0.20 0.01 0.71
## prjhack_mostmug_devx2:rural.ses.med31[1] 0.23 0.19 0.01 0.69
## prjhack_mostmug_devx2:rural.ses.med31[2] 0.23 0.19 0.01 0.68
## prjhack_mostmug_devx2:rural.ses.med31[3] 0.24 0.19 0.01 0.69
## prjhack_mostmug_devx2:rural.ses.med31[4] 0.30 0.22 0.01 0.77
## prjhack_mostmug_devx2:rural.ses.med41[1] 0.24 0.19 0.01 0.69
## prjhack_mostmug_devx2:rural.ses.med41[2] 0.23 0.18 0.01 0.66
## prjhack_mostmug_devx2:rural.ses.med41[3] 0.24 0.19 0.01 0.67
## prjhack_mostmug_devx2:rural.ses.med41[4] 0.30 0.21 0.01 0.78
## Rhat Bulk_ESS Tail_ESS
## prjthflt5_mostmug_devx21[1] 1.00 8031 2284
## prjthflt5_mostmug_devx21[2] 1.00 6573 2778
## prjthflt5_mostmug_devx21[3] 1.00 6545 2862
## prjthflt5_mostmug_devx21[4] 1.00 7110 2711
## prjthflt5_mostmug_av12x21[1] 1.00 5692 2680
## prjthflt5_mostmug_av12x21[2] 1.00 8280 2743
## prjthflt5_mostmug_av12x21[3] 1.00 7280 2614
## prjthflt5_mostmug_av12x21[4] 1.00 9993 2960
## prjthflt5_mostmug_av12x21[5] 1.00 6827 2392
## prjthflt5_mostmug_av12x21[6] 1.00 7895 2713
## prjthflt5_mostmug_av12x21[7] 1.00 7285 2647
## prjthflt5_mostmug_av12x21[8] 1.00 8855 2850
## prjthflt5_mostmug_devx2:rural.ses.med21[1] 1.00 5210 2462
## prjthflt5_mostmug_devx2:rural.ses.med21[2] 1.00 5881 2660
## prjthflt5_mostmug_devx2:rural.ses.med21[3] 1.00 5507 2618
## prjthflt5_mostmug_devx2:rural.ses.med21[4] 1.00 5168 2633
## prjthflt5_mostmug_devx2:rural.ses.med31[1] 1.00 4283 2620
## prjthflt5_mostmug_devx2:rural.ses.med31[2] 1.00 5302 2781
## prjthflt5_mostmug_devx2:rural.ses.med31[3] 1.00 5815 3061
## prjthflt5_mostmug_devx2:rural.ses.med31[4] 1.00 5779 2507
## prjthflt5_mostmug_devx2:rural.ses.med41[1] 1.00 4910 2775
## prjthflt5_mostmug_devx2:rural.ses.med41[2] 1.00 5728 2542
## prjthflt5_mostmug_devx2:rural.ses.med41[3] 1.00 4958 2419
## prjthflt5_mostmug_devx2:rural.ses.med41[4] 1.00 5582 2981
## prjthfgt5_mostmug_devx21[1] 1.00 8221 2824
## prjthfgt5_mostmug_devx21[2] 1.00 5935 3008
## prjthfgt5_mostmug_devx21[3] 1.00 5563 3135
## prjthfgt5_mostmug_devx21[4] 1.00 6977 2911
## prjthfgt5_mostmug_av12x21[1] 1.00 6404 2535
## prjthfgt5_mostmug_av12x21[2] 1.00 6310 2450
## prjthfgt5_mostmug_av12x21[3] 1.00 7374 2792
## prjthfgt5_mostmug_av12x21[4] 1.00 6929 2432
## prjthfgt5_mostmug_av12x21[5] 1.00 6858 2286
## prjthfgt5_mostmug_av12x21[6] 1.00 7205 2769
## prjthfgt5_mostmug_av12x21[7] 1.00 9768 2379
## prjthfgt5_mostmug_av12x21[8] 1.00 6018 2681
## prjthfgt5_mostmug_devx2:rural.ses.med21[1] 1.00 6000 2789
## prjthfgt5_mostmug_devx2:rural.ses.med21[2] 1.00 5938 2432
## prjthfgt5_mostmug_devx2:rural.ses.med21[3] 1.00 5810 2795
## prjthfgt5_mostmug_devx2:rural.ses.med21[4] 1.00 6207 2855
## prjthfgt5_mostmug_devx2:rural.ses.med31[1] 1.00 5774 2986
## prjthfgt5_mostmug_devx2:rural.ses.med31[2] 1.00 6035 2659
## prjthfgt5_mostmug_devx2:rural.ses.med31[3] 1.00 4772 3032
## prjthfgt5_mostmug_devx2:rural.ses.med31[4] 1.00 5466 3302
## prjthfgt5_mostmug_devx2:rural.ses.med41[1] 1.00 5946 2878
## prjthfgt5_mostmug_devx2:rural.ses.med41[2] 1.00 6035 3026
## prjthfgt5_mostmug_devx2:rural.ses.med41[3] 1.00 5259 2755
## prjthfgt5_mostmug_devx2:rural.ses.med41[4] 1.00 5815 2475
## prjthreat_mostmug_devx21[1] 1.00 7022 2923
## prjthreat_mostmug_devx21[2] 1.00 7772 2959
## prjthreat_mostmug_devx21[3] 1.00 6373 3275
## prjthreat_mostmug_devx21[4] 1.00 9105 2980
## prjthreat_mostmug_av12x21[1] 1.00 8181 2571
## prjthreat_mostmug_av12x21[2] 1.00 6996 2467
## prjthreat_mostmug_av12x21[3] 1.00 8249 2506
## prjthreat_mostmug_av12x21[4] 1.00 6863 2403
## prjthreat_mostmug_av12x21[5] 1.00 8816 2636
## prjthreat_mostmug_av12x21[6] 1.00 7383 2480
## prjthreat_mostmug_av12x21[7] 1.00 7315 2482
## prjthreat_mostmug_av12x21[8] 1.00 6595 2559
## prjthreat_mostmug_devx2:rural.ses.med21[1] 1.00 5948 2454
## prjthreat_mostmug_devx2:rural.ses.med21[2] 1.00 5805 2807
## prjthreat_mostmug_devx2:rural.ses.med21[3] 1.00 5640 2574
## prjthreat_mostmug_devx2:rural.ses.med21[4] 1.00 5639 2542
## prjthreat_mostmug_devx2:rural.ses.med31[1] 1.00 6587 2343
## prjthreat_mostmug_devx2:rural.ses.med31[2] 1.00 6724 2730
## prjthreat_mostmug_devx2:rural.ses.med31[3] 1.00 5665 2172
## prjthreat_mostmug_devx2:rural.ses.med31[4] 1.00 5607 2804
## prjthreat_mostmug_devx2:rural.ses.med41[1] 1.00 5618 2498
## prjthreat_mostmug_devx2:rural.ses.med41[2] 1.00 6511 2666
## prjthreat_mostmug_devx2:rural.ses.med41[3] 1.00 5393 2610
## prjthreat_mostmug_devx2:rural.ses.med41[4] 1.00 5642 2812
## prjharm_mostmug_devx21[1] 1.00 7339 2911
## prjharm_mostmug_devx21[2] 1.00 8401 2926
## prjharm_mostmug_devx21[3] 1.00 4581 2946
## prjharm_mostmug_devx21[4] 1.00 8157 3024
## prjharm_mostmug_av12x21[1] 1.00 6356 2147
## prjharm_mostmug_av12x21[2] 1.00 8491 2297
## prjharm_mostmug_av12x21[3] 1.00 7313 2770
## prjharm_mostmug_av12x21[4] 1.00 7160 2546
## prjharm_mostmug_av12x21[5] 1.00 6993 2730
## prjharm_mostmug_av12x21[6] 1.00 7647 2066
## prjharm_mostmug_av12x21[7] 1.00 7694 2805
## prjharm_mostmug_av12x21[8] 1.00 7661 2574
## prjharm_mostmug_devx2:rural.ses.med21[1] 1.00 5662 2840
## prjharm_mostmug_devx2:rural.ses.med21[2] 1.00 5350 2989
## prjharm_mostmug_devx2:rural.ses.med21[3] 1.00 5892 2473
## prjharm_mostmug_devx2:rural.ses.med21[4] 1.00 5482 3317
## prjharm_mostmug_devx2:rural.ses.med31[1] 1.00 6906 2825
## prjharm_mostmug_devx2:rural.ses.med31[2] 1.00 5568 2662
## prjharm_mostmug_devx2:rural.ses.med31[3] 1.00 5571 2522
## prjharm_mostmug_devx2:rural.ses.med31[4] 1.00 5737 3239
## prjharm_mostmug_devx2:rural.ses.med41[1] 1.00 6880 2478
## prjharm_mostmug_devx2:rural.ses.med41[2] 1.00 6054 2589
## prjharm_mostmug_devx2:rural.ses.med41[3] 1.00 6702 2560
## prjharm_mostmug_devx2:rural.ses.med41[4] 1.00 5330 3100
## prjusedrg_mostmug_devx21[1] 1.00 7049 3081
## prjusedrg_mostmug_devx21[2] 1.00 7461 2678
## prjusedrg_mostmug_devx21[3] 1.00 4667 3640
## prjusedrg_mostmug_devx21[4] 1.00 7494 2551
## prjusedrg_mostmug_av12x21[1] 1.00 7900 2115
## prjusedrg_mostmug_av12x21[2] 1.00 7274 2474
## prjusedrg_mostmug_av12x21[3] 1.00 6768 2245
## prjusedrg_mostmug_av12x21[4] 1.00 6384 2204
## prjusedrg_mostmug_av12x21[5] 1.00 8128 2452
## prjusedrg_mostmug_av12x21[6] 1.00 7306 2437
## prjusedrg_mostmug_av12x21[7] 1.00 6804 2943
## prjusedrg_mostmug_av12x21[8] 1.00 6599 2285
## prjusedrg_mostmug_devx2:rural.ses.med21[1] 1.00 7258 2935
## prjusedrg_mostmug_devx2:rural.ses.med21[2] 1.00 6158 2509
## prjusedrg_mostmug_devx2:rural.ses.med21[3] 1.00 7100 2542
## prjusedrg_mostmug_devx2:rural.ses.med21[4] 1.00 5194 2535
## prjusedrg_mostmug_devx2:rural.ses.med31[1] 1.00 7164 2990
## prjusedrg_mostmug_devx2:rural.ses.med31[2] 1.00 6471 2640
## prjusedrg_mostmug_devx2:rural.ses.med31[3] 1.00 6039 2656
## prjusedrg_mostmug_devx2:rural.ses.med31[4] 1.00 6666 2632
## prjusedrg_mostmug_devx2:rural.ses.med41[1] 1.00 6059 2495
## prjusedrg_mostmug_devx2:rural.ses.med41[2] 1.00 5714 2704
## prjusedrg_mostmug_devx2:rural.ses.med41[3] 1.00 3768 3050
## prjusedrg_mostmug_devx2:rural.ses.med41[4] 1.00 4853 2519
## prjhack_mostmug_devx21[1] 1.00 7506 2969
## prjhack_mostmug_devx21[2] 1.00 9142 2935
## prjhack_mostmug_devx21[3] 1.00 8376 2677
## prjhack_mostmug_devx21[4] 1.00 7240 2615
## prjhack_mostmug_av12x21[1] 1.00 6601 2395
## prjhack_mostmug_av12x21[2] 1.00 6174 2284
## prjhack_mostmug_av12x21[3] 1.00 8560 2569
## prjhack_mostmug_av12x21[4] 1.00 6944 2547
## prjhack_mostmug_av12x21[5] 1.00 8402 2731
## prjhack_mostmug_av12x21[6] 1.00 7109 2670
## prjhack_mostmug_av12x21[7] 1.00 7032 2771
## prjhack_mostmug_av12x21[8] 1.00 7120 2824
## prjhack_mostmug_devx2:rural.ses.med21[1] 1.00 6510 2822
## prjhack_mostmug_devx2:rural.ses.med21[2] 1.00 6902 2784
## prjhack_mostmug_devx2:rural.ses.med21[3] 1.00 5390 2904
## prjhack_mostmug_devx2:rural.ses.med21[4] 1.00 5803 2759
## prjhack_mostmug_devx2:rural.ses.med31[1] 1.00 7159 2428
## prjhack_mostmug_devx2:rural.ses.med31[2] 1.00 5534 2300
## prjhack_mostmug_devx2:rural.ses.med31[3] 1.00 5422 2944
## prjhack_mostmug_devx2:rural.ses.med31[4] 1.00 6351 2697
## prjhack_mostmug_devx2:rural.ses.med41[1] 1.00 5596 2678
## prjhack_mostmug_devx2:rural.ses.med41[2] 1.00 7852 2668
## prjhack_mostmug_devx2:rural.ses.med41[3] 1.00 6035 3048
## prjhack_mostmug_devx2:rural.ses.med41[4] 1.00 5603 2626
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.1.7.4 Prior summary
out.chg.prjcrime.stmug.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## group resp dpar nlpar lb ub source
## default
## prjhack user
## prjhack user
## prjhack user
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjhack (vectorized)
## prjharm user
## prjharm user
## prjharm user
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjharm (vectorized)
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 user
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthfgt5 (vectorized)
## prjthflt5 user
## prjthflt5 user
## prjthflt5 user
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthflt5 (vectorized)
## prjthreat user
## prjthreat user
## prjthreat user
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjthreat (vectorized)
## prjusedrg user
## prjusedrg user
## prjusedrg user
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## prjusedrg (vectorized)
## default
## prjhack user
## prjharm user
## prjthfgt5 user
## prjthflt5 user
## prjthreat user
## prjusedrg user
## prjhack 0 default
## prjharm 0 default
## prjthfgt5 0 default
## prjthflt5 0 default
## prjthreat 0 default
## prjusedrg 0 default
## id prjhack 0 (vectorized)
## id prjhack 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjharm 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthfgt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthflt5 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjthreat 0 (vectorized)
## id prjusedrg 0 (vectorized)
## id prjusedrg 0 (vectorized)
## prjhack user
## prjhack default
## prjhack default
## prjhack default
## prjhack user
## prjharm user
## prjharm default
## prjharm default
## prjharm default
## prjharm user
## prjthfgt5 user
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 default
## prjthfgt5 user
## prjthflt5 user
## prjthflt5 default
## prjthflt5 default
## prjthflt5 default
## prjthflt5 user
## prjthreat user
## prjthreat default
## prjthreat default
## prjthreat default
## prjthreat user
## prjusedrg user
## prjusedrg default
## prjusedrg default
## prjusedrg default
## prjusedrg user8.1.1.8 Corr X Community: stress & “any” crim intent
#Community Change: any criminal intent ~ mo(stress)
#Create function for repetitive prior settings
setmyprior <- function(mochgcoefname, moavcoefname,
simochgcoefname, simoavcoefname) {
c(
set_prior('normal(0, 2)', class = 'Intercept'),
set_prior('normal(0, 1)', class = 'b'),
set_prior('normal(0, 0.25)', class = 'b', coef = mochgcoefname), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = moavcoefname), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = simochgcoefname),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = simoavcoefname))
}
#Update function to call all ppchecks for bivar any crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit)
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95)
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95)
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, plotcoefs, plotcoefs2)
return(allchecks)
}
myprior <- setmyprior('mostmony_devx2', 'mostmony_av12x2',
'mostmony_devx21', 'mostmony_av12x21')
chg.anyprjcrime.stmony.comm.fit <- brm(
prjany ~ 1 +
mo(stmony_devx2) + mo(stmony_av12x2) +
rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stmony_comm_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stmony.comm.fit <- ppchecks(chg.anyprjcrime.stmony.comm.fit)
myprior <- setmyprior('mosttran_devx2', 'mosttran_av12x2',
'mosttran_devx21', 'mosttran_av12x21')
chg.anyprjcrime.sttran.comm.fit <- brm(
prjany ~ 1 +
mo(sttran_devx2) + mo(sttran_av12x2) +
rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_sttran_comm_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.sttran.comm.fit <- ppchecks(chg.anyprjcrime.sttran.comm.fit)
myprior <- setmyprior('mostresp_devx2', 'mostresp_av12x2',
'mostresp_devx21', 'mostresp_av12x21')
chg.anyprjcrime.stresp.comm.fit <- brm(
prjany ~ 1 +
mo(stresp_devx2) + mo(stresp_av12x2) +
rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stresp_comm_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stresp.comm.fit <- ppchecks(chg.anyprjcrime.stresp.comm.fit)
myprior <- setmyprior('mostfair_devx2', 'mostfair_av12x2',
'mostfair_devx21', 'mostfair_av12x21')
chg.anyprjcrime.stfair.comm.fit <- brm(
prjany ~ 1 +
mo(stfair_devx2) + mo(stfair_av12x2) +
rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stfair_comm_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stfair.comm.fit <- ppchecks(chg.anyprjcrime.stfair.comm.fit)
myprior <- setmyprior('mostjob_devx2', 'mostjob_av12x2',
'mostjob_devx21', 'mostjob_av12x21')
chg.anyprjcrime.stjob.comm.fit <- brm(
prjany ~ 1 +
mo(stjob_devx2) + mo(stjob_av12x2) +
rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stjob_comm_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stjob.comm.fit <- ppchecks(chg.anyprjcrime.stjob.comm.fit)
myprior <- setmyprior('mostthft_devx2', 'mostthft_av12x2',
'mostthft_devx21', 'mostthft_av12x21')
chg.anyprjcrime.stthft.comm.fit <- brm(
prjany ~ 1 +
mo(stthft_devx2) + mo(stthft_av12x2) +
rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stthft_comm_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stthft.comm.fit <- ppchecks(chg.anyprjcrime.stthft.comm.fit)
myprior <- setmyprior('mostmug_devx2', 'mostmug_av12x2',
'mostmug_devx21', 'mostmug_av12x21')
chg.anyprjcrime.stmug.comm.fit <- brm(
prjany ~ 1 +
mo(stmug_devx2) + mo(stmug_av12x2) +
rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id),
data = stress.long, family = "bernoulli", prior = myprior,
cores = nCoresphys, chains = 4, backend = "cmdstanr", seed = 8675309,
file = "Models/chg_anyprjcrime_stmug_comm_fit", file_refit = "on_change"
)
out.chg.anyprjcrime.stmug.comm.fit<- ppchecks(chg.anyprjcrime.stmug.comm.fit)8.1.1.8.1 Coefficient plot (intervals)
p1 <- out.chg.anyprjcrime.stmony.comm.fit[[5]]
p2 <- out.chg.anyprjcrime.sttran.comm.fit[[5]]
p3 <- out.chg.anyprjcrime.stresp.comm.fit[[5]]
p4 <- out.chg.anyprjcrime.stfair.comm.fit[[5]]
p5 <- out.chg.anyprjcrime.stjob.comm.fit[[5]]
p6 <- out.chg.anyprjcrime.stthft.comm.fit[[5]]
p7 <- out.chg.anyprjcrime.stmug.comm.fit[[5]]
playout <- '
AB
CD
E#
FG
'
p1 + p2 + p3 + p4 + p5 + p6 + p7 +
plot_layout(design = playout) +
plot_annotation(
title = 'Coefficient plot',
subtitle = 'Posterior intervals for monotonic ordinal within-person ("devx2") and between-person ("av12x2") stress\ncoefficients predicting "any criminal intent" with medians, 80% (thick line), and 95% (thin line) intervals')
8.1.1.8.2 PPcheck (density)
p1 <- out.chg.anyprjcrime.stmony.comm.fit[[3]] + labs(title = "Any crime intent/stmony (chg)")
p2 <- out.chg.anyprjcrime.sttran.comm.fit[[3]] + labs(title = "Any crime intent/sttran (chg)")
p3 <- out.chg.anyprjcrime.stresp.comm.fit[[3]] + labs(title = "Any crime intent/stresp (chg)")
p4 <- out.chg.anyprjcrime.stfair.comm.fit[[3]] + labs(title = "Any crime intent/stfair (chg)")
p5 <- out.chg.anyprjcrime.stjob.comm.fit[[3]] + labs(title = "Any crime intent/stjob (chg)")
p6 <- out.chg.anyprjcrime.stthft.comm.fit[[3]] + labs(title = "Any crime intent/stthft (chg)")
p7 <- out.chg.anyprjcrime.stmug.comm.fit[[3]] + labs(title = "Any crime intent/stmug (chg)")
(p1 + p2) / (p3 + p4) / (p5 + plot_spacer()) / (p6 + p7)
8.1.1.9 Fit + prior summaries (Chg Any past crime/all stress)
8.1.1.9.1 stmony
out.chg.anyprjcrime.stmony.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.16 0.43 2.41 4.09 1.00 1316 1982
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept -4.31 0.71 -5.82 -2.97 1.00
## rural.ses.med2 -0.32 0.79 -1.84 1.25 1.00
## rural.ses.med3 0.59 0.78 -0.96 2.07 1.00
## rural.ses.med4 1.43 0.95 -0.61 3.10 1.00
## mostmony_devx2 0.14 0.18 -0.24 0.49 1.00
## mostmony_av12x2 -0.06 0.08 -0.22 0.09 1.00
## mostmony_devx2:rural.ses.med2 -0.41 0.44 -1.35 0.41 1.00
## mostmony_devx2:rural.ses.med3 0.39 0.36 -0.33 1.14 1.00
## mostmony_devx2:rural.ses.med4 0.54 0.44 -0.28 1.45 1.00
## Bulk_ESS Tail_ESS
## Intercept 1938 2719
## rural.ses.med2 2354 2144
## rural.ses.med3 2022 2674
## rural.ses.med4 1745 2397
## mostmony_devx2 2488 2165
## mostmony_av12x2 1903 2714
## mostmony_devx2:rural.ses.med2 1879 1993
## mostmony_devx2:rural.ses.med3 1577 2016
## mostmony_devx2:rural.ses.med4 1043 1850
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## mostmony_devx21[1] 0.25 0.14 0.04 0.58 1.00
## mostmony_devx21[2] 0.25 0.14 0.04 0.56 1.00
## mostmony_devx21[3] 0.26 0.15 0.04 0.59 1.00
## mostmony_devx21[4] 0.25 0.14 0.04 0.57 1.00
## mostmony_av12x21[1] 0.12 0.08 0.02 0.32 1.00
## mostmony_av12x21[2] 0.13 0.09 0.02 0.34 1.00
## mostmony_av12x21[3] 0.13 0.08 0.02 0.33 1.00
## mostmony_av12x21[4] 0.13 0.08 0.02 0.34 1.00
## mostmony_av12x21[5] 0.12 0.08 0.02 0.32 1.00
## mostmony_av12x21[6] 0.12 0.08 0.01 0.31 1.00
## mostmony_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## mostmony_av12x21[8] 0.12 0.08 0.02 0.32 1.00
## mostmony_devx2:rural.ses.med21[1] 0.27 0.20 0.01 0.71 1.00
## mostmony_devx2:rural.ses.med21[2] 0.21 0.18 0.01 0.66 1.00
## mostmony_devx2:rural.ses.med21[3] 0.24 0.18 0.01 0.69 1.00
## mostmony_devx2:rural.ses.med21[4] 0.28 0.21 0.01 0.75 1.00
## mostmony_devx2:rural.ses.med31[1] 0.26 0.20 0.01 0.73 1.00
## mostmony_devx2:rural.ses.med31[2] 0.31 0.20 0.01 0.75 1.00
## mostmony_devx2:rural.ses.med31[3] 0.17 0.16 0.00 0.59 1.00
## mostmony_devx2:rural.ses.med31[4] 0.26 0.20 0.01 0.71 1.00
## mostmony_devx2:rural.ses.med41[1] 0.43 0.23 0.02 0.85 1.00
## mostmony_devx2:rural.ses.med41[2] 0.17 0.15 0.00 0.57 1.00
## mostmony_devx2:rural.ses.med41[3] 0.16 0.15 0.01 0.56 1.00
## mostmony_devx2:rural.ses.med41[4] 0.24 0.19 0.01 0.68 1.00
## Bulk_ESS Tail_ESS
## mostmony_devx21[1] 4586 2795
## mostmony_devx21[2] 5091 2675
## mostmony_devx21[3] 4602 2946
## mostmony_devx21[4] 4852 2686
## mostmony_av12x21[1] 4901 2342
## mostmony_av12x21[2] 4234 2320
## mostmony_av12x21[3] 5143 2634
## mostmony_av12x21[4] 4326 2398
## mostmony_av12x21[5] 5155 2498
## mostmony_av12x21[6] 4619 2763
## mostmony_av12x21[7] 4678 3010
## mostmony_av12x21[8] 4750 3191
## mostmony_devx2:rural.ses.med21[1] 3452 1894
## mostmony_devx2:rural.ses.med21[2] 3731 1946
## mostmony_devx2:rural.ses.med21[3] 4455 2275
## mostmony_devx2:rural.ses.med21[4] 4023 2641
## mostmony_devx2:rural.ses.med31[1] 3535 2207
## mostmony_devx2:rural.ses.med31[2] 2811 2391
## mostmony_devx2:rural.ses.med31[3] 2724 3055
## mostmony_devx2:rural.ses.med31[4] 3289 2424
## mostmony_devx2:rural.ses.med41[1] 1802 1968
## mostmony_devx2:rural.ses.med41[2] 2464 2107
## mostmony_devx2:rural.ses.med41[3] 2879 2668
## mostmony_devx2:rural.ses.med41[4] 3495 2978
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stmony.comm.fit[[2]]## prior class coef
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## group resp dpar nlpar lb ub source
## user
## user
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## id 0 (vectorized)
## id 0 (vectorized)
## user
## default
## default
## default
## user8.1.1.9.2 sttran
out.chg.anyprjcrime.sttran.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.12 0.42 2.38 4.04 1.00 1102 2204
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept -3.57 0.70 -4.95 -2.20 1.00
## rural.ses.med2 -0.07 0.83 -1.65 1.66 1.00
## rural.ses.med3 0.84 0.74 -0.75 2.17 1.00
## rural.ses.med4 1.42 0.92 -0.47 3.04 1.00
## mosttran_devx2 -0.03 0.17 -0.38 0.31 1.00
## mosttran_av12x2 -0.13 0.08 -0.29 0.03 1.00
## mosttran_devx2:rural.ses.med2 -0.65 0.46 -1.63 0.19 1.00
## mosttran_devx2:rural.ses.med3 0.26 0.36 -0.42 1.05 1.00
## mosttran_devx2:rural.ses.med4 0.61 0.42 -0.12 1.49 1.00
## Bulk_ESS Tail_ESS
## Intercept 1935 2587
## rural.ses.med2 3204 2750
## rural.ses.med3 2111 2513
## rural.ses.med4 2353 3105
## mosttran_devx2 2049 2462
## mosttran_av12x2 2282 2980
## mosttran_devx2:rural.ses.med2 2426 2248
## mosttran_devx2:rural.ses.med3 1766 2032
## mosttran_devx2:rural.ses.med4 1514 2415
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## mosttran_devx21[1] 0.27 0.15 0.04 0.61 1.00
## mosttran_devx21[2] 0.24 0.14 0.04 0.57 1.00
## mosttran_devx21[3] 0.24 0.14 0.03 0.56 1.00
## mosttran_devx21[4] 0.26 0.15 0.04 0.59 1.00
## mosttran_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## mosttran_av12x21[2] 0.13 0.08 0.02 0.32 1.00
## mosttran_av12x21[3] 0.14 0.08 0.02 0.34 1.00
## mosttran_av12x21[4] 0.16 0.09 0.02 0.37 1.00
## mosttran_av12x21[5] 0.13 0.08 0.02 0.32 1.00
## mosttran_av12x21[6] 0.11 0.07 0.01 0.28 1.00
## mosttran_av12x21[7] 0.11 0.07 0.01 0.29 1.00
## mosttran_av12x21[8] 0.11 0.07 0.01 0.30 1.00
## mosttran_devx2:rural.ses.med21[1] 0.30 0.20 0.01 0.74 1.00
## mosttran_devx2:rural.ses.med21[2] 0.17 0.15 0.00 0.55 1.00
## mosttran_devx2:rural.ses.med21[3] 0.26 0.19 0.01 0.70 1.00
## mosttran_devx2:rural.ses.med21[4] 0.27 0.20 0.01 0.71 1.00
## mosttran_devx2:rural.ses.med31[1] 0.29 0.20 0.01 0.73 1.00
## mosttran_devx2:rural.ses.med31[2] 0.24 0.18 0.01 0.68 1.00
## mosttran_devx2:rural.ses.med31[3] 0.19 0.17 0.00 0.63 1.00
## mosttran_devx2:rural.ses.med31[4] 0.28 0.21 0.01 0.75 1.00
## mosttran_devx2:rural.ses.med41[1] 0.42 0.23 0.02 0.84 1.00
## mosttran_devx2:rural.ses.med41[2] 0.14 0.14 0.00 0.52 1.00
## mosttran_devx2:rural.ses.med41[3] 0.15 0.14 0.00 0.52 1.00
## mosttran_devx2:rural.ses.med41[4] 0.28 0.20 0.01 0.73 1.00
## Bulk_ESS Tail_ESS
## mosttran_devx21[1] 4556 2565
## mosttran_devx21[2] 3970 2961
## mosttran_devx21[3] 5047 2717
## mosttran_devx21[4] 4540 2794
## mosttran_av12x21[1] 4104 2219
## mosttran_av12x21[2] 4543 2341
## mosttran_av12x21[3] 3998 3027
## mosttran_av12x21[4] 4063 2388
## mosttran_av12x21[5] 4576 2663
## mosttran_av12x21[6] 4210 2406
## mosttran_av12x21[7] 4211 2674
## mosttran_av12x21[8] 4489 2820
## mosttran_devx2:rural.ses.med21[1] 3730 2129
## mosttran_devx2:rural.ses.med21[2] 3348 2523
## mosttran_devx2:rural.ses.med21[3] 4323 2159
## mosttran_devx2:rural.ses.med21[4] 3206 1999
## mosttran_devx2:rural.ses.med31[1] 3420 1912
## mosttran_devx2:rural.ses.med31[2] 3830 2626
## mosttran_devx2:rural.ses.med31[3] 3358 2374
## mosttran_devx2:rural.ses.med31[4] 3397 2422
## mosttran_devx2:rural.ses.med41[1] 2357 2072
## mosttran_devx2:rural.ses.med41[2] 2690 2246
## mosttran_devx2:rural.ses.med41[3] 3201 2210
## mosttran_devx2:rural.ses.med41[4] 2977 2383
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.sttran.comm.fit[[2]]## prior class coef
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## group resp dpar nlpar lb ub source
## user
## user
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## id 0 (vectorized)
## id 0 (vectorized)
## user
## default
## default
## default
## user8.1.1.9.3 stresp
out.chg.anyprjcrime.stresp.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.01 0.40 2.28 3.82 1.00 998 2399
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept -4.33 0.66 -5.67 -3.05 1.00
## rural.ses.med2 -0.56 0.74 -2.01 0.95 1.00
## rural.ses.med3 1.07 0.65 -0.25 2.31 1.00
## rural.ses.med4 1.44 0.86 -0.40 2.94 1.00
## mostresp_devx2 -0.24 0.19 -0.62 0.11 1.00
## mostresp_av12x2 0.15 0.07 0.02 0.28 1.00
## mostresp_devx2:rural.ses.med2 -0.26 0.41 -1.18 0.48 1.00
## mostresp_devx2:rural.ses.med3 0.04 0.34 -0.75 0.62 1.00
## mostresp_devx2:rural.ses.med4 0.42 0.35 -0.24 1.11 1.00
## Bulk_ESS Tail_ESS
## Intercept 1910 2307
## rural.ses.med2 2468 2858
## rural.ses.med3 2290 2704
## rural.ses.med4 2250 2901
## mostresp_devx2 2592 2916
## mostresp_av12x2 1838 2654
## mostresp_devx2:rural.ses.med2 2391 2372
## mostresp_devx2:rural.ses.med3 1752 1699
## mostresp_devx2:rural.ses.med4 1760 2643
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## mostresp_devx21[1] 0.26 0.15 0.04 0.59 1.00
## mostresp_devx21[2] 0.24 0.14 0.04 0.54 1.00
## mostresp_devx21[3] 0.23 0.13 0.03 0.53 1.00
## mostresp_devx21[4] 0.27 0.15 0.05 0.59 1.00
## mostresp_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## mostresp_av12x21[2] 0.12 0.07 0.02 0.30 1.00
## mostresp_av12x21[3] 0.13 0.08 0.02 0.34 1.00
## mostresp_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## mostresp_av12x21[5] 0.13 0.08 0.02 0.32 1.00
## mostresp_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## mostresp_av12x21[7] 0.12 0.08 0.02 0.31 1.00
## mostresp_av12x21[8] 0.11 0.07 0.02 0.29 1.00
## mostresp_devx2:rural.ses.med21[1] 0.27 0.20 0.01 0.72 1.00
## mostresp_devx2:rural.ses.med21[2] 0.23 0.18 0.01 0.67 1.00
## mostresp_devx2:rural.ses.med21[3] 0.20 0.17 0.01 0.63 1.00
## mostresp_devx2:rural.ses.med21[4] 0.29 0.21 0.01 0.75 1.00
## mostresp_devx2:rural.ses.med31[1] 0.26 0.19 0.01 0.70 1.00
## mostresp_devx2:rural.ses.med31[2] 0.24 0.19 0.01 0.70 1.00
## mostresp_devx2:rural.ses.med31[3] 0.22 0.18 0.01 0.64 1.00
## mostresp_devx2:rural.ses.med31[4] 0.29 0.22 0.01 0.79 1.00
## mostresp_devx2:rural.ses.med41[1] 0.41 0.23 0.02 0.84 1.00
## mostresp_devx2:rural.ses.med41[2] 0.19 0.17 0.00 0.62 1.00
## mostresp_devx2:rural.ses.med41[3] 0.17 0.15 0.00 0.57 1.00
## mostresp_devx2:rural.ses.med41[4] 0.23 0.18 0.01 0.65 1.00
## Bulk_ESS Tail_ESS
## mostresp_devx21[1] 4858 2482
## mostresp_devx21[2] 4422 2753
## mostresp_devx21[3] 5162 3256
## mostresp_devx21[4] 4993 3354
## mostresp_av12x21[1] 4901 2717
## mostresp_av12x21[2] 4846 2602
## mostresp_av12x21[3] 5105 2584
## mostresp_av12x21[4] 4386 2135
## mostresp_av12x21[5] 4854 2551
## mostresp_av12x21[6] 4607 2829
## mostresp_av12x21[7] 4741 3020
## mostresp_av12x21[8] 4858 3165
## mostresp_devx2:rural.ses.med21[1] 3565 2328
## mostresp_devx2:rural.ses.med21[2] 4449 2908
## mostresp_devx2:rural.ses.med21[3] 3914 2523
## mostresp_devx2:rural.ses.med21[4] 3993 2741
## mostresp_devx2:rural.ses.med31[1] 4280 2047
## mostresp_devx2:rural.ses.med31[2] 3180 2138
## mostresp_devx2:rural.ses.med31[3] 3911 2206
## mostresp_devx2:rural.ses.med31[4] 2590 2329
## mostresp_devx2:rural.ses.med41[1] 2495 2414
## mostresp_devx2:rural.ses.med41[2] 3840 2451
## mostresp_devx2:rural.ses.med41[3] 3306 2443
## mostresp_devx2:rural.ses.med41[4] 4295 2850
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stresp.comm.fit[[2]]## prior class coef
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## group resp dpar nlpar lb ub source
## user
## user
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## id 0 (vectorized)
## id 0 (vectorized)
## user
## default
## default
## default
## user8.1.1.9.4 stfair
out.chg.anyprjcrime.stfair.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.05 0.40 2.34 3.92 1.00 991 1781
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept -4.72 0.69 -6.10 -3.33 1.00
## rural.ses.med2 -0.31 0.79 -1.83 1.32 1.00
## rural.ses.med3 0.63 0.70 -0.79 1.92 1.00
## rural.ses.med4 1.19 0.90 -0.76 2.79 1.00
## mostfair_devx2 -0.07 0.19 -0.45 0.28 1.00
## mostfair_av12x2 0.14 0.07 0.01 0.28 1.00
## mostfair_devx2:rural.ses.med2 -0.44 0.45 -1.43 0.37 1.00
## mostfair_devx2:rural.ses.med3 0.32 0.31 -0.26 0.97 1.00
## mostfair_devx2:rural.ses.med4 0.58 0.38 -0.10 1.36 1.00
## Bulk_ESS Tail_ESS
## Intercept 1915 2812
## rural.ses.med2 2935 2142
## rural.ses.med3 2691 2917
## rural.ses.med4 2721 3028
## mostfair_devx2 3004 2574
## mostfair_av12x2 1987 2661
## mostfair_devx2:rural.ses.med2 2438 2269
## mostfair_devx2:rural.ses.med3 2517 2465
## mostfair_devx2:rural.ses.med4 2038 2709
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## mostfair_devx21[1] 0.27 0.15 0.04 0.60 1.00
## mostfair_devx21[2] 0.24 0.14 0.04 0.56 1.00
## mostfair_devx21[3] 0.23 0.14 0.03 0.54 1.00
## mostfair_devx21[4] 0.26 0.15 0.04 0.58 1.00
## mostfair_av12x21[1] 0.12 0.08 0.02 0.31 1.00
## mostfair_av12x21[2] 0.11 0.07 0.02 0.29 1.00
## mostfair_av12x21[3] 0.14 0.09 0.02 0.35 1.00
## mostfair_av12x21[4] 0.12 0.08 0.02 0.32 1.00
## mostfair_av12x21[5] 0.12 0.07 0.01 0.30 1.00
## mostfair_av12x21[6] 0.12 0.08 0.01 0.30 1.00
## mostfair_av12x21[7] 0.14 0.08 0.02 0.34 1.00
## mostfair_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## mostfair_devx2:rural.ses.med21[1] 0.27 0.19 0.01 0.70 1.00
## mostfair_devx2:rural.ses.med21[2] 0.21 0.17 0.01 0.63 1.00
## mostfair_devx2:rural.ses.med21[3] 0.23 0.18 0.01 0.66 1.00
## mostfair_devx2:rural.ses.med21[4] 0.29 0.21 0.01 0.74 1.00
## mostfair_devx2:rural.ses.med31[1] 0.28 0.20 0.01 0.71 1.00
## mostfair_devx2:rural.ses.med31[2] 0.25 0.19 0.01 0.68 1.00
## mostfair_devx2:rural.ses.med31[3] 0.22 0.17 0.01 0.63 1.00
## mostfair_devx2:rural.ses.med31[4] 0.26 0.19 0.01 0.70 1.00
## mostfair_devx2:rural.ses.med41[1] 0.38 0.22 0.02 0.80 1.00
## mostfair_devx2:rural.ses.med41[2] 0.20 0.16 0.01 0.61 1.00
## mostfair_devx2:rural.ses.med41[3] 0.20 0.16 0.01 0.59 1.00
## mostfair_devx2:rural.ses.med41[4] 0.22 0.18 0.01 0.66 1.00
## Bulk_ESS Tail_ESS
## mostfair_devx21[1] 4917 2446
## mostfair_devx21[2] 4912 2940
## mostfair_devx21[3] 5171 2809
## mostfair_devx21[4] 4814 3072
## mostfair_av12x21[1] 4466 2373
## mostfair_av12x21[2] 4371 2493
## mostfair_av12x21[3] 4244 2301
## mostfair_av12x21[4] 5129 2348
## mostfair_av12x21[5] 4455 2326
## mostfair_av12x21[6] 5080 2558
## mostfair_av12x21[7] 4852 2727
## mostfair_av12x21[8] 5089 2824
## mostfair_devx2:rural.ses.med21[1] 3673 1890
## mostfair_devx2:rural.ses.med21[2] 4107 2085
## mostfair_devx2:rural.ses.med21[3] 4701 2556
## mostfair_devx2:rural.ses.med21[4] 4255 2910
## mostfair_devx2:rural.ses.med31[1] 2807 1497
## mostfair_devx2:rural.ses.med31[2] 3468 1781
## mostfair_devx2:rural.ses.med31[3] 4046 3010
## mostfair_devx2:rural.ses.med31[4] 3628 2682
## mostfair_devx2:rural.ses.med41[1] 2960 2092
## mostfair_devx2:rural.ses.med41[2] 3269 2442
## mostfair_devx2:rural.ses.med41[3] 3856 2497
## mostfair_devx2:rural.ses.med41[4] 4104 2596
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stfair.comm.fit[[2]]## prior class coef
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## group resp dpar nlpar lb ub source
## user
## user
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## id 0 (vectorized)
## id 0 (vectorized)
## user
## default
## default
## default
## user8.1.1.9.5 stjob
out.chg.anyprjcrime.stjob.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 2.95 0.39 2.26 3.80 1.00 1150 2124
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept -4.88 0.69 -6.30 -3.58 1.00 2007
## rural.ses.med2 -0.29 0.77 -1.71 1.27 1.00 2843
## rural.ses.med3 0.38 0.74 -1.17 1.74 1.00 3269
## rural.ses.med4 1.77 0.98 -0.40 3.38 1.00 1574
## mostjob_devx2 0.03 0.18 -0.33 0.38 1.00 2526
## mostjob_av12x2 0.17 0.06 0.05 0.29 1.00 1977
## mostjob_devx2:rural.ses.med2 -0.46 0.43 -1.38 0.28 1.00 2528
## mostjob_devx2:rural.ses.med3 0.39 0.31 -0.20 1.01 1.00 2638
## mostjob_devx2:rural.ses.med4 0.19 0.52 -0.90 1.14 1.00 1203
## Tail_ESS
## Intercept 2074
## rural.ses.med2 2391
## rural.ses.med3 2790
## rural.ses.med4 2475
## mostjob_devx2 2644
## mostjob_av12x2 2745
## mostjob_devx2:rural.ses.med2 1949
## mostjob_devx2:rural.ses.med3 2654
## mostjob_devx2:rural.ses.med4 2286
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## mostjob_devx21[1] 0.26 0.15 0.04 0.59 1.00
## mostjob_devx21[2] 0.25 0.15 0.04 0.59 1.00
## mostjob_devx21[3] 0.24 0.14 0.03 0.56 1.00
## mostjob_devx21[4] 0.26 0.14 0.04 0.58 1.00
## mostjob_av12x21[1] 0.11 0.07 0.01 0.27 1.00
## mostjob_av12x21[2] 0.11 0.07 0.01 0.29 1.00
## mostjob_av12x21[3] 0.11 0.07 0.01 0.28 1.00
## mostjob_av12x21[4] 0.12 0.08 0.02 0.30 1.00
## mostjob_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## mostjob_av12x21[6] 0.15 0.09 0.02 0.36 1.00
## mostjob_av12x21[7] 0.15 0.09 0.02 0.35 1.00
## mostjob_av12x21[8] 0.12 0.08 0.02 0.30 1.00
## mostjob_devx2:rural.ses.med21[1] 0.26 0.18 0.01 0.67 1.00
## mostjob_devx2:rural.ses.med21[2] 0.22 0.17 0.01 0.64 1.00
## mostjob_devx2:rural.ses.med21[3] 0.25 0.18 0.01 0.68 1.00
## mostjob_devx2:rural.ses.med21[4] 0.28 0.20 0.01 0.73 1.00
## mostjob_devx2:rural.ses.med31[1] 0.28 0.20 0.02 0.71 1.00
## mostjob_devx2:rural.ses.med31[2] 0.28 0.19 0.01 0.70 1.00
## mostjob_devx2:rural.ses.med31[3] 0.19 0.16 0.01 0.59 1.00
## mostjob_devx2:rural.ses.med31[4] 0.25 0.19 0.01 0.67 1.00
## mostjob_devx2:rural.ses.med41[1] 0.36 0.25 0.01 0.85 1.00
## mostjob_devx2:rural.ses.med41[2] 0.18 0.16 0.00 0.62 1.00
## mostjob_devx2:rural.ses.med41[3] 0.20 0.19 0.00 0.68 1.00
## mostjob_devx2:rural.ses.med41[4] 0.26 0.20 0.01 0.74 1.00
## Bulk_ESS Tail_ESS
## mostjob_devx21[1] 5612 2699
## mostjob_devx21[2] 4955 2832
## mostjob_devx21[3] 5308 2806
## mostjob_devx21[4] 5200 2615
## mostjob_av12x21[1] 4393 2373
## mostjob_av12x21[2] 5115 2226
## mostjob_av12x21[3] 4628 2552
## mostjob_av12x21[4] 5704 2826
## mostjob_av12x21[5] 4674 2738
## mostjob_av12x21[6] 5072 2922
## mostjob_av12x21[7] 5395 3266
## mostjob_av12x21[8] 4844 3091
## mostjob_devx2:rural.ses.med21[1] 4419 2247
## mostjob_devx2:rural.ses.med21[2] 3656 2403
## mostjob_devx2:rural.ses.med21[3] 4346 2842
## mostjob_devx2:rural.ses.med21[4] 3046 2370
## mostjob_devx2:rural.ses.med31[1] 3934 2506
## mostjob_devx2:rural.ses.med31[2] 3525 2319
## mostjob_devx2:rural.ses.med31[3] 4039 2904
## mostjob_devx2:rural.ses.med31[4] 4544 2844
## mostjob_devx2:rural.ses.med41[1] 1680 2799
## mostjob_devx2:rural.ses.med41[2] 4137 2090
## mostjob_devx2:rural.ses.med41[3] 1970 2937
## mostjob_devx2:rural.ses.med41[4] 3936 2965
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stjob.comm.fit[[2]]## prior class coef
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## group resp dpar nlpar lb ub source
## user
## user
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## id 0 (vectorized)
## id 0 (vectorized)
## user
## default
## default
## default
## user8.1.1.9.6 stthft
out.chg.anyprjcrime.stthft.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.10 0.42 2.36 4.00 1.00 1113 2120
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept -3.74 0.64 -5.03 -2.55 1.00
## rural.ses.med2 -0.39 0.82 -1.92 1.21 1.00
## rural.ses.med3 0.46 0.73 -1.08 1.82 1.00
## rural.ses.med4 1.76 0.77 0.14 3.17 1.00
## mostthft_devx2 -0.25 0.19 -0.62 0.13 1.00
## mostthft_av12x2 0.03 0.08 -0.11 0.19 1.00
## mostthft_devx2:rural.ses.med2 -0.38 0.50 -1.43 0.55 1.00
## mostthft_devx2:rural.ses.med3 0.43 0.35 -0.20 1.20 1.00
## mostthft_devx2:rural.ses.med4 0.37 0.34 -0.30 1.03 1.00
## Bulk_ESS Tail_ESS
## Intercept 2055 2717
## rural.ses.med2 2896 3094
## rural.ses.med3 2480 2730
## rural.ses.med4 2220 2678
## mostthft_devx2 2281 2524
## mostthft_av12x2 2070 2625
## mostthft_devx2:rural.ses.med2 2107 2076
## mostthft_devx2:rural.ses.med3 2163 2719
## mostthft_devx2:rural.ses.med4 1737 1926
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## mostthft_devx21[1] 0.25 0.14 0.04 0.56 1.00
## mostthft_devx21[2] 0.30 0.15 0.05 0.62 1.00
## mostthft_devx21[3] 0.21 0.13 0.03 0.51 1.00
## mostthft_devx21[4] 0.24 0.14 0.04 0.57 1.00
## mostthft_av12x21[1] 0.13 0.08 0.02 0.33 1.00
## mostthft_av12x21[2] 0.12 0.08 0.02 0.31 1.00
## mostthft_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## mostthft_av12x21[4] 0.12 0.08 0.01 0.32 1.00
## mostthft_av12x21[5] 0.13 0.08 0.02 0.33 1.00
## mostthft_av12x21[6] 0.13 0.08 0.02 0.33 1.00
## mostthft_av12x21[7] 0.13 0.08 0.02 0.32 1.00
## mostthft_av12x21[8] 0.12 0.08 0.02 0.31 1.00
## mostthft_devx2:rural.ses.med21[1] 0.27 0.20 0.01 0.72 1.00
## mostthft_devx2:rural.ses.med21[2] 0.23 0.18 0.01 0.66 1.00
## mostthft_devx2:rural.ses.med21[3] 0.20 0.17 0.01 0.64 1.00
## mostthft_devx2:rural.ses.med21[4] 0.29 0.21 0.01 0.76 1.00
## mostthft_devx2:rural.ses.med31[1] 0.30 0.20 0.01 0.74 1.00
## mostthft_devx2:rural.ses.med31[2] 0.23 0.18 0.01 0.67 1.00
## mostthft_devx2:rural.ses.med31[3] 0.17 0.15 0.01 0.57 1.00
## mostthft_devx2:rural.ses.med31[4] 0.30 0.20 0.01 0.74 1.00
## mostthft_devx2:rural.ses.med41[1] 0.35 0.22 0.02 0.78 1.00
## mostthft_devx2:rural.ses.med41[2] 0.19 0.16 0.01 0.63 1.00
## mostthft_devx2:rural.ses.med41[3] 0.25 0.18 0.01 0.67 1.00
## mostthft_devx2:rural.ses.med41[4] 0.22 0.18 0.01 0.68 1.00
## Bulk_ESS Tail_ESS
## mostthft_devx21[1] 4371 2655
## mostthft_devx21[2] 3926 2468
## mostthft_devx21[3] 4613 2993
## mostthft_devx21[4] 4965 2834
## mostthft_av12x21[1] 4799 2176
## mostthft_av12x21[2] 4430 2142
## mostthft_av12x21[3] 4782 2416
## mostthft_av12x21[4] 5660 2308
## mostthft_av12x21[5] 4601 2036
## mostthft_av12x21[6] 5658 2557
## mostthft_av12x21[7] 5684 3212
## mostthft_av12x21[8] 5106 3118
## mostthft_devx2:rural.ses.med21[1] 3840 2363
## mostthft_devx2:rural.ses.med21[2] 4175 2545
## mostthft_devx2:rural.ses.med21[3] 4814 2903
## mostthft_devx2:rural.ses.med21[4] 4330 2785
## mostthft_devx2:rural.ses.med31[1] 3388 2291
## mostthft_devx2:rural.ses.med31[2] 3433 2377
## mostthft_devx2:rural.ses.med31[3] 3700 2872
## mostthft_devx2:rural.ses.med31[4] 3291 2581
## mostthft_devx2:rural.ses.med41[1] 2635 2396
## mostthft_devx2:rural.ses.med41[2] 3425 2611
## mostthft_devx2:rural.ses.med41[3] 3766 2451
## mostthft_devx2:rural.ses.med41[4] 3374 2250
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stthft.comm.fit[[2]]## prior class coef
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## group resp dpar nlpar lb ub source
## user
## user
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## id 0 (vectorized)
## id 0 (vectorized)
## user
## default
## default
## default
## user8.1.1.9.7 stmug
out.chg.anyprjcrime.stmug.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.06 0.41 2.31 3.93 1.00 1098 1502
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept -3.98 0.63 -5.24 -2.74 1.00 1848
## rural.ses.med2 -1.05 0.74 -2.45 0.48 1.00 2697
## rural.ses.med3 1.49 0.74 -0.13 2.89 1.00 2069
## rural.ses.med4 1.63 0.90 -0.32 3.18 1.00 1974
## mostmug_devx2 -0.03 0.21 -0.43 0.40 1.00 2111
## mostmug_av12x2 -0.05 0.08 -0.21 0.11 1.00 2098
## mostmug_devx2:rural.ses.med2 0.12 0.41 -0.65 0.94 1.00 1941
## mostmug_devx2:rural.ses.med3 -0.19 0.43 -1.06 0.72 1.00 1458
## mostmug_devx2:rural.ses.med4 0.42 0.40 -0.34 1.24 1.00 1496
## Tail_ESS
## Intercept 2512
## rural.ses.med2 2556
## rural.ses.med3 2148
## rural.ses.med4 2240
## mostmug_devx2 2845
## mostmug_av12x2 2878
## mostmug_devx2:rural.ses.med2 2680
## mostmug_devx2:rural.ses.med3 1692
## mostmug_devx2:rural.ses.med4 2111
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## mostmug_devx21[1] 0.26 0.15 0.04 0.59 1.00
## mostmug_devx21[2] 0.25 0.15 0.03 0.60 1.00
## mostmug_devx21[3] 0.24 0.15 0.03 0.59 1.00
## mostmug_devx21[4] 0.26 0.14 0.04 0.58 1.00
## mostmug_av12x21[1] 0.12 0.08 0.01 0.31 1.00
## mostmug_av12x21[2] 0.12 0.08 0.02 0.32 1.00
## mostmug_av12x21[3] 0.12 0.08 0.02 0.31 1.00
## mostmug_av12x21[4] 0.13 0.08 0.02 0.33 1.00
## mostmug_av12x21[5] 0.13 0.08 0.02 0.32 1.00
## mostmug_av12x21[6] 0.12 0.08 0.02 0.31 1.00
## mostmug_av12x21[7] 0.12 0.08 0.02 0.32 1.00
## mostmug_av12x21[8] 0.13 0.08 0.02 0.33 1.00
## mostmug_devx2:rural.ses.med21[1] 0.24 0.19 0.01 0.69 1.00
## mostmug_devx2:rural.ses.med21[2] 0.21 0.18 0.01 0.65 1.00
## mostmug_devx2:rural.ses.med21[3] 0.25 0.19 0.01 0.71 1.00
## mostmug_devx2:rural.ses.med21[4] 0.30 0.22 0.01 0.76 1.00
## mostmug_devx2:rural.ses.med31[1] 0.23 0.19 0.01 0.70 1.00
## mostmug_devx2:rural.ses.med31[2] 0.23 0.19 0.01 0.68 1.00
## mostmug_devx2:rural.ses.med31[3] 0.24 0.19 0.01 0.70 1.00
## mostmug_devx2:rural.ses.med31[4] 0.30 0.21 0.01 0.77 1.00
## mostmug_devx2:rural.ses.med41[1] 0.39 0.24 0.01 0.83 1.00
## mostmug_devx2:rural.ses.med41[2] 0.17 0.16 0.00 0.60 1.00
## mostmug_devx2:rural.ses.med41[3] 0.20 0.17 0.01 0.64 1.00
## mostmug_devx2:rural.ses.med41[4] 0.24 0.19 0.01 0.68 1.00
## Bulk_ESS Tail_ESS
## mostmug_devx21[1] 4322 2590
## mostmug_devx21[2] 3892 2557
## mostmug_devx21[3] 3515 3192
## mostmug_devx21[4] 5219 2841
## mostmug_av12x21[1] 4488 2130
## mostmug_av12x21[2] 4559 2385
## mostmug_av12x21[3] 5039 2454
## mostmug_av12x21[4] 4354 2385
## mostmug_av12x21[5] 4849 2852
## mostmug_av12x21[6] 4898 2503
## mostmug_av12x21[7] 5174 2738
## mostmug_av12x21[8] 5217 3079
## mostmug_devx2:rural.ses.med21[1] 3376 2510
## mostmug_devx2:rural.ses.med21[2] 4386 2706
## mostmug_devx2:rural.ses.med21[3] 4382 3153
## mostmug_devx2:rural.ses.med21[4] 4234 3058
## mostmug_devx2:rural.ses.med31[1] 2454 2169
## mostmug_devx2:rural.ses.med31[2] 3188 2485
## mostmug_devx2:rural.ses.med31[3] 3342 2653
## mostmug_devx2:rural.ses.med31[4] 3558 2822
## mostmug_devx2:rural.ses.med41[1] 2064 1501
## mostmug_devx2:rural.ses.med41[2] 3423 2319
## mostmug_devx2:rural.ses.med41[3] 3507 2150
## mostmug_devx2:rural.ses.med41[4] 3852 2917
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).out.chg.anyprjcrime.stmug.comm.fit[[2]]## prior class coef
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## group resp dpar nlpar lb ub source
## user
## user
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## id 0 (vectorized)
## id 0 (vectorized)
## user
## default
## default
## default
## user8.1.2 Negative Emotions & Stress x Community Correlation Models
8.1.2.1 Corr X Community: stmony & negative emotions
#Community Change: negative emotions items ~ mo(stmony)
#Vectorize priors:
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmony_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmony_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmony_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmony_av12x21',
resp = depdv_names)
)
chg.alldepress.stmony.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) +
rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stmony_comm_fit",
file_refit = "on_change"
)
##Update function to call all ppchecks for bivar depressive symptom chg models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="depcantgo")
ppcheckdv2 <-pp_check(modelfit, resp="depeffort")
ppcheckdv3 <-pp_check(modelfit, resp="deplonely")
ppcheckdv4 <-pp_check(modelfit, resp="depblues")
ppcheckdv5 <-pp_check(modelfit, resp="depunfair")
ppcheckdv6 <-pp_check(modelfit, resp="depmistrt")
ppcheckdv7 <-pp_check(modelfit, resp="depbetray")
plotcoefs <- mcmc_areas(modelfit, regex_pars = "^bsp_", prob = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior distributions for monotonic ordinal stress coefficients \nwith medians and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = "^bsp_", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for monotonic ordinal stress coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2, ppcheckdv3,
ppcheckdv4, ppcheckdv5, ppcheckdv6, ppcheckdv7,
plotcoefs, plotcoefs2)
return(allchecks)
}
out.chg.alldepress.stmony.comm.fit <- ppchecks(chg.alldepress.stmony.comm.fit)8.1.2.1.1 Coefficient plot (intervals)
out.chg.alldepress.stmony.comm.fit[[11]]
8.1.2.1.2 PPcheck (density)
p1 <- out.chg.alldepress.stmony.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stmony.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stmony.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stmony.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stmony.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stmony.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stmony.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
8.1.2.1.3 Fit summary
out.chg.alldepress.stmony.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## depeffort ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## deplonely ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## depblues ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## depunfair ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## depmistrt ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## depbetray ~ 1 + mo(stmony_devx2) + mo(stmony_av12x2) + rural.ses.med + mo(stmony_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.30 0.19 0.01 0.70 1.00 689
## sd(depeffort_Intercept) 0.46 0.27 0.03 1.02 1.01 546
## sd(deplonely_Intercept) 0.46 0.24 0.03 0.91 1.01 462
## sd(depblues_Intercept) 0.68 0.32 0.07 1.27 1.00 532
## sd(depunfair_Intercept) 0.22 0.16 0.01 0.59 1.00 889
## sd(depmistrt_Intercept) 0.32 0.22 0.02 0.80 1.00 917
## sd(depbetray_Intercept) 0.41 0.25 0.02 0.95 1.01 439
## Tail_ESS
## sd(depcantgo_Intercept) 1676
## sd(depeffort_Intercept) 1366
## sd(deplonely_Intercept) 1146
## sd(depblues_Intercept) 923
## sd(depunfair_Intercept) 1823
## sd(depmistrt_Intercept) 1779
## sd(depbetray_Intercept) 1283
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_Intercept -1.18 0.35 -1.91 -0.57
## depeffort_Intercept -2.22 0.44 -3.14 -1.42
## deplonely_Intercept -1.24 0.40 -2.06 -0.49
## depblues_Intercept -2.17 0.47 -3.08 -1.21
## depunfair_Intercept -2.33 0.35 -3.09 -1.69
## depmistrt_Intercept -2.65 0.41 -3.48 -1.85
## depbetray_Intercept -2.99 0.45 -3.92 -2.13
## depcantgo_rural.ses.med2 0.30 0.53 -0.94 1.26
## depcantgo_rural.ses.med3 0.34 0.49 -0.60 1.38
## depcantgo_rural.ses.med4 -0.07 0.52 -1.06 1.05
## depeffort_rural.ses.med2 0.16 0.56 -0.99 1.30
## depeffort_rural.ses.med3 0.38 0.58 -0.90 1.47
## depeffort_rural.ses.med4 0.17 0.58 -0.98 1.31
## deplonely_rural.ses.med2 -0.20 0.56 -1.37 0.92
## deplonely_rural.ses.med3 0.34 0.53 -0.66 1.52
## deplonely_rural.ses.med4 -0.27 0.55 -1.53 0.69
## depblues_rural.ses.med2 0.18 0.61 -0.92 1.60
## depblues_rural.ses.med3 -0.15 0.58 -1.38 0.92
## depblues_rural.ses.med4 0.80 0.59 -0.40 1.99
## depunfair_rural.ses.med2 0.01 0.51 -1.01 1.04
## depunfair_rural.ses.med3 0.73 0.50 -0.28 1.71
## depunfair_rural.ses.med4 0.76 0.59 -0.60 1.79
## depmistrt_rural.ses.med2 0.44 0.54 -0.63 1.52
## depmistrt_rural.ses.med3 0.56 0.55 -0.60 1.62
## depmistrt_rural.ses.med4 0.22 0.60 -1.12 1.24
## depbetray_rural.ses.med2 0.05 0.58 -1.18 1.11
## depbetray_rural.ses.med3 1.16 0.46 0.29 2.11
## depbetray_rural.ses.med4 0.62 0.65 -0.83 1.77
## depcantgo_mostmony_devx2 0.25 0.13 -0.00 0.53
## depcantgo_mostmony_av12x2 0.05 0.03 -0.01 0.12
## depcantgo_mostmony_devx2:rural.ses.med2 0.01 0.30 -0.45 0.79
## depcantgo_mostmony_devx2:rural.ses.med3 -0.16 0.24 -0.65 0.31
## depcantgo_mostmony_devx2:rural.ses.med4 0.11 0.31 -0.45 0.80
## depeffort_mostmony_devx2 0.10 0.16 -0.22 0.41
## depeffort_mostmony_av12x2 0.00 0.04 -0.08 0.08
## depeffort_mostmony_devx2:rural.ses.med2 -0.11 0.32 -0.88 0.45
## depeffort_mostmony_devx2:rural.ses.med3 0.05 0.30 -0.50 0.70
## depeffort_mostmony_devx2:rural.ses.med4 0.20 0.29 -0.40 0.73
## deplonely_mostmony_devx2 0.11 0.15 -0.22 0.38
## deplonely_mostmony_av12x2 -0.01 0.04 -0.08 0.06
## deplonely_mostmony_devx2:rural.ses.med2 -0.15 0.34 -0.95 0.45
## deplonely_mostmony_devx2:rural.ses.med3 -0.26 0.32 -1.00 0.27
## deplonely_mostmony_devx2:rural.ses.med4 0.45 0.35 -0.08 1.32
## depblues_mostmony_devx2 -0.17 0.16 -0.51 0.14
## depblues_mostmony_av12x2 0.06 0.05 -0.03 0.15
## depblues_mostmony_devx2:rural.ses.med2 0.07 0.32 -0.56 0.77
## depblues_mostmony_devx2:rural.ses.med3 0.24 0.28 -0.32 0.81
## depblues_mostmony_devx2:rural.ses.med4 -0.14 0.34 -0.90 0.48
## depunfair_mostmony_devx2 0.25 0.13 -0.01 0.50
## depunfair_mostmony_av12x2 0.07 0.04 0.00 0.15
## depunfair_mostmony_devx2:rural.ses.med2 0.17 0.26 -0.33 0.70
## depunfair_mostmony_devx2:rural.ses.med3 -0.03 0.25 -0.51 0.47
## depunfair_mostmony_devx2:rural.ses.med4 0.07 0.29 -0.55 0.65
## depmistrt_mostmony_devx2 0.08 0.15 -0.21 0.37
## depmistrt_mostmony_av12x2 0.10 0.04 0.02 0.19
## depmistrt_mostmony_devx2:rural.ses.med2 0.06 0.27 -0.52 0.56
## depmistrt_mostmony_devx2:rural.ses.med3 -0.04 0.29 -0.67 0.49
## depmistrt_mostmony_devx2:rural.ses.med4 0.37 0.37 -0.21 1.27
## depbetray_mostmony_devx2 0.06 0.16 -0.27 0.38
## depbetray_mostmony_av12x2 0.14 0.05 0.06 0.24
## depbetray_mostmony_devx2:rural.ses.med2 0.20 0.29 -0.39 0.73
## depbetray_mostmony_devx2:rural.ses.med3 -0.40 0.34 -1.18 0.19
## depbetray_mostmony_devx2:rural.ses.med4 0.18 0.31 -0.45 0.80
## Rhat Bulk_ESS Tail_ESS
## depcantgo_Intercept 1.00 3924 3067
## depeffort_Intercept 1.00 3664 3136
## deplonely_Intercept 1.00 2797 2898
## depblues_Intercept 1.00 2997 2966
## depunfair_Intercept 1.00 4153 3072
## depmistrt_Intercept 1.00 3884 2840
## depbetray_Intercept 1.00 3264 3213
## depcantgo_rural.ses.med2 1.00 2860 2375
## depcantgo_rural.ses.med3 1.00 3482 2831
## depcantgo_rural.ses.med4 1.00 3469 2792
## depeffort_rural.ses.med2 1.00 4149 3034
## depeffort_rural.ses.med3 1.00 2845 2633
## depeffort_rural.ses.med4 1.00 3281 2367
## deplonely_rural.ses.med2 1.00 2558 2532
## deplonely_rural.ses.med3 1.00 3282 2850
## deplonely_rural.ses.med4 1.00 3578 2780
## depblues_rural.ses.med2 1.00 3698 2722
## depblues_rural.ses.med3 1.00 4057 2906
## depblues_rural.ses.med4 1.00 3490 2448
## depunfair_rural.ses.med2 1.00 3607 2494
## depunfair_rural.ses.med3 1.00 3322 2910
## depunfair_rural.ses.med4 1.00 3020 2276
## depmistrt_rural.ses.med2 1.00 3759 2843
## depmistrt_rural.ses.med3 1.00 3837 2716
## depmistrt_rural.ses.med4 1.00 3388 2610
## depbetray_rural.ses.med2 1.00 3399 2814
## depbetray_rural.ses.med3 1.00 4024 2569
## depbetray_rural.ses.med4 1.00 3332 2980
## depcantgo_mostmony_devx2 1.00 3105 3093
## depcantgo_mostmony_av12x2 1.00 6651 3128
## depcantgo_mostmony_devx2:rural.ses.med2 1.00 2296 1964
## depcantgo_mostmony_devx2:rural.ses.med3 1.00 3374 2835
## depcantgo_mostmony_devx2:rural.ses.med4 1.00 2622 2129
## depeffort_mostmony_devx2 1.00 3965 3186
## depeffort_mostmony_av12x2 1.00 6441 3014
## depeffort_mostmony_devx2:rural.ses.med2 1.00 3438 2552
## depeffort_mostmony_devx2:rural.ses.med3 1.00 2532 2293
## depeffort_mostmony_devx2:rural.ses.med4 1.00 3140 2377
## deplonely_mostmony_devx2 1.00 2741 2953
## deplonely_mostmony_av12x2 1.00 6687 3238
## deplonely_mostmony_devx2:rural.ses.med2 1.00 2168 2144
## deplonely_mostmony_devx2:rural.ses.med3 1.00 2949 2689
## deplonely_mostmony_devx2:rural.ses.med4 1.00 2607 1933
## depblues_mostmony_devx2 1.00 3474 3038
## depblues_mostmony_av12x2 1.00 5978 2843
## depblues_mostmony_devx2:rural.ses.med2 1.00 3229 2340
## depblues_mostmony_devx2:rural.ses.med3 1.00 3408 2361
## depblues_mostmony_devx2:rural.ses.med4 1.00 3118 2779
## depunfair_mostmony_devx2 1.00 3629 2658
## depunfair_mostmony_av12x2 1.00 6582 2977
## depunfair_mostmony_devx2:rural.ses.med2 1.00 3142 2629
## depunfair_mostmony_devx2:rural.ses.med3 1.00 2977 2480
## depunfair_mostmony_devx2:rural.ses.med4 1.00 2565 2681
## depmistrt_mostmony_devx2 1.00 3388 2933
## depmistrt_mostmony_av12x2 1.00 6002 3085
## depmistrt_mostmony_devx2:rural.ses.med2 1.00 3182 2262
## depmistrt_mostmony_devx2:rural.ses.med3 1.00 3241 2153
## depmistrt_mostmony_devx2:rural.ses.med4 1.00 2912 2261
## depbetray_mostmony_devx2 1.00 3083 2982
## depbetray_mostmony_av12x2 1.00 6475 3227
## depbetray_mostmony_devx2:rural.ses.med2 1.00 2930 2399
## depbetray_mostmony_devx2:rural.ses.med3 1.00 2760 2553
## depbetray_mostmony_devx2:rural.ses.med4 1.00 2640 2506
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## depcantgo_mostmony_devx21[1] 0.23 0.13 0.03
## depcantgo_mostmony_devx21[2] 0.29 0.13 0.06
## depcantgo_mostmony_devx21[3] 0.20 0.11 0.03
## depcantgo_mostmony_devx21[4] 0.28 0.15 0.05
## depcantgo_mostmony_av12x21[1] 0.12 0.08 0.01
## depcantgo_mostmony_av12x21[2] 0.13 0.08 0.02
## depcantgo_mostmony_av12x21[3] 0.13 0.08 0.02
## depcantgo_mostmony_av12x21[4] 0.14 0.09 0.02
## depcantgo_mostmony_av12x21[5] 0.12 0.08 0.01
## depcantgo_mostmony_av12x21[6] 0.12 0.08 0.01
## depcantgo_mostmony_av12x21[7] 0.12 0.08 0.02
## depcantgo_mostmony_av12x21[8] 0.12 0.08 0.02
## depcantgo_mostmony_devx2:rural.ses.med21[1] 0.26 0.20 0.01
## depcantgo_mostmony_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## depcantgo_mostmony_devx2:rural.ses.med21[3] 0.22 0.18 0.01
## depcantgo_mostmony_devx2:rural.ses.med21[4] 0.30 0.22 0.01
## depcantgo_mostmony_devx2:rural.ses.med31[1] 0.27 0.19 0.01
## depcantgo_mostmony_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## depcantgo_mostmony_devx2:rural.ses.med31[3] 0.22 0.17 0.01
## depcantgo_mostmony_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## depcantgo_mostmony_devx2:rural.ses.med41[1] 0.25 0.20 0.01
## depcantgo_mostmony_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## depcantgo_mostmony_devx2:rural.ses.med41[3] 0.21 0.17 0.01
## depcantgo_mostmony_devx2:rural.ses.med41[4] 0.32 0.23 0.01
## depeffort_mostmony_devx21[1] 0.25 0.14 0.04
## depeffort_mostmony_devx21[2] 0.27 0.15 0.04
## depeffort_mostmony_devx21[3] 0.22 0.13 0.03
## depeffort_mostmony_devx21[4] 0.26 0.15 0.04
## depeffort_mostmony_av12x21[1] 0.13 0.08 0.02
## depeffort_mostmony_av12x21[2] 0.13 0.08 0.02
## depeffort_mostmony_av12x21[3] 0.12 0.08 0.02
## depeffort_mostmony_av12x21[4] 0.12 0.08 0.02
## depeffort_mostmony_av12x21[5] 0.12 0.08 0.02
## depeffort_mostmony_av12x21[6] 0.12 0.08 0.02
## depeffort_mostmony_av12x21[7] 0.12 0.08 0.02
## depeffort_mostmony_av12x21[8] 0.13 0.09 0.01
## depeffort_mostmony_devx2:rural.ses.med21[1] 0.25 0.19 0.01
## depeffort_mostmony_devx2:rural.ses.med21[2] 0.21 0.17 0.01
## depeffort_mostmony_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## depeffort_mostmony_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## depeffort_mostmony_devx2:rural.ses.med31[1] 0.26 0.20 0.01
## depeffort_mostmony_devx2:rural.ses.med31[2] 0.21 0.17 0.01
## depeffort_mostmony_devx2:rural.ses.med31[3] 0.22 0.19 0.01
## depeffort_mostmony_devx2:rural.ses.med31[4] 0.30 0.21 0.01
## depeffort_mostmony_devx2:rural.ses.med41[1] 0.23 0.18 0.01
## depeffort_mostmony_devx2:rural.ses.med41[2] 0.28 0.20 0.01
## depeffort_mostmony_devx2:rural.ses.med41[3] 0.22 0.17 0.01
## depeffort_mostmony_devx2:rural.ses.med41[4] 0.27 0.20 0.01
## deplonely_mostmony_devx21[1] 0.24 0.14 0.04
## deplonely_mostmony_devx21[2] 0.32 0.17 0.04
## deplonely_mostmony_devx21[3] 0.21 0.13 0.03
## deplonely_mostmony_devx21[4] 0.23 0.14 0.03
## deplonely_mostmony_av12x21[1] 0.13 0.09 0.02
## deplonely_mostmony_av12x21[2] 0.13 0.08 0.02
## deplonely_mostmony_av12x21[3] 0.13 0.08 0.02
## deplonely_mostmony_av12x21[4] 0.12 0.08 0.02
## deplonely_mostmony_av12x21[5] 0.12 0.08 0.02
## deplonely_mostmony_av12x21[6] 0.12 0.08 0.02
## deplonely_mostmony_av12x21[7] 0.13 0.08 0.02
## deplonely_mostmony_av12x21[8] 0.13 0.08 0.02
## deplonely_mostmony_devx2:rural.ses.med21[1] 0.24 0.19 0.01
## deplonely_mostmony_devx2:rural.ses.med21[2] 0.19 0.18 0.00
## deplonely_mostmony_devx2:rural.ses.med21[3] 0.25 0.19 0.01
## deplonely_mostmony_devx2:rural.ses.med21[4] 0.32 0.22 0.01
## deplonely_mostmony_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## deplonely_mostmony_devx2:rural.ses.med31[2] 0.19 0.16 0.01
## deplonely_mostmony_devx2:rural.ses.med31[3] 0.22 0.17 0.01
## deplonely_mostmony_devx2:rural.ses.med31[4] 0.33 0.22 0.01
## deplonely_mostmony_devx2:rural.ses.med41[1] 0.23 0.18 0.01
## deplonely_mostmony_devx2:rural.ses.med41[2] 0.21 0.16 0.01
## deplonely_mostmony_devx2:rural.ses.med41[3] 0.23 0.17 0.01
## deplonely_mostmony_devx2:rural.ses.med41[4] 0.33 0.22 0.01
## depblues_mostmony_devx21[1] 0.28 0.15 0.04
## depblues_mostmony_devx21[2] 0.24 0.13 0.04
## depblues_mostmony_devx21[3] 0.23 0.13 0.03
## depblues_mostmony_devx21[4] 0.25 0.14 0.04
## depblues_mostmony_av12x21[1] 0.13 0.08 0.02
## depblues_mostmony_av12x21[2] 0.13 0.08 0.02
## depblues_mostmony_av12x21[3] 0.13 0.08 0.02
## depblues_mostmony_av12x21[4] 0.13 0.08 0.02
## depblues_mostmony_av12x21[5] 0.12 0.08 0.02
## depblues_mostmony_av12x21[6] 0.12 0.08 0.02
## depblues_mostmony_av12x21[7] 0.13 0.08 0.02
## depblues_mostmony_av12x21[8] 0.13 0.08 0.02
## depblues_mostmony_devx2:rural.ses.med21[1] 0.26 0.21 0.01
## depblues_mostmony_devx2:rural.ses.med21[2] 0.22 0.17 0.01
## depblues_mostmony_devx2:rural.ses.med21[3] 0.22 0.18 0.01
## depblues_mostmony_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## depblues_mostmony_devx2:rural.ses.med31[1] 0.26 0.19 0.01
## depblues_mostmony_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## depblues_mostmony_devx2:rural.ses.med31[3] 0.24 0.18 0.01
## depblues_mostmony_devx2:rural.ses.med31[4] 0.28 0.20 0.01
## depblues_mostmony_devx2:rural.ses.med41[1] 0.26 0.20 0.01
## depblues_mostmony_devx2:rural.ses.med41[2] 0.20 0.17 0.01
## depblues_mostmony_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## depblues_mostmony_devx2:rural.ses.med41[4] 0.31 0.22 0.01
## depunfair_mostmony_devx21[1] 0.20 0.12 0.03
## depunfair_mostmony_devx21[2] 0.26 0.13 0.05
## depunfair_mostmony_devx21[3] 0.33 0.14 0.07
## depunfair_mostmony_devx21[4] 0.22 0.13 0.03
## depunfair_mostmony_av12x21[1] 0.13 0.08 0.02
## depunfair_mostmony_av12x21[2] 0.13 0.08 0.02
## depunfair_mostmony_av12x21[3] 0.10 0.07 0.01
## depunfair_mostmony_av12x21[4] 0.11 0.07 0.02
## depunfair_mostmony_av12x21[5] 0.12 0.08 0.02
## depunfair_mostmony_av12x21[6] 0.12 0.08 0.02
## depunfair_mostmony_av12x21[7] 0.14 0.08 0.02
## depunfair_mostmony_av12x21[8] 0.15 0.09 0.02
## depunfair_mostmony_devx2:rural.ses.med21[1] 0.23 0.18 0.01
## depunfair_mostmony_devx2:rural.ses.med21[2] 0.24 0.18 0.01
## depunfair_mostmony_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## depunfair_mostmony_devx2:rural.ses.med21[4] 0.29 0.20 0.01
## depunfair_mostmony_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## depunfair_mostmony_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## depunfair_mostmony_devx2:rural.ses.med31[3] 0.22 0.17 0.01
## depunfair_mostmony_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## depunfair_mostmony_devx2:rural.ses.med41[1] 0.28 0.21 0.01
## depunfair_mostmony_devx2:rural.ses.med41[2] 0.22 0.17 0.01
## depunfair_mostmony_devx2:rural.ses.med41[3] 0.21 0.17 0.01
## depunfair_mostmony_devx2:rural.ses.med41[4] 0.29 0.21 0.01
## depmistrt_mostmony_devx21[1] 0.25 0.15 0.03
## depmistrt_mostmony_devx21[2] 0.22 0.13 0.03
## depmistrt_mostmony_devx21[3] 0.26 0.14 0.04
## depmistrt_mostmony_devx21[4] 0.27 0.15 0.04
## depmistrt_mostmony_av12x21[1] 0.13 0.08 0.02
## depmistrt_mostmony_av12x21[2] 0.14 0.09 0.02
## depmistrt_mostmony_av12x21[3] 0.13 0.08 0.02
## depmistrt_mostmony_av12x21[4] 0.12 0.07 0.02
## depmistrt_mostmony_av12x21[5] 0.10 0.06 0.01
## depmistrt_mostmony_av12x21[6] 0.10 0.07 0.01
## depmistrt_mostmony_av12x21[7] 0.16 0.09 0.03
## depmistrt_mostmony_av12x21[8] 0.13 0.08 0.02
## depmistrt_mostmony_devx2:rural.ses.med21[1] 0.25 0.19 0.01
## depmistrt_mostmony_devx2:rural.ses.med21[2] 0.22 0.17 0.01
## depmistrt_mostmony_devx2:rural.ses.med21[3] 0.25 0.18 0.01
## depmistrt_mostmony_devx2:rural.ses.med21[4] 0.28 0.21 0.01
## depmistrt_mostmony_devx2:rural.ses.med31[1] 0.26 0.20 0.01
## depmistrt_mostmony_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## depmistrt_mostmony_devx2:rural.ses.med31[3] 0.22 0.17 0.01
## depmistrt_mostmony_devx2:rural.ses.med31[4] 0.30 0.22 0.01
## depmistrt_mostmony_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depmistrt_mostmony_devx2:rural.ses.med41[2] 0.17 0.16 0.01
## depmistrt_mostmony_devx2:rural.ses.med41[3] 0.22 0.17 0.01
## depmistrt_mostmony_devx2:rural.ses.med41[4] 0.35 0.23 0.01
## depbetray_mostmony_devx21[1] 0.27 0.15 0.04
## depbetray_mostmony_devx21[2] 0.24 0.13 0.04
## depbetray_mostmony_devx21[3] 0.24 0.14 0.04
## depbetray_mostmony_devx21[4] 0.25 0.14 0.04
## depbetray_mostmony_av12x21[1] 0.12 0.08 0.02
## depbetray_mostmony_av12x21[2] 0.13 0.08 0.02
## depbetray_mostmony_av12x21[3] 0.11 0.07 0.02
## depbetray_mostmony_av12x21[4] 0.12 0.07 0.02
## depbetray_mostmony_av12x21[5] 0.11 0.07 0.02
## depbetray_mostmony_av12x21[6] 0.10 0.07 0.01
## depbetray_mostmony_av12x21[7] 0.16 0.09 0.03
## depbetray_mostmony_av12x21[8] 0.15 0.09 0.02
## depbetray_mostmony_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## depbetray_mostmony_devx2:rural.ses.med21[2] 0.20 0.17 0.01
## depbetray_mostmony_devx2:rural.ses.med21[3] 0.27 0.19 0.01
## depbetray_mostmony_devx2:rural.ses.med21[4] 0.26 0.20 0.01
## depbetray_mostmony_devx2:rural.ses.med31[1] 0.17 0.16 0.00
## depbetray_mostmony_devx2:rural.ses.med31[2] 0.18 0.15 0.01
## depbetray_mostmony_devx2:rural.ses.med31[3] 0.34 0.21 0.02
## depbetray_mostmony_devx2:rural.ses.med31[4] 0.31 0.21 0.01
## depbetray_mostmony_devx2:rural.ses.med41[1] 0.28 0.20 0.01
## depbetray_mostmony_devx2:rural.ses.med41[2] 0.25 0.19 0.01
## depbetray_mostmony_devx2:rural.ses.med41[3] 0.18 0.16 0.01
## depbetray_mostmony_devx2:rural.ses.med41[4] 0.28 0.20 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## depcantgo_mostmony_devx21[1] 0.54 1.00 5918 3185
## depcantgo_mostmony_devx21[2] 0.57 1.00 5330 2734
## depcantgo_mostmony_devx21[3] 0.45 1.00 5772 2822
## depcantgo_mostmony_devx21[4] 0.60 1.00 7156 2821
## depcantgo_mostmony_av12x21[1] 0.31 1.00 7540 2424
## depcantgo_mostmony_av12x21[2] 0.32 1.00 8507 2624
## depcantgo_mostmony_av12x21[3] 0.31 1.00 7770 2515
## depcantgo_mostmony_av12x21[4] 0.34 1.00 6780 2773
## depcantgo_mostmony_av12x21[5] 0.30 1.00 7451 2174
## depcantgo_mostmony_av12x21[6] 0.31 1.00 7731 2615
## depcantgo_mostmony_av12x21[7] 0.30 1.00 6348 2751
## depcantgo_mostmony_av12x21[8] 0.31 1.00 6567 2859
## depcantgo_mostmony_devx2:rural.ses.med21[1] 0.73 1.00 6284 2481
## depcantgo_mostmony_devx2:rural.ses.med21[2] 0.67 1.00 4511 2574
## depcantgo_mostmony_devx2:rural.ses.med21[3] 0.65 1.00 4251 2746
## depcantgo_mostmony_devx2:rural.ses.med21[4] 0.80 1.00 4588 2915
## depcantgo_mostmony_devx2:rural.ses.med31[1] 0.72 1.00 4508 2070
## depcantgo_mostmony_devx2:rural.ses.med31[2] 0.66 1.00 4827 2792
## depcantgo_mostmony_devx2:rural.ses.med31[3] 0.64 1.00 4751 2592
## depcantgo_mostmony_devx2:rural.ses.med31[4] 0.75 1.00 5720 2731
## depcantgo_mostmony_devx2:rural.ses.med41[1] 0.71 1.00 5002 2496
## depcantgo_mostmony_devx2:rural.ses.med41[2] 0.64 1.00 5419 2839
## depcantgo_mostmony_devx2:rural.ses.med41[3] 0.63 1.00 4822 2738
## depcantgo_mostmony_devx2:rural.ses.med41[4] 0.82 1.00 4467 2344
## depeffort_mostmony_devx21[1] 0.57 1.00 6615 2412
## depeffort_mostmony_devx21[2] 0.59 1.00 6172 2297
## depeffort_mostmony_devx21[3] 0.52 1.00 6373 2713
## depeffort_mostmony_devx21[4] 0.59 1.00 7883 2750
## depeffort_mostmony_av12x21[1] 0.33 1.00 7280 2385
## depeffort_mostmony_av12x21[2] 0.33 1.00 6873 2640
## depeffort_mostmony_av12x21[3] 0.31 1.00 7312 2320
## depeffort_mostmony_av12x21[4] 0.32 1.00 9145 2636
## depeffort_mostmony_av12x21[5] 0.32 1.00 8284 2202
## depeffort_mostmony_av12x21[6] 0.31 1.00 6748 2647
## depeffort_mostmony_av12x21[7] 0.31 1.00 8366 2248
## depeffort_mostmony_av12x21[8] 0.34 1.00 7020 2596
## depeffort_mostmony_devx2:rural.ses.med21[1] 0.70 1.00 6720 2903
## depeffort_mostmony_devx2:rural.ses.med21[2] 0.65 1.00 5563 2110
## depeffort_mostmony_devx2:rural.ses.med21[3] 0.67 1.00 5268 2943
## depeffort_mostmony_devx2:rural.ses.med21[4] 0.79 1.00 4816 2172
## depeffort_mostmony_devx2:rural.ses.med31[1] 0.71 1.00 5909 2661
## depeffort_mostmony_devx2:rural.ses.med31[2] 0.64 1.00 6360 2349
## depeffort_mostmony_devx2:rural.ses.med31[3] 0.68 1.00 3924 2874
## depeffort_mostmony_devx2:rural.ses.med31[4] 0.76 1.00 6472 3256
## depeffort_mostmony_devx2:rural.ses.med41[1] 0.67 1.00 6180 2359
## depeffort_mostmony_devx2:rural.ses.med41[2] 0.72 1.00 4533 2442
## depeffort_mostmony_devx2:rural.ses.med41[3] 0.64 1.00 4468 2518
## depeffort_mostmony_devx2:rural.ses.med41[4] 0.76 1.00 5295 2726
## deplonely_mostmony_devx21[1] 0.57 1.00 6814 2945
## deplonely_mostmony_devx21[2] 0.65 1.00 3622 2937
## deplonely_mostmony_devx21[3] 0.51 1.00 4909 2734
## deplonely_mostmony_devx21[4] 0.56 1.00 6527 3155
## deplonely_mostmony_av12x21[1] 0.33 1.00 7601 2486
## deplonely_mostmony_av12x21[2] 0.33 1.00 7121 2493
## deplonely_mostmony_av12x21[3] 0.31 1.00 8613 2895
## deplonely_mostmony_av12x21[4] 0.30 1.00 7452 2476
## deplonely_mostmony_av12x21[5] 0.30 1.00 7777 2556
## deplonely_mostmony_av12x21[6] 0.32 1.00 7273 2379
## deplonely_mostmony_av12x21[7] 0.32 1.00 6649 2563
## deplonely_mostmony_av12x21[8] 0.33 1.00 5944 3010
## deplonely_mostmony_devx2:rural.ses.med21[1] 0.71 1.00 5043 2121
## deplonely_mostmony_devx2:rural.ses.med21[2] 0.64 1.00 3597 2022
## deplonely_mostmony_devx2:rural.ses.med21[3] 0.69 1.00 4348 3163
## deplonely_mostmony_devx2:rural.ses.med21[4] 0.80 1.00 5105 2985
## deplonely_mostmony_devx2:rural.ses.med31[1] 0.69 1.00 5374 2617
## deplonely_mostmony_devx2:rural.ses.med31[2] 0.60 1.00 4971 2620
## deplonely_mostmony_devx2:rural.ses.med31[3] 0.64 1.00 4271 3195
## deplonely_mostmony_devx2:rural.ses.med31[4] 0.80 1.00 4796 2742
## deplonely_mostmony_devx2:rural.ses.med41[1] 0.67 1.00 5378 2547
## deplonely_mostmony_devx2:rural.ses.med41[2] 0.63 1.00 5601 2522
## deplonely_mostmony_devx2:rural.ses.med41[3] 0.63 1.00 3790 2664
## deplonely_mostmony_devx2:rural.ses.med41[4] 0.78 1.00 5025 2853
## depblues_mostmony_devx21[1] 0.62 1.00 7083 2756
## depblues_mostmony_devx21[2] 0.56 1.00 7582 2648
## depblues_mostmony_devx21[3] 0.53 1.00 6891 2919
## depblues_mostmony_devx21[4] 0.57 1.00 7344 2800
## depblues_mostmony_av12x21[1] 0.32 1.00 7017 2428
## depblues_mostmony_av12x21[2] 0.32 1.00 7719 2027
## depblues_mostmony_av12x21[3] 0.32 1.00 6719 2090
## depblues_mostmony_av12x21[4] 0.32 1.00 7302 2667
## depblues_mostmony_av12x21[5] 0.31 1.00 7754 2511
## depblues_mostmony_av12x21[6] 0.30 1.00 7546 2439
## depblues_mostmony_av12x21[7] 0.32 1.00 6389 3004
## depblues_mostmony_av12x21[8] 0.34 1.00 7364 2995
## depblues_mostmony_devx2:rural.ses.med21[1] 0.75 1.00 4832 2361
## depblues_mostmony_devx2:rural.ses.med21[2] 0.64 1.00 6082 2951
## depblues_mostmony_devx2:rural.ses.med21[3] 0.66 1.00 5095 2334
## depblues_mostmony_devx2:rural.ses.med21[4] 0.80 1.00 5596 2954
## depblues_mostmony_devx2:rural.ses.med31[1] 0.68 1.00 5542 2773
## depblues_mostmony_devx2:rural.ses.med31[2] 0.66 1.00 6176 2613
## depblues_mostmony_devx2:rural.ses.med31[3] 0.66 1.00 5607 2620
## depblues_mostmony_devx2:rural.ses.med31[4] 0.73 1.00 5535 2724
## depblues_mostmony_devx2:rural.ses.med41[1] 0.72 1.00 5611 2151
## depblues_mostmony_devx2:rural.ses.med41[2] 0.62 1.00 5539 2553
## depblues_mostmony_devx2:rural.ses.med41[3] 0.66 1.00 5696 2373
## depblues_mostmony_devx2:rural.ses.med41[4] 0.80 1.00 4882 2714
## depunfair_mostmony_devx21[1] 0.47 1.00 5611 2828
## depunfair_mostmony_devx21[2] 0.55 1.00 6035 2441
## depunfair_mostmony_devx21[3] 0.63 1.00 4908 2876
## depunfair_mostmony_devx21[4] 0.52 1.00 6722 3066
## depunfair_mostmony_av12x21[1] 0.33 1.00 7431 2050
## depunfair_mostmony_av12x21[2] 0.32 1.00 7278 2642
## depunfair_mostmony_av12x21[3] 0.27 1.00 7006 2628
## depunfair_mostmony_av12x21[4] 0.27 1.00 8049 2593
## depunfair_mostmony_av12x21[5] 0.31 1.00 7266 1998
## depunfair_mostmony_av12x21[6] 0.31 1.00 7811 2651
## depunfair_mostmony_av12x21[7] 0.34 1.00 6975 2592
## depunfair_mostmony_av12x21[8] 0.36 1.00 6520 3210
## depunfair_mostmony_devx2:rural.ses.med21[1] 0.67 1.00 5183 2081
## depunfair_mostmony_devx2:rural.ses.med21[2] 0.66 1.00 5648 2228
## depunfair_mostmony_devx2:rural.ses.med21[3] 0.65 1.00 5358 3021
## depunfair_mostmony_devx2:rural.ses.med21[4] 0.73 1.00 6344 3046
## depunfair_mostmony_devx2:rural.ses.med31[1] 0.72 1.00 4931 2090
## depunfair_mostmony_devx2:rural.ses.med31[2] 0.66 1.00 5765 2539
## depunfair_mostmony_devx2:rural.ses.med31[3] 0.66 1.00 5867 2605
## depunfair_mostmony_devx2:rural.ses.med31[4] 0.75 1.00 6342 2551
## depunfair_mostmony_devx2:rural.ses.med41[1] 0.74 1.00 5308 2705
## depunfair_mostmony_devx2:rural.ses.med41[2] 0.65 1.00 5519 2586
## depunfair_mostmony_devx2:rural.ses.med41[3] 0.65 1.00 5338 2884
## depunfair_mostmony_devx2:rural.ses.med41[4] 0.77 1.00 5553 2716
## depmistrt_mostmony_devx21[1] 0.59 1.00 7159 2299
## depmistrt_mostmony_devx21[2] 0.53 1.00 7814 2747
## depmistrt_mostmony_devx21[3] 0.58 1.00 5704 3122
## depmistrt_mostmony_devx21[4] 0.60 1.00 7454 2944
## depmistrt_mostmony_av12x21[1] 0.31 1.00 7084 2561
## depmistrt_mostmony_av12x21[2] 0.35 1.00 7072 2414
## depmistrt_mostmony_av12x21[3] 0.31 1.00 7532 1958
## depmistrt_mostmony_av12x21[4] 0.28 1.00 8679 2989
## depmistrt_mostmony_av12x21[5] 0.25 1.00 6888 2838
## depmistrt_mostmony_av12x21[6] 0.27 1.00 6822 2779
## depmistrt_mostmony_av12x21[7] 0.37 1.00 5839 2469
## depmistrt_mostmony_av12x21[8] 0.33 1.01 6724 2151
## depmistrt_mostmony_devx2:rural.ses.med21[1] 0.68 1.00 6058 2627
## depmistrt_mostmony_devx2:rural.ses.med21[2] 0.64 1.00 6335 2816
## depmistrt_mostmony_devx2:rural.ses.med21[3] 0.67 1.00 4944 2702
## depmistrt_mostmony_devx2:rural.ses.med21[4] 0.76 1.00 5525 2944
## depmistrt_mostmony_devx2:rural.ses.med31[1] 0.72 1.00 5327 2218
## depmistrt_mostmony_devx2:rural.ses.med31[2] 0.65 1.00 6178 2671
## depmistrt_mostmony_devx2:rural.ses.med31[3] 0.66 1.00 5354 2857
## depmistrt_mostmony_devx2:rural.ses.med31[4] 0.78 1.00 5138 2788
## depmistrt_mostmony_devx2:rural.ses.med41[1] 0.68 1.00 5830 2731
## depmistrt_mostmony_devx2:rural.ses.med41[2] 0.60 1.00 4758 2985
## depmistrt_mostmony_devx2:rural.ses.med41[3] 0.65 1.00 4880 3476
## depmistrt_mostmony_devx2:rural.ses.med41[4] 0.82 1.00 4156 2817
## depbetray_mostmony_devx21[1] 0.61 1.00 6928 2690
## depbetray_mostmony_devx21[2] 0.55 1.00 8355 2584
## depbetray_mostmony_devx21[3] 0.56 1.00 7080 2704
## depbetray_mostmony_devx21[4] 0.57 1.00 7217 2878
## depbetray_mostmony_av12x21[1] 0.31 1.00 7778 2589
## depbetray_mostmony_av12x21[2] 0.32 1.00 7172 2953
## depbetray_mostmony_av12x21[3] 0.27 1.00 6798 2481
## depbetray_mostmony_av12x21[4] 0.29 1.00 7889 2715
## depbetray_mostmony_av12x21[5] 0.27 1.00 6680 2968
## depbetray_mostmony_av12x21[6] 0.26 1.00 7917 2831
## depbetray_mostmony_av12x21[7] 0.36 1.00 7644 3044
## depbetray_mostmony_av12x21[8] 0.36 1.00 7886 2609
## depbetray_mostmony_devx2:rural.ses.med21[1] 0.71 1.00 6238 2697
## depbetray_mostmony_devx2:rural.ses.med21[2] 0.62 1.00 5838 2550
## depbetray_mostmony_devx2:rural.ses.med21[3] 0.70 1.00 4447 3065
## depbetray_mostmony_devx2:rural.ses.med21[4] 0.71 1.00 5020 2781
## depbetray_mostmony_devx2:rural.ses.med31[1] 0.58 1.00 4894 2705
## depbetray_mostmony_devx2:rural.ses.med31[2] 0.57 1.00 6087 2475
## depbetray_mostmony_devx2:rural.ses.med31[3] 0.77 1.00 4041 2481
## depbetray_mostmony_devx2:rural.ses.med31[4] 0.77 1.00 4970 2804
## depbetray_mostmony_devx2:rural.ses.med41[1] 0.73 1.00 6013 2811
## depbetray_mostmony_devx2:rural.ses.med41[2] 0.70 1.00 4852 2568
## depbetray_mostmony_devx2:rural.ses.med41[3] 0.58 1.00 4691 2497
## depbetray_mostmony_devx2:rural.ses.med41[4] 0.75 1.00 5726 2482
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.2.1.4 Prior summary
out.chg.alldepress.stmony.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmony_av12x2
## normal(0, 0.25) b mostmony_devx2
## normal(0, 1) b mostmony_devx2:rural.ses.med2
## normal(0, 1) b mostmony_devx2:rural.ses.med3
## normal(0, 1) b mostmony_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmony_av12x21
## dirichlet(1) simo mostmony_devx2:rural.ses.med21
## dirichlet(1) simo mostmony_devx2:rural.ses.med31
## dirichlet(1) simo mostmony_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmony_devx21
## group resp dpar nlpar lb ub source
## default
## depbetray user
## depbetray user
## depbetray user
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depblues user
## depblues user
## depblues user
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depcantgo user
## depcantgo user
## depcantgo user
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depeffort user
## depeffort user
## depeffort user
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## deplonely user
## deplonely user
## deplonely user
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## depmistrt user
## depmistrt user
## depmistrt user
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depunfair user
## depunfair user
## depunfair user
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## default
## depbetray user
## depblues user
## depcantgo user
## depeffort user
## deplonely user
## depmistrt user
## depunfair user
## depbetray 0 default
## depblues 0 default
## depcantgo 0 default
## depeffort 0 default
## deplonely 0 default
## depmistrt 0 default
## depunfair 0 default
## id depbetray 0 (vectorized)
## id depbetray 0 (vectorized)
## id depblues 0 (vectorized)
## id depblues 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depeffort 0 (vectorized)
## id depeffort 0 (vectorized)
## id deplonely 0 (vectorized)
## id deplonely 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depunfair 0 (vectorized)
## id depunfair 0 (vectorized)
## depbetray user
## depbetray default
## depbetray default
## depbetray default
## depbetray user
## depblues user
## depblues default
## depblues default
## depblues default
## depblues user
## depcantgo user
## depcantgo default
## depcantgo default
## depcantgo default
## depcantgo user
## depeffort user
## depeffort default
## depeffort default
## depeffort default
## depeffort user
## deplonely user
## deplonely default
## deplonely default
## deplonely default
## deplonely user
## depmistrt user
## depmistrt default
## depmistrt default
## depmistrt default
## depmistrt user
## depunfair user
## depunfair default
## depunfair default
## depunfair default
## depunfair user8.1.2.2 Corr X Community: sttran & negative emotions
#Community Change: negative emotions items ~ mo(sttran)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mosttran_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mosttran_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mosttran_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mosttran_av12x21',
resp = depdv_names)
)
chg.alldepress.sttran.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) +
rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_sttran_comm_fit",
file_refit = "on_change"
)
out.chg.alldepress.sttran.comm.fit <- ppchecks(chg.alldepress.sttran.comm.fit)8.1.2.2.1 Coefficient plot (intervals)
out.chg.alldepress.sttran.comm.fit[[11]]
8.1.2.2.2 PPcheck (density)
p1 <- out.chg.alldepress.sttran.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.sttran.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.sttran.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.sttran.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.sttran.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.sttran.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.sttran.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
8.1.2.2.3 Fit summary
out.chg.alldepress.sttran.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## depeffort ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## deplonely ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## depblues ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## depunfair ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## depmistrt ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## depbetray ~ 1 + mo(sttran_devx2) + mo(sttran_av12x2) + rural.ses.med + mo(sttran_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.35 0.21 0.02 0.77 1.01 629
## sd(depeffort_Intercept) 0.43 0.27 0.02 0.98 1.01 475
## sd(deplonely_Intercept) 0.43 0.24 0.03 0.90 1.01 517
## sd(depblues_Intercept) 0.69 0.32 0.06 1.29 1.01 451
## sd(depunfair_Intercept) 0.23 0.17 0.01 0.62 1.00 990
## sd(depmistrt_Intercept) 0.34 0.22 0.01 0.80 1.00 851
## sd(depbetray_Intercept) 0.45 0.26 0.02 0.97 1.01 595
## Tail_ESS
## sd(depcantgo_Intercept) 1499
## sd(depeffort_Intercept) 1343
## sd(deplonely_Intercept) 1138
## sd(depblues_Intercept) 753
## sd(depunfair_Intercept) 1969
## sd(depmistrt_Intercept) 1716
## sd(depbetray_Intercept) 1429
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_Intercept -1.06 0.34 -1.76 -0.41
## depeffort_Intercept -1.95 0.42 -2.77 -1.11
## deplonely_Intercept -1.02 0.37 -1.69 -0.22
## depblues_Intercept -2.34 0.43 -3.25 -1.51
## depunfair_Intercept -2.33 0.37 -3.08 -1.60
## depmistrt_Intercept -2.50 0.42 -3.32 -1.66
## depbetray_Intercept -2.62 0.42 -3.47 -1.77
## depcantgo_rural.ses.med2 -0.35 0.52 -1.48 0.55
## depcantgo_rural.ses.med3 0.57 0.44 -0.23 1.51
## depcantgo_rural.ses.med4 -0.17 0.47 -1.14 0.80
## depeffort_rural.ses.med2 -0.26 0.68 -1.59 1.13
## depeffort_rural.ses.med3 0.59 0.54 -0.57 1.63
## depeffort_rural.ses.med4 0.61 0.55 -0.58 1.68
## deplonely_rural.ses.med2 -0.05 0.61 -1.11 1.32
## deplonely_rural.ses.med3 0.04 0.52 -1.05 1.12
## deplonely_rural.ses.med4 0.03 0.43 -0.78 0.96
## depblues_rural.ses.med2 0.43 0.58 -0.77 1.58
## depblues_rural.ses.med3 0.37 0.54 -0.77 1.43
## depblues_rural.ses.med4 0.51 0.60 -0.82 1.58
## depunfair_rural.ses.med2 -0.24 0.67 -1.54 1.11
## depunfair_rural.ses.med3 0.46 0.56 -0.79 1.52
## depunfair_rural.ses.med4 0.75 0.50 -0.18 1.87
## depmistrt_rural.ses.med2 0.85 0.57 -0.22 2.11
## depmistrt_rural.ses.med3 0.91 0.57 -0.13 2.15
## depmistrt_rural.ses.med4 0.73 0.57 -0.41 1.96
## depbetray_rural.ses.med2 0.84 0.58 -0.26 2.08
## depbetray_rural.ses.med3 0.75 0.56 -0.43 1.80
## depbetray_rural.ses.med4 1.10 0.52 -0.02 2.10
## depcantgo_mosttran_devx2 0.19 0.12 -0.04 0.44
## depcantgo_mosttran_av12x2 0.05 0.04 -0.02 0.12
## depcantgo_mosttran_devx2:rural.ses.med2 0.40 0.32 -0.10 1.16
## depcantgo_mosttran_devx2:rural.ses.med3 -0.34 0.30 -1.08 0.11
## depcantgo_mosttran_devx2:rural.ses.med4 0.12 0.24 -0.33 0.65
## depeffort_mosttran_devx2 0.04 0.16 -0.28 0.35
## depeffort_mosttran_av12x2 -0.03 0.04 -0.12 0.06
## depeffort_mosttran_devx2:rural.ses.med2 0.12 0.34 -0.63 0.73
## depeffort_mosttran_devx2:rural.ses.med3 -0.10 0.32 -0.82 0.49
## depeffort_mosttran_devx2:rural.ses.med4 0.01 0.30 -0.48 0.73
## deplonely_mosttran_devx2 0.07 0.15 -0.25 0.34
## deplonely_mosttran_av12x2 -0.03 0.04 -0.11 0.05
## deplonely_mosttran_devx2:rural.ses.med2 -0.19 0.34 -1.00 0.37
## deplonely_mosttran_devx2:rural.ses.med3 -0.01 0.31 -0.52 0.72
## deplonely_mosttran_devx2:rural.ses.med4 0.29 0.24 -0.21 0.78
## depblues_mosttran_devx2 -0.08 0.16 -0.39 0.25
## depblues_mosttran_av12x2 0.05 0.05 -0.05 0.15
## depblues_mosttran_devx2:rural.ses.med2 -0.20 0.37 -1.00 0.48
## depblues_mosttran_devx2:rural.ses.med3 -0.07 0.31 -0.73 0.54
## depblues_mosttran_devx2:rural.ses.med4 0.01 0.30 -0.54 0.70
## depunfair_mosttran_devx2 0.23 0.14 -0.07 0.48
## depunfair_mosttran_av12x2 0.07 0.04 -0.01 0.15
## depunfair_mosttran_devx2:rural.ses.med2 0.26 0.31 -0.45 0.81
## depunfair_mosttran_devx2:rural.ses.med3 0.10 0.27 -0.46 0.65
## depunfair_mosttran_devx2:rural.ses.med4 0.17 0.26 -0.37 0.72
## depmistrt_mosttran_devx2 0.13 0.16 -0.24 0.41
## depmistrt_mosttran_av12x2 0.06 0.05 -0.03 0.17
## depmistrt_mosttran_devx2:rural.ses.med2 -0.20 0.33 -0.98 0.37
## depmistrt_mosttran_devx2:rural.ses.med3 -0.24 0.32 -1.00 0.33
## depmistrt_mosttran_devx2:rural.ses.med4 -0.05 0.32 -0.79 0.46
## depbetray_mosttran_devx2 0.04 0.16 -0.27 0.35
## depbetray_mosttran_av12x2 0.08 0.05 -0.01 0.17
## depbetray_mosttran_devx2:rural.ses.med2 -0.27 0.35 -1.08 0.35
## depbetray_mosttran_devx2:rural.ses.med3 -0.08 0.32 -0.75 0.54
## depbetray_mosttran_devx2:rural.ses.med4 -0.05 0.25 -0.54 0.48
## Rhat Bulk_ESS Tail_ESS
## depcantgo_Intercept 1.00 3917 2947
## depeffort_Intercept 1.00 2857 2725
## deplonely_Intercept 1.00 2803 2366
## depblues_Intercept 1.00 3231 2997
## depunfair_Intercept 1.00 3050 2496
## depmistrt_Intercept 1.00 2797 2427
## depbetray_Intercept 1.00 3381 3357
## depcantgo_rural.ses.med2 1.00 3754 3036
## depcantgo_rural.ses.med3 1.00 3740 2755
## depcantgo_rural.ses.med4 1.00 3260 2781
## depeffort_rural.ses.med2 1.00 3204 2528
## depeffort_rural.ses.med3 1.00 3044 2892
## depeffort_rural.ses.med4 1.00 3028 2550
## deplonely_rural.ses.med2 1.00 2606 2513
## deplonely_rural.ses.med3 1.00 2595 2189
## deplonely_rural.ses.med4 1.00 2622 1993
## depblues_rural.ses.med2 1.00 3698 3012
## depblues_rural.ses.med3 1.00 3574 2628
## depblues_rural.ses.med4 1.00 3357 2466
## depunfair_rural.ses.med2 1.00 2498 2019
## depunfair_rural.ses.med3 1.00 2525 2395
## depunfair_rural.ses.med4 1.00 2496 1797
## depmistrt_rural.ses.med2 1.00 3265 2301
## depmistrt_rural.ses.med3 1.00 3107 2790
## depmistrt_rural.ses.med4 1.00 3173 2772
## depbetray_rural.ses.med2 1.00 3775 2589
## depbetray_rural.ses.med3 1.00 3154 2412
## depbetray_rural.ses.med4 1.00 3592 2730
## depcantgo_mosttran_devx2 1.00 3712 2959
## depcantgo_mosttran_av12x2 1.00 5526 2886
## depcantgo_mosttran_devx2:rural.ses.med2 1.00 3081 2169
## depcantgo_mosttran_devx2:rural.ses.med3 1.00 2820 2049
## depcantgo_mosttran_devx2:rural.ses.med4 1.00 3297 2655
## depeffort_mosttran_devx2 1.00 2865 2699
## depeffort_mosttran_av12x2 1.00 5889 3281
## depeffort_mosttran_devx2:rural.ses.med2 1.00 2965 2964
## depeffort_mosttran_devx2:rural.ses.med3 1.00 2756 2228
## depeffort_mosttran_devx2:rural.ses.med4 1.00 2359 2487
## deplonely_mosttran_devx2 1.00 2550 2523
## deplonely_mosttran_av12x2 1.00 4790 3037
## deplonely_mosttran_devx2:rural.ses.med2 1.00 2546 2220
## deplonely_mosttran_devx2:rural.ses.med3 1.00 2131 1588
## deplonely_mosttran_devx2:rural.ses.med4 1.00 2237 1648
## depblues_mosttran_devx2 1.00 3913 3103
## depblues_mosttran_av12x2 1.00 5067 3139
## depblues_mosttran_devx2:rural.ses.med2 1.00 3299 2650
## depblues_mosttran_devx2:rural.ses.med3 1.00 3239 2434
## depblues_mosttran_devx2:rural.ses.med4 1.00 2984 2549
## depunfair_mosttran_devx2 1.00 2343 2371
## depunfair_mosttran_av12x2 1.00 6050 3069
## depunfair_mosttran_devx2:rural.ses.med2 1.00 2380 1823
## depunfair_mosttran_devx2:rural.ses.med3 1.00 2376 2575
## depunfair_mosttran_devx2:rural.ses.med4 1.00 2160 1910
## depmistrt_mosttran_devx2 1.00 2116 1960
## depmistrt_mosttran_av12x2 1.00 5456 2969
## depmistrt_mosttran_devx2:rural.ses.med2 1.00 2990 2194
## depmistrt_mosttran_devx2:rural.ses.med3 1.00 2489 2196
## depmistrt_mosttran_devx2:rural.ses.med4 1.00 2460 2466
## depbetray_mosttran_devx2 1.00 3197 3000
## depbetray_mosttran_av12x2 1.00 6035 3273
## depbetray_mosttran_devx2:rural.ses.med2 1.00 3209 2280
## depbetray_mosttran_devx2:rural.ses.med3 1.00 2623 2204
## depbetray_mosttran_devx2:rural.ses.med4 1.00 3338 2685
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## depcantgo_mosttran_devx21[1] 0.23 0.14 0.03
## depcantgo_mosttran_devx21[2] 0.29 0.14 0.06
## depcantgo_mosttran_devx21[3] 0.26 0.13 0.04
## depcantgo_mosttran_devx21[4] 0.22 0.13 0.03
## depcantgo_mosttran_av12x21[1] 0.12 0.08 0.01
## depcantgo_mosttran_av12x21[2] 0.13 0.08 0.02
## depcantgo_mosttran_av12x21[3] 0.11 0.07 0.01
## depcantgo_mosttran_av12x21[4] 0.12 0.08 0.01
## depcantgo_mosttran_av12x21[5] 0.13 0.08 0.02
## depcantgo_mosttran_av12x21[6] 0.12 0.08 0.02
## depcantgo_mosttran_av12x21[7] 0.14 0.09 0.02
## depcantgo_mosttran_av12x21[8] 0.14 0.09 0.02
## depcantgo_mosttran_devx2:rural.ses.med21[1] 0.23 0.18 0.01
## depcantgo_mosttran_devx2:rural.ses.med21[2] 0.22 0.16 0.01
## depcantgo_mosttran_devx2:rural.ses.med21[3] 0.23 0.17 0.01
## depcantgo_mosttran_devx2:rural.ses.med21[4] 0.32 0.22 0.02
## depcantgo_mosttran_devx2:rural.ses.med31[1] 0.21 0.17 0.01
## depcantgo_mosttran_devx2:rural.ses.med31[2] 0.22 0.17 0.01
## depcantgo_mosttran_devx2:rural.ses.med31[3] 0.24 0.17 0.01
## depcantgo_mosttran_devx2:rural.ses.med31[4] 0.32 0.22 0.01
## depcantgo_mosttran_devx2:rural.ses.med41[1] 0.25 0.20 0.01
## depcantgo_mosttran_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## depcantgo_mosttran_devx2:rural.ses.med41[3] 0.22 0.17 0.01
## depcantgo_mosttran_devx2:rural.ses.med41[4] 0.30 0.21 0.01
## depeffort_mosttran_devx21[1] 0.26 0.15 0.04
## depeffort_mosttran_devx21[2] 0.24 0.14 0.04
## depeffort_mosttran_devx21[3] 0.24 0.14 0.03
## depeffort_mosttran_devx21[4] 0.27 0.15 0.04
## depeffort_mosttran_av12x21[1] 0.13 0.08 0.02
## depeffort_mosttran_av12x21[2] 0.13 0.08 0.02
## depeffort_mosttran_av12x21[3] 0.13 0.08 0.02
## depeffort_mosttran_av12x21[4] 0.12 0.08 0.01
## depeffort_mosttran_av12x21[5] 0.12 0.08 0.02
## depeffort_mosttran_av12x21[6] 0.12 0.08 0.02
## depeffort_mosttran_av12x21[7] 0.12 0.08 0.01
## depeffort_mosttran_av12x21[8] 0.13 0.08 0.02
## depeffort_mosttran_devx2:rural.ses.med21[1] 0.23 0.19 0.01
## depeffort_mosttran_devx2:rural.ses.med21[2] 0.29 0.22 0.01
## depeffort_mosttran_devx2:rural.ses.med21[3] 0.20 0.17 0.01
## depeffort_mosttran_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## depeffort_mosttran_devx2:rural.ses.med31[1] 0.25 0.20 0.01
## depeffort_mosttran_devx2:rural.ses.med31[2] 0.20 0.17 0.01
## depeffort_mosttran_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## depeffort_mosttran_devx2:rural.ses.med31[4] 0.31 0.22 0.01
## depeffort_mosttran_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depeffort_mosttran_devx2:rural.ses.med41[2] 0.22 0.19 0.01
## depeffort_mosttran_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## depeffort_mosttran_devx2:rural.ses.med41[4] 0.31 0.23 0.01
## deplonely_mosttran_devx21[1] 0.24 0.14 0.04
## deplonely_mosttran_devx21[2] 0.23 0.13 0.03
## deplonely_mosttran_devx21[3] 0.27 0.15 0.04
## deplonely_mosttran_devx21[4] 0.26 0.15 0.04
## deplonely_mosttran_av12x21[1] 0.13 0.08 0.02
## deplonely_mosttran_av12x21[2] 0.13 0.08 0.02
## deplonely_mosttran_av12x21[3] 0.13 0.08 0.02
## deplonely_mosttran_av12x21[4] 0.12 0.08 0.02
## deplonely_mosttran_av12x21[5] 0.11 0.07 0.01
## deplonely_mosttran_av12x21[6] 0.12 0.08 0.01
## deplonely_mosttran_av12x21[7] 0.12 0.08 0.02
## deplonely_mosttran_av12x21[8] 0.13 0.08 0.02
## deplonely_mosttran_devx2:rural.ses.med21[1] 0.28 0.20 0.01
## deplonely_mosttran_devx2:rural.ses.med21[2] 0.19 0.16 0.01
## deplonely_mosttran_devx2:rural.ses.med21[3] 0.21 0.17 0.01
## deplonely_mosttran_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## deplonely_mosttran_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## deplonely_mosttran_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## deplonely_mosttran_devx2:rural.ses.med31[3] 0.21 0.17 0.01
## deplonely_mosttran_devx2:rural.ses.med31[4] 0.30 0.22 0.01
## deplonely_mosttran_devx2:rural.ses.med41[1] 0.18 0.16 0.00
## deplonely_mosttran_devx2:rural.ses.med41[2] 0.18 0.15 0.01
## deplonely_mosttran_devx2:rural.ses.med41[3] 0.37 0.21 0.02
## deplonely_mosttran_devx2:rural.ses.med41[4] 0.28 0.19 0.01
## depblues_mosttran_devx21[1] 0.25 0.15 0.03
## depblues_mosttran_devx21[2] 0.23 0.13 0.03
## depblues_mosttran_devx21[3] 0.26 0.15 0.04
## depblues_mosttran_devx21[4] 0.26 0.15 0.04
## depblues_mosttran_av12x21[1] 0.13 0.08 0.02
## depblues_mosttran_av12x21[2] 0.13 0.08 0.02
## depblues_mosttran_av12x21[3] 0.12 0.08 0.02
## depblues_mosttran_av12x21[4] 0.13 0.08 0.02
## depblues_mosttran_av12x21[5] 0.11 0.07 0.01
## depblues_mosttran_av12x21[6] 0.12 0.08 0.01
## depblues_mosttran_av12x21[7] 0.13 0.08 0.02
## depblues_mosttran_av12x21[8] 0.13 0.08 0.02
## depblues_mosttran_devx2:rural.ses.med21[1] 0.23 0.18 0.01
## depblues_mosttran_devx2:rural.ses.med21[2] 0.18 0.17 0.00
## depblues_mosttran_devx2:rural.ses.med21[3] 0.28 0.20 0.01
## depblues_mosttran_devx2:rural.ses.med21[4] 0.30 0.22 0.01
## depblues_mosttran_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## depblues_mosttran_devx2:rural.ses.med31[2] 0.20 0.17 0.00
## depblues_mosttran_devx2:rural.ses.med31[3] 0.23 0.18 0.01
## depblues_mosttran_devx2:rural.ses.med31[4] 0.31 0.22 0.01
## depblues_mosttran_devx2:rural.ses.med41[1] 0.26 0.20 0.01
## depblues_mosttran_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## depblues_mosttran_devx2:rural.ses.med41[3] 0.23 0.19 0.01
## depblues_mosttran_devx2:rural.ses.med41[4] 0.29 0.21 0.01
## depunfair_mosttran_devx21[1] 0.18 0.12 0.03
## depunfair_mosttran_devx21[2] 0.31 0.14 0.06
## depunfair_mosttran_devx21[3] 0.29 0.13 0.06
## depunfair_mosttran_devx21[4] 0.21 0.12 0.03
## depunfair_mosttran_av12x21[1] 0.12 0.08 0.02
## depunfair_mosttran_av12x21[2] 0.12 0.08 0.02
## depunfair_mosttran_av12x21[3] 0.11 0.07 0.01
## depunfair_mosttran_av12x21[4] 0.11 0.07 0.01
## depunfair_mosttran_av12x21[5] 0.11 0.07 0.02
## depunfair_mosttran_av12x21[6] 0.13 0.08 0.02
## depunfair_mosttran_av12x21[7] 0.16 0.09 0.02
## depunfair_mosttran_av12x21[8] 0.14 0.08 0.02
## depunfair_mosttran_devx2:rural.ses.med21[1] 0.20 0.17 0.01
## depunfair_mosttran_devx2:rural.ses.med21[2] 0.36 0.22 0.01
## depunfair_mosttran_devx2:rural.ses.med21[3] 0.18 0.15 0.01
## depunfair_mosttran_devx2:rural.ses.med21[4] 0.26 0.19 0.01
## depunfair_mosttran_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## depunfair_mosttran_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## depunfair_mosttran_devx2:rural.ses.med31[3] 0.21 0.17 0.01
## depunfair_mosttran_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## depunfair_mosttran_devx2:rural.ses.med41[1] 0.22 0.19 0.01
## depunfair_mosttran_devx2:rural.ses.med41[2] 0.22 0.17 0.01
## depunfair_mosttran_devx2:rural.ses.med41[3] 0.27 0.19 0.01
## depunfair_mosttran_devx2:rural.ses.med41[4] 0.29 0.20 0.01
## depmistrt_mosttran_devx21[1] 0.22 0.14 0.03
## depmistrt_mosttran_devx21[2] 0.21 0.13 0.03
## depmistrt_mosttran_devx21[3] 0.33 0.17 0.04
## depmistrt_mosttran_devx21[4] 0.24 0.14 0.03
## depmistrt_mosttran_av12x21[1] 0.13 0.08 0.02
## depmistrt_mosttran_av12x21[2] 0.14 0.09 0.02
## depmistrt_mosttran_av12x21[3] 0.13 0.08 0.02
## depmistrt_mosttran_av12x21[4] 0.11 0.07 0.02
## depmistrt_mosttran_av12x21[5] 0.10 0.07 0.01
## depmistrt_mosttran_av12x21[6] 0.10 0.07 0.01
## depmistrt_mosttran_av12x21[7] 0.15 0.09 0.02
## depmistrt_mosttran_av12x21[8] 0.14 0.09 0.02
## depmistrt_mosttran_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## depmistrt_mosttran_devx2:rural.ses.med21[2] 0.21 0.17 0.01
## depmistrt_mosttran_devx2:rural.ses.med21[3] 0.22 0.17 0.01
## depmistrt_mosttran_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## depmistrt_mosttran_devx2:rural.ses.med31[1] 0.25 0.19 0.01
## depmistrt_mosttran_devx2:rural.ses.med31[2] 0.25 0.19 0.01
## depmistrt_mosttran_devx2:rural.ses.med31[3] 0.19 0.17 0.01
## depmistrt_mosttran_devx2:rural.ses.med31[4] 0.31 0.22 0.01
## depmistrt_mosttran_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depmistrt_mosttran_devx2:rural.ses.med41[2] 0.22 0.18 0.01
## depmistrt_mosttran_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## depmistrt_mosttran_devx2:rural.ses.med41[4] 0.30 0.22 0.01
## depbetray_mosttran_devx21[1] 0.26 0.15 0.04
## depbetray_mosttran_devx21[2] 0.24 0.14 0.04
## depbetray_mosttran_devx21[3] 0.24 0.14 0.04
## depbetray_mosttran_devx21[4] 0.27 0.15 0.04
## depbetray_mosttran_av12x21[1] 0.13 0.08 0.02
## depbetray_mosttran_av12x21[2] 0.13 0.08 0.02
## depbetray_mosttran_av12x21[3] 0.12 0.08 0.02
## depbetray_mosttran_av12x21[4] 0.11 0.07 0.02
## depbetray_mosttran_av12x21[5] 0.10 0.07 0.01
## depbetray_mosttran_av12x21[6] 0.13 0.08 0.02
## depbetray_mosttran_av12x21[7] 0.16 0.10 0.02
## depbetray_mosttran_av12x21[8] 0.12 0.08 0.02
## depbetray_mosttran_devx2:rural.ses.med21[1] 0.25 0.19 0.01
## depbetray_mosttran_devx2:rural.ses.med21[2] 0.19 0.16 0.01
## depbetray_mosttran_devx2:rural.ses.med21[3] 0.25 0.19 0.01
## depbetray_mosttran_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## depbetray_mosttran_devx2:rural.ses.med31[1] 0.25 0.20 0.01
## depbetray_mosttran_devx2:rural.ses.med31[2] 0.20 0.17 0.01
## depbetray_mosttran_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## depbetray_mosttran_devx2:rural.ses.med31[4] 0.31 0.22 0.02
## depbetray_mosttran_devx2:rural.ses.med41[1] 0.26 0.20 0.01
## depbetray_mosttran_devx2:rural.ses.med41[2] 0.23 0.18 0.01
## depbetray_mosttran_devx2:rural.ses.med41[3] 0.23 0.18 0.01
## depbetray_mosttran_devx2:rural.ses.med41[4] 0.29 0.21 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## depcantgo_mosttran_devx21[1] 0.54 1.00 7256 2660
## depcantgo_mosttran_devx21[2] 0.59 1.00 7064 2716
## depcantgo_mosttran_devx21[3] 0.55 1.00 6126 2725
## depcantgo_mosttran_devx21[4] 0.52 1.00 7770 3273
## depcantgo_mosttran_av12x21[1] 0.31 1.00 9181 2258
## depcantgo_mosttran_av12x21[2] 0.32 1.00 8982 2192
## depcantgo_mosttran_av12x21[3] 0.29 1.00 7417 2515
## depcantgo_mosttran_av12x21[4] 0.31 1.00 7923 2070
## depcantgo_mosttran_av12x21[5] 0.32 1.00 7963 2621
## depcantgo_mosttran_av12x21[6] 0.30 1.00 8110 2733
## depcantgo_mosttran_av12x21[7] 0.34 1.00 6055 2877
## depcantgo_mosttran_av12x21[8] 0.35 1.00 7297 2831
## depcantgo_mosttran_devx2:rural.ses.med21[1] 0.67 1.00 5159 2118
## depcantgo_mosttran_devx2:rural.ses.med21[2] 0.62 1.00 4942 2581
## depcantgo_mosttran_devx2:rural.ses.med21[3] 0.63 1.00 4003 2388
## depcantgo_mosttran_devx2:rural.ses.med21[4] 0.78 1.00 5044 2720
## depcantgo_mosttran_devx2:rural.ses.med31[1] 0.63 1.00 4665 2692
## depcantgo_mosttran_devx2:rural.ses.med31[2] 0.62 1.00 4768 2797
## depcantgo_mosttran_devx2:rural.ses.med31[3] 0.66 1.00 4594 2859
## depcantgo_mosttran_devx2:rural.ses.med31[4] 0.79 1.00 4806 2567
## depcantgo_mosttran_devx2:rural.ses.med41[1] 0.72 1.00 5028 2524
## depcantgo_mosttran_devx2:rural.ses.med41[2] 0.67 1.00 5612 2585
## depcantgo_mosttran_devx2:rural.ses.med41[3] 0.63 1.00 5446 2968
## depcantgo_mosttran_devx2:rural.ses.med41[4] 0.76 1.00 5512 3097
## depeffort_mosttran_devx21[1] 0.59 1.00 7269 2839
## depeffort_mosttran_devx21[2] 0.56 1.00 7292 2774
## depeffort_mosttran_devx21[3] 0.55 1.00 7099 2704
## depeffort_mosttran_devx21[4] 0.60 1.00 7483 2783
## depeffort_mosttran_av12x21[1] 0.33 1.00 8011 2805
## depeffort_mosttran_av12x21[2] 0.33 1.00 7467 2416
## depeffort_mosttran_av12x21[3] 0.31 1.00 7850 2992
## depeffort_mosttran_av12x21[4] 0.32 1.00 7017 2133
## depeffort_mosttran_av12x21[5] 0.31 1.00 7880 2532
## depeffort_mosttran_av12x21[6] 0.30 1.00 7812 2678
## depeffort_mosttran_av12x21[7] 0.30 1.00 7250 2484
## depeffort_mosttran_av12x21[8] 0.33 1.00 7989 2309
## depeffort_mosttran_devx2:rural.ses.med21[1] 0.69 1.00 4661 2329
## depeffort_mosttran_devx2:rural.ses.med21[2] 0.77 1.00 3849 3151
## depeffort_mosttran_devx2:rural.ses.med21[3] 0.65 1.00 5385 3016
## depeffort_mosttran_devx2:rural.ses.med21[4] 0.74 1.00 5872 2832
## depeffort_mosttran_devx2:rural.ses.med31[1] 0.72 1.00 4419 2463
## depeffort_mosttran_devx2:rural.ses.med31[2] 0.63 1.00 5800 2739
## depeffort_mosttran_devx2:rural.ses.med31[3] 0.67 1.00 4827 3250
## depeffort_mosttran_devx2:rural.ses.med31[4] 0.79 1.00 4875 2912
## depeffort_mosttran_devx2:rural.ses.med41[1] 0.70 1.00 6310 2424
## depeffort_mosttran_devx2:rural.ses.med41[2] 0.69 1.00 5080 2384
## depeffort_mosttran_devx2:rural.ses.med41[3] 0.65 1.00 3893 2610
## depeffort_mosttran_devx2:rural.ses.med41[4] 0.81 1.00 3899 2905
## deplonely_mosttran_devx21[1] 0.58 1.00 4770 3111
## deplonely_mosttran_devx21[2] 0.54 1.00 6309 2296
## deplonely_mosttran_devx21[3] 0.59 1.00 3658 2557
## deplonely_mosttran_devx21[4] 0.59 1.00 6330 2472
## deplonely_mosttran_av12x21[1] 0.32 1.01 7517 2433
## deplonely_mosttran_av12x21[2] 0.32 1.00 7137 1879
## deplonely_mosttran_av12x21[3] 0.33 1.00 7773 2194
## deplonely_mosttran_av12x21[4] 0.32 1.00 8480 2602
## deplonely_mosttran_av12x21[5] 0.29 1.00 6523 2845
## deplonely_mosttran_av12x21[6] 0.30 1.00 7858 2737
## deplonely_mosttran_av12x21[7] 0.32 1.00 8172 2775
## deplonely_mosttran_av12x21[8] 0.33 1.00 6076 2586
## deplonely_mosttran_devx2:rural.ses.med21[1] 0.75 1.00 5861 2530
## deplonely_mosttran_devx2:rural.ses.med21[2] 0.62 1.00 4779 2859
## deplonely_mosttran_devx2:rural.ses.med21[3] 0.62 1.00 4568 2691
## deplonely_mosttran_devx2:rural.ses.med21[4] 0.78 1.00 4441 2419
## deplonely_mosttran_devx2:rural.ses.med31[1] 0.72 1.00 5619 2541
## deplonely_mosttran_devx2:rural.ses.med31[2] 0.65 1.00 4423 2876
## deplonely_mosttran_devx2:rural.ses.med31[3] 0.64 1.00 3985 2805
## deplonely_mosttran_devx2:rural.ses.med31[4] 0.80 1.00 4328 2377
## deplonely_mosttran_devx2:rural.ses.med41[1] 0.61 1.00 3912 2861
## deplonely_mosttran_devx2:rural.ses.med41[2] 0.55 1.00 5228 2839
## deplonely_mosttran_devx2:rural.ses.med41[3] 0.78 1.00 3418 2236
## deplonely_mosttran_devx2:rural.ses.med41[4] 0.70 1.00 6604 2934
## depblues_mosttran_devx21[1] 0.59 1.00 4981 2084
## depblues_mosttran_devx21[2] 0.53 1.00 6783 2739
## depblues_mosttran_devx21[3] 0.59 1.00 5037 2822
## depblues_mosttran_devx21[4] 0.59 1.00 7085 2646
## depblues_mosttran_av12x21[1] 0.32 1.00 7471 2677
## depblues_mosttran_av12x21[2] 0.33 1.01 7306 2222
## depblues_mosttran_av12x21[3] 0.31 1.00 7403 2507
## depblues_mosttran_av12x21[4] 0.33 1.00 8072 2542
## depblues_mosttran_av12x21[5] 0.29 1.00 6951 2334
## depblues_mosttran_av12x21[6] 0.30 1.00 7849 2091
## depblues_mosttran_av12x21[7] 0.33 1.00 7655 2857
## depblues_mosttran_av12x21[8] 0.33 1.00 7311 2702
## depblues_mosttran_devx2:rural.ses.med21[1] 0.68 1.00 5661 2485
## depblues_mosttran_devx2:rural.ses.med21[2] 0.62 1.00 4974 2366
## depblues_mosttran_devx2:rural.ses.med21[3] 0.75 1.00 4921 3134
## depblues_mosttran_devx2:rural.ses.med21[4] 0.77 1.00 6081 2869
## depblues_mosttran_devx2:rural.ses.med31[1] 0.69 1.00 5314 2063
## depblues_mosttran_devx2:rural.ses.med31[2] 0.63 1.00 5959 2739
## depblues_mosttran_devx2:rural.ses.med31[3] 0.69 1.00 4331 2649
## depblues_mosttran_devx2:rural.ses.med31[4] 0.78 1.00 5391 2834
## depblues_mosttran_devx2:rural.ses.med41[1] 0.74 1.00 5013 1885
## depblues_mosttran_devx2:rural.ses.med41[2] 0.65 1.00 4714 2351
## depblues_mosttran_devx2:rural.ses.med41[3] 0.69 1.00 4743 2943
## depblues_mosttran_devx2:rural.ses.med41[4] 0.75 1.00 4639 2625
## depunfair_mosttran_devx21[1] 0.47 1.00 5015 2941
## depunfair_mosttran_devx21[2] 0.61 1.00 5313 2657
## depunfair_mosttran_devx21[3] 0.58 1.00 5198 2827
## depunfair_mosttran_devx21[4] 0.51 1.00 7296 2853
## depunfair_mosttran_av12x21[1] 0.32 1.00 6551 2257
## depunfair_mosttran_av12x21[2] 0.31 1.00 6552 2254
## depunfair_mosttran_av12x21[3] 0.29 1.00 7506 2401
## depunfair_mosttran_av12x21[4] 0.29 1.00 9009 2897
## depunfair_mosttran_av12x21[5] 0.29 1.00 7400 2889
## depunfair_mosttran_av12x21[6] 0.32 1.00 7083 2311
## depunfair_mosttran_av12x21[7] 0.38 1.00 7614 3018
## depunfair_mosttran_av12x21[8] 0.33 1.00 6687 2836
## depunfair_mosttran_devx2:rural.ses.med21[1] 0.66 1.00 4968 2456
## depunfair_mosttran_devx2:rural.ses.med21[2] 0.79 1.00 2998 2189
## depunfair_mosttran_devx2:rural.ses.med21[3] 0.56 1.00 5113 2788
## depunfair_mosttran_devx2:rural.ses.med21[4] 0.71 1.00 5432 2572
## depunfair_mosttran_devx2:rural.ses.med31[1] 0.74 1.00 6063 2940
## depunfair_mosttran_devx2:rural.ses.med31[2] 0.66 1.00 5469 2781
## depunfair_mosttran_devx2:rural.ses.med31[3] 0.62 1.00 5146 2344
## depunfair_mosttran_devx2:rural.ses.med31[4] 0.76 1.00 5669 2990
## depunfair_mosttran_devx2:rural.ses.med41[1] 0.69 1.00 4088 2710
## depunfair_mosttran_devx2:rural.ses.med41[2] 0.63 1.00 6782 2729
## depunfair_mosttran_devx2:rural.ses.med41[3] 0.71 1.00 3769 2595
## depunfair_mosttran_devx2:rural.ses.med41[4] 0.74 1.00 6163 2195
## depmistrt_mosttran_devx21[1] 0.56 1.00 4024 3056
## depmistrt_mosttran_devx21[2] 0.50 1.00 8454 2639
## depmistrt_mosttran_devx21[3] 0.66 1.00 2849 2654
## depmistrt_mosttran_devx21[4] 0.56 1.00 7923 2931
## depmistrt_mosttran_av12x21[1] 0.33 1.00 7373 2625
## depmistrt_mosttran_av12x21[2] 0.35 1.00 7178 2601
## depmistrt_mosttran_av12x21[3] 0.33 1.00 7387 2781
## depmistrt_mosttran_av12x21[4] 0.28 1.00 8335 2734
## depmistrt_mosttran_av12x21[5] 0.27 1.00 7167 2739
## depmistrt_mosttran_av12x21[6] 0.28 1.00 6428 3233
## depmistrt_mosttran_av12x21[7] 0.36 1.00 7184 2921
## depmistrt_mosttran_av12x21[8] 0.34 1.00 7323 2396
## depmistrt_mosttran_devx2:rural.ses.med21[1] 0.71 1.00 5790 2540
## depmistrt_mosttran_devx2:rural.ses.med21[2] 0.62 1.00 5437 2790
## depmistrt_mosttran_devx2:rural.ses.med21[3] 0.65 1.00 4196 2296
## depmistrt_mosttran_devx2:rural.ses.med21[4] 0.79 1.00 4929 2637
## depmistrt_mosttran_devx2:rural.ses.med31[1] 0.68 1.00 5295 2010
## depmistrt_mosttran_devx2:rural.ses.med31[2] 0.69 1.00 4883 2620
## depmistrt_mosttran_devx2:rural.ses.med31[3] 0.64 1.00 5025 2690
## depmistrt_mosttran_devx2:rural.ses.med31[4] 0.79 1.00 4831 2821
## depmistrt_mosttran_devx2:rural.ses.med41[1] 0.70 1.00 6738 2491
## depmistrt_mosttran_devx2:rural.ses.med41[2] 0.68 1.00 4803 2669
## depmistrt_mosttran_devx2:rural.ses.med41[3] 0.66 1.00 4722 2958
## depmistrt_mosttran_devx2:rural.ses.med41[4] 0.78 1.00 4969 2673
## depbetray_mosttran_devx21[1] 0.58 1.00 7298 2757
## depbetray_mosttran_devx21[2] 0.55 1.00 7551 2578
## depbetray_mosttran_devx21[3] 0.55 1.00 7642 3101
## depbetray_mosttran_devx21[4] 0.60 1.00 7933 2634
## depbetray_mosttran_av12x21[1] 0.31 1.00 8182 2750
## depbetray_mosttran_av12x21[2] 0.33 1.00 7994 2338
## depbetray_mosttran_av12x21[3] 0.31 1.00 8665 2553
## depbetray_mosttran_av12x21[4] 0.29 1.00 8010 2701
## depbetray_mosttran_av12x21[5] 0.26 1.00 6796 2916
## depbetray_mosttran_av12x21[6] 0.31 1.00 8284 2603
## depbetray_mosttran_av12x21[7] 0.39 1.00 6375 3094
## depbetray_mosttran_av12x21[8] 0.30 1.00 8611 2754
## depbetray_mosttran_devx2:rural.ses.med21[1] 0.69 1.00 5678 2668
## depbetray_mosttran_devx2:rural.ses.med21[2] 0.61 1.00 6190 2627
## depbetray_mosttran_devx2:rural.ses.med21[3] 0.68 1.00 5178 2914
## depbetray_mosttran_devx2:rural.ses.med21[4] 0.77 1.00 4946 2360
## depbetray_mosttran_devx2:rural.ses.med31[1] 0.71 1.00 5024 2519
## depbetray_mosttran_devx2:rural.ses.med31[2] 0.65 1.00 5855 3089
## depbetray_mosttran_devx2:rural.ses.med31[3] 0.71 1.00 4119 3193
## depbetray_mosttran_devx2:rural.ses.med31[4] 0.78 1.00 5519 3219
## depbetray_mosttran_devx2:rural.ses.med41[1] 0.70 1.00 5200 2818
## depbetray_mosttran_devx2:rural.ses.med41[2] 0.66 1.00 5188 2255
## depbetray_mosttran_devx2:rural.ses.med41[3] 0.64 1.00 5140 2767
## depbetray_mosttran_devx2:rural.ses.med41[4] 0.76 1.00 6337 2726
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.2.2.4 Prior summary
out.chg.alldepress.sttran.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mosttran_av12x2
## normal(0, 0.25) b mosttran_devx2
## normal(0, 1) b mosttran_devx2:rural.ses.med2
## normal(0, 1) b mosttran_devx2:rural.ses.med3
## normal(0, 1) b mosttran_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mosttran_av12x21
## dirichlet(1) simo mosttran_devx2:rural.ses.med21
## dirichlet(1) simo mosttran_devx2:rural.ses.med31
## dirichlet(1) simo mosttran_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mosttran_devx21
## group resp dpar nlpar lb ub source
## default
## depbetray user
## depbetray user
## depbetray user
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depblues user
## depblues user
## depblues user
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depcantgo user
## depcantgo user
## depcantgo user
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depeffort user
## depeffort user
## depeffort user
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## deplonely user
## deplonely user
## deplonely user
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## depmistrt user
## depmistrt user
## depmistrt user
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depunfair user
## depunfair user
## depunfair user
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## default
## depbetray user
## depblues user
## depcantgo user
## depeffort user
## deplonely user
## depmistrt user
## depunfair user
## depbetray 0 default
## depblues 0 default
## depcantgo 0 default
## depeffort 0 default
## deplonely 0 default
## depmistrt 0 default
## depunfair 0 default
## id depbetray 0 (vectorized)
## id depbetray 0 (vectorized)
## id depblues 0 (vectorized)
## id depblues 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depeffort 0 (vectorized)
## id depeffort 0 (vectorized)
## id deplonely 0 (vectorized)
## id deplonely 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depunfair 0 (vectorized)
## id depunfair 0 (vectorized)
## depbetray user
## depbetray default
## depbetray default
## depbetray default
## depbetray user
## depblues user
## depblues default
## depblues default
## depblues default
## depblues user
## depcantgo user
## depcantgo default
## depcantgo default
## depcantgo default
## depcantgo user
## depeffort user
## depeffort default
## depeffort default
## depeffort default
## depeffort user
## deplonely user
## deplonely default
## deplonely default
## deplonely default
## deplonely user
## depmistrt user
## depmistrt default
## depmistrt default
## depmistrt default
## depmistrt user
## depunfair user
## depunfair default
## depunfair default
## depunfair default
## depunfair user8.1.2.3 Corr X Community: stresp & negative emotions
#Community Change: negative emotions items ~ mo(stresp)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostresp_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostresp_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostresp_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostresp_av12x21',
resp = depdv_names)
)
chg.alldepress.stresp.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) +
rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stresp_comm_fit",
file_refit = "on_change"
)
out.chg.alldepress.stresp.comm.fit <- ppchecks(chg.alldepress.stresp.comm.fit)8.1.2.3.1 Coefficient plot (intervals)
out.chg.alldepress.stresp.comm.fit[[11]]
8.1.2.3.2 PPcheck (density)
p1 <- out.chg.alldepress.stresp.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stresp.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stresp.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stresp.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stresp.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stresp.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stresp.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
8.1.2.3.3 Fit summary
out.chg.alldepress.stresp.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## depeffort ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## deplonely ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## depblues ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## depunfair ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## depmistrt ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## depbetray ~ 1 + mo(stresp_devx2) + mo(stresp_av12x2) + rural.ses.med + mo(stresp_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.31 0.20 0.02 0.73 1.01 525
## sd(depeffort_Intercept) 0.44 0.27 0.02 0.99 1.00 759
## sd(deplonely_Intercept) 0.41 0.24 0.02 0.89 1.01 543
## sd(depblues_Intercept) 0.68 0.33 0.06 1.29 1.00 560
## sd(depunfair_Intercept) 0.23 0.17 0.01 0.62 1.01 967
## sd(depmistrt_Intercept) 0.30 0.21 0.01 0.74 1.00 850
## sd(depbetray_Intercept) 0.41 0.25 0.03 0.94 1.01 608
## Tail_ESS
## sd(depcantgo_Intercept) 1379
## sd(depeffort_Intercept) 1575
## sd(deplonely_Intercept) 1178
## sd(depblues_Intercept) 763
## sd(depunfair_Intercept) 2032
## sd(depmistrt_Intercept) 1570
## sd(depbetray_Intercept) 1720
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_Intercept -0.53 0.32 -1.18 0.11
## depeffort_Intercept -2.36 0.43 -3.27 -1.58
## deplonely_Intercept -1.05 0.34 -1.74 -0.36
## depblues_Intercept -2.16 0.43 -3.02 -1.28
## depunfair_Intercept -2.02 0.36 -2.71 -1.29
## depmistrt_Intercept -2.58 0.45 -3.46 -1.64
## depbetray_Intercept -2.92 0.42 -3.75 -2.10
## depcantgo_rural.ses.med2 0.18 0.45 -0.68 1.14
## depcantgo_rural.ses.med3 0.03 0.47 -1.08 0.86
## depcantgo_rural.ses.med4 0.44 0.53 -0.47 1.59
## depeffort_rural.ses.med2 -0.43 0.58 -1.63 0.62
## depeffort_rural.ses.med3 0.13 0.57 -1.16 1.15
## depeffort_rural.ses.med4 0.68 0.51 -0.32 1.69
## deplonely_rural.ses.med2 -0.56 0.48 -1.49 0.47
## deplonely_rural.ses.med3 -0.27 0.46 -1.25 0.61
## deplonely_rural.ses.med4 0.26 0.44 -0.61 1.20
## depblues_rural.ses.med2 0.31 0.59 -0.75 1.57
## depblues_rural.ses.med3 0.33 0.54 -0.87 1.29
## depblues_rural.ses.med4 0.22 0.55 -0.95 1.24
## depunfair_rural.ses.med2 -0.70 0.53 -1.80 0.27
## depunfair_rural.ses.med3 0.65 0.45 -0.31 1.52
## depunfair_rural.ses.med4 0.62 0.47 -0.39 1.47
## depmistrt_rural.ses.med2 0.14 0.63 -1.23 1.28
## depmistrt_rural.ses.med3 0.30 0.50 -0.69 1.30
## depmistrt_rural.ses.med4 0.18 0.57 -0.97 1.25
## depbetray_rural.ses.med2 0.37 0.58 -0.92 1.45
## depbetray_rural.ses.med3 0.98 0.51 0.02 2.08
## depbetray_rural.ses.med4 0.83 0.54 -0.31 1.82
## depcantgo_mostresp_devx2 0.10 0.14 -0.18 0.37
## depcantgo_mostresp_av12x2 -0.03 0.03 -0.08 0.02
## depcantgo_mostresp_devx2:rural.ses.med2 0.04 0.23 -0.49 0.44
## depcantgo_mostresp_devx2:rural.ses.med3 0.03 0.21 -0.34 0.51
## depcantgo_mostresp_devx2:rural.ses.med4 -0.14 0.23 -0.59 0.29
## depeffort_mostresp_devx2 0.08 0.16 -0.23 0.41
## depeffort_mostresp_av12x2 0.04 0.03 -0.03 0.11
## depeffort_mostresp_devx2:rural.ses.med2 0.22 0.27 -0.29 0.78
## depeffort_mostresp_devx2:rural.ses.med3 0.18 0.26 -0.27 0.74
## depeffort_mostresp_devx2:rural.ses.med4 -0.10 0.24 -0.59 0.38
## deplonely_mostresp_devx2 0.01 0.14 -0.27 0.30
## deplonely_mostresp_av12x2 0.01 0.03 -0.05 0.06
## deplonely_mostresp_devx2:rural.ses.med2 0.12 0.23 -0.37 0.56
## deplonely_mostresp_devx2:rural.ses.med3 0.10 0.21 -0.32 0.51
## deplonely_mostresp_devx2:rural.ses.med4 0.11 0.22 -0.32 0.53
## depblues_mostresp_devx2 -0.11 0.16 -0.43 0.21
## depblues_mostresp_av12x2 0.02 0.04 -0.06 0.09
## depblues_mostresp_devx2:rural.ses.med2 -0.02 0.29 -0.62 0.52
## depblues_mostresp_devx2:rural.ses.med3 -0.04 0.31 -0.77 0.50
## depblues_mostresp_devx2:rural.ses.med4 0.23 0.27 -0.27 0.78
## depunfair_mostresp_devx2 0.10 0.15 -0.23 0.37
## depunfair_mostresp_av12x2 0.05 0.03 -0.00 0.11
## depunfair_mostresp_devx2:rural.ses.med2 0.52 0.23 0.11 0.99
## depunfair_mostresp_devx2:rural.ses.med3 0.05 0.21 -0.35 0.48
## depunfair_mostresp_devx2:rural.ses.med4 0.20 0.21 -0.21 0.64
## depmistrt_mostresp_devx2 -0.07 0.18 -0.42 0.28
## depmistrt_mostresp_av12x2 0.13 0.03 0.07 0.20
## depmistrt_mostresp_devx2:rural.ses.med2 0.40 0.43 -0.31 1.27
## depmistrt_mostresp_devx2:rural.ses.med3 0.18 0.24 -0.28 0.65
## depmistrt_mostresp_devx2:rural.ses.med4 0.24 0.24 -0.28 0.69
## depbetray_mostresp_devx2 0.08 0.16 -0.23 0.38
## depbetray_mostresp_av12x2 0.10 0.03 0.04 0.17
## depbetray_mostresp_devx2:rural.ses.med2 0.02 0.30 -0.67 0.55
## depbetray_mostresp_devx2:rural.ses.med3 -0.09 0.24 -0.58 0.38
## depbetray_mostresp_devx2:rural.ses.med4 0.09 0.24 -0.36 0.59
## Rhat Bulk_ESS Tail_ESS
## depcantgo_Intercept 1.00 2962 2980
## depeffort_Intercept 1.00 3267 2956
## deplonely_Intercept 1.00 2853 2986
## depblues_Intercept 1.00 2963 2903
## depunfair_Intercept 1.00 3012 2648
## depmistrt_Intercept 1.00 2643 2819
## depbetray_Intercept 1.00 2743 2715
## depcantgo_rural.ses.med2 1.00 3458 2849
## depcantgo_rural.ses.med3 1.00 2601 2578
## depcantgo_rural.ses.med4 1.00 3023 2898
## depeffort_rural.ses.med2 1.00 3944 2990
## depeffort_rural.ses.med3 1.00 3293 2783
## depeffort_rural.ses.med4 1.00 4181 2856
## deplonely_rural.ses.med2 1.00 3058 2569
## deplonely_rural.ses.med3 1.00 2453 2461
## deplonely_rural.ses.med4 1.00 3378 3065
## depblues_rural.ses.med2 1.00 3248 2689
## depblues_rural.ses.med3 1.00 3304 2648
## depblues_rural.ses.med4 1.00 3789 2942
## depunfair_rural.ses.med2 1.00 3776 2590
## depunfair_rural.ses.med3 1.00 3110 2319
## depunfair_rural.ses.med4 1.00 3083 2344
## depmistrt_rural.ses.med2 1.00 2315 2794
## depmistrt_rural.ses.med3 1.00 3561 2776
## depmistrt_rural.ses.med4 1.00 2991 2987
## depbetray_rural.ses.med2 1.00 3334 2679
## depbetray_rural.ses.med3 1.00 2995 2906
## depbetray_rural.ses.med4 1.00 3255 2964
## depcantgo_mostresp_devx2 1.00 2550 2752
## depcantgo_mostresp_av12x2 1.00 7460 3242
## depcantgo_mostresp_devx2:rural.ses.med2 1.00 2857 2768
## depcantgo_mostresp_devx2:rural.ses.med3 1.00 2321 2381
## depcantgo_mostresp_devx2:rural.ses.med4 1.00 2819 3216
## depeffort_mostresp_devx2 1.00 3127 3315
## depeffort_mostresp_av12x2 1.00 6638 3070
## depeffort_mostresp_devx2:rural.ses.med2 1.00 3850 3134
## depeffort_mostresp_devx2:rural.ses.med3 1.00 3299 3369
## depeffort_mostresp_devx2:rural.ses.med4 1.00 4137 3091
## deplonely_mostresp_devx2 1.00 2788 3094
## deplonely_mostresp_av12x2 1.00 6035 3108
## deplonely_mostresp_devx2:rural.ses.med2 1.00 2779 2309
## deplonely_mostresp_devx2:rural.ses.med3 1.00 2621 2727
## deplonely_mostresp_devx2:rural.ses.med4 1.00 3200 2719
## depblues_mostresp_devx2 1.00 3048 3140
## depblues_mostresp_av12x2 1.00 5719 3191
## depblues_mostresp_devx2:rural.ses.med2 1.00 3194 3078
## depblues_mostresp_devx2:rural.ses.med3 1.00 2594 2388
## depblues_mostresp_devx2:rural.ses.med4 1.00 3575 3023
## depunfair_mostresp_devx2 1.00 2790 2657
## depunfair_mostresp_av12x2 1.00 6670 3194
## depunfair_mostresp_devx2:rural.ses.med2 1.00 2761 2748
## depunfair_mostresp_devx2:rural.ses.med3 1.00 3032 2654
## depunfair_mostresp_devx2:rural.ses.med4 1.00 3258 2790
## depmistrt_mostresp_devx2 1.00 2145 3034
## depmistrt_mostresp_av12x2 1.00 6116 2882
## depmistrt_mostresp_devx2:rural.ses.med2 1.00 1940 2629
## depmistrt_mostresp_devx2:rural.ses.med3 1.00 2738 2969
## depmistrt_mostresp_devx2:rural.ses.med4 1.00 2567 2505
## depbetray_mostresp_devx2 1.00 2730 3071
## depbetray_mostresp_av12x2 1.00 6319 3219
## depbetray_mostresp_devx2:rural.ses.med2 1.00 2752 2210
## depbetray_mostresp_devx2:rural.ses.med3 1.00 2778 2896
## depbetray_mostresp_devx2:rural.ses.med4 1.00 3183 3059
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## depcantgo_mostresp_devx21[1] 0.26 0.14 0.04
## depcantgo_mostresp_devx21[2] 0.21 0.13 0.03
## depcantgo_mostresp_devx21[3] 0.27 0.14 0.04
## depcantgo_mostresp_devx21[4] 0.26 0.15 0.04
## depcantgo_mostresp_av12x21[1] 0.13 0.08 0.02
## depcantgo_mostresp_av12x21[2] 0.13 0.09 0.01
## depcantgo_mostresp_av12x21[3] 0.12 0.08 0.01
## depcantgo_mostresp_av12x21[4] 0.11 0.07 0.01
## depcantgo_mostresp_av12x21[5] 0.12 0.08 0.02
## depcantgo_mostresp_av12x21[6] 0.12 0.08 0.02
## depcantgo_mostresp_av12x21[7] 0.12 0.08 0.02
## depcantgo_mostresp_av12x21[8] 0.14 0.09 0.02
## depcantgo_mostresp_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## depcantgo_mostresp_devx2:rural.ses.med21[2] 0.21 0.17 0.01
## depcantgo_mostresp_devx2:rural.ses.med21[3] 0.25 0.20 0.01
## depcantgo_mostresp_devx2:rural.ses.med21[4] 0.28 0.21 0.01
## depcantgo_mostresp_devx2:rural.ses.med31[1] 0.28 0.21 0.01
## depcantgo_mostresp_devx2:rural.ses.med31[2] 0.22 0.19 0.01
## depcantgo_mostresp_devx2:rural.ses.med31[3] 0.21 0.17 0.01
## depcantgo_mostresp_devx2:rural.ses.med31[4] 0.29 0.21 0.01
## depcantgo_mostresp_devx2:rural.ses.med41[1] 0.28 0.20 0.01
## depcantgo_mostresp_devx2:rural.ses.med41[2] 0.27 0.20 0.01
## depcantgo_mostresp_devx2:rural.ses.med41[3] 0.19 0.17 0.00
## depcantgo_mostresp_devx2:rural.ses.med41[4] 0.25 0.19 0.01
## depeffort_mostresp_devx21[1] 0.27 0.15 0.04
## depeffort_mostresp_devx21[2] 0.25 0.14 0.04
## depeffort_mostresp_devx21[3] 0.22 0.14 0.03
## depeffort_mostresp_devx21[4] 0.26 0.15 0.04
## depeffort_mostresp_av12x21[1] 0.12 0.08 0.02
## depeffort_mostresp_av12x21[2] 0.14 0.09 0.02
## depeffort_mostresp_av12x21[3] 0.12 0.07 0.02
## depeffort_mostresp_av12x21[4] 0.12 0.08 0.02
## depeffort_mostresp_av12x21[5] 0.12 0.08 0.02
## depeffort_mostresp_av12x21[6] 0.13 0.08 0.02
## depeffort_mostresp_av12x21[7] 0.13 0.08 0.02
## depeffort_mostresp_av12x21[8] 0.13 0.08 0.02
## depeffort_mostresp_devx2:rural.ses.med21[1] 0.24 0.19 0.01
## depeffort_mostresp_devx2:rural.ses.med21[2] 0.28 0.20 0.01
## depeffort_mostresp_devx2:rural.ses.med21[3] 0.20 0.17 0.01
## depeffort_mostresp_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## depeffort_mostresp_devx2:rural.ses.med31[1] 0.29 0.20 0.01
## depeffort_mostresp_devx2:rural.ses.med31[2] 0.21 0.17 0.01
## depeffort_mostresp_devx2:rural.ses.med31[3] 0.19 0.17 0.00
## depeffort_mostresp_devx2:rural.ses.med31[4] 0.32 0.21 0.01
## depeffort_mostresp_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depeffort_mostresp_devx2:rural.ses.med41[2] 0.23 0.19 0.01
## depeffort_mostresp_devx2:rural.ses.med41[3] 0.23 0.18 0.01
## depeffort_mostresp_devx2:rural.ses.med41[4] 0.28 0.21 0.01
## deplonely_mostresp_devx21[1] 0.26 0.15 0.04
## deplonely_mostresp_devx21[2] 0.24 0.14 0.04
## deplonely_mostresp_devx21[3] 0.24 0.14 0.04
## deplonely_mostresp_devx21[4] 0.26 0.15 0.04
## deplonely_mostresp_av12x21[1] 0.13 0.08 0.02
## deplonely_mostresp_av12x21[2] 0.13 0.08 0.02
## deplonely_mostresp_av12x21[3] 0.12 0.08 0.02
## deplonely_mostresp_av12x21[4] 0.12 0.08 0.02
## deplonely_mostresp_av12x21[5] 0.12 0.08 0.02
## deplonely_mostresp_av12x21[6] 0.12 0.08 0.02
## deplonely_mostresp_av12x21[7] 0.12 0.08 0.02
## deplonely_mostresp_av12x21[8] 0.13 0.08 0.02
## deplonely_mostresp_devx2:rural.ses.med21[1] 0.24 0.19 0.01
## deplonely_mostresp_devx2:rural.ses.med21[2] 0.20 0.16 0.01
## deplonely_mostresp_devx2:rural.ses.med21[3] 0.28 0.20 0.01
## deplonely_mostresp_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## deplonely_mostresp_devx2:rural.ses.med31[1] 0.27 0.19 0.01
## deplonely_mostresp_devx2:rural.ses.med31[2] 0.25 0.19 0.01
## deplonely_mostresp_devx2:rural.ses.med31[3] 0.21 0.17 0.01
## deplonely_mostresp_devx2:rural.ses.med31[4] 0.28 0.20 0.01
## deplonely_mostresp_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## deplonely_mostresp_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## deplonely_mostresp_devx2:rural.ses.med41[3] 0.27 0.19 0.01
## deplonely_mostresp_devx2:rural.ses.med41[4] 0.28 0.20 0.01
## depblues_mostresp_devx21[1] 0.26 0.14 0.04
## depblues_mostresp_devx21[2] 0.24 0.14 0.04
## depblues_mostresp_devx21[3] 0.24 0.14 0.04
## depblues_mostresp_devx21[4] 0.26 0.15 0.04
## depblues_mostresp_av12x21[1] 0.12 0.08 0.02
## depblues_mostresp_av12x21[2] 0.12 0.08 0.02
## depblues_mostresp_av12x21[3] 0.12 0.08 0.02
## depblues_mostresp_av12x21[4] 0.12 0.08 0.02
## depblues_mostresp_av12x21[5] 0.12 0.08 0.01
## depblues_mostresp_av12x21[6] 0.13 0.09 0.02
## depblues_mostresp_av12x21[7] 0.13 0.08 0.02
## depblues_mostresp_av12x21[8] 0.13 0.08 0.02
## depblues_mostresp_devx2:rural.ses.med21[1] 0.27 0.21 0.01
## depblues_mostresp_devx2:rural.ses.med21[2] 0.22 0.19 0.01
## depblues_mostresp_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## depblues_mostresp_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## depblues_mostresp_devx2:rural.ses.med31[1] 0.26 0.20 0.01
## depblues_mostresp_devx2:rural.ses.med31[2] 0.21 0.17 0.01
## depblues_mostresp_devx2:rural.ses.med31[3] 0.22 0.17 0.01
## depblues_mostresp_devx2:rural.ses.med31[4] 0.31 0.22 0.01
## depblues_mostresp_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depblues_mostresp_devx2:rural.ses.med41[2] 0.19 0.17 0.00
## depblues_mostresp_devx2:rural.ses.med41[3] 0.26 0.19 0.01
## depblues_mostresp_devx2:rural.ses.med41[4] 0.29 0.20 0.01
## depunfair_mostresp_devx21[1] 0.23 0.14 0.03
## depunfair_mostresp_devx21[2] 0.25 0.14 0.04
## depunfair_mostresp_devx21[3] 0.25 0.14 0.04
## depunfair_mostresp_devx21[4] 0.26 0.14 0.04
## depunfair_mostresp_av12x21[1] 0.11 0.07 0.01
## depunfair_mostresp_av12x21[2] 0.13 0.08 0.02
## depunfair_mostresp_av12x21[3] 0.13 0.08 0.02
## depunfair_mostresp_av12x21[4] 0.12 0.08 0.02
## depunfair_mostresp_av12x21[5] 0.12 0.08 0.01
## depunfair_mostresp_av12x21[6] 0.12 0.08 0.02
## depunfair_mostresp_av12x21[7] 0.14 0.08 0.02
## depunfair_mostresp_av12x21[8] 0.14 0.08 0.02
## depunfair_mostresp_devx2:rural.ses.med21[1] 0.12 0.11 0.00
## depunfair_mostresp_devx2:rural.ses.med21[2] 0.43 0.18 0.07
## depunfair_mostresp_devx2:rural.ses.med21[3] 0.21 0.14 0.01
## depunfair_mostresp_devx2:rural.ses.med21[4] 0.24 0.16 0.01
## depunfair_mostresp_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## depunfair_mostresp_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## depunfair_mostresp_devx2:rural.ses.med31[3] 0.22 0.17 0.01
## depunfair_mostresp_devx2:rural.ses.med31[4] 0.28 0.20 0.01
## depunfair_mostresp_devx2:rural.ses.med41[1] 0.26 0.18 0.01
## depunfair_mostresp_devx2:rural.ses.med41[2] 0.25 0.18 0.01
## depunfair_mostresp_devx2:rural.ses.med41[3] 0.23 0.16 0.01
## depunfair_mostresp_devx2:rural.ses.med41[4] 0.27 0.19 0.01
## depmistrt_mostresp_devx21[1] 0.29 0.16 0.04
## depmistrt_mostresp_devx21[2] 0.22 0.14 0.03
## depmistrt_mostresp_devx21[3] 0.24 0.14 0.04
## depmistrt_mostresp_devx21[4] 0.25 0.15 0.04
## depmistrt_mostresp_av12x21[1] 0.09 0.06 0.01
## depmistrt_mostresp_av12x21[2] 0.12 0.07 0.02
## depmistrt_mostresp_av12x21[3] 0.14 0.08 0.02
## depmistrt_mostresp_av12x21[4] 0.15 0.09 0.02
## depmistrt_mostresp_av12x21[5] 0.12 0.07 0.02
## depmistrt_mostresp_av12x21[6] 0.13 0.08 0.02
## depmistrt_mostresp_av12x21[7] 0.13 0.08 0.02
## depmistrt_mostresp_av12x21[8] 0.13 0.08 0.02
## depmistrt_mostresp_devx2:rural.ses.med21[1] 0.25 0.18 0.01
## depmistrt_mostresp_devx2:rural.ses.med21[2] 0.15 0.15 0.00
## depmistrt_mostresp_devx2:rural.ses.med21[3] 0.15 0.16 0.00
## depmistrt_mostresp_devx2:rural.ses.med21[4] 0.45 0.26 0.02
## depmistrt_mostresp_devx2:rural.ses.med31[1] 0.24 0.18 0.01
## depmistrt_mostresp_devx2:rural.ses.med31[2] 0.25 0.18 0.01
## depmistrt_mostresp_devx2:rural.ses.med31[3] 0.22 0.17 0.01
## depmistrt_mostresp_devx2:rural.ses.med31[4] 0.30 0.20 0.01
## depmistrt_mostresp_devx2:rural.ses.med41[1] 0.21 0.18 0.01
## depmistrt_mostresp_devx2:rural.ses.med41[2] 0.35 0.22 0.02
## depmistrt_mostresp_devx2:rural.ses.med41[3] 0.20 0.16 0.01
## depmistrt_mostresp_devx2:rural.ses.med41[4] 0.24 0.18 0.01
## depbetray_mostresp_devx21[1] 0.25 0.14 0.04
## depbetray_mostresp_devx21[2] 0.24 0.14 0.04
## depbetray_mostresp_devx21[3] 0.24 0.14 0.04
## depbetray_mostresp_devx21[4] 0.26 0.14 0.04
## depbetray_mostresp_av12x21[1] 0.11 0.07 0.01
## depbetray_mostresp_av12x21[2] 0.13 0.08 0.02
## depbetray_mostresp_av12x21[3] 0.13 0.08 0.02
## depbetray_mostresp_av12x21[4] 0.14 0.08 0.02
## depbetray_mostresp_av12x21[5] 0.13 0.08 0.02
## depbetray_mostresp_av12x21[6] 0.14 0.09 0.02
## depbetray_mostresp_av12x21[7] 0.13 0.08 0.02
## depbetray_mostresp_av12x21[8] 0.10 0.07 0.01
## depbetray_mostresp_devx2:rural.ses.med21[1] 0.27 0.20 0.01
## depbetray_mostresp_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## depbetray_mostresp_devx2:rural.ses.med21[3] 0.21 0.17 0.01
## depbetray_mostresp_devx2:rural.ses.med21[4] 0.29 0.22 0.01
## depbetray_mostresp_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## depbetray_mostresp_devx2:rural.ses.med31[2] 0.20 0.17 0.01
## depbetray_mostresp_devx2:rural.ses.med31[3] 0.26 0.19 0.01
## depbetray_mostresp_devx2:rural.ses.med31[4] 0.27 0.20 0.01
## depbetray_mostresp_devx2:rural.ses.med41[1] 0.28 0.20 0.01
## depbetray_mostresp_devx2:rural.ses.med41[2] 0.21 0.18 0.01
## depbetray_mostresp_devx2:rural.ses.med41[3] 0.23 0.18 0.01
## depbetray_mostresp_devx2:rural.ses.med41[4] 0.27 0.20 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## depcantgo_mostresp_devx21[1] 0.58 1.00 6901 2593
## depcantgo_mostresp_devx21[2] 0.53 1.00 7431 3070
## depcantgo_mostresp_devx21[3] 0.57 1.00 5555 2670
## depcantgo_mostresp_devx21[4] 0.58 1.00 7779 3140
## depcantgo_mostresp_av12x21[1] 0.33 1.00 8749 2821
## depcantgo_mostresp_av12x21[2] 0.33 1.00 9293 2270
## depcantgo_mostresp_av12x21[3] 0.31 1.00 7196 2139
## depcantgo_mostresp_av12x21[4] 0.30 1.00 7618 2699
## depcantgo_mostresp_av12x21[5] 0.30 1.00 7833 3104
## depcantgo_mostresp_av12x21[6] 0.31 1.00 7337 2553
## depcantgo_mostresp_av12x21[7] 0.31 1.00 7224 2678
## depcantgo_mostresp_av12x21[8] 0.35 1.00 6692 2878
## depcantgo_mostresp_devx2:rural.ses.med21[1] 0.69 1.00 5838 2448
## depcantgo_mostresp_devx2:rural.ses.med21[2] 0.64 1.00 7078 2971
## depcantgo_mostresp_devx2:rural.ses.med21[3] 0.70 1.00 4158 2827
## depcantgo_mostresp_devx2:rural.ses.med21[4] 0.75 1.00 5177 2896
## depcantgo_mostresp_devx2:rural.ses.med31[1] 0.74 1.00 4748 2674
## depcantgo_mostresp_devx2:rural.ses.med31[2] 0.68 1.00 3971 2755
## depcantgo_mostresp_devx2:rural.ses.med31[3] 0.63 1.00 5211 2740
## depcantgo_mostresp_devx2:rural.ses.med31[4] 0.73 1.00 5586 2917
## depcantgo_mostresp_devx2:rural.ses.med41[1] 0.72 1.00 5754 2212
## depcantgo_mostresp_devx2:rural.ses.med41[2] 0.73 1.00 5515 2699
## depcantgo_mostresp_devx2:rural.ses.med41[3] 0.64 1.00 4622 3191
## depcantgo_mostresp_devx2:rural.ses.med41[4] 0.71 1.00 6605 3031
## depeffort_mostresp_devx21[1] 0.60 1.00 7610 3079
## depeffort_mostresp_devx21[2] 0.57 1.00 5891 2915
## depeffort_mostresp_devx21[3] 0.54 1.00 5364 3305
## depeffort_mostresp_devx21[4] 0.59 1.00 6308 2460
## depeffort_mostresp_av12x21[1] 0.31 1.00 7769 2600
## depeffort_mostresp_av12x21[2] 0.35 1.00 8315 2679
## depeffort_mostresp_av12x21[3] 0.30 1.00 6366 2063
## depeffort_mostresp_av12x21[4] 0.30 1.00 7891 2388
## depeffort_mostresp_av12x21[5] 0.31 1.00 6758 2578
## depeffort_mostresp_av12x21[6] 0.33 1.00 6912 2121
## depeffort_mostresp_av12x21[7] 0.31 1.00 7242 2640
## depeffort_mostresp_av12x21[8] 0.33 1.00 8575 2517
## depeffort_mostresp_devx2:rural.ses.med21[1] 0.67 1.00 6654 2647
## depeffort_mostresp_devx2:rural.ses.med21[2] 0.72 1.00 6093 2352
## depeffort_mostresp_devx2:rural.ses.med21[3] 0.62 1.00 6012 2567
## depeffort_mostresp_devx2:rural.ses.med21[4] 0.73 1.00 6253 2897
## depeffort_mostresp_devx2:rural.ses.med31[1] 0.73 1.00 5334 2227
## depeffort_mostresp_devx2:rural.ses.med31[2] 0.64 1.00 5260 2474
## depeffort_mostresp_devx2:rural.ses.med31[3] 0.64 1.00 4734 2649
## depeffort_mostresp_devx2:rural.ses.med31[4] 0.77 1.00 5301 2873
## depeffort_mostresp_devx2:rural.ses.med41[1] 0.72 1.00 7871 2516
## depeffort_mostresp_devx2:rural.ses.med41[2] 0.69 1.00 6697 2004
## depeffort_mostresp_devx2:rural.ses.med41[3] 0.67 1.00 6708 2414
## depeffort_mostresp_devx2:rural.ses.med41[4] 0.73 1.00 5723 2321
## deplonely_mostresp_devx21[1] 0.59 1.00 7148 2080
## deplonely_mostresp_devx21[2] 0.56 1.00 7702 2976
## deplonely_mostresp_devx21[3] 0.55 1.00 6149 2800
## deplonely_mostresp_devx21[4] 0.60 1.01 8966 2546
## deplonely_mostresp_av12x21[1] 0.33 1.00 8180 3064
## deplonely_mostresp_av12x21[2] 0.33 1.00 7168 2700
## deplonely_mostresp_av12x21[3] 0.31 1.00 6874 2356
## deplonely_mostresp_av12x21[4] 0.32 1.00 8168 2466
## deplonely_mostresp_av12x21[5] 0.31 1.00 8251 2596
## deplonely_mostresp_av12x21[6] 0.32 1.00 7607 2871
## deplonely_mostresp_av12x21[7] 0.32 1.00 7458 2765
## deplonely_mostresp_av12x21[8] 0.33 1.00 6701 3092
## deplonely_mostresp_devx2:rural.ses.med21[1] 0.70 1.00 5388 2733
## deplonely_mostresp_devx2:rural.ses.med21[2] 0.60 1.00 4671 2580
## deplonely_mostresp_devx2:rural.ses.med21[3] 0.71 1.00 4000 2886
## deplonely_mostresp_devx2:rural.ses.med21[4] 0.73 1.00 6086 3021
## deplonely_mostresp_devx2:rural.ses.med31[1] 0.71 1.00 5792 2306
## deplonely_mostresp_devx2:rural.ses.med31[2] 0.69 1.00 6764 2623
## deplonely_mostresp_devx2:rural.ses.med31[3] 0.62 1.00 5026 2965
## deplonely_mostresp_devx2:rural.ses.med31[4] 0.73 1.00 6335 2649
## deplonely_mostresp_devx2:rural.ses.med41[1] 0.70 1.00 6100 2233
## deplonely_mostresp_devx2:rural.ses.med41[2] 0.62 1.00 7297 2611
## deplonely_mostresp_devx2:rural.ses.med41[3] 0.69 1.00 5253 2966
## deplonely_mostresp_devx2:rural.ses.med41[4] 0.73 1.00 6368 2547
## depblues_mostresp_devx21[1] 0.58 1.00 7396 2830
## depblues_mostresp_devx21[2] 0.56 1.00 8032 3021
## depblues_mostresp_devx21[3] 0.55 1.00 7088 3012
## depblues_mostresp_devx21[4] 0.58 1.00 7533 2171
## depblues_mostresp_av12x21[1] 0.31 1.00 7677 2732
## depblues_mostresp_av12x21[2] 0.30 1.00 7589 2600
## depblues_mostresp_av12x21[3] 0.32 1.00 8558 2509
## depblues_mostresp_av12x21[4] 0.31 1.00 8820 2502
## depblues_mostresp_av12x21[5] 0.32 1.00 8548 2193
## depblues_mostresp_av12x21[6] 0.34 1.00 7696 2912
## depblues_mostresp_av12x21[7] 0.33 1.00 7682 3176
## depblues_mostresp_av12x21[8] 0.32 1.00 8207 2763
## depblues_mostresp_devx2:rural.ses.med21[1] 0.73 1.00 5095 2528
## depblues_mostresp_devx2:rural.ses.med21[2] 0.68 1.00 4479 2624
## depblues_mostresp_devx2:rural.ses.med21[3] 0.68 1.00 5615 2744
## depblues_mostresp_devx2:rural.ses.med21[4] 0.74 1.00 6414 2477
## depblues_mostresp_devx2:rural.ses.med31[1] 0.73 1.00 4422 2661
## depblues_mostresp_devx2:rural.ses.med31[2] 0.64 1.00 5398 2662
## depblues_mostresp_devx2:rural.ses.med31[3] 0.65 1.00 5234 3009
## depblues_mostresp_devx2:rural.ses.med31[4] 0.80 1.00 4464 2749
## depblues_mostresp_devx2:rural.ses.med41[1] 0.69 1.00 5060 1945
## depblues_mostresp_devx2:rural.ses.med41[2] 0.64 1.00 6596 2762
## depblues_mostresp_devx2:rural.ses.med41[3] 0.70 1.00 5925 2533
## depblues_mostresp_devx2:rural.ses.med41[4] 0.75 1.00 6510 2641
## depunfair_mostresp_devx21[1] 0.56 1.00 4875 2994
## depunfair_mostresp_devx21[2] 0.56 1.00 5826 2081
## depunfair_mostresp_devx21[3] 0.56 1.00 5243 3034
## depunfair_mostresp_devx21[4] 0.58 1.00 7773 3156
## depunfair_mostresp_av12x21[1] 0.29 1.00 7875 2429
## depunfair_mostresp_av12x21[2] 0.33 1.00 7197 1925
## depunfair_mostresp_av12x21[3] 0.33 1.00 7058 2658
## depunfair_mostresp_av12x21[4] 0.30 1.00 7235 2322
## depunfair_mostresp_av12x21[5] 0.31 1.00 8298 2058
## depunfair_mostresp_av12x21[6] 0.31 1.00 7932 2653
## depunfair_mostresp_av12x21[7] 0.35 1.00 7086 2678
## depunfair_mostresp_av12x21[8] 0.33 1.00 6581 2680
## depunfair_mostresp_devx2:rural.ses.med21[1] 0.41 1.00 5893 2520
## depunfair_mostresp_devx2:rural.ses.med21[2] 0.78 1.00 4534 2160
## depunfair_mostresp_devx2:rural.ses.med21[3] 0.54 1.00 4618 2753
## depunfair_mostresp_devx2:rural.ses.med21[4] 0.61 1.00 6487 2901
## depunfair_mostresp_devx2:rural.ses.med31[1] 0.70 1.00 5462 2158
## depunfair_mostresp_devx2:rural.ses.med31[2] 0.64 1.00 6582 2738
## depunfair_mostresp_devx2:rural.ses.med31[3] 0.65 1.00 5808 2899
## depunfair_mostresp_devx2:rural.ses.med31[4] 0.74 1.00 5833 2529
## depunfair_mostresp_devx2:rural.ses.med41[1] 0.68 1.00 5823 2214
## depunfair_mostresp_devx2:rural.ses.med41[2] 0.68 1.00 7225 2704
## depunfair_mostresp_devx2:rural.ses.med41[3] 0.62 1.00 5745 2771
## depunfair_mostresp_devx2:rural.ses.med41[4] 0.69 1.00 7380 3026
## depmistrt_mostresp_devx21[1] 0.64 1.00 4919 2881
## depmistrt_mostresp_devx21[2] 0.55 1.00 5568 2752
## depmistrt_mostresp_devx21[3] 0.56 1.00 5676 3183
## depmistrt_mostresp_devx21[4] 0.59 1.00 7064 2618
## depmistrt_mostresp_av12x21[1] 0.23 1.00 6992 2090
## depmistrt_mostresp_av12x21[2] 0.30 1.00 6673 2328
## depmistrt_mostresp_av12x21[3] 0.33 1.00 7378 2461
## depmistrt_mostresp_av12x21[4] 0.35 1.00 6790 2113
## depmistrt_mostresp_av12x21[5] 0.29 1.00 6394 2292
## depmistrt_mostresp_av12x21[6] 0.33 1.00 8571 2480
## depmistrt_mostresp_av12x21[7] 0.32 1.00 7276 2676
## depmistrt_mostresp_av12x21[8] 0.31 1.00 7053 2210
## depmistrt_mostresp_devx2:rural.ses.med21[1] 0.68 1.00 5081 2413
## depmistrt_mostresp_devx2:rural.ses.med21[2] 0.58 1.00 3723 3347
## depmistrt_mostresp_devx2:rural.ses.med21[3] 0.60 1.00 2939 2876
## depmistrt_mostresp_devx2:rural.ses.med21[4] 0.88 1.00 2570 2995
## depmistrt_mostresp_devx2:rural.ses.med31[1] 0.68 1.00 5691 2679
## depmistrt_mostresp_devx2:rural.ses.med31[2] 0.66 1.00 5514 2548
## depmistrt_mostresp_devx2:rural.ses.med31[3] 0.63 1.00 5815 2788
## depmistrt_mostresp_devx2:rural.ses.med31[4] 0.72 1.00 5949 3121
## depmistrt_mostresp_devx2:rural.ses.med41[1] 0.64 1.00 5856 2164
## depmistrt_mostresp_devx2:rural.ses.med41[2] 0.79 1.00 3666 2915
## depmistrt_mostresp_devx2:rural.ses.med41[3] 0.62 1.00 5752 2899
## depmistrt_mostresp_devx2:rural.ses.med41[4] 0.67 1.00 6307 2564
## depbetray_mostresp_devx21[1] 0.58 1.00 5784 2780
## depbetray_mostresp_devx21[2] 0.55 1.00 8502 2758
## depbetray_mostresp_devx21[3] 0.55 1.00 6983 2800
## depbetray_mostresp_devx21[4] 0.58 1.00 8191 2632
## depbetray_mostresp_av12x21[1] 0.27 1.00 6674 2195
## depbetray_mostresp_av12x21[2] 0.31 1.00 8197 2525
## depbetray_mostresp_av12x21[3] 0.34 1.00 7109 2371
## depbetray_mostresp_av12x21[4] 0.34 1.00 7891 2164
## depbetray_mostresp_av12x21[5] 0.32 1.00 7993 2528
## depbetray_mostresp_av12x21[6] 0.34 1.00 6892 2652
## depbetray_mostresp_av12x21[7] 0.32 1.00 6540 2905
## depbetray_mostresp_av12x21[8] 0.27 1.00 6389 2511
## depbetray_mostresp_devx2:rural.ses.med21[1] 0.72 1.00 4828 2524
## depbetray_mostresp_devx2:rural.ses.med21[2] 0.67 1.00 5930 2618
## depbetray_mostresp_devx2:rural.ses.med21[3] 0.65 1.00 4833 2827
## depbetray_mostresp_devx2:rural.ses.med21[4] 0.77 1.00 4390 2807
## depbetray_mostresp_devx2:rural.ses.med31[1] 0.73 1.00 6188 2637
## depbetray_mostresp_devx2:rural.ses.med31[2] 0.65 1.00 5113 2654
## depbetray_mostresp_devx2:rural.ses.med31[3] 0.69 1.00 4366 2580
## depbetray_mostresp_devx2:rural.ses.med31[4] 0.73 1.00 5473 2554
## depbetray_mostresp_devx2:rural.ses.med41[1] 0.73 1.00 6196 2324
## depbetray_mostresp_devx2:rural.ses.med41[2] 0.66 1.00 6257 2818
## depbetray_mostresp_devx2:rural.ses.med41[3] 0.67 1.00 5351 2814
## depbetray_mostresp_devx2:rural.ses.med41[4] 0.73 1.00 5893 2416
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.2.3.4 Prior summary
out.chg.alldepress.stresp.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostresp_av12x2
## normal(0, 0.25) b mostresp_devx2
## normal(0, 1) b mostresp_devx2:rural.ses.med2
## normal(0, 1) b mostresp_devx2:rural.ses.med3
## normal(0, 1) b mostresp_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostresp_av12x21
## dirichlet(1) simo mostresp_devx2:rural.ses.med21
## dirichlet(1) simo mostresp_devx2:rural.ses.med31
## dirichlet(1) simo mostresp_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostresp_devx21
## group resp dpar nlpar lb ub source
## default
## depbetray user
## depbetray user
## depbetray user
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depblues user
## depblues user
## depblues user
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depcantgo user
## depcantgo user
## depcantgo user
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depeffort user
## depeffort user
## depeffort user
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## deplonely user
## deplonely user
## deplonely user
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## depmistrt user
## depmistrt user
## depmistrt user
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depunfair user
## depunfair user
## depunfair user
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## default
## depbetray user
## depblues user
## depcantgo user
## depeffort user
## deplonely user
## depmistrt user
## depunfair user
## depbetray 0 default
## depblues 0 default
## depcantgo 0 default
## depeffort 0 default
## deplonely 0 default
## depmistrt 0 default
## depunfair 0 default
## id depbetray 0 (vectorized)
## id depbetray 0 (vectorized)
## id depblues 0 (vectorized)
## id depblues 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depeffort 0 (vectorized)
## id depeffort 0 (vectorized)
## id deplonely 0 (vectorized)
## id deplonely 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depunfair 0 (vectorized)
## id depunfair 0 (vectorized)
## depbetray user
## depbetray default
## depbetray default
## depbetray default
## depbetray user
## depblues user
## depblues default
## depblues default
## depblues default
## depblues user
## depcantgo user
## depcantgo default
## depcantgo default
## depcantgo default
## depcantgo user
## depeffort user
## depeffort default
## depeffort default
## depeffort default
## depeffort user
## deplonely user
## deplonely default
## deplonely default
## deplonely default
## deplonely user
## depmistrt user
## depmistrt default
## depmistrt default
## depmistrt default
## depmistrt user
## depunfair user
## depunfair default
## depunfair default
## depunfair default
## depunfair user8.1.2.4 Corr X Community: stfair & negative emotions
# Community Change: negative emotions items ~ mo(stfair)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostfair_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostfair_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostfair_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostfair_av12x21',
resp = depdv_names)
)
chg.alldepress.stfair.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) +
rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stfair_comm_fit",
file_refit = "on_change"
)
out.chg.alldepress.stfair.comm.fit <- ppchecks(chg.alldepress.stfair.comm.fit)8.1.2.4.1 Coefficient plot (intervals)
out.chg.alldepress.stfair.comm.fit[[11]]
8.1.2.4.2 PPcheck (density)
p1 <- out.chg.alldepress.stfair.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stfair.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stfair.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stfair.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stfair.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stfair.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stfair.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
8.1.2.4.3 Fit summary
out.chg.alldepress.stfair.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## depeffort ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## deplonely ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## depblues ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## depunfair ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## depmistrt ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## depbetray ~ 1 + mo(stfair_devx2) + mo(stfair_av12x2) + rural.ses.med + mo(stfair_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.32 0.20 0.01 0.72 1.00 565
## sd(depeffort_Intercept) 0.42 0.27 0.02 0.97 1.01 614
## sd(deplonely_Intercept) 0.42 0.24 0.02 0.88 1.01 530
## sd(depblues_Intercept) 0.71 0.31 0.06 1.29 1.01 558
## sd(depunfair_Intercept) 0.23 0.17 0.01 0.61 1.00 1034
## sd(depmistrt_Intercept) 0.30 0.21 0.01 0.75 1.00 795
## sd(depbetray_Intercept) 0.42 0.27 0.02 0.97 1.00 519
## Tail_ESS
## sd(depcantgo_Intercept) 1318
## sd(depeffort_Intercept) 1129
## sd(deplonely_Intercept) 818
## sd(depblues_Intercept) 672
## sd(depunfair_Intercept) 1859
## sd(depmistrt_Intercept) 1416
## sd(depbetray_Intercept) 1169
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_Intercept -0.77 0.37 -1.57 -0.11
## depeffort_Intercept -2.45 0.45 -3.36 -1.61
## deplonely_Intercept -1.20 0.35 -1.95 -0.55
## depblues_Intercept -2.38 0.44 -3.31 -1.54
## depunfair_Intercept -2.15 0.37 -2.93 -1.47
## depmistrt_Intercept -2.74 0.42 -3.55 -1.90
## depbetray_Intercept -3.09 0.47 -4.06 -2.22
## depcantgo_rural.ses.med2 -0.53 0.56 -1.78 0.43
## depcantgo_rural.ses.med3 0.23 0.47 -0.75 1.16
## depcantgo_rural.ses.med4 0.44 0.49 -0.47 1.50
## depeffort_rural.ses.med2 -0.27 0.57 -1.45 0.85
## depeffort_rural.ses.med3 -0.20 0.66 -1.57 1.01
## depeffort_rural.ses.med4 0.58 0.59 -0.64 1.74
## deplonely_rural.ses.med2 -0.68 0.52 -1.75 0.39
## deplonely_rural.ses.med3 0.04 0.49 -1.06 0.93
## deplonely_rural.ses.med4 0.42 0.52 -0.65 1.44
## depblues_rural.ses.med2 0.19 0.60 -1.10 1.29
## depblues_rural.ses.med3 0.02 0.57 -1.21 1.05
## depblues_rural.ses.med4 0.56 0.58 -0.62 1.75
## depunfair_rural.ses.med2 -0.26 0.60 -1.45 0.91
## depunfair_rural.ses.med3 0.73 0.50 -0.38 1.66
## depunfair_rural.ses.med4 0.75 0.53 -0.37 1.81
## depmistrt_rural.ses.med2 -0.01 0.63 -1.36 1.07
## depmistrt_rural.ses.med3 0.02 0.57 -1.22 1.06
## depmistrt_rural.ses.med4 0.49 0.57 -0.77 1.55
## depbetray_rural.ses.med2 0.19 0.60 -1.14 1.28
## depbetray_rural.ses.med3 0.46 0.65 -0.95 1.59
## depbetray_rural.ses.med4 0.77 0.63 -0.74 1.84
## depcantgo_mostfair_devx2 0.16 0.15 -0.11 0.48
## depcantgo_mostfair_av12x2 -0.01 0.02 -0.06 0.04
## depcantgo_mostfair_devx2:rural.ses.med2 0.48 0.35 -0.04 1.34
## depcantgo_mostfair_devx2:rural.ses.med3 -0.06 0.22 -0.48 0.42
## depcantgo_mostfair_devx2:rural.ses.med4 -0.19 0.24 -0.69 0.30
## depeffort_mostfair_devx2 0.14 0.17 -0.20 0.47
## depeffort_mostfair_av12x2 0.03 0.03 -0.03 0.10
## depeffort_mostfair_devx2:rural.ses.med2 0.18 0.30 -0.39 0.85
## depeffort_mostfair_devx2:rural.ses.med3 0.33 0.31 -0.27 0.95
## depeffort_mostfair_devx2:rural.ses.med4 -0.07 0.32 -0.78 0.50
## deplonely_mostfair_devx2 0.10 0.15 -0.19 0.41
## deplonely_mostfair_av12x2 0.00 0.03 -0.06 0.05
## deplonely_mostfair_devx2:rural.ses.med2 0.15 0.26 -0.41 0.62
## deplonely_mostfair_devx2:rural.ses.med3 -0.05 0.23 -0.45 0.50
## deplonely_mostfair_devx2:rural.ses.med4 0.01 0.26 -0.50 0.57
## depblues_mostfair_devx2 -0.02 0.17 -0.39 0.30
## depblues_mostfair_av12x2 0.02 0.04 -0.05 0.09
## depblues_mostfair_devx2:rural.ses.med2 -0.01 0.34 -0.75 0.59
## depblues_mostfair_devx2:rural.ses.med3 0.15 0.27 -0.42 0.65
## depblues_mostfair_devx2:rural.ses.med4 0.01 0.32 -0.66 0.58
## depunfair_mostfair_devx2 0.21 0.14 -0.08 0.51
## depunfair_mostfair_av12x2 0.03 0.03 -0.03 0.09
## depunfair_mostfair_devx2:rural.ses.med2 0.29 0.28 -0.28 0.86
## depunfair_mostfair_devx2:rural.ses.med3 -0.01 0.24 -0.45 0.52
## depunfair_mostfair_devx2:rural.ses.med4 0.15 0.27 -0.37 0.70
## depmistrt_mostfair_devx2 0.01 0.16 -0.33 0.31
## depmistrt_mostfair_av12x2 0.13 0.03 0.06 0.19
## depmistrt_mostfair_devx2:rural.ses.med2 0.31 0.31 -0.24 0.97
## depmistrt_mostfair_devx2:rural.ses.med3 0.27 0.25 -0.28 0.73
## depmistrt_mostfair_devx2:rural.ses.med4 0.06 0.29 -0.54 0.60
## depbetray_mostfair_devx2 0.11 0.17 -0.22 0.45
## depbetray_mostfair_av12x2 0.11 0.04 0.05 0.19
## depbetray_mostfair_devx2:rural.ses.med2 0.17 0.31 -0.40 0.85
## depbetray_mostfair_devx2:rural.ses.med3 0.10 0.32 -0.58 0.64
## depbetray_mostfair_devx2:rural.ses.med4 0.12 0.31 -0.41 0.79
## Rhat Bulk_ESS Tail_ESS
## depcantgo_Intercept 1.00 2543 2338
## depeffort_Intercept 1.00 2537 3253
## deplonely_Intercept 1.00 3096 2901
## depblues_Intercept 1.00 2428 2837
## depunfair_Intercept 1.00 3427 2916
## depmistrt_Intercept 1.00 3280 3057
## depbetray_Intercept 1.00 2681 3107
## depcantgo_rural.ses.med2 1.00 3286 2650
## depcantgo_rural.ses.med3 1.00 2186 2138
## depcantgo_rural.ses.med4 1.00 2978 2701
## depeffort_rural.ses.med2 1.00 3205 2777
## depeffort_rural.ses.med3 1.00 2759 2914
## depeffort_rural.ses.med4 1.00 3303 2794
## deplonely_rural.ses.med2 1.00 2926 2495
## deplonely_rural.ses.med3 1.00 2733 2053
## deplonely_rural.ses.med4 1.00 3078 2735
## depblues_rural.ses.med2 1.00 3842 3069
## depblues_rural.ses.med3 1.00 3024 3000
## depblues_rural.ses.med4 1.00 3656 2786
## depunfair_rural.ses.med2 1.00 2647 2128
## depunfair_rural.ses.med3 1.00 2983 2583
## depunfair_rural.ses.med4 1.00 3019 2682
## depmistrt_rural.ses.med2 1.00 3354 2837
## depmistrt_rural.ses.med3 1.00 3038 2405
## depmistrt_rural.ses.med4 1.00 3180 2520
## depbetray_rural.ses.med2 1.00 3015 2670
## depbetray_rural.ses.med3 1.00 2292 3089
## depbetray_rural.ses.med4 1.00 2955 2588
## depcantgo_mostfair_devx2 1.00 2626 2729
## depcantgo_mostfair_av12x2 1.00 7142 3351
## depcantgo_mostfair_devx2:rural.ses.med2 1.00 2616 2103
## depcantgo_mostfair_devx2:rural.ses.med3 1.00 2001 2569
## depcantgo_mostfair_devx2:rural.ses.med4 1.00 2928 2417
## depeffort_mostfair_devx2 1.00 2802 3333
## depeffort_mostfair_av12x2 1.00 6284 3100
## depeffort_mostfair_devx2:rural.ses.med2 1.00 3287 2809
## depeffort_mostfair_devx2:rural.ses.med3 1.00 2473 3002
## depeffort_mostfair_devx2:rural.ses.med4 1.00 2753 2266
## deplonely_mostfair_devx2 1.00 3010 3056
## deplonely_mostfair_av12x2 1.00 7150 2980
## deplonely_mostfair_devx2:rural.ses.med2 1.00 2910 1931
## deplonely_mostfair_devx2:rural.ses.med3 1.00 2771 2328
## deplonely_mostfair_devx2:rural.ses.med4 1.00 2936 2636
## depblues_mostfair_devx2 1.00 3007 3070
## depblues_mostfair_av12x2 1.00 4909 2826
## depblues_mostfair_devx2:rural.ses.med2 1.00 3364 2691
## depblues_mostfair_devx2:rural.ses.med3 1.00 2599 1825
## depblues_mostfair_devx2:rural.ses.med4 1.00 2919 2469
## depunfair_mostfair_devx2 1.00 2903 3167
## depunfair_mostfair_av12x2 1.00 6352 2809
## depunfair_mostfair_devx2:rural.ses.med2 1.00 2350 2005
## depunfair_mostfair_devx2:rural.ses.med3 1.00 2784 2901
## depunfair_mostfair_devx2:rural.ses.med4 1.00 2788 2787
## depmistrt_mostfair_devx2 1.00 2930 3114
## depmistrt_mostfair_av12x2 1.00 6472 3089
## depmistrt_mostfair_devx2:rural.ses.med2 1.00 2867 2705
## depmistrt_mostfair_devx2:rural.ses.med3 1.00 2664 1790
## depmistrt_mostfair_devx2:rural.ses.med4 1.00 2752 2273
## depbetray_mostfair_devx2 1.00 2500 3294
## depbetray_mostfair_av12x2 1.00 6023 3094
## depbetray_mostfair_devx2:rural.ses.med2 1.00 2736 2175
## depbetray_mostfair_devx2:rural.ses.med3 1.00 2018 2432
## depbetray_mostfair_devx2:rural.ses.med4 1.00 2695 2970
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## depcantgo_mostfair_devx21[1] 0.26 0.15 0.04
## depcantgo_mostfair_devx21[2] 0.25 0.13 0.04
## depcantgo_mostfair_devx21[3] 0.21 0.13 0.03
## depcantgo_mostfair_devx21[4] 0.28 0.15 0.04
## depcantgo_mostfair_av12x21[1] 0.13 0.08 0.02
## depcantgo_mostfair_av12x21[2] 0.13 0.08 0.02
## depcantgo_mostfair_av12x21[3] 0.12 0.08 0.02
## depcantgo_mostfair_av12x21[4] 0.12 0.08 0.02
## depcantgo_mostfair_av12x21[5] 0.13 0.08 0.02
## depcantgo_mostfair_av12x21[6] 0.12 0.08 0.02
## depcantgo_mostfair_av12x21[7] 0.13 0.08 0.02
## depcantgo_mostfair_av12x21[8] 0.13 0.08 0.02
## depcantgo_mostfair_devx2:rural.ses.med21[1] 0.22 0.17 0.01
## depcantgo_mostfair_devx2:rural.ses.med21[2] 0.29 0.18 0.02
## depcantgo_mostfair_devx2:rural.ses.med21[3] 0.14 0.13 0.00
## depcantgo_mostfair_devx2:rural.ses.med21[4] 0.35 0.22 0.01
## depcantgo_mostfair_devx2:rural.ses.med31[1] 0.26 0.20 0.01
## depcantgo_mostfair_devx2:rural.ses.med31[2] 0.26 0.20 0.01
## depcantgo_mostfair_devx2:rural.ses.med31[3] 0.20 0.17 0.01
## depcantgo_mostfair_devx2:rural.ses.med31[4] 0.28 0.20 0.01
## depcantgo_mostfair_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depcantgo_mostfair_devx2:rural.ses.med41[2] 0.22 0.18 0.01
## depcantgo_mostfair_devx2:rural.ses.med41[3] 0.25 0.18 0.01
## depcantgo_mostfair_devx2:rural.ses.med41[4] 0.27 0.20 0.01
## depeffort_mostfair_devx21[1] 0.24 0.14 0.03
## depeffort_mostfair_devx21[2] 0.30 0.16 0.05
## depeffort_mostfair_devx21[3] 0.20 0.13 0.03
## depeffort_mostfair_devx21[4] 0.26 0.15 0.04
## depeffort_mostfair_av12x21[1] 0.13 0.08 0.02
## depeffort_mostfair_av12x21[2] 0.13 0.08 0.02
## depeffort_mostfair_av12x21[3] 0.13 0.08 0.02
## depeffort_mostfair_av12x21[4] 0.12 0.08 0.02
## depeffort_mostfair_av12x21[5] 0.12 0.07 0.02
## depeffort_mostfair_av12x21[6] 0.12 0.08 0.02
## depeffort_mostfair_av12x21[7] 0.12 0.08 0.02
## depeffort_mostfair_av12x21[8] 0.13 0.09 0.02
## depeffort_mostfair_devx2:rural.ses.med21[1] 0.24 0.19 0.01
## depeffort_mostfair_devx2:rural.ses.med21[2] 0.21 0.17 0.01
## depeffort_mostfair_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## depeffort_mostfair_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## depeffort_mostfair_devx2:rural.ses.med31[1] 0.24 0.18 0.01
## depeffort_mostfair_devx2:rural.ses.med31[2] 0.32 0.20 0.01
## depeffort_mostfair_devx2:rural.ses.med31[3] 0.14 0.14 0.00
## depeffort_mostfair_devx2:rural.ses.med31[4] 0.31 0.20 0.01
## depeffort_mostfair_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depeffort_mostfair_devx2:rural.ses.med41[2] 0.22 0.19 0.01
## depeffort_mostfair_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## depeffort_mostfair_devx2:rural.ses.med41[4] 0.30 0.22 0.01
## deplonely_mostfair_devx21[1] 0.27 0.15 0.05
## deplonely_mostfair_devx21[2] 0.22 0.13 0.03
## deplonely_mostfair_devx21[3] 0.24 0.14 0.03
## deplonely_mostfair_devx21[4] 0.27 0.15 0.04
## deplonely_mostfair_av12x21[1] 0.13 0.08 0.02
## deplonely_mostfair_av12x21[2] 0.13 0.08 0.02
## deplonely_mostfair_av12x21[3] 0.12 0.08 0.01
## deplonely_mostfair_av12x21[4] 0.12 0.08 0.02
## deplonely_mostfair_av12x21[5] 0.12 0.08 0.01
## deplonely_mostfair_av12x21[6] 0.12 0.08 0.01
## deplonely_mostfair_av12x21[7] 0.13 0.08 0.02
## deplonely_mostfair_av12x21[8] 0.13 0.08 0.02
## deplonely_mostfair_devx2:rural.ses.med21[1] 0.25 0.18 0.01
## deplonely_mostfair_devx2:rural.ses.med21[2] 0.23 0.18 0.01
## deplonely_mostfair_devx2:rural.ses.med21[3] 0.26 0.18 0.01
## deplonely_mostfair_devx2:rural.ses.med21[4] 0.27 0.20 0.01
## deplonely_mostfair_devx2:rural.ses.med31[1] 0.26 0.20 0.01
## deplonely_mostfair_devx2:rural.ses.med31[2] 0.24 0.19 0.01
## deplonely_mostfair_devx2:rural.ses.med31[3] 0.22 0.18 0.01
## deplonely_mostfair_devx2:rural.ses.med31[4] 0.28 0.20 0.01
## deplonely_mostfair_devx2:rural.ses.med41[1] 0.27 0.19 0.01
## deplonely_mostfair_devx2:rural.ses.med41[2] 0.22 0.18 0.01
## deplonely_mostfair_devx2:rural.ses.med41[3] 0.22 0.17 0.01
## deplonely_mostfair_devx2:rural.ses.med41[4] 0.29 0.21 0.01
## depblues_mostfair_devx21[1] 0.26 0.15 0.04
## depblues_mostfair_devx21[2] 0.24 0.14 0.03
## depblues_mostfair_devx21[3] 0.24 0.14 0.04
## depblues_mostfair_devx21[4] 0.27 0.15 0.04
## depblues_mostfair_av12x21[1] 0.12 0.08 0.02
## depblues_mostfair_av12x21[2] 0.12 0.08 0.02
## depblues_mostfair_av12x21[3] 0.12 0.08 0.02
## depblues_mostfair_av12x21[4] 0.12 0.08 0.02
## depblues_mostfair_av12x21[5] 0.12 0.08 0.02
## depblues_mostfair_av12x21[6] 0.13 0.08 0.02
## depblues_mostfair_av12x21[7] 0.13 0.08 0.02
## depblues_mostfair_av12x21[8] 0.14 0.09 0.02
## depblues_mostfair_devx2:rural.ses.med21[1] 0.25 0.19 0.01
## depblues_mostfair_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## depblues_mostfair_devx2:rural.ses.med21[3] 0.23 0.18 0.01
## depblues_mostfair_devx2:rural.ses.med21[4] 0.30 0.22 0.01
## depblues_mostfair_devx2:rural.ses.med31[1] 0.26 0.20 0.01
## depblues_mostfair_devx2:rural.ses.med31[2] 0.24 0.18 0.01
## depblues_mostfair_devx2:rural.ses.med31[3] 0.23 0.18 0.01
## depblues_mostfair_devx2:rural.ses.med31[4] 0.27 0.20 0.01
## depblues_mostfair_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depblues_mostfair_devx2:rural.ses.med41[2] 0.22 0.19 0.01
## depblues_mostfair_devx2:rural.ses.med41[3] 0.23 0.18 0.01
## depblues_mostfair_devx2:rural.ses.med41[4] 0.30 0.22 0.01
## depunfair_mostfair_devx21[1] 0.21 0.12 0.03
## depunfair_mostfair_devx21[2] 0.30 0.14 0.05
## depunfair_mostfair_devx21[3] 0.24 0.12 0.04
## depunfair_mostfair_devx21[4] 0.25 0.13 0.04
## depunfair_mostfair_av12x21[1] 0.12 0.08 0.02
## depunfair_mostfair_av12x21[2] 0.12 0.08 0.02
## depunfair_mostfair_av12x21[3] 0.12 0.08 0.02
## depunfair_mostfair_av12x21[4] 0.12 0.08 0.02
## depunfair_mostfair_av12x21[5] 0.12 0.08 0.02
## depunfair_mostfair_av12x21[6] 0.12 0.08 0.02
## depunfair_mostfair_av12x21[7] 0.14 0.08 0.02
## depunfair_mostfair_av12x21[8] 0.13 0.08 0.02
## depunfair_mostfair_devx2:rural.ses.med21[1] 0.20 0.17 0.01
## depunfair_mostfair_devx2:rural.ses.med21[2] 0.35 0.21 0.01
## depunfair_mostfair_devx2:rural.ses.med21[3] 0.18 0.15 0.01
## depunfair_mostfair_devx2:rural.ses.med21[4] 0.28 0.20 0.01
## depunfair_mostfair_devx2:rural.ses.med31[1] 0.26 0.19 0.01
## depunfair_mostfair_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## depunfair_mostfair_devx2:rural.ses.med31[3] 0.24 0.19 0.01
## depunfair_mostfair_devx2:rural.ses.med31[4] 0.28 0.20 0.01
## depunfair_mostfair_devx2:rural.ses.med41[1] 0.26 0.19 0.01
## depunfair_mostfair_devx2:rural.ses.med41[2] 0.20 0.17 0.01
## depunfair_mostfair_devx2:rural.ses.med41[3] 0.26 0.19 0.01
## depunfair_mostfair_devx2:rural.ses.med41[4] 0.28 0.20 0.01
## depmistrt_mostfair_devx21[1] 0.27 0.15 0.04
## depmistrt_mostfair_devx21[2] 0.23 0.13 0.03
## depmistrt_mostfair_devx21[3] 0.24 0.14 0.03
## depmistrt_mostfair_devx21[4] 0.26 0.15 0.04
## depmistrt_mostfair_av12x21[1] 0.10 0.06 0.01
## depmistrt_mostfair_av12x21[2] 0.12 0.08 0.02
## depmistrt_mostfair_av12x21[3] 0.13 0.08 0.02
## depmistrt_mostfair_av12x21[4] 0.15 0.09 0.02
## depmistrt_mostfair_av12x21[5] 0.13 0.08 0.02
## depmistrt_mostfair_av12x21[6] 0.11 0.07 0.02
## depmistrt_mostfair_av12x21[7] 0.17 0.09 0.03
## depmistrt_mostfair_av12x21[8] 0.11 0.07 0.02
## depmistrt_mostfair_devx2:rural.ses.med21[1] 0.24 0.19 0.01
## depmistrt_mostfair_devx2:rural.ses.med21[2] 0.27 0.19 0.01
## depmistrt_mostfair_devx2:rural.ses.med21[3] 0.17 0.15 0.00
## depmistrt_mostfair_devx2:rural.ses.med21[4] 0.32 0.21 0.01
## depmistrt_mostfair_devx2:rural.ses.med31[1] 0.27 0.19 0.01
## depmistrt_mostfair_devx2:rural.ses.med31[2] 0.17 0.15 0.00
## depmistrt_mostfair_devx2:rural.ses.med31[3] 0.34 0.21 0.02
## depmistrt_mostfair_devx2:rural.ses.med31[4] 0.22 0.18 0.01
## depmistrt_mostfair_devx2:rural.ses.med41[1] 0.27 0.20 0.01
## depmistrt_mostfair_devx2:rural.ses.med41[2] 0.22 0.18 0.01
## depmistrt_mostfair_devx2:rural.ses.med41[3] 0.22 0.17 0.01
## depmistrt_mostfair_devx2:rural.ses.med41[4] 0.29 0.22 0.01
## depbetray_mostfair_devx21[1] 0.26 0.15 0.04
## depbetray_mostfair_devx21[2] 0.27 0.14 0.04
## depbetray_mostfair_devx21[3] 0.21 0.13 0.03
## depbetray_mostfair_devx21[4] 0.25 0.14 0.04
## depbetray_mostfair_av12x21[1] 0.12 0.07 0.02
## depbetray_mostfair_av12x21[2] 0.14 0.08 0.02
## depbetray_mostfair_av12x21[3] 0.14 0.09 0.02
## depbetray_mostfair_av12x21[4] 0.15 0.09 0.02
## depbetray_mostfair_av12x21[5] 0.10 0.07 0.01
## depbetray_mostfair_av12x21[6] 0.10 0.06 0.01
## depbetray_mostfair_av12x21[7] 0.12 0.08 0.02
## depbetray_mostfair_av12x21[8] 0.13 0.08 0.02
## depbetray_mostfair_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## depbetray_mostfair_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## depbetray_mostfair_devx2:rural.ses.med21[3] 0.20 0.18 0.01
## depbetray_mostfair_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## depbetray_mostfair_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## depbetray_mostfair_devx2:rural.ses.med31[2] 0.27 0.21 0.01
## depbetray_mostfair_devx2:rural.ses.med31[3] 0.19 0.17 0.01
## depbetray_mostfair_devx2:rural.ses.med31[4] 0.27 0.21 0.01
## depbetray_mostfair_devx2:rural.ses.med41[1] 0.28 0.20 0.01
## depbetray_mostfair_devx2:rural.ses.med41[2] 0.21 0.17 0.01
## depbetray_mostfair_devx2:rural.ses.med41[3] 0.21 0.18 0.01
## depbetray_mostfair_devx2:rural.ses.med41[4] 0.30 0.21 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## depcantgo_mostfair_devx21[1] 0.59 1.00 5746 2882
## depcantgo_mostfair_devx21[2] 0.55 1.00 5569 2820
## depcantgo_mostfair_devx21[3] 0.51 1.00 5552 3232
## depcantgo_mostfair_devx21[4] 0.61 1.00 7078 2774
## depcantgo_mostfair_av12x21[1] 0.32 1.00 7154 2377
## depcantgo_mostfair_av12x21[2] 0.32 1.00 7627 2116
## depcantgo_mostfair_av12x21[3] 0.32 1.00 7633 2091
## depcantgo_mostfair_av12x21[4] 0.32 1.00 7209 2564
## depcantgo_mostfair_av12x21[5] 0.32 1.00 8755 2486
## depcantgo_mostfair_av12x21[6] 0.30 1.00 7666 2904
## depcantgo_mostfair_av12x21[7] 0.32 1.00 7325 2399
## depcantgo_mostfair_av12x21[8] 0.33 1.00 8387 3193
## depcantgo_mostfair_devx2:rural.ses.med21[1] 0.62 1.00 5611 2796
## depcantgo_mostfair_devx2:rural.ses.med21[2] 0.71 1.00 4570 2135
## depcantgo_mostfair_devx2:rural.ses.med21[3] 0.49 1.00 4534 2686
## depcantgo_mostfair_devx2:rural.ses.med21[4] 0.78 1.00 4225 3090
## depcantgo_mostfair_devx2:rural.ses.med31[1] 0.72 1.00 4773 2541
## depcantgo_mostfair_devx2:rural.ses.med31[2] 0.71 1.00 4361 2609
## depcantgo_mostfair_devx2:rural.ses.med31[3] 0.62 1.00 4713 3003
## depcantgo_mostfair_devx2:rural.ses.med31[4] 0.74 1.00 5163 2594
## depcantgo_mostfair_devx2:rural.ses.med41[1] 0.71 1.00 5488 2365
## depcantgo_mostfair_devx2:rural.ses.med41[2] 0.64 1.00 4995 2442
## depcantgo_mostfair_devx2:rural.ses.med41[3] 0.68 1.00 5099 2810
## depcantgo_mostfair_devx2:rural.ses.med41[4] 0.73 1.00 5899 2461
## depeffort_mostfair_devx21[1] 0.58 1.00 6710 2613
## depeffort_mostfair_devx21[2] 0.63 1.00 5055 2619
## depeffort_mostfair_devx21[3] 0.50 1.00 5510 2804
## depeffort_mostfair_devx21[4] 0.59 1.00 8336 2857
## depeffort_mostfair_av12x21[1] 0.32 1.00 6422 2520
## depeffort_mostfair_av12x21[2] 0.34 1.00 7248 2161
## depeffort_mostfair_av12x21[3] 0.33 1.00 7624 2764
## depeffort_mostfair_av12x21[4] 0.32 1.00 7569 2307
## depeffort_mostfair_av12x21[5] 0.29 1.00 7051 2891
## depeffort_mostfair_av12x21[6] 0.30 1.00 6964 2527
## depeffort_mostfair_av12x21[7] 0.31 1.00 6770 2937
## depeffort_mostfair_av12x21[8] 0.34 1.00 7214 2896
## depeffort_mostfair_devx2:rural.ses.med21[1] 0.68 1.00 5136 2561
## depeffort_mostfair_devx2:rural.ses.med21[2] 0.65 1.00 6379 2559
## depeffort_mostfair_devx2:rural.ses.med21[3] 0.67 1.00 4559 2614
## depeffort_mostfair_devx2:rural.ses.med21[4] 0.77 1.00 5757 2512
## depeffort_mostfair_devx2:rural.ses.med31[1] 0.68 1.00 5859 2771
## depeffort_mostfair_devx2:rural.ses.med31[2] 0.75 1.00 4294 2731
## depeffort_mostfair_devx2:rural.ses.med31[3] 0.55 1.00 3521 2942
## depeffort_mostfair_devx2:rural.ses.med31[4] 0.72 1.00 5700 3059
## depeffort_mostfair_devx2:rural.ses.med41[1] 0.70 1.00 5472 2576
## depeffort_mostfair_devx2:rural.ses.med41[2] 0.69 1.00 4531 2525
## depeffort_mostfair_devx2:rural.ses.med41[3] 0.66 1.00 4895 2770
## depeffort_mostfair_devx2:rural.ses.med41[4] 0.79 1.00 5373 3140
## deplonely_mostfair_devx21[1] 0.59 1.00 6248 2735
## deplonely_mostfair_devx21[2] 0.53 1.00 6281 2367
## deplonely_mostfair_devx21[3] 0.57 1.00 6046 2711
## deplonely_mostfair_devx21[4] 0.59 1.00 6581 2946
## deplonely_mostfair_av12x21[1] 0.33 1.00 7421 2484
## deplonely_mostfair_av12x21[2] 0.32 1.00 7650 2528
## deplonely_mostfair_av12x21[3] 0.32 1.00 7150 2232
## deplonely_mostfair_av12x21[4] 0.32 1.00 8775 2390
## deplonely_mostfair_av12x21[5] 0.32 1.00 7217 1947
## deplonely_mostfair_av12x21[6] 0.32 1.00 6564 2344
## deplonely_mostfair_av12x21[7] 0.32 1.00 6779 2944
## deplonely_mostfair_av12x21[8] 0.32 1.00 7996 3101
## deplonely_mostfair_devx2:rural.ses.med21[1] 0.66 1.00 5752 2746
## deplonely_mostfair_devx2:rural.ses.med21[2] 0.67 1.00 5904 2940
## deplonely_mostfair_devx2:rural.ses.med21[3] 0.69 1.00 4864 2831
## deplonely_mostfair_devx2:rural.ses.med21[4] 0.73 1.00 4261 3055
## deplonely_mostfair_devx2:rural.ses.med31[1] 0.72 1.00 5327 2318
## deplonely_mostfair_devx2:rural.ses.med31[2] 0.68 1.00 4358 2484
## deplonely_mostfair_devx2:rural.ses.med31[3] 0.66 1.00 4246 2616
## deplonely_mostfair_devx2:rural.ses.med31[4] 0.73 1.00 5267 2851
## deplonely_mostfair_devx2:rural.ses.med41[1] 0.71 1.00 6090 2472
## deplonely_mostfair_devx2:rural.ses.med41[2] 0.66 1.00 5144 2629
## deplonely_mostfair_devx2:rural.ses.med41[3] 0.62 1.00 4511 2251
## deplonely_mostfair_devx2:rural.ses.med41[4] 0.75 1.00 5760 2844
## depblues_mostfair_devx21[1] 0.59 1.00 6973 2678
## depblues_mostfair_devx21[2] 0.57 1.00 5934 2694
## depblues_mostfair_devx21[3] 0.55 1.00 7109 3013
## depblues_mostfair_devx21[4] 0.61 1.00 6907 2817
## depblues_mostfair_av12x21[1] 0.32 1.00 7024 2594
## depblues_mostfair_av12x21[2] 0.31 1.00 7236 2304
## depblues_mostfair_av12x21[3] 0.30 1.00 8494 2986
## depblues_mostfair_av12x21[4] 0.30 1.00 6039 2472
## depblues_mostfair_av12x21[5] 0.30 1.00 7908 2561
## depblues_mostfair_av12x21[6] 0.32 1.00 7167 2663
## depblues_mostfair_av12x21[7] 0.33 1.00 8168 2683
## depblues_mostfair_av12x21[8] 0.34 1.00 6157 2901
## depblues_mostfair_devx2:rural.ses.med21[1] 0.69 1.00 5801 2958
## depblues_mostfair_devx2:rural.ses.med21[2] 0.66 1.00 4867 2686
## depblues_mostfair_devx2:rural.ses.med21[3] 0.67 1.00 5419 3140
## depblues_mostfair_devx2:rural.ses.med21[4] 0.78 1.00 4891 2722
## depblues_mostfair_devx2:rural.ses.med31[1] 0.72 1.00 4849 2401
## depblues_mostfair_devx2:rural.ses.med31[2] 0.68 1.00 5155 2281
## depblues_mostfair_devx2:rural.ses.med31[3] 0.66 1.00 4494 2737
## depblues_mostfair_devx2:rural.ses.med31[4] 0.74 1.00 4111 2519
## depblues_mostfair_devx2:rural.ses.med41[1] 0.69 1.00 5870 2729
## depblues_mostfair_devx2:rural.ses.med41[2] 0.69 1.00 6134 2352
## depblues_mostfair_devx2:rural.ses.med41[3] 0.67 1.00 4171 2678
## depblues_mostfair_devx2:rural.ses.med41[4] 0.79 1.00 5218 2784
## depunfair_mostfair_devx21[1] 0.50 1.00 6565 2357
## depunfair_mostfair_devx21[2] 0.61 1.00 5154 2414
## depunfair_mostfair_devx21[3] 0.52 1.00 5985 2952
## depunfair_mostfair_devx21[4] 0.55 1.00 6631 2768
## depunfair_mostfair_av12x21[1] 0.30 1.00 8246 2297
## depunfair_mostfair_av12x21[2] 0.33 1.00 7189 2485
## depunfair_mostfair_av12x21[3] 0.32 1.00 6611 2359
## depunfair_mostfair_av12x21[4] 0.32 1.00 7153 2296
## depunfair_mostfair_av12x21[5] 0.31 1.00 5683 2056
## depunfair_mostfair_av12x21[6] 0.32 1.00 6900 2434
## depunfair_mostfair_av12x21[7] 0.34 1.00 7163 3053
## depunfair_mostfair_av12x21[8] 0.33 1.00 7725 2210
## depunfair_mostfair_devx2:rural.ses.med21[1] 0.65 1.00 4761 2532
## depunfair_mostfair_devx2:rural.ses.med21[2] 0.78 1.00 3956 2545
## depunfair_mostfair_devx2:rural.ses.med21[3] 0.58 1.00 4808 2900
## depunfair_mostfair_devx2:rural.ses.med21[4] 0.71 1.00 6029 3027
## depunfair_mostfair_devx2:rural.ses.med31[1] 0.69 1.00 4813 2544
## depunfair_mostfair_devx2:rural.ses.med31[2] 0.69 1.00 6605 2137
## depunfair_mostfair_devx2:rural.ses.med31[3] 0.68 1.00 4083 2760
## depunfair_mostfair_devx2:rural.ses.med31[4] 0.72 1.00 4982 2745
## depunfair_mostfair_devx2:rural.ses.med41[1] 0.71 1.00 5150 2570
## depunfair_mostfair_devx2:rural.ses.med41[2] 0.63 1.00 5333 2272
## depunfair_mostfair_devx2:rural.ses.med41[3] 0.69 1.00 4704 3113
## depunfair_mostfair_devx2:rural.ses.med41[4] 0.74 1.00 5432 2615
## depmistrt_mostfair_devx21[1] 0.61 1.00 8524 2811
## depmistrt_mostfair_devx21[2] 0.54 1.00 8106 2715
## depmistrt_mostfair_devx21[3] 0.56 1.00 5524 3037
## depmistrt_mostfair_devx21[4] 0.60 1.00 6416 2878
## depmistrt_mostfair_av12x21[1] 0.25 1.00 7773 2576
## depmistrt_mostfair_av12x21[2] 0.31 1.00 6590 2124
## depmistrt_mostfair_av12x21[3] 0.31 1.00 8067 2624
## depmistrt_mostfair_av12x21[4] 0.36 1.00 7249 2752
## depmistrt_mostfair_av12x21[5] 0.30 1.00 7291 2160
## depmistrt_mostfair_av12x21[6] 0.29 1.00 8257 2475
## depmistrt_mostfair_av12x21[7] 0.38 1.00 6827 3166
## depmistrt_mostfair_av12x21[8] 0.26 1.00 7591 3002
## depmistrt_mostfair_devx2:rural.ses.med21[1] 0.68 1.00 5420 2190
## depmistrt_mostfair_devx2:rural.ses.med21[2] 0.70 1.00 4133 2202
## depmistrt_mostfair_devx2:rural.ses.med21[3] 0.58 1.00 4397 2153
## depmistrt_mostfair_devx2:rural.ses.med21[4] 0.76 1.00 5310 2877
## depmistrt_mostfair_devx2:rural.ses.med31[1] 0.70 1.00 4725 2388
## depmistrt_mostfair_devx2:rural.ses.med31[2] 0.57 1.00 5911 2603
## depmistrt_mostfair_devx2:rural.ses.med31[3] 0.76 1.00 3418 2522
## depmistrt_mostfair_devx2:rural.ses.med31[4] 0.69 1.00 4643 2396
## depmistrt_mostfair_devx2:rural.ses.med41[1] 0.73 1.00 4745 2453
## depmistrt_mostfair_devx2:rural.ses.med41[2] 0.66 1.00 6867 2493
## depmistrt_mostfair_devx2:rural.ses.med41[3] 0.64 1.00 5856 2509
## depmistrt_mostfair_devx2:rural.ses.med41[4] 0.77 1.00 5123 2751
## depbetray_mostfair_devx21[1] 0.58 1.00 7536 2694
## depbetray_mostfair_devx21[2] 0.58 1.00 5017 2466
## depbetray_mostfair_devx21[3] 0.52 1.00 6159 3032
## depbetray_mostfair_devx21[4] 0.58 1.00 8438 2831
## depbetray_mostfair_av12x21[1] 0.30 1.00 7388 2761
## depbetray_mostfair_av12x21[2] 0.34 1.00 6397 2477
## depbetray_mostfair_av12x21[3] 0.34 1.00 6986 2737
## depbetray_mostfair_av12x21[4] 0.37 1.00 6869 2718
## depbetray_mostfair_av12x21[5] 0.27 1.00 7947 2781
## depbetray_mostfair_av12x21[6] 0.26 1.00 6903 2575
## depbetray_mostfair_av12x21[7] 0.31 1.00 6569 2872
## depbetray_mostfair_av12x21[8] 0.31 1.00 6214 2835
## depbetray_mostfair_devx2:rural.ses.med21[1] 0.70 1.00 5745 2732
## depbetray_mostfair_devx2:rural.ses.med21[2] 0.65 1.00 5789 2758
## depbetray_mostfair_devx2:rural.ses.med21[3] 0.65 1.00 4515 2941
## depbetray_mostfair_devx2:rural.ses.med21[4] 0.75 1.00 5310 2946
## depbetray_mostfair_devx2:rural.ses.med31[1] 0.72 1.00 4844 2796
## depbetray_mostfair_devx2:rural.ses.med31[2] 0.73 1.00 3476 3053
## depbetray_mostfair_devx2:rural.ses.med31[3] 0.63 1.00 4284 2916
## depbetray_mostfair_devx2:rural.ses.med31[4] 0.75 1.00 4038 2617
## depbetray_mostfair_devx2:rural.ses.med41[1] 0.74 1.00 4098 2522
## depbetray_mostfair_devx2:rural.ses.med41[2] 0.64 1.00 4436 2279
## depbetray_mostfair_devx2:rural.ses.med41[3] 0.67 1.00 4637 3152
## depbetray_mostfair_devx2:rural.ses.med41[4] 0.75 1.00 5832 3315
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.2.4.4 Prior summary
out.chg.alldepress.stfair.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostfair_av12x2
## normal(0, 0.25) b mostfair_devx2
## normal(0, 1) b mostfair_devx2:rural.ses.med2
## normal(0, 1) b mostfair_devx2:rural.ses.med3
## normal(0, 1) b mostfair_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostfair_av12x21
## dirichlet(1) simo mostfair_devx2:rural.ses.med21
## dirichlet(1) simo mostfair_devx2:rural.ses.med31
## dirichlet(1) simo mostfair_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostfair_devx21
## group resp dpar nlpar lb ub source
## default
## depbetray user
## depbetray user
## depbetray user
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depblues user
## depblues user
## depblues user
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depcantgo user
## depcantgo user
## depcantgo user
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depeffort user
## depeffort user
## depeffort user
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## deplonely user
## deplonely user
## deplonely user
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## depmistrt user
## depmistrt user
## depmistrt user
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depunfair user
## depunfair user
## depunfair user
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## default
## depbetray user
## depblues user
## depcantgo user
## depeffort user
## deplonely user
## depmistrt user
## depunfair user
## depbetray 0 default
## depblues 0 default
## depcantgo 0 default
## depeffort 0 default
## deplonely 0 default
## depmistrt 0 default
## depunfair 0 default
## id depbetray 0 (vectorized)
## id depbetray 0 (vectorized)
## id depblues 0 (vectorized)
## id depblues 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depeffort 0 (vectorized)
## id depeffort 0 (vectorized)
## id deplonely 0 (vectorized)
## id deplonely 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depunfair 0 (vectorized)
## id depunfair 0 (vectorized)
## depbetray user
## depbetray default
## depbetray default
## depbetray default
## depbetray user
## depblues user
## depblues default
## depblues default
## depblues default
## depblues user
## depcantgo user
## depcantgo default
## depcantgo default
## depcantgo default
## depcantgo user
## depeffort user
## depeffort default
## depeffort default
## depeffort default
## depeffort user
## deplonely user
## deplonely default
## deplonely default
## deplonely default
## deplonely user
## depmistrt user
## depmistrt default
## depmistrt default
## depmistrt default
## depmistrt user
## depunfair user
## depunfair default
## depunfair default
## depunfair default
## depunfair user8.1.2.5 Corr X Community: stjob & negative emotions
#Community Change: negative emotions items ~ mo(stjob)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostjob_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostjob_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostjob_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostjob_av12x21',
resp = depdv_names)
)
chg.alldepress.stjob.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) +
rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stjob_comm_fit",
file_refit = "on_change"
)
out.chg.alldepress.stjob.comm.fit <- ppchecks(chg.alldepress.stjob.comm.fit)8.1.2.5.1 Coefficient plot (intervals)
out.chg.alldepress.stjob.comm.fit[[11]]
8.1.2.5.2 PPcheck (density)
p1 <- out.chg.alldepress.stjob.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stjob.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stjob.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stjob.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stjob.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stjob.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stjob.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
8.1.2.5.3 Fit summary
out.chg.alldepress.stjob.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## depeffort ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## deplonely ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## depblues ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## depunfair ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## depmistrt ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## depbetray ~ 1 + mo(stjob_devx2) + mo(stjob_av12x2) + rural.ses.med + mo(stjob_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.30 0.20 0.01 0.72 1.00 637
## sd(depeffort_Intercept) 0.44 0.27 0.02 1.00 1.00 603
## sd(deplonely_Intercept) 0.41 0.24 0.03 0.87 1.00 536
## sd(depblues_Intercept) 0.70 0.33 0.06 1.31 1.01 521
## sd(depunfair_Intercept) 0.22 0.16 0.01 0.58 1.00 1106
## sd(depmistrt_Intercept) 0.34 0.23 0.01 0.82 1.00 681
## sd(depbetray_Intercept) 0.45 0.27 0.03 1.00 1.01 637
## Tail_ESS
## sd(depcantgo_Intercept) 1392
## sd(depeffort_Intercept) 1430
## sd(deplonely_Intercept) 1203
## sd(depblues_Intercept) 906
## sd(depunfair_Intercept) 2019
## sd(depmistrt_Intercept) 1750
## sd(depbetray_Intercept) 1479
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_Intercept -0.61 0.29 -1.20 -0.05
## depeffort_Intercept -2.26 0.40 -3.11 -1.50
## deplonely_Intercept -1.28 0.31 -1.92 -0.69
## depblues_Intercept -2.17 0.42 -3.01 -1.34
## depunfair_Intercept -2.08 0.37 -2.82 -1.37
## depmistrt_Intercept -2.31 0.44 -3.13 -1.33
## depbetray_Intercept -3.02 0.43 -3.86 -2.18
## depcantgo_rural.ses.med2 -0.12 0.54 -1.37 0.83
## depcantgo_rural.ses.med3 0.81 0.41 0.02 1.64
## depcantgo_rural.ses.med4 0.17 0.59 -0.81 1.60
## depeffort_rural.ses.med2 -0.15 0.56 -1.33 0.90
## depeffort_rural.ses.med3 0.30 0.59 -1.01 1.30
## depeffort_rural.ses.med4 0.24 0.53 -0.88 1.31
## deplonely_rural.ses.med2 -0.02 0.48 -0.94 1.01
## deplonely_rural.ses.med3 0.39 0.44 -0.44 1.30
## deplonely_rural.ses.med4 0.50 0.52 -0.50 1.59
## depblues_rural.ses.med2 0.06 0.55 -1.06 1.14
## depblues_rural.ses.med3 0.18 0.57 -1.12 1.18
## depblues_rural.ses.med4 0.35 0.68 -1.12 1.63
## depunfair_rural.ses.med2 -0.32 0.54 -1.53 0.62
## depunfair_rural.ses.med3 0.84 0.48 -0.15 1.74
## depunfair_rural.ses.med4 0.27 0.46 -0.73 1.04
## depmistrt_rural.ses.med2 0.32 0.57 -0.83 1.46
## depmistrt_rural.ses.med3 0.96 0.54 -0.04 2.11
## depmistrt_rural.ses.med4 0.62 0.55 -0.48 1.79
## depbetray_rural.ses.med2 -0.32 0.51 -1.40 0.61
## depbetray_rural.ses.med3 0.85 0.50 -0.19 1.78
## depbetray_rural.ses.med4 1.08 0.58 -0.19 2.15
## depcantgo_mostjob_devx2 0.23 0.13 -0.04 0.48
## depcantgo_mostjob_av12x2 -0.06 0.02 -0.11 -0.01
## depcantgo_mostjob_devx2:rural.ses.med2 0.22 0.27 -0.25 0.79
## depcantgo_mostjob_devx2:rural.ses.med3 -0.35 0.18 -0.70 0.02
## depcantgo_mostjob_devx2:rural.ses.med4 0.04 0.29 -0.57 0.59
## depeffort_mostjob_devx2 0.08 0.15 -0.23 0.37
## depeffort_mostjob_av12x2 0.02 0.03 -0.04 0.09
## depeffort_mostjob_devx2:rural.ses.med2 0.11 0.28 -0.43 0.72
## depeffort_mostjob_devx2:rural.ses.med3 0.03 0.31 -0.68 0.53
## depeffort_mostjob_devx2:rural.ses.med4 0.23 0.31 -0.31 0.92
## deplonely_mostjob_devx2 0.18 0.13 -0.08 0.44
## deplonely_mostjob_av12x2 -0.01 0.03 -0.06 0.05
## deplonely_mostjob_devx2:rural.ses.med2 -0.29 0.32 -1.06 0.23
## deplonely_mostjob_devx2:rural.ses.med3 -0.25 0.20 -0.67 0.14
## deplonely_mostjob_devx2:rural.ses.med4 -0.03 0.24 -0.50 0.47
## depblues_mostjob_devx2 -0.14 0.16 -0.45 0.17
## depblues_mostjob_av12x2 0.02 0.04 -0.05 0.10
## depblues_mostjob_devx2:rural.ses.med2 0.12 0.30 -0.53 0.65
## depblues_mostjob_devx2:rural.ses.med3 0.06 0.30 -0.61 0.60
## depblues_mostjob_devx2:rural.ses.med4 0.20 0.41 -0.52 1.10
## depunfair_mostjob_devx2 0.09 0.15 -0.21 0.38
## depunfair_mostjob_av12x2 0.07 0.03 0.02 0.12
## depunfair_mostjob_devx2:rural.ses.med2 0.37 0.23 -0.09 0.84
## depunfair_mostjob_devx2:rural.ses.med3 -0.08 0.22 -0.53 0.38
## depunfair_mostjob_devx2:rural.ses.med4 0.55 0.25 0.08 1.07
## depmistrt_mostjob_devx2 0.01 0.18 -0.40 0.32
## depmistrt_mostjob_av12x2 0.05 0.03 -0.01 0.11
## depmistrt_mostjob_devx2:rural.ses.med2 0.12 0.29 -0.50 0.61
## depmistrt_mostjob_devx2:rural.ses.med3 -0.28 0.35 -1.08 0.26
## depmistrt_mostjob_devx2:rural.ses.med4 0.09 0.28 -0.47 0.66
## depbetray_mostjob_devx2 0.12 0.16 -0.21 0.43
## depbetray_mostjob_av12x2 0.11 0.04 0.04 0.18
## depbetray_mostjob_devx2:rural.ses.med2 0.51 0.23 0.04 0.98
## depbetray_mostjob_devx2:rural.ses.med3 -0.07 0.24 -0.55 0.38
## depbetray_mostjob_devx2:rural.ses.med4 -0.06 0.30 -0.65 0.53
## Rhat Bulk_ESS Tail_ESS
## depcantgo_Intercept 1.00 3387 3076
## depeffort_Intercept 1.00 2370 2564
## deplonely_Intercept 1.00 3464 3172
## depblues_Intercept 1.00 2415 3196
## depunfair_Intercept 1.00 3087 3094
## depmistrt_Intercept 1.00 2362 2633
## depbetray_Intercept 1.00 3539 3394
## depcantgo_rural.ses.med2 1.00 2686 2560
## depcantgo_rural.ses.med3 1.00 3095 2579
## depcantgo_rural.ses.med4 1.00 2676 2443
## depeffort_rural.ses.med2 1.00 3865 3156
## depeffort_rural.ses.med3 1.00 3451 2924
## depeffort_rural.ses.med4 1.00 4060 3156
## deplonely_rural.ses.med2 1.00 3538 2862
## deplonely_rural.ses.med3 1.00 3462 3249
## deplonely_rural.ses.med4 1.00 3607 3162
## depblues_rural.ses.med2 1.00 3637 2745
## depblues_rural.ses.med3 1.00 3567 2854
## depblues_rural.ses.med4 1.00 3389 3087
## depunfair_rural.ses.med2 1.00 2980 2920
## depunfair_rural.ses.med3 1.00 2426 2071
## depunfair_rural.ses.med4 1.00 4017 3215
## depmistrt_rural.ses.med2 1.00 2944 2256
## depmistrt_rural.ses.med3 1.00 3064 3231
## depmistrt_rural.ses.med4 1.00 2736 2713
## depbetray_rural.ses.med2 1.00 4052 2901
## depbetray_rural.ses.med3 1.00 3237 3131
## depbetray_rural.ses.med4 1.00 3307 2754
## depcantgo_mostjob_devx2 1.00 2317 2442
## depcantgo_mostjob_av12x2 1.00 6378 3407
## depcantgo_mostjob_devx2:rural.ses.med2 1.00 2473 2690
## depcantgo_mostjob_devx2:rural.ses.med3 1.00 2388 2360
## depcantgo_mostjob_devx2:rural.ses.med4 1.00 2440 2582
## depeffort_mostjob_devx2 1.00 2262 2812
## depeffort_mostjob_av12x2 1.00 5948 3269
## depeffort_mostjob_devx2:rural.ses.med2 1.00 3190 2859
## depeffort_mostjob_devx2:rural.ses.med3 1.00 2918 2006
## depeffort_mostjob_devx2:rural.ses.med4 1.00 3603 2936
## deplonely_mostjob_devx2 1.00 3142 2846
## deplonely_mostjob_av12x2 1.00 6639 3670
## deplonely_mostjob_devx2:rural.ses.med2 1.00 3013 2064
## deplonely_mostjob_devx2:rural.ses.med3 1.00 3243 3162
## deplonely_mostjob_devx2:rural.ses.med4 1.00 3648 3049
## depblues_mostjob_devx2 1.00 3344 3151
## depblues_mostjob_av12x2 1.00 6001 3363
## depblues_mostjob_devx2:rural.ses.med2 1.00 2996 2280
## depblues_mostjob_devx2:rural.ses.med3 1.00 2767 1914
## depblues_mostjob_devx2:rural.ses.med4 1.00 2942 3003
## depunfair_mostjob_devx2 1.00 2723 3025
## depunfair_mostjob_av12x2 1.00 7377 3132
## depunfair_mostjob_devx2:rural.ses.med2 1.00 2906 2761
## depunfair_mostjob_devx2:rural.ses.med3 1.00 2522 2078
## depunfair_mostjob_devx2:rural.ses.med4 1.00 3525 3161
## depmistrt_mostjob_devx2 1.00 1928 2707
## depmistrt_mostjob_av12x2 1.00 8263 3414
## depmistrt_mostjob_devx2:rural.ses.med2 1.00 2489 1863
## depmistrt_mostjob_devx2:rural.ses.med3 1.00 2525 2653
## depmistrt_mostjob_devx2:rural.ses.med4 1.00 2743 2739
## depbetray_mostjob_devx2 1.00 3110 3137
## depbetray_mostjob_av12x2 1.00 7693 3600
## depbetray_mostjob_devx2:rural.ses.med2 1.00 2920 2602
## depbetray_mostjob_devx2:rural.ses.med3 1.00 3179 3213
## depbetray_mostjob_devx2:rural.ses.med4 1.00 3111 2778
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_mostjob_devx21[1] 0.20 0.12 0.03 0.48
## depcantgo_mostjob_devx21[2] 0.20 0.12 0.03 0.48
## depcantgo_mostjob_devx21[3] 0.36 0.15 0.08 0.65
## depcantgo_mostjob_devx21[4] 0.24 0.13 0.03 0.53
## depcantgo_mostjob_av12x21[1] 0.10 0.07 0.01 0.26
## depcantgo_mostjob_av12x21[2] 0.10 0.07 0.01 0.27
## depcantgo_mostjob_av12x21[3] 0.11 0.07 0.01 0.28
## depcantgo_mostjob_av12x21[4] 0.12 0.08 0.02 0.31
## depcantgo_mostjob_av12x21[5] 0.14 0.08 0.02 0.34
## depcantgo_mostjob_av12x21[6] 0.16 0.10 0.02 0.38
## depcantgo_mostjob_av12x21[7] 0.14 0.09 0.02 0.35
## depcantgo_mostjob_av12x21[8] 0.13 0.08 0.02 0.31
## depcantgo_mostjob_devx2:rural.ses.med21[1] 0.29 0.20 0.01 0.73
## depcantgo_mostjob_devx2:rural.ses.med21[2] 0.17 0.16 0.00 0.61
## depcantgo_mostjob_devx2:rural.ses.med21[3] 0.27 0.19 0.01 0.70
## depcantgo_mostjob_devx2:rural.ses.med21[4] 0.28 0.20 0.01 0.72
## depcantgo_mostjob_devx2:rural.ses.med31[1] 0.17 0.14 0.01 0.54
## depcantgo_mostjob_devx2:rural.ses.med31[2] 0.35 0.18 0.03 0.72
## depcantgo_mostjob_devx2:rural.ses.med31[3] 0.28 0.17 0.02 0.64
## depcantgo_mostjob_devx2:rural.ses.med31[4] 0.21 0.16 0.01 0.59
## depcantgo_mostjob_devx2:rural.ses.med41[1] 0.26 0.21 0.01 0.75
## depcantgo_mostjob_devx2:rural.ses.med41[2] 0.21 0.17 0.01 0.62
## depcantgo_mostjob_devx2:rural.ses.med41[3] 0.26 0.20 0.01 0.73
## depcantgo_mostjob_devx2:rural.ses.med41[4] 0.28 0.20 0.01 0.73
## depeffort_mostjob_devx21[1] 0.25 0.15 0.04 0.58
## depeffort_mostjob_devx21[2] 0.26 0.14 0.04 0.58
## depeffort_mostjob_devx21[3] 0.23 0.13 0.04 0.53
## depeffort_mostjob_devx21[4] 0.26 0.15 0.04 0.59
## depeffort_mostjob_av12x21[1] 0.12 0.08 0.02 0.31
## depeffort_mostjob_av12x21[2] 0.13 0.08 0.02 0.33
## depeffort_mostjob_av12x21[3] 0.12 0.08 0.02 0.30
## depeffort_mostjob_av12x21[4] 0.12 0.08 0.02 0.31
## depeffort_mostjob_av12x21[5] 0.12 0.08 0.01 0.32
## depeffort_mostjob_av12x21[6] 0.12 0.08 0.02 0.31
## depeffort_mostjob_av12x21[7] 0.13 0.08 0.02 0.32
## depeffort_mostjob_av12x21[8] 0.13 0.08 0.02 0.34
## depeffort_mostjob_devx2:rural.ses.med21[1] 0.25 0.19 0.01 0.70
## depeffort_mostjob_devx2:rural.ses.med21[2] 0.22 0.18 0.01 0.65
## depeffort_mostjob_devx2:rural.ses.med21[3] 0.23 0.18 0.01 0.66
## depeffort_mostjob_devx2:rural.ses.med21[4] 0.30 0.22 0.01 0.77
## depeffort_mostjob_devx2:rural.ses.med31[1] 0.28 0.21 0.01 0.75
## depeffort_mostjob_devx2:rural.ses.med31[2] 0.22 0.18 0.01 0.68
## depeffort_mostjob_devx2:rural.ses.med31[3] 0.21 0.17 0.01 0.64
## depeffort_mostjob_devx2:rural.ses.med31[4] 0.29 0.22 0.01 0.81
## depeffort_mostjob_devx2:rural.ses.med41[1] 0.23 0.18 0.01 0.66
## depeffort_mostjob_devx2:rural.ses.med41[2] 0.19 0.17 0.01 0.63
## depeffort_mostjob_devx2:rural.ses.med41[3] 0.24 0.18 0.01 0.68
## depeffort_mostjob_devx2:rural.ses.med41[4] 0.34 0.23 0.01 0.79
## deplonely_mostjob_devx21[1] 0.21 0.13 0.03 0.52
## deplonely_mostjob_devx21[2] 0.25 0.13 0.04 0.54
## deplonely_mostjob_devx21[3] 0.29 0.14 0.05 0.60
## deplonely_mostjob_devx21[4] 0.26 0.14 0.04 0.58
## deplonely_mostjob_av12x21[1] 0.13 0.08 0.02 0.31
## deplonely_mostjob_av12x21[2] 0.13 0.08 0.02 0.33
## deplonely_mostjob_av12x21[3] 0.13 0.08 0.02 0.32
## deplonely_mostjob_av12x21[4] 0.12 0.08 0.02 0.32
## deplonely_mostjob_av12x21[5] 0.12 0.08 0.02 0.31
## deplonely_mostjob_av12x21[6] 0.12 0.08 0.01 0.32
## deplonely_mostjob_av12x21[7] 0.13 0.08 0.02 0.33
## deplonely_mostjob_av12x21[8] 0.13 0.08 0.02 0.33
## deplonely_mostjob_devx2:rural.ses.med21[1] 0.23 0.18 0.01 0.66
## deplonely_mostjob_devx2:rural.ses.med21[2] 0.19 0.16 0.01 0.59
## deplonely_mostjob_devx2:rural.ses.med21[3] 0.25 0.18 0.01 0.67
## deplonely_mostjob_devx2:rural.ses.med21[4] 0.33 0.22 0.01 0.80
## deplonely_mostjob_devx2:rural.ses.med31[1] 0.24 0.18 0.01 0.66
## deplonely_mostjob_devx2:rural.ses.med31[2] 0.25 0.18 0.01 0.68
## deplonely_mostjob_devx2:rural.ses.med31[3] 0.26 0.18 0.01 0.66
## deplonely_mostjob_devx2:rural.ses.med31[4] 0.25 0.19 0.01 0.68
## deplonely_mostjob_devx2:rural.ses.med41[1] 0.26 0.20 0.01 0.73
## deplonely_mostjob_devx2:rural.ses.med41[2] 0.23 0.18 0.01 0.65
## deplonely_mostjob_devx2:rural.ses.med41[3] 0.23 0.18 0.01 0.66
## deplonely_mostjob_devx2:rural.ses.med41[4] 0.28 0.20 0.01 0.74
## depblues_mostjob_devx21[1] 0.24 0.14 0.04 0.57
## depblues_mostjob_devx21[2] 0.26 0.14 0.05 0.58
## depblues_mostjob_devx21[3] 0.24 0.14 0.04 0.55
## depblues_mostjob_devx21[4] 0.25 0.14 0.04 0.57
## depblues_mostjob_av12x21[1] 0.13 0.08 0.02 0.33
## depblues_mostjob_av12x21[2] 0.12 0.08 0.02 0.32
## depblues_mostjob_av12x21[3] 0.12 0.08 0.02 0.32
## depblues_mostjob_av12x21[4] 0.13 0.08 0.02 0.32
## depblues_mostjob_av12x21[5] 0.13 0.08 0.02 0.32
## depblues_mostjob_av12x21[6] 0.13 0.08 0.02 0.32
## depblues_mostjob_av12x21[7] 0.13 0.08 0.02 0.32
## depblues_mostjob_av12x21[8] 0.12 0.08 0.02 0.31
## depblues_mostjob_devx2:rural.ses.med21[1] 0.24 0.19 0.01 0.70
## depblues_mostjob_devx2:rural.ses.med21[2] 0.22 0.18 0.01 0.67
## depblues_mostjob_devx2:rural.ses.med21[3] 0.27 0.20 0.01 0.71
## depblues_mostjob_devx2:rural.ses.med21[4] 0.28 0.21 0.01 0.77
## depblues_mostjob_devx2:rural.ses.med31[1] 0.27 0.20 0.01 0.73
## depblues_mostjob_devx2:rural.ses.med31[2] 0.21 0.17 0.01 0.63
## depblues_mostjob_devx2:rural.ses.med31[3] 0.23 0.18 0.01 0.66
## depblues_mostjob_devx2:rural.ses.med31[4] 0.29 0.22 0.01 0.79
## depblues_mostjob_devx2:rural.ses.med41[1] 0.26 0.19 0.01 0.72
## depblues_mostjob_devx2:rural.ses.med41[2] 0.19 0.17 0.00 0.61
## depblues_mostjob_devx2:rural.ses.med41[3] 0.20 0.19 0.00 0.66
## depblues_mostjob_devx2:rural.ses.med41[4] 0.35 0.24 0.01 0.84
## depunfair_mostjob_devx21[1] 0.25 0.14 0.04 0.58
## depunfair_mostjob_devx21[2] 0.22 0.13 0.03 0.54
## depunfair_mostjob_devx21[3] 0.27 0.15 0.04 0.60
## depunfair_mostjob_devx21[4] 0.25 0.14 0.04 0.58
## depunfair_mostjob_av12x21[1] 0.11 0.07 0.02 0.30
## depunfair_mostjob_av12x21[2] 0.12 0.08 0.01 0.31
## depunfair_mostjob_av12x21[3] 0.13 0.08 0.02 0.32
## depunfair_mostjob_av12x21[4] 0.13 0.08 0.02 0.33
## depunfair_mostjob_av12x21[5] 0.12 0.08 0.02 0.31
## depunfair_mostjob_av12x21[6] 0.11 0.07 0.01 0.28
## depunfair_mostjob_av12x21[7] 0.15 0.09 0.02 0.36
## depunfair_mostjob_av12x21[8] 0.13 0.08 0.02 0.32
## depunfair_mostjob_devx2:rural.ses.med21[1] 0.24 0.18 0.01 0.66
## depunfair_mostjob_devx2:rural.ses.med21[2] 0.20 0.15 0.01 0.58
## depunfair_mostjob_devx2:rural.ses.med21[3] 0.33 0.19 0.02 0.72
## depunfair_mostjob_devx2:rural.ses.med21[4] 0.23 0.18 0.01 0.65
## depunfair_mostjob_devx2:rural.ses.med31[1] 0.25 0.19 0.01 0.72
## depunfair_mostjob_devx2:rural.ses.med31[2] 0.24 0.19 0.01 0.70
## depunfair_mostjob_devx2:rural.ses.med31[3] 0.23 0.18 0.01 0.68
## depunfair_mostjob_devx2:rural.ses.med31[4] 0.28 0.20 0.01 0.72
## depunfair_mostjob_devx2:rural.ses.med41[1] 0.15 0.13 0.00 0.49
## depunfair_mostjob_devx2:rural.ses.med41[2] 0.15 0.12 0.01 0.44
## depunfair_mostjob_devx2:rural.ses.med41[3] 0.48 0.18 0.09 0.81
## depunfair_mostjob_devx2:rural.ses.med41[4] 0.22 0.16 0.01 0.61
## depmistrt_mostjob_devx21[1] 0.26 0.15 0.04 0.63
## depmistrt_mostjob_devx21[2] 0.23 0.13 0.04 0.54
## depmistrt_mostjob_devx21[3] 0.25 0.15 0.03 0.58
## depmistrt_mostjob_devx21[4] 0.26 0.15 0.04 0.59
## depmistrt_mostjob_av12x21[1] 0.14 0.09 0.02 0.36
## depmistrt_mostjob_av12x21[2] 0.13 0.08 0.02 0.33
## depmistrt_mostjob_av12x21[3] 0.13 0.08 0.02 0.32
## depmistrt_mostjob_av12x21[4] 0.13 0.08 0.02 0.33
## depmistrt_mostjob_av12x21[5] 0.12 0.08 0.02 0.30
## depmistrt_mostjob_av12x21[6] 0.12 0.07 0.02 0.30
## depmistrt_mostjob_av12x21[7] 0.12 0.08 0.02 0.31
## depmistrt_mostjob_av12x21[8] 0.12 0.08 0.02 0.32
## depmistrt_mostjob_devx2:rural.ses.med21[1] 0.25 0.19 0.01 0.72
## depmistrt_mostjob_devx2:rural.ses.med21[2] 0.23 0.17 0.01 0.64
## depmistrt_mostjob_devx2:rural.ses.med21[3] 0.25 0.19 0.01 0.69
## depmistrt_mostjob_devx2:rural.ses.med21[4] 0.27 0.21 0.01 0.77
## depmistrt_mostjob_devx2:rural.ses.med31[1] 0.26 0.19 0.01 0.70
## depmistrt_mostjob_devx2:rural.ses.med31[2] 0.20 0.17 0.01 0.63
## depmistrt_mostjob_devx2:rural.ses.med31[3] 0.18 0.17 0.00 0.62
## depmistrt_mostjob_devx2:rural.ses.med31[4] 0.35 0.23 0.02 0.82
## depmistrt_mostjob_devx2:rural.ses.med41[1] 0.24 0.19 0.01 0.69
## depmistrt_mostjob_devx2:rural.ses.med41[2] 0.23 0.18 0.01 0.67
## depmistrt_mostjob_devx2:rural.ses.med41[3] 0.25 0.19 0.01 0.68
## depmistrt_mostjob_devx2:rural.ses.med41[4] 0.29 0.21 0.01 0.75
## depbetray_mostjob_devx21[1] 0.24 0.14 0.03 0.57
## depbetray_mostjob_devx21[2] 0.25 0.14 0.04 0.58
## depbetray_mostjob_devx21[3] 0.27 0.15 0.04 0.59
## depbetray_mostjob_devx21[4] 0.24 0.14 0.04 0.55
## depbetray_mostjob_av12x21[1] 0.14 0.09 0.02 0.35
## depbetray_mostjob_av12x21[2] 0.12 0.08 0.02 0.30
## depbetray_mostjob_av12x21[3] 0.11 0.07 0.02 0.29
## depbetray_mostjob_av12x21[4] 0.11 0.07 0.01 0.28
## depbetray_mostjob_av12x21[5] 0.11 0.07 0.01 0.29
## depbetray_mostjob_av12x21[6] 0.10 0.07 0.02 0.26
## depbetray_mostjob_av12x21[7] 0.13 0.08 0.02 0.32
## depbetray_mostjob_av12x21[8] 0.18 0.10 0.03 0.40
## depbetray_mostjob_devx2:rural.ses.med21[1] 0.16 0.13 0.00 0.49
## depbetray_mostjob_devx2:rural.ses.med21[2] 0.19 0.15 0.01 0.55
## depbetray_mostjob_devx2:rural.ses.med21[3] 0.44 0.19 0.07 0.80
## depbetray_mostjob_devx2:rural.ses.med21[4] 0.22 0.16 0.01 0.60
## depbetray_mostjob_devx2:rural.ses.med31[1] 0.25 0.20 0.01 0.71
## depbetray_mostjob_devx2:rural.ses.med31[2] 0.22 0.17 0.01 0.64
## depbetray_mostjob_devx2:rural.ses.med31[3] 0.24 0.18 0.01 0.68
## depbetray_mostjob_devx2:rural.ses.med31[4] 0.29 0.21 0.01 0.76
## depbetray_mostjob_devx2:rural.ses.med41[1] 0.24 0.18 0.01 0.68
## depbetray_mostjob_devx2:rural.ses.med41[2] 0.22 0.18 0.01 0.64
## depbetray_mostjob_devx2:rural.ses.med41[3] 0.26 0.20 0.01 0.71
## depbetray_mostjob_devx2:rural.ses.med41[4] 0.28 0.20 0.01 0.73
## Rhat Bulk_ESS Tail_ESS
## depcantgo_mostjob_devx21[1] 1.00 6542 2622
## depcantgo_mostjob_devx21[2] 1.00 6442 2541
## depcantgo_mostjob_devx21[3] 1.00 3769 2808
## depcantgo_mostjob_devx21[4] 1.00 7426 2552
## depcantgo_mostjob_av12x21[1] 1.00 7753 2666
## depcantgo_mostjob_av12x21[2] 1.00 7798 2648
## depcantgo_mostjob_av12x21[3] 1.00 7035 2386
## depcantgo_mostjob_av12x21[4] 1.00 8586 2413
## depcantgo_mostjob_av12x21[5] 1.00 7222 2662
## depcantgo_mostjob_av12x21[6] 1.00 8315 2397
## depcantgo_mostjob_av12x21[7] 1.00 6512 2698
## depcantgo_mostjob_av12x21[8] 1.00 7840 2932
## depcantgo_mostjob_devx2:rural.ses.med21[1] 1.00 4841 2674
## depcantgo_mostjob_devx2:rural.ses.med21[2] 1.00 4930 2818
## depcantgo_mostjob_devx2:rural.ses.med21[3] 1.00 4976 2977
## depcantgo_mostjob_devx2:rural.ses.med21[4] 1.00 6075 2838
## depcantgo_mostjob_devx2:rural.ses.med31[1] 1.00 5375 2618
## depcantgo_mostjob_devx2:rural.ses.med31[2] 1.00 5533 2213
## depcantgo_mostjob_devx2:rural.ses.med31[3] 1.00 4499 2924
## depcantgo_mostjob_devx2:rural.ses.med31[4] 1.00 5875 2600
## depcantgo_mostjob_devx2:rural.ses.med41[1] 1.00 3729 2524
## depcantgo_mostjob_devx2:rural.ses.med41[2] 1.00 5591 2594
## depcantgo_mostjob_devx2:rural.ses.med41[3] 1.00 3666 2739
## depcantgo_mostjob_devx2:rural.ses.med41[4] 1.00 5971 2827
## depeffort_mostjob_devx21[1] 1.00 6273 2488
## depeffort_mostjob_devx21[2] 1.00 5364 2795
## depeffort_mostjob_devx21[3] 1.00 5991 2695
## depeffort_mostjob_devx21[4] 1.00 7004 2888
## depeffort_mostjob_av12x21[1] 1.00 8657 2641
## depeffort_mostjob_av12x21[2] 1.00 8637 2458
## depeffort_mostjob_av12x21[3] 1.00 8303 2462
## depeffort_mostjob_av12x21[4] 1.00 7556 2502
## depeffort_mostjob_av12x21[5] 1.00 7487 2636
## depeffort_mostjob_av12x21[6] 1.00 8872 3100
## depeffort_mostjob_av12x21[7] 1.00 7874 2780
## depeffort_mostjob_av12x21[8] 1.00 6443 2866
## depeffort_mostjob_devx2:rural.ses.med21[1] 1.00 5871 2276
## depeffort_mostjob_devx2:rural.ses.med21[2] 1.00 6141 2741
## depeffort_mostjob_devx2:rural.ses.med21[3] 1.00 6087 3170
## depeffort_mostjob_devx2:rural.ses.med21[4] 1.00 5510 2874
## depeffort_mostjob_devx2:rural.ses.med31[1] 1.00 4353 2539
## depeffort_mostjob_devx2:rural.ses.med31[2] 1.00 4969 2611
## depeffort_mostjob_devx2:rural.ses.med31[3] 1.00 5047 2745
## depeffort_mostjob_devx2:rural.ses.med31[4] 1.00 4261 2932
## depeffort_mostjob_devx2:rural.ses.med41[1] 1.00 5612 2196
## depeffort_mostjob_devx2:rural.ses.med41[2] 1.00 5356 2479
## depeffort_mostjob_devx2:rural.ses.med41[3] 1.00 4090 2813
## depeffort_mostjob_devx2:rural.ses.med41[4] 1.00 4990 2289
## deplonely_mostjob_devx21[1] 1.00 6529 3073
## deplonely_mostjob_devx21[2] 1.00 8605 2345
## deplonely_mostjob_devx21[3] 1.00 5341 3090
## deplonely_mostjob_devx21[4] 1.00 7684 2677
## deplonely_mostjob_av12x21[1] 1.00 7040 2450
## deplonely_mostjob_av12x21[2] 1.00 7177 2305
## deplonely_mostjob_av12x21[3] 1.00 8345 2401
## deplonely_mostjob_av12x21[4] 1.00 6661 2425
## deplonely_mostjob_av12x21[5] 1.00 8498 2932
## deplonely_mostjob_av12x21[6] 1.00 9048 2642
## deplonely_mostjob_av12x21[7] 1.00 7353 2469
## deplonely_mostjob_av12x21[8] 1.00 7345 2619
## deplonely_mostjob_devx2:rural.ses.med21[1] 1.00 5041 2538
## deplonely_mostjob_devx2:rural.ses.med21[2] 1.00 5743 2755
## deplonely_mostjob_devx2:rural.ses.med21[3] 1.00 5450 3014
## deplonely_mostjob_devx2:rural.ses.med21[4] 1.00 4699 2622
## deplonely_mostjob_devx2:rural.ses.med31[1] 1.00 6146 2324
## deplonely_mostjob_devx2:rural.ses.med31[2] 1.00 6126 2544
## deplonely_mostjob_devx2:rural.ses.med31[3] 1.00 5951 3012
## deplonely_mostjob_devx2:rural.ses.med31[4] 1.00 6828 2625
## deplonely_mostjob_devx2:rural.ses.med41[1] 1.00 6021 2513
## deplonely_mostjob_devx2:rural.ses.med41[2] 1.00 6493 2429
## deplonely_mostjob_devx2:rural.ses.med41[3] 1.00 5997 2128
## deplonely_mostjob_devx2:rural.ses.med41[4] 1.00 5849 3315
## depblues_mostjob_devx21[1] 1.00 8029 2682
## depblues_mostjob_devx21[2] 1.00 6185 2604
## depblues_mostjob_devx21[3] 1.00 8739 3199
## depblues_mostjob_devx21[4] 1.00 7402 2528
## depblues_mostjob_av12x21[1] 1.00 8172 1970
## depblues_mostjob_av12x21[2] 1.00 7514 2605
## depblues_mostjob_av12x21[3] 1.00 8255 2937
## depblues_mostjob_av12x21[4] 1.00 8178 2937
## depblues_mostjob_av12x21[5] 1.00 9025 2769
## depblues_mostjob_av12x21[6] 1.00 7641 2556
## depblues_mostjob_av12x21[7] 1.00 7327 2994
## depblues_mostjob_av12x21[8] 1.00 7272 3279
## depblues_mostjob_devx2:rural.ses.med21[1] 1.00 5945 2561
## depblues_mostjob_devx2:rural.ses.med21[2] 1.00 7266 2115
## depblues_mostjob_devx2:rural.ses.med21[3] 1.00 4463 3217
## depblues_mostjob_devx2:rural.ses.med21[4] 1.00 5255 2298
## depblues_mostjob_devx2:rural.ses.med31[1] 1.00 4742 2342
## depblues_mostjob_devx2:rural.ses.med31[2] 1.00 5833 2572
## depblues_mostjob_devx2:rural.ses.med31[3] 1.00 5465 2548
## depblues_mostjob_devx2:rural.ses.med31[4] 1.00 4788 2368
## depblues_mostjob_devx2:rural.ses.med41[1] 1.00 5258 2895
## depblues_mostjob_devx2:rural.ses.med41[2] 1.00 3890 2772
## depblues_mostjob_devx2:rural.ses.med41[3] 1.00 3813 3272
## depblues_mostjob_devx2:rural.ses.med41[4] 1.00 3909 3159
## depunfair_mostjob_devx21[1] 1.00 7160 2895
## depunfair_mostjob_devx21[2] 1.00 7721 2837
## depunfair_mostjob_devx21[3] 1.00 4884 2965
## depunfair_mostjob_devx21[4] 1.00 6549 2740
## depunfair_mostjob_av12x21[1] 1.00 6917 2965
## depunfair_mostjob_av12x21[2] 1.00 7670 2387
## depunfair_mostjob_av12x21[3] 1.00 7204 2314
## depunfair_mostjob_av12x21[4] 1.00 7804 2709
## depunfair_mostjob_av12x21[5] 1.00 8569 2692
## depunfair_mostjob_av12x21[6] 1.00 7374 2866
## depunfair_mostjob_av12x21[7] 1.00 7158 2919
## depunfair_mostjob_av12x21[8] 1.00 8388 2611
## depunfair_mostjob_devx2:rural.ses.med21[1] 1.00 4766 2166
## depunfair_mostjob_devx2:rural.ses.med21[2] 1.00 5547 2496
## depunfair_mostjob_devx2:rural.ses.med21[3] 1.00 4174 2610
## depunfair_mostjob_devx2:rural.ses.med21[4] 1.00 6167 2595
## depunfair_mostjob_devx2:rural.ses.med31[1] 1.00 4047 2325
## depunfair_mostjob_devx2:rural.ses.med31[2] 1.00 5007 2197
## depunfair_mostjob_devx2:rural.ses.med31[3] 1.00 5116 2848
## depunfair_mostjob_devx2:rural.ses.med31[4] 1.00 6646 2691
## depunfair_mostjob_devx2:rural.ses.med41[1] 1.00 5272 2586
## depunfair_mostjob_devx2:rural.ses.med41[2] 1.00 5362 2585
## depunfair_mostjob_devx2:rural.ses.med41[3] 1.00 4458 2816
## depunfair_mostjob_devx2:rural.ses.med41[4] 1.00 6007 2543
## depmistrt_mostjob_devx21[1] 1.00 4060 2933
## depmistrt_mostjob_devx21[2] 1.00 7712 2992
## depmistrt_mostjob_devx21[3] 1.00 3423 2795
## depmistrt_mostjob_devx21[4] 1.00 6540 2945
## depmistrt_mostjob_av12x21[1] 1.00 7699 2292
## depmistrt_mostjob_av12x21[2] 1.00 7367 2113
## depmistrt_mostjob_av12x21[3] 1.00 6966 2353
## depmistrt_mostjob_av12x21[4] 1.00 7594 2320
## depmistrt_mostjob_av12x21[5] 1.00 6548 2411
## depmistrt_mostjob_av12x21[6] 1.00 7298 2609
## depmistrt_mostjob_av12x21[7] 1.00 7205 3015
## depmistrt_mostjob_av12x21[8] 1.00 6601 2574
## depmistrt_mostjob_devx2:rural.ses.med21[1] 1.00 5586 2298
## depmistrt_mostjob_devx2:rural.ses.med21[2] 1.00 5956 2899
## depmistrt_mostjob_devx2:rural.ses.med21[3] 1.00 3732 2753
## depmistrt_mostjob_devx2:rural.ses.med21[4] 1.00 3810 2361
## depmistrt_mostjob_devx2:rural.ses.med31[1] 1.00 5648 2795
## depmistrt_mostjob_devx2:rural.ses.med31[2] 1.00 5677 2490
## depmistrt_mostjob_devx2:rural.ses.med31[3] 1.00 3847 2555
## depmistrt_mostjob_devx2:rural.ses.med31[4] 1.00 4236 3059
## depmistrt_mostjob_devx2:rural.ses.med41[1] 1.00 5532 2298
## depmistrt_mostjob_devx2:rural.ses.med41[2] 1.00 6275 2247
## depmistrt_mostjob_devx2:rural.ses.med41[3] 1.00 5846 2679
## depmistrt_mostjob_devx2:rural.ses.med41[4] 1.00 5574 2715
## depbetray_mostjob_devx21[1] 1.00 6501 2422
## depbetray_mostjob_devx21[2] 1.00 7030 2578
## depbetray_mostjob_devx21[3] 1.00 6025 3055
## depbetray_mostjob_devx21[4] 1.00 6846 3243
## depbetray_mostjob_av12x21[1] 1.00 7897 2284
## depbetray_mostjob_av12x21[2] 1.00 7994 2375
## depbetray_mostjob_av12x21[3] 1.00 9269 2438
## depbetray_mostjob_av12x21[4] 1.00 6690 2694
## depbetray_mostjob_av12x21[5] 1.00 9584 2223
## depbetray_mostjob_av12x21[6] 1.00 7133 2875
## depbetray_mostjob_av12x21[7] 1.00 8021 2917
## depbetray_mostjob_av12x21[8] 1.00 7770 2801
## depbetray_mostjob_devx2:rural.ses.med21[1] 1.00 4879 2406
## depbetray_mostjob_devx2:rural.ses.med21[2] 1.00 6108 2867
## depbetray_mostjob_devx2:rural.ses.med21[3] 1.00 4408 2558
## depbetray_mostjob_devx2:rural.ses.med21[4] 1.00 5661 2591
## depbetray_mostjob_devx2:rural.ses.med31[1] 1.00 6368 2104
## depbetray_mostjob_devx2:rural.ses.med31[2] 1.00 7188 2462
## depbetray_mostjob_devx2:rural.ses.med31[3] 1.00 5447 3070
## depbetray_mostjob_devx2:rural.ses.med31[4] 1.00 6576 2650
## depbetray_mostjob_devx2:rural.ses.med41[1] 1.00 5220 2455
## depbetray_mostjob_devx2:rural.ses.med41[2] 1.00 5927 2546
## depbetray_mostjob_devx2:rural.ses.med41[3] 1.00 3975 3241
## depbetray_mostjob_devx2:rural.ses.med41[4] 1.00 7829 3000
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.2.5.4 Prior summary
out.chg.alldepress.stjob.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostjob_av12x2
## normal(0, 0.25) b mostjob_devx2
## normal(0, 1) b mostjob_devx2:rural.ses.med2
## normal(0, 1) b mostjob_devx2:rural.ses.med3
## normal(0, 1) b mostjob_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostjob_av12x21
## dirichlet(1) simo mostjob_devx2:rural.ses.med21
## dirichlet(1) simo mostjob_devx2:rural.ses.med31
## dirichlet(1) simo mostjob_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostjob_devx21
## group resp dpar nlpar lb ub source
## default
## depbetray user
## depbetray user
## depbetray user
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depblues user
## depblues user
## depblues user
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depcantgo user
## depcantgo user
## depcantgo user
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depeffort user
## depeffort user
## depeffort user
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## deplonely user
## deplonely user
## deplonely user
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## depmistrt user
## depmistrt user
## depmistrt user
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depunfair user
## depunfair user
## depunfair user
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## default
## depbetray user
## depblues user
## depcantgo user
## depeffort user
## deplonely user
## depmistrt user
## depunfair user
## depbetray 0 default
## depblues 0 default
## depcantgo 0 default
## depeffort 0 default
## deplonely 0 default
## depmistrt 0 default
## depunfair 0 default
## id depbetray 0 (vectorized)
## id depbetray 0 (vectorized)
## id depblues 0 (vectorized)
## id depblues 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depeffort 0 (vectorized)
## id depeffort 0 (vectorized)
## id deplonely 0 (vectorized)
## id deplonely 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depunfair 0 (vectorized)
## id depunfair 0 (vectorized)
## depbetray user
## depbetray default
## depbetray default
## depbetray default
## depbetray user
## depblues user
## depblues default
## depblues default
## depblues default
## depblues user
## depcantgo user
## depcantgo default
## depcantgo default
## depcantgo default
## depcantgo user
## depeffort user
## depeffort default
## depeffort default
## depeffort default
## depeffort user
## deplonely user
## deplonely default
## deplonely default
## deplonely default
## deplonely user
## depmistrt user
## depmistrt default
## depmistrt default
## depmistrt default
## depmistrt user
## depunfair user
## depunfair default
## depunfair default
## depunfair default
## depunfair user8.1.2.6 Corr X Community: stthft & negative emotions
#Community Change: negative emotions items ~ mo(stthft)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostthft_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostthft_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostthft_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostthft_av12x21',
resp = depdv_names)
)
chg.alldepress.stthft.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) +
rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stthft_comm_fit",
file_refit = "on_change"
)
out.chg.alldepress.stthft.comm.fit <- ppchecks(chg.alldepress.stthft.comm.fit)8.1.2.6.1 Coefficient plot (intervals)
out.chg.alldepress.stthft.comm.fit[[11]]
8.1.2.6.2 PPcheck (density)
p1 <- out.chg.alldepress.stthft.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stthft.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stthft.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stthft.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stthft.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stthft.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stthft.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
8.1.2.6.3 Fit summary
out.chg.alldepress.stthft.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## depeffort ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## deplonely ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## depblues ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## depunfair ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## depmistrt ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## depbetray ~ 1 + mo(stthft_devx2) + mo(stthft_av12x2) + rural.ses.med + mo(stthft_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.32 0.20 0.02 0.73 1.00 662
## sd(depeffort_Intercept) 0.43 0.27 0.02 0.99 1.01 506
## sd(deplonely_Intercept) 0.44 0.24 0.02 0.90 1.00 579
## sd(depblues_Intercept) 0.72 0.31 0.10 1.31 1.02 387
## sd(depunfair_Intercept) 0.23 0.16 0.01 0.60 1.00 832
## sd(depmistrt_Intercept) 0.35 0.23 0.02 0.86 1.00 619
## sd(depbetray_Intercept) 0.53 0.29 0.03 1.09 1.01 365
## Tail_ESS
## sd(depcantgo_Intercept) 1436
## sd(depeffort_Intercept) 937
## sd(deplonely_Intercept) 1158
## sd(depblues_Intercept) 714
## sd(depunfair_Intercept) 1619
## sd(depmistrt_Intercept) 1298
## sd(depbetray_Intercept) 908
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_Intercept -0.81 0.33 -1.48 -0.17
## depeffort_Intercept -1.97 0.39 -2.79 -1.23
## deplonely_Intercept -1.54 0.36 -2.26 -0.85
## depblues_Intercept -2.39 0.45 -3.33 -1.56
## depunfair_Intercept -1.64 0.39 -2.44 -0.90
## depmistrt_Intercept -2.29 0.41 -3.06 -1.45
## depbetray_Intercept -2.49 0.46 -3.37 -1.53
## depcantgo_rural.ses.med2 -0.46 0.53 -1.68 0.43
## depcantgo_rural.ses.med3 0.34 0.51 -0.56 1.52
## depcantgo_rural.ses.med4 -0.02 0.47 -0.99 0.89
## depeffort_rural.ses.med2 0.31 0.59 -0.87 1.59
## depeffort_rural.ses.med3 0.04 0.58 -1.21 1.10
## depeffort_rural.ses.med4 0.27 0.54 -0.87 1.29
## deplonely_rural.ses.med2 -0.74 0.59 -1.91 0.51
## deplonely_rural.ses.med3 0.45 0.55 -0.53 1.64
## deplonely_rural.ses.med4 0.42 0.54 -0.53 1.61
## depblues_rural.ses.med2 0.30 0.67 -1.16 1.52
## depblues_rural.ses.med3 0.44 0.57 -0.76 1.54
## depblues_rural.ses.med4 -0.21 0.60 -1.54 0.85
## depunfair_rural.ses.med2 0.13 0.61 -1.19 1.25
## depunfair_rural.ses.med3 -0.40 0.61 -1.63 0.82
## depunfair_rural.ses.med4 0.42 0.48 -0.64 1.24
## depmistrt_rural.ses.med2 -0.10 0.63 -1.43 1.02
## depmistrt_rural.ses.med3 0.88 0.50 -0.09 1.88
## depmistrt_rural.ses.med4 0.54 0.53 -0.51 1.64
## depbetray_rural.ses.med2 -0.61 0.69 -2.02 0.65
## depbetray_rural.ses.med3 -0.06 0.55 -1.19 0.91
## depbetray_rural.ses.med4 0.83 0.54 -0.27 1.84
## depcantgo_mostthft_devx2 0.12 0.15 -0.18 0.42
## depcantgo_mostthft_av12x2 0.05 0.03 -0.01 0.12
## depcantgo_mostthft_devx2:rural.ses.med2 0.47 0.33 -0.05 1.28
## depcantgo_mostthft_devx2:rural.ses.med3 -0.18 0.26 -0.79 0.26
## depcantgo_mostthft_devx2:rural.ses.med4 0.02 0.23 -0.39 0.52
## depeffort_mostthft_devx2 0.02 0.16 -0.29 0.36
## depeffort_mostthft_av12x2 -0.03 0.04 -0.11 0.04
## depeffort_mostthft_devx2:rural.ses.med2 -0.21 0.38 -1.05 0.50
## depeffort_mostthft_devx2:rural.ses.med3 0.19 0.27 -0.42 0.67
## depeffort_mostthft_devx2:rural.ses.med4 0.21 0.28 -0.29 0.79
## deplonely_mostthft_devx2 0.26 0.15 -0.06 0.54
## deplonely_mostthft_av12x2 0.02 0.03 -0.05 0.09
## deplonely_mostthft_devx2:rural.ses.med2 0.22 0.36 -0.44 1.00
## deplonely_mostthft_devx2:rural.ses.med3 -0.39 0.38 -1.27 0.19
## deplonely_mostthft_devx2:rural.ses.med4 0.03 0.28 -0.46 0.67
## depblues_mostthft_devx2 0.08 0.18 -0.28 0.44
## depblues_mostthft_av12x2 -0.07 0.04 -0.15 0.02
## depblues_mostthft_devx2:rural.ses.med2 -0.05 0.44 -0.94 0.79
## depblues_mostthft_devx2:rural.ses.med3 -0.01 0.30 -0.69 0.54
## depblues_mostthft_devx2:rural.ses.med4 0.61 0.32 0.02 1.27
## depunfair_mostthft_devx2 -0.01 0.17 -0.34 0.32
## depunfair_mostthft_av12x2 0.00 0.03 -0.06 0.07
## depunfair_mostthft_devx2:rural.ses.med2 0.13 0.34 -0.54 0.86
## depunfair_mostthft_devx2:rural.ses.med3 0.45 0.22 -0.03 0.89
## depunfair_mostthft_devx2:rural.ses.med4 0.37 0.21 -0.04 0.81
## depmistrt_mostthft_devx2 0.08 0.19 -0.32 0.41
## depmistrt_mostthft_av12x2 0.04 0.04 -0.04 0.11
## depmistrt_mostthft_devx2:rural.ses.med2 0.38 0.34 -0.23 1.14
## depmistrt_mostthft_devx2:rural.ses.med3 -0.24 0.32 -0.95 0.29
## depmistrt_mostthft_devx2:rural.ses.med4 0.13 0.27 -0.46 0.63
## depbetray_mostthft_devx2 -0.01 0.19 -0.41 0.34
## depbetray_mostthft_av12x2 0.05 0.04 -0.03 0.14
## depbetray_mostthft_devx2:rural.ses.med2 0.59 0.36 -0.03 1.44
## depbetray_mostthft_devx2:rural.ses.med3 0.44 0.23 0.01 0.91
## depbetray_mostthft_devx2:rural.ses.med4 0.17 0.25 -0.36 0.63
## Rhat Bulk_ESS Tail_ESS
## depcantgo_Intercept 1.00 3021 2863
## depeffort_Intercept 1.00 2696 2714
## deplonely_Intercept 1.00 2519 2366
## depblues_Intercept 1.00 1760 2506
## depunfair_Intercept 1.00 2055 2323
## depmistrt_Intercept 1.00 1813 2655
## depbetray_Intercept 1.00 2320 2794
## depcantgo_rural.ses.med2 1.00 2468 1838
## depcantgo_rural.ses.med3 1.00 2311 2342
## depcantgo_rural.ses.med4 1.00 2632 2539
## depeffort_rural.ses.med2 1.00 3002 2841
## depeffort_rural.ses.med3 1.00 2449 2634
## depeffort_rural.ses.med4 1.00 2838 2574
## deplonely_rural.ses.med2 1.00 2332 1933
## deplonely_rural.ses.med3 1.00 2771 3071
## deplonely_rural.ses.med4 1.00 2410 2913
## depblues_rural.ses.med2 1.00 2519 2558
## depblues_rural.ses.med3 1.00 2646 2422
## depblues_rural.ses.med4 1.00 3560 2635
## depunfair_rural.ses.med2 1.00 2867 2694
## depunfair_rural.ses.med3 1.00 1890 1212
## depunfair_rural.ses.med4 1.00 2484 2901
## depmistrt_rural.ses.med2 1.00 2835 2701
## depmistrt_rural.ses.med3 1.00 2968 2775
## depmistrt_rural.ses.med4 1.00 2489 2472
## depbetray_rural.ses.med2 1.00 3285 2597
## depbetray_rural.ses.med3 1.00 3171 3108
## depbetray_rural.ses.med4 1.00 2677 2753
## depcantgo_mostthft_devx2 1.00 2688 2784
## depcantgo_mostthft_av12x2 1.00 5986 2878
## depcantgo_mostthft_devx2:rural.ses.med2 1.00 1998 1676
## depcantgo_mostthft_devx2:rural.ses.med3 1.00 2344 2495
## depcantgo_mostthft_devx2:rural.ses.med4 1.00 2585 2558
## depeffort_mostthft_devx2 1.00 2803 2883
## depeffort_mostthft_av12x2 1.00 6861 3395
## depeffort_mostthft_devx2:rural.ses.med2 1.00 2406 2480
## depeffort_mostthft_devx2:rural.ses.med3 1.00 2213 1590
## depeffort_mostthft_devx2:rural.ses.med4 1.00 2610 2814
## deplonely_mostthft_devx2 1.00 2026 1973
## deplonely_mostthft_av12x2 1.00 5598 2728
## deplonely_mostthft_devx2:rural.ses.med2 1.00 1908 1699
## deplonely_mostthft_devx2:rural.ses.med3 1.00 2042 2262
## deplonely_mostthft_devx2:rural.ses.med4 1.00 2089 2586
## depblues_mostthft_devx2 1.00 2703 2515
## depblues_mostthft_av12x2 1.00 4457 3202
## depblues_mostthft_devx2:rural.ses.med2 1.00 2440 2755
## depblues_mostthft_devx2:rural.ses.med3 1.00 2364 2233
## depblues_mostthft_devx2:rural.ses.med4 1.00 2982 3007
## depunfair_mostthft_devx2 1.00 2042 1790
## depunfair_mostthft_av12x2 1.00 6583 2730
## depunfair_mostthft_devx2:rural.ses.med2 1.00 2243 2381
## depunfair_mostthft_devx2:rural.ses.med3 1.00 1949 1276
## depunfair_mostthft_devx2:rural.ses.med4 1.00 2359 2657
## depmistrt_mostthft_devx2 1.00 1706 2477
## depmistrt_mostthft_av12x2 1.00 5966 2767
## depmistrt_mostthft_devx2:rural.ses.med2 1.00 2513 2384
## depmistrt_mostthft_devx2:rural.ses.med3 1.00 2275 2185
## depmistrt_mostthft_devx2:rural.ses.med4 1.00 1901 2128
## depbetray_mostthft_devx2 1.00 2254 2623
## depbetray_mostthft_av12x2 1.00 5355 3353
## depbetray_mostthft_devx2:rural.ses.med2 1.00 2526 2134
## depbetray_mostthft_devx2:rural.ses.med3 1.00 2905 2880
## depbetray_mostthft_devx2:rural.ses.med4 1.00 2291 2415
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI
## depcantgo_mostthft_devx21[1] 0.24 0.14 0.03
## depcantgo_mostthft_devx21[2] 0.26 0.14 0.04
## depcantgo_mostthft_devx21[3] 0.24 0.13 0.04
## depcantgo_mostthft_devx21[4] 0.27 0.15 0.04
## depcantgo_mostthft_av12x21[1] 0.12 0.08 0.02
## depcantgo_mostthft_av12x21[2] 0.12 0.07 0.02
## depcantgo_mostthft_av12x21[3] 0.11 0.07 0.02
## depcantgo_mostthft_av12x21[4] 0.14 0.09 0.02
## depcantgo_mostthft_av12x21[5] 0.14 0.08 0.02
## depcantgo_mostthft_av12x21[6] 0.13 0.08 0.02
## depcantgo_mostthft_av12x21[7] 0.12 0.08 0.01
## depcantgo_mostthft_av12x21[8] 0.12 0.08 0.01
## depcantgo_mostthft_devx2:rural.ses.med21[1] 0.20 0.16 0.01
## depcantgo_mostthft_devx2:rural.ses.med21[2] 0.24 0.17 0.01
## depcantgo_mostthft_devx2:rural.ses.med21[3] 0.27 0.17 0.02
## depcantgo_mostthft_devx2:rural.ses.med21[4] 0.29 0.21 0.01
## depcantgo_mostthft_devx2:rural.ses.med31[1] 0.29 0.20 0.01
## depcantgo_mostthft_devx2:rural.ses.med31[2] 0.19 0.17 0.01
## depcantgo_mostthft_devx2:rural.ses.med31[3] 0.21 0.17 0.01
## depcantgo_mostthft_devx2:rural.ses.med31[4] 0.31 0.21 0.01
## depcantgo_mostthft_devx2:rural.ses.med41[1] 0.26 0.19 0.01
## depcantgo_mostthft_devx2:rural.ses.med41[2] 0.22 0.17 0.01
## depcantgo_mostthft_devx2:rural.ses.med41[3] 0.23 0.18 0.01
## depcantgo_mostthft_devx2:rural.ses.med41[4] 0.29 0.21 0.01
## depeffort_mostthft_devx21[1] 0.26 0.15 0.04
## depeffort_mostthft_devx21[2] 0.24 0.14 0.04
## depeffort_mostthft_devx21[3] 0.24 0.14 0.03
## depeffort_mostthft_devx21[4] 0.27 0.15 0.04
## depeffort_mostthft_av12x21[1] 0.13 0.08 0.02
## depeffort_mostthft_av12x21[2] 0.13 0.08 0.02
## depeffort_mostthft_av12x21[3] 0.12 0.08 0.02
## depeffort_mostthft_av12x21[4] 0.12 0.08 0.02
## depeffort_mostthft_av12x21[5] 0.12 0.08 0.02
## depeffort_mostthft_av12x21[6] 0.12 0.08 0.02
## depeffort_mostthft_av12x21[7] 0.13 0.08 0.02
## depeffort_mostthft_av12x21[8] 0.13 0.08 0.02
## depeffort_mostthft_devx2:rural.ses.med21[1] 0.24 0.18 0.01
## depeffort_mostthft_devx2:rural.ses.med21[2] 0.19 0.16 0.00
## depeffort_mostthft_devx2:rural.ses.med21[3] 0.27 0.20 0.01
## depeffort_mostthft_devx2:rural.ses.med21[4] 0.30 0.21 0.01
## depeffort_mostthft_devx2:rural.ses.med31[1] 0.27 0.20 0.01
## depeffort_mostthft_devx2:rural.ses.med31[2] 0.27 0.19 0.01
## depeffort_mostthft_devx2:rural.ses.med31[3] 0.21 0.16 0.01
## depeffort_mostthft_devx2:rural.ses.med31[4] 0.25 0.19 0.01
## depeffort_mostthft_devx2:rural.ses.med41[1] 0.26 0.19 0.01
## depeffort_mostthft_devx2:rural.ses.med41[2] 0.20 0.17 0.01
## depeffort_mostthft_devx2:rural.ses.med41[3] 0.21 0.17 0.01
## depeffort_mostthft_devx2:rural.ses.med41[4] 0.33 0.22 0.01
## deplonely_mostthft_devx21[1] 0.17 0.12 0.02
## deplonely_mostthft_devx21[2] 0.35 0.15 0.08
## deplonely_mostthft_devx21[3] 0.22 0.12 0.04
## deplonely_mostthft_devx21[4] 0.26 0.14 0.05
## deplonely_mostthft_av12x21[1] 0.12 0.08 0.01
## deplonely_mostthft_av12x21[2] 0.12 0.08 0.02
## deplonely_mostthft_av12x21[3] 0.12 0.08 0.02
## deplonely_mostthft_av12x21[4] 0.12 0.08 0.02
## deplonely_mostthft_av12x21[5] 0.12 0.08 0.02
## deplonely_mostthft_av12x21[6] 0.13 0.08 0.01
## deplonely_mostthft_av12x21[7] 0.13 0.08 0.02
## deplonely_mostthft_av12x21[8] 0.13 0.08 0.02
## deplonely_mostthft_devx2:rural.ses.med21[1] 0.22 0.18 0.01
## deplonely_mostthft_devx2:rural.ses.med21[2] 0.22 0.17 0.01
## deplonely_mostthft_devx2:rural.ses.med21[3] 0.24 0.19 0.01
## deplonely_mostthft_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## deplonely_mostthft_devx2:rural.ses.med31[1] 0.26 0.18 0.01
## deplonely_mostthft_devx2:rural.ses.med31[2] 0.15 0.15 0.00
## deplonely_mostthft_devx2:rural.ses.med31[3] 0.21 0.17 0.01
## deplonely_mostthft_devx2:rural.ses.med31[4] 0.38 0.23 0.02
## deplonely_mostthft_devx2:rural.ses.med41[1] 0.26 0.21 0.01
## deplonely_mostthft_devx2:rural.ses.med41[2] 0.22 0.17 0.01
## deplonely_mostthft_devx2:rural.ses.med41[3] 0.22 0.18 0.01
## deplonely_mostthft_devx2:rural.ses.med41[4] 0.30 0.23 0.01
## depblues_mostthft_devx21[1] 0.25 0.15 0.04
## depblues_mostthft_devx21[2] 0.25 0.14 0.04
## depblues_mostthft_devx21[3] 0.23 0.14 0.03
## depblues_mostthft_devx21[4] 0.26 0.15 0.04
## depblues_mostthft_av12x21[1] 0.14 0.09 0.02
## depblues_mostthft_av12x21[2] 0.14 0.08 0.02
## depblues_mostthft_av12x21[3] 0.13 0.08 0.02
## depblues_mostthft_av12x21[4] 0.12 0.08 0.02
## depblues_mostthft_av12x21[5] 0.11 0.07 0.01
## depblues_mostthft_av12x21[6] 0.11 0.07 0.01
## depblues_mostthft_av12x21[7] 0.12 0.08 0.02
## depblues_mostthft_av12x21[8] 0.12 0.08 0.02
## depblues_mostthft_devx2:rural.ses.med21[1] 0.23 0.18 0.01
## depblues_mostthft_devx2:rural.ses.med21[2] 0.21 0.18 0.00
## depblues_mostthft_devx2:rural.ses.med21[3] 0.25 0.20 0.01
## depblues_mostthft_devx2:rural.ses.med21[4] 0.31 0.22 0.01
## depblues_mostthft_devx2:rural.ses.med31[1] 0.26 0.20 0.01
## depblues_mostthft_devx2:rural.ses.med31[2] 0.22 0.18 0.01
## depblues_mostthft_devx2:rural.ses.med31[3] 0.22 0.18 0.01
## depblues_mostthft_devx2:rural.ses.med31[4] 0.30 0.22 0.01
## depblues_mostthft_devx2:rural.ses.med41[1] 0.22 0.17 0.01
## depblues_mostthft_devx2:rural.ses.med41[2] 0.18 0.15 0.01
## depblues_mostthft_devx2:rural.ses.med41[3] 0.17 0.14 0.01
## depblues_mostthft_devx2:rural.ses.med41[4] 0.43 0.22 0.03
## depunfair_mostthft_devx21[1] 0.26 0.15 0.04
## depunfair_mostthft_devx21[2] 0.25 0.14 0.04
## depunfair_mostthft_devx21[3] 0.23 0.14 0.04
## depunfair_mostthft_devx21[4] 0.26 0.15 0.04
## depunfair_mostthft_av12x21[1] 0.12 0.08 0.02
## depunfair_mostthft_av12x21[2] 0.12 0.08 0.02
## depunfair_mostthft_av12x21[3] 0.12 0.08 0.02
## depunfair_mostthft_av12x21[4] 0.12 0.08 0.02
## depunfair_mostthft_av12x21[5] 0.12 0.08 0.01
## depunfair_mostthft_av12x21[6] 0.13 0.08 0.02
## depunfair_mostthft_av12x21[7] 0.13 0.08 0.02
## depunfair_mostthft_av12x21[8] 0.13 0.08 0.02
## depunfair_mostthft_devx2:rural.ses.med21[1] 0.26 0.19 0.01
## depunfair_mostthft_devx2:rural.ses.med21[2] 0.22 0.18 0.01
## depunfair_mostthft_devx2:rural.ses.med21[3] 0.20 0.17 0.01
## depunfair_mostthft_devx2:rural.ses.med21[4] 0.32 0.22 0.01
## depunfair_mostthft_devx2:rural.ses.med31[1] 0.23 0.16 0.01
## depunfair_mostthft_devx2:rural.ses.med31[2] 0.50 0.20 0.07
## depunfair_mostthft_devx2:rural.ses.med31[3] 0.08 0.09 0.00
## depunfair_mostthft_devx2:rural.ses.med31[4] 0.19 0.15 0.01
## depunfair_mostthft_devx2:rural.ses.med41[1] 0.22 0.17 0.01
## depunfair_mostthft_devx2:rural.ses.med41[2] 0.20 0.15 0.01
## depunfair_mostthft_devx2:rural.ses.med41[3] 0.35 0.19 0.03
## depunfair_mostthft_devx2:rural.ses.med41[4] 0.23 0.17 0.01
## depmistrt_mostthft_devx21[1] 0.24 0.14 0.04
## depmistrt_mostthft_devx21[2] 0.22 0.12 0.03
## depmistrt_mostthft_devx21[3] 0.29 0.16 0.04
## depmistrt_mostthft_devx21[4] 0.25 0.15 0.03
## depmistrt_mostthft_av12x21[1] 0.13 0.08 0.02
## depmistrt_mostthft_av12x21[2] 0.12 0.08 0.02
## depmistrt_mostthft_av12x21[3] 0.12 0.08 0.02
## depmistrt_mostthft_av12x21[4] 0.11 0.08 0.01
## depmistrt_mostthft_av12x21[5] 0.12 0.08 0.01
## depmistrt_mostthft_av12x21[6] 0.12 0.08 0.02
## depmistrt_mostthft_av12x21[7] 0.14 0.09 0.02
## depmistrt_mostthft_av12x21[8] 0.13 0.09 0.02
## depmistrt_mostthft_devx2:rural.ses.med21[1] 0.23 0.18 0.01
## depmistrt_mostthft_devx2:rural.ses.med21[2] 0.26 0.18 0.01
## depmistrt_mostthft_devx2:rural.ses.med21[3] 0.22 0.17 0.01
## depmistrt_mostthft_devx2:rural.ses.med21[4] 0.29 0.21 0.01
## depmistrt_mostthft_devx2:rural.ses.med31[1] 0.22 0.18 0.01
## depmistrt_mostthft_devx2:rural.ses.med31[2] 0.23 0.18 0.01
## depmistrt_mostthft_devx2:rural.ses.med31[3] 0.21 0.17 0.01
## depmistrt_mostthft_devx2:rural.ses.med31[4] 0.33 0.23 0.01
## depmistrt_mostthft_devx2:rural.ses.med41[1] 0.23 0.18 0.01
## depmistrt_mostthft_devx2:rural.ses.med41[2] 0.20 0.16 0.01
## depmistrt_mostthft_devx2:rural.ses.med41[3] 0.33 0.22 0.01
## depmistrt_mostthft_devx2:rural.ses.med41[4] 0.25 0.19 0.01
## depbetray_mostthft_devx21[1] 0.26 0.15 0.04
## depbetray_mostthft_devx21[2] 0.24 0.15 0.03
## depbetray_mostthft_devx21[3] 0.24 0.14 0.04
## depbetray_mostthft_devx21[4] 0.26 0.15 0.04
## depbetray_mostthft_av12x21[1] 0.12 0.08 0.02
## depbetray_mostthft_av12x21[2] 0.12 0.08 0.02
## depbetray_mostthft_av12x21[3] 0.11 0.07 0.01
## depbetray_mostthft_av12x21[4] 0.11 0.07 0.02
## depbetray_mostthft_av12x21[5] 0.12 0.08 0.02
## depbetray_mostthft_av12x21[6] 0.13 0.08 0.02
## depbetray_mostthft_av12x21[7] 0.15 0.09 0.02
## depbetray_mostthft_av12x21[8] 0.13 0.08 0.02
## depbetray_mostthft_devx2:rural.ses.med21[1] 0.17 0.15 0.00
## depbetray_mostthft_devx2:rural.ses.med21[2] 0.38 0.19 0.03
## depbetray_mostthft_devx2:rural.ses.med21[3] 0.18 0.14 0.01
## depbetray_mostthft_devx2:rural.ses.med21[4] 0.27 0.19 0.01
## depbetray_mostthft_devx2:rural.ses.med31[1] 0.18 0.15 0.00
## depbetray_mostthft_devx2:rural.ses.med31[2] 0.34 0.18 0.03
## depbetray_mostthft_devx2:rural.ses.med31[3] 0.26 0.16 0.02
## depbetray_mostthft_devx2:rural.ses.med31[4] 0.21 0.16 0.01
## depbetray_mostthft_devx2:rural.ses.med41[1] 0.25 0.19 0.01
## depbetray_mostthft_devx2:rural.ses.med41[2] 0.23 0.17 0.01
## depbetray_mostthft_devx2:rural.ses.med41[3] 0.28 0.20 0.01
## depbetray_mostthft_devx2:rural.ses.med41[4] 0.24 0.19 0.01
## u-95% CI Rhat Bulk_ESS Tail_ESS
## depcantgo_mostthft_devx21[1] 0.57 1.00 4990 2724
## depcantgo_mostthft_devx21[2] 0.56 1.00 4872 2722
## depcantgo_mostthft_devx21[3] 0.54 1.00 6444 3097
## depcantgo_mostthft_devx21[4] 0.60 1.00 6770 2745
## depcantgo_mostthft_av12x21[1] 0.31 1.00 7920 2641
## depcantgo_mostthft_av12x21[2] 0.29 1.00 7153 2643
## depcantgo_mostthft_av12x21[3] 0.29 1.00 7826 2677
## depcantgo_mostthft_av12x21[4] 0.35 1.00 6911 2297
## depcantgo_mostthft_av12x21[5] 0.34 1.00 6524 1787
## depcantgo_mostthft_av12x21[6] 0.33 1.00 6452 2930
## depcantgo_mostthft_av12x21[7] 0.31 1.00 6837 2332
## depcantgo_mostthft_av12x21[8] 0.31 1.00 7081 2405
## depcantgo_mostthft_devx2:rural.ses.med21[1] 0.60 1.00 5283 3005
## depcantgo_mostthft_devx2:rural.ses.med21[2] 0.63 1.00 4539 2374
## depcantgo_mostthft_devx2:rural.ses.med21[3] 0.66 1.00 3607 2805
## depcantgo_mostthft_devx2:rural.ses.med21[4] 0.74 1.00 3506 2489
## depcantgo_mostthft_devx2:rural.ses.med31[1] 0.74 1.00 4124 2162
## depcantgo_mostthft_devx2:rural.ses.med31[2] 0.60 1.00 4295 2687
## depcantgo_mostthft_devx2:rural.ses.med31[3] 0.63 1.00 4959 2573
## depcantgo_mostthft_devx2:rural.ses.med31[4] 0.76 1.00 5142 2725
## depcantgo_mostthft_devx2:rural.ses.med41[1] 0.71 1.00 5638 2603
## depcantgo_mostthft_devx2:rural.ses.med41[2] 0.64 1.00 5407 1948
## depcantgo_mostthft_devx2:rural.ses.med41[3] 0.67 1.00 3820 3090
## depcantgo_mostthft_devx2:rural.ses.med41[4] 0.75 1.00 4377 2452
## depeffort_mostthft_devx21[1] 0.61 1.00 6631 2902
## depeffort_mostthft_devx21[2] 0.55 1.00 5710 2767
## depeffort_mostthft_devx21[3] 0.55 1.00 6556 2694
## depeffort_mostthft_devx21[4] 0.60 1.00 6022 2865
## depeffort_mostthft_av12x21[1] 0.32 1.00 7131 2462
## depeffort_mostthft_av12x21[2] 0.33 1.00 8238 2491
## depeffort_mostthft_av12x21[3] 0.32 1.00 7229 1990
## depeffort_mostthft_av12x21[4] 0.30 1.00 6459 3188
## depeffort_mostthft_av12x21[5] 0.30 1.00 6290 2526
## depeffort_mostthft_av12x21[6] 0.31 1.00 7026 3181
## depeffort_mostthft_av12x21[7] 0.31 1.00 7257 3156
## depeffort_mostthft_av12x21[8] 0.32 1.00 6547 2852
## depeffort_mostthft_devx2:rural.ses.med21[1] 0.66 1.00 4280 1800
## depeffort_mostthft_devx2:rural.ses.med21[2] 0.61 1.00 4740 2546
## depeffort_mostthft_devx2:rural.ses.med21[3] 0.72 1.00 4681 2699
## depeffort_mostthft_devx2:rural.ses.med21[4] 0.76 1.00 4717 2547
## depeffort_mostthft_devx2:rural.ses.med31[1] 0.71 1.00 4894 2717
## depeffort_mostthft_devx2:rural.ses.med31[2] 0.69 1.00 4074 1865
## depeffort_mostthft_devx2:rural.ses.med31[3] 0.61 1.00 4434 2455
## depeffort_mostthft_devx2:rural.ses.med31[4] 0.72 1.00 3673 1922
## depeffort_mostthft_devx2:rural.ses.med41[1] 0.71 1.00 6348 2269
## depeffort_mostthft_devx2:rural.ses.med41[2] 0.60 1.00 5944 2789
## depeffort_mostthft_devx2:rural.ses.med41[3] 0.64 1.00 5024 2377
## depeffort_mostthft_devx2:rural.ses.med41[4] 0.78 1.00 4550 2752
## deplonely_mostthft_devx21[1] 0.46 1.00 4238 2598
## deplonely_mostthft_devx21[2] 0.64 1.00 4788 2431
## deplonely_mostthft_devx21[3] 0.48 1.00 5677 3137
## deplonely_mostthft_devx21[4] 0.55 1.00 6821 2573
## deplonely_mostthft_av12x21[1] 0.32 1.00 7354 2681
## deplonely_mostthft_av12x21[2] 0.32 1.00 7891 2204
## deplonely_mostthft_av12x21[3] 0.32 1.00 6121 2377
## deplonely_mostthft_av12x21[4] 0.32 1.00 7284 2205
## deplonely_mostthft_av12x21[5] 0.32 1.00 7541 2378
## deplonely_mostthft_av12x21[6] 0.33 1.00 6964 2324
## deplonely_mostthft_av12x21[7] 0.32 1.00 6904 3196
## deplonely_mostthft_av12x21[8] 0.32 1.00 6325 2815
## deplonely_mostthft_devx2:rural.ses.med21[1] 0.66 1.00 4285 2338
## deplonely_mostthft_devx2:rural.ses.med21[2] 0.64 1.00 4498 2585
## deplonely_mostthft_devx2:rural.ses.med21[3] 0.68 1.00 4054 2827
## deplonely_mostthft_devx2:rural.ses.med21[4] 0.78 1.00 4012 2534
## deplonely_mostthft_devx2:rural.ses.med31[1] 0.66 1.00 5215 2692
## deplonely_mostthft_devx2:rural.ses.med31[2] 0.57 1.00 3597 2071
## deplonely_mostthft_devx2:rural.ses.med31[3] 0.62 1.00 4071 2544
## deplonely_mostthft_devx2:rural.ses.med31[4] 0.84 1.00 3834 2694
## deplonely_mostthft_devx2:rural.ses.med41[1] 0.73 1.00 3676 2735
## deplonely_mostthft_devx2:rural.ses.med41[2] 0.64 1.00 5742 2247
## deplonely_mostthft_devx2:rural.ses.med41[3] 0.64 1.00 4254 2868
## deplonely_mostthft_devx2:rural.ses.med41[4] 0.81 1.00 3132 2809
## depblues_mostthft_devx21[1] 0.57 1.00 5926 2458
## depblues_mostthft_devx21[2] 0.56 1.00 5953 2908
## depblues_mostthft_devx21[3] 0.56 1.00 4889 2096
## depblues_mostthft_devx21[4] 0.60 1.00 6853 2994
## depblues_mostthft_av12x21[1] 0.35 1.00 6397 2236
## depblues_mostthft_av12x21[2] 0.33 1.00 7684 2575
## depblues_mostthft_av12x21[3] 0.33 1.00 5914 2366
## depblues_mostthft_av12x21[4] 0.32 1.00 6108 2471
## depblues_mostthft_av12x21[5] 0.28 1.00 5876 2470
## depblues_mostthft_av12x21[6] 0.29 1.00 7835 2873
## depblues_mostthft_av12x21[7] 0.30 1.00 7360 2839
## depblues_mostthft_av12x21[8] 0.32 1.00 5785 2527
## depblues_mostthft_devx2:rural.ses.med21[1] 0.67 1.00 5813 2202
## depblues_mostthft_devx2:rural.ses.med21[2] 0.67 1.00 3864 2767
## depblues_mostthft_devx2:rural.ses.med21[3] 0.73 1.00 3613 2285
## depblues_mostthft_devx2:rural.ses.med21[4] 0.78 1.00 5712 2908
## depblues_mostthft_devx2:rural.ses.med31[1] 0.71 1.00 4328 2504
## depblues_mostthft_devx2:rural.ses.med31[2] 0.64 1.00 4453 2357
## depblues_mostthft_devx2:rural.ses.med31[3] 0.66 1.00 4621 2180
## depblues_mostthft_devx2:rural.ses.med31[4] 0.78 1.00 4330 2380
## depblues_mostthft_devx2:rural.ses.med41[1] 0.62 1.00 4617 2495
## depblues_mostthft_devx2:rural.ses.med41[2] 0.56 1.00 4888 2372
## depblues_mostthft_devx2:rural.ses.med41[3] 0.54 1.00 4937 2507
## depblues_mostthft_devx2:rural.ses.med41[4] 0.81 1.00 4634 2573
## depunfair_mostthft_devx21[1] 0.59 1.00 5302 2405
## depunfair_mostthft_devx21[2] 0.57 1.00 4461 2376
## depunfair_mostthft_devx21[3] 0.55 1.00 5933 3109
## depunfair_mostthft_devx21[4] 0.59 1.00 6347 3045
## depunfair_mostthft_av12x21[1] 0.32 1.00 6058 2609
## depunfair_mostthft_av12x21[2] 0.31 1.00 7198 2545
## depunfair_mostthft_av12x21[3] 0.30 1.00 5958 2404
## depunfair_mostthft_av12x21[4] 0.31 1.00 6255 2728
## depunfair_mostthft_av12x21[5] 0.32 1.00 6107 2070
## depunfair_mostthft_av12x21[6] 0.33 1.01 6294 2550
## depunfair_mostthft_av12x21[7] 0.33 1.00 6107 2855
## depunfair_mostthft_av12x21[8] 0.33 1.00 5448 2728
## depunfair_mostthft_devx2:rural.ses.med21[1] 0.69 1.00 5270 2305
## depunfair_mostthft_devx2:rural.ses.med21[2] 0.66 1.00 4540 2413
## depunfair_mostthft_devx2:rural.ses.med21[3] 0.62 1.00 4676 2742
## depunfair_mostthft_devx2:rural.ses.med21[4] 0.80 1.00 4681 2936
## depunfair_mostthft_devx2:rural.ses.med31[1] 0.61 1.00 4364 2430
## depunfair_mostthft_devx2:rural.ses.med31[2] 0.84 1.00 2701 1325
## depunfair_mostthft_devx2:rural.ses.med31[3] 0.32 1.00 3444 2147
## depunfair_mostthft_devx2:rural.ses.med31[4] 0.56 1.00 4678 2935
## depunfair_mostthft_devx2:rural.ses.med41[1] 0.63 1.00 4938 2387
## depunfair_mostthft_devx2:rural.ses.med41[2] 0.58 1.00 5714 2093
## depunfair_mostthft_devx2:rural.ses.med41[3] 0.75 1.00 4431 2767
## depunfair_mostthft_devx2:rural.ses.med41[4] 0.63 1.00 5959 2455
## depmistrt_mostthft_devx21[1] 0.58 1.00 5087 2876
## depmistrt_mostthft_devx21[2] 0.51 1.00 7196 2846
## depmistrt_mostthft_devx21[3] 0.64 1.00 2661 2634
## depmistrt_mostthft_devx21[4] 0.59 1.00 5202 2785
## depmistrt_mostthft_av12x21[1] 0.32 1.00 6903 1854
## depmistrt_mostthft_av12x21[2] 0.32 1.00 7020 2158
## depmistrt_mostthft_av12x21[3] 0.31 1.00 7743 2359
## depmistrt_mostthft_av12x21[4] 0.31 1.00 7262 2407
## depmistrt_mostthft_av12x21[5] 0.32 1.00 6054 2347
## depmistrt_mostthft_av12x21[6] 0.31 1.00 7138 2306
## depmistrt_mostthft_av12x21[7] 0.34 1.00 6156 2598
## depmistrt_mostthft_av12x21[8] 0.34 1.00 6339 2753
## depmistrt_mostthft_devx2:rural.ses.med21[1] 0.65 1.00 5689 2432
## depmistrt_mostthft_devx2:rural.ses.med21[2] 0.68 1.00 4709 2163
## depmistrt_mostthft_devx2:rural.ses.med21[3] 0.63 1.00 4637 3189
## depmistrt_mostthft_devx2:rural.ses.med21[4] 0.75 1.00 4633 2922
## depmistrt_mostthft_devx2:rural.ses.med31[1] 0.66 1.00 4352 2603
## depmistrt_mostthft_devx2:rural.ses.med31[2] 0.66 1.00 5937 2810
## depmistrt_mostthft_devx2:rural.ses.med31[3] 0.64 1.00 5231 3052
## depmistrt_mostthft_devx2:rural.ses.med31[4] 0.81 1.00 4463 2794
## depmistrt_mostthft_devx2:rural.ses.med41[1] 0.66 1.00 6355 2423
## depmistrt_mostthft_devx2:rural.ses.med41[2] 0.62 1.00 5235 2552
## depmistrt_mostthft_devx2:rural.ses.med41[3] 0.78 1.00 2682 2938
## depmistrt_mostthft_devx2:rural.ses.med41[4] 0.70 1.00 3934 2927
## depbetray_mostthft_devx21[1] 0.61 1.00 5557 2838
## depbetray_mostthft_devx21[2] 0.58 1.00 5196 2651
## depbetray_mostthft_devx21[3] 0.54 1.00 5798 2685
## depbetray_mostthft_devx21[4] 0.59 1.00 6391 3021
## depbetray_mostthft_av12x21[1] 0.30 1.00 7472 2384
## depbetray_mostthft_av12x21[2] 0.32 1.00 7347 2113
## depbetray_mostthft_av12x21[3] 0.28 1.00 6345 1971
## depbetray_mostthft_av12x21[4] 0.29 1.00 7631 3085
## depbetray_mostthft_av12x21[5] 0.32 1.00 7235 2539
## depbetray_mostthft_av12x21[6] 0.33 1.00 7050 2543
## depbetray_mostthft_av12x21[7] 0.37 1.00 5657 2935
## depbetray_mostthft_av12x21[8] 0.34 1.00 6322 2668
## depbetray_mostthft_devx2:rural.ses.med21[1] 0.55 1.00 4973 2580
## depbetray_mostthft_devx2:rural.ses.med21[2] 0.76 1.00 4034 2081
## depbetray_mostthft_devx2:rural.ses.med21[3] 0.52 1.00 3716 2676
## depbetray_mostthft_devx2:rural.ses.med21[4] 0.71 1.00 3837 2756
## depbetray_mostthft_devx2:rural.ses.med31[1] 0.55 1.00 4339 2023
## depbetray_mostthft_devx2:rural.ses.med31[2] 0.72 1.00 5077 2481
## depbetray_mostthft_devx2:rural.ses.med31[3] 0.63 1.00 4737 2655
## depbetray_mostthft_devx2:rural.ses.med31[4] 0.58 1.00 5951 2332
## depbetray_mostthft_devx2:rural.ses.med41[1] 0.68 1.00 5498 2387
## depbetray_mostthft_devx2:rural.ses.med41[2] 0.65 1.00 6608 2611
## depbetray_mostthft_devx2:rural.ses.med41[3] 0.72 1.00 4444 2822
## depbetray_mostthft_devx2:rural.ses.med41[4] 0.70 1.00 5539 2730
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.2.6.4 Prior summary
out.chg.alldepress.stthft.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostthft_av12x2
## normal(0, 0.25) b mostthft_devx2
## normal(0, 1) b mostthft_devx2:rural.ses.med2
## normal(0, 1) b mostthft_devx2:rural.ses.med3
## normal(0, 1) b mostthft_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostthft_av12x21
## dirichlet(1) simo mostthft_devx2:rural.ses.med21
## dirichlet(1) simo mostthft_devx2:rural.ses.med31
## dirichlet(1) simo mostthft_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostthft_devx21
## group resp dpar nlpar lb ub source
## default
## depbetray user
## depbetray user
## depbetray user
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depblues user
## depblues user
## depblues user
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depcantgo user
## depcantgo user
## depcantgo user
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depeffort user
## depeffort user
## depeffort user
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## deplonely user
## deplonely user
## deplonely user
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## depmistrt user
## depmistrt user
## depmistrt user
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depunfair user
## depunfair user
## depunfair user
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## default
## depbetray user
## depblues user
## depcantgo user
## depeffort user
## deplonely user
## depmistrt user
## depunfair user
## depbetray 0 default
## depblues 0 default
## depcantgo 0 default
## depeffort 0 default
## deplonely 0 default
## depmistrt 0 default
## depunfair 0 default
## id depbetray 0 (vectorized)
## id depbetray 0 (vectorized)
## id depblues 0 (vectorized)
## id depblues 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depeffort 0 (vectorized)
## id depeffort 0 (vectorized)
## id deplonely 0 (vectorized)
## id deplonely 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depunfair 0 (vectorized)
## id depunfair 0 (vectorized)
## depbetray user
## depbetray default
## depbetray default
## depbetray default
## depbetray user
## depblues user
## depblues default
## depblues default
## depblues default
## depblues user
## depcantgo user
## depcantgo default
## depcantgo default
## depcantgo default
## depcantgo user
## depeffort user
## depeffort default
## depeffort default
## depeffort default
## depeffort user
## deplonely user
## deplonely default
## deplonely default
## deplonely default
## deplonely user
## depmistrt user
## depmistrt default
## depmistrt default
## depmistrt default
## depmistrt user
## depunfair user
## depunfair default
## depunfair default
## depunfair default
## depunfair user8.1.2.7 Corr X Community: stmug & negative emotions
#Bivariate Change: negative emotions items ~ mo(stmug)
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names),
set_prior('normal(0, 0.25)', class = 'b', coef = 'mostmug_devx2',
resp = depdv_names), #norm(0,1)/4 thresholds
set_prior('normal(0, 0.125)', class = 'b', coef = 'mostmug_av12x2',
resp = depdv_names), #norm(0,1)/8 thresh
set_prior('dirichlet(2, 2, 2, 2)', class = 'simo', coef = 'mostmug_devx21',
resp = depdv_names),
set_prior('dirichlet(2, 2, 2, 2, 2, 2, 2, 2)', class = 'simo', coef = 'mostmug_av12x21',
resp = depdv_names)
)
chg.alldepress.stmug.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) +
rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id),
data = stress.long,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stmug_comm_fit",
file_refit = "on_change"
)
out.chg.alldepress.stmug.comm.fit <- ppchecks(chg.alldepress.stmug.comm.fit)8.1.2.7.1 Coefficient plot (intervals)
out.chg.alldepress.stmug.comm.fit[[11]]
8.1.2.7.2 PPcheck (density)
p1 <- out.chg.alldepress.stmug.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stmug.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stmug.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stmug.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stmug.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stmug.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stmug.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
8.1.2.7.3 Fit summary
out.chg.alldepress.stmug.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## depeffort ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## deplonely ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## depblues ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## depunfair ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## depmistrt ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## depbetray ~ 1 + mo(stmug_devx2) + mo(stmug_av12x2) + rural.ses.med + mo(stmug_devx2):rural.ses.med + (1 | id)
## Data: stress.long (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.31 0.20 0.01 0.71 1.00 529
## sd(depeffort_Intercept) 0.46 0.27 0.02 1.01 1.01 418
## sd(deplonely_Intercept) 0.40 0.23 0.02 0.86 1.01 469
## sd(depblues_Intercept) 0.73 0.32 0.08 1.30 1.00 499
## sd(depunfair_Intercept) 0.22 0.16 0.01 0.58 1.00 840
## sd(depmistrt_Intercept) 0.31 0.21 0.02 0.78 1.01 838
## sd(depbetray_Intercept) 0.47 0.27 0.03 1.04 1.01 512
## Tail_ESS
## sd(depcantgo_Intercept) 1245
## sd(depeffort_Intercept) 1306
## sd(deplonely_Intercept) 926
## sd(depblues_Intercept) 624
## sd(depunfair_Intercept) 1223
## sd(depmistrt_Intercept) 1546
## sd(depbetray_Intercept) 1427
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_Intercept -0.71 0.39 -1.52 0.01
## depeffort_Intercept -2.09 0.40 -2.89 -1.28
## deplonely_Intercept -1.08 0.39 -1.78 -0.24
## depblues_Intercept -2.20 0.42 -3.07 -1.38
## depunfair_Intercept -1.54 0.35 -2.24 -0.86
## depmistrt_Intercept -2.06 0.44 -2.84 -1.15
## depbetray_Intercept -2.46 0.44 -3.30 -1.52
## depcantgo_rural.ses.med2 0.06 0.67 -1.17 1.40
## depcantgo_rural.ses.med3 -0.05 0.55 -1.24 1.00
## depcantgo_rural.ses.med4 0.03 0.50 -0.99 1.02
## depeffort_rural.ses.med2 0.60 0.56 -0.50 1.72
## depeffort_rural.ses.med3 0.07 0.64 -1.34 1.21
## depeffort_rural.ses.med4 0.45 0.57 -0.70 1.57
## deplonely_rural.ses.med2 -0.95 0.46 -1.91 -0.11
## deplonely_rural.ses.med3 -0.43 0.50 -1.36 0.71
## deplonely_rural.ses.med4 -0.14 0.45 -1.07 0.71
## depblues_rural.ses.med2 0.71 0.49 -0.31 1.68
## depblues_rural.ses.med3 0.15 0.60 -1.15 1.32
## depblues_rural.ses.med4 0.50 0.57 -0.65 1.65
## depunfair_rural.ses.med2 0.35 0.50 -0.71 1.32
## depunfair_rural.ses.med3 0.26 0.59 -1.12 1.27
## depunfair_rural.ses.med4 0.48 0.55 -0.72 1.41
## depmistrt_rural.ses.med2 0.55 0.54 -0.45 1.75
## depmistrt_rural.ses.med3 0.85 0.54 -0.16 2.00
## depmistrt_rural.ses.med4 -0.08 0.60 -1.45 0.94
## depbetray_rural.ses.med2 0.41 0.59 -0.71 1.71
## depbetray_rural.ses.med3 -0.02 0.61 -1.32 1.07
## depbetray_rural.ses.med4 0.14 0.67 -1.33 1.27
## depcantgo_mostmug_devx2 0.08 0.17 -0.26 0.41
## depcantgo_mostmug_av12x2 0.02 0.04 -0.05 0.09
## depcantgo_mostmug_devx2:rural.ses.med2 -0.01 0.44 -1.13 0.54
## depcantgo_mostmug_devx2:rural.ses.med3 0.09 0.31 -0.44 0.85
## depcantgo_mostmug_devx2:rural.ses.med4 0.02 0.24 -0.45 0.53
## depeffort_mostmug_devx2 0.07 0.18 -0.30 0.41
## depeffort_mostmug_av12x2 -0.01 0.04 -0.10 0.08
## depeffort_mostmug_devx2:rural.ses.med2 -0.31 0.28 -0.86 0.26
## depeffort_mostmug_devx2:rural.ses.med3 0.16 0.32 -0.53 0.76
## depeffort_mostmug_devx2:rural.ses.med4 0.03 0.34 -0.75 0.60
## deplonely_mostmug_devx2 -0.02 0.18 -0.37 0.32
## deplonely_mostmug_av12x2 0.08 0.04 0.01 0.15
## deplonely_mostmug_devx2:rural.ses.med2 0.38 0.26 -0.11 0.92
## deplonely_mostmug_devx2:rural.ses.med3 0.23 0.29 -0.43 0.76
## deplonely_mostmug_devx2:rural.ses.med4 0.37 0.27 -0.12 0.94
## depblues_mostmug_devx2 -0.05 0.18 -0.41 0.29
## depblues_mostmug_av12x2 -0.00 0.05 -0.09 0.09
## depblues_mostmug_devx2:rural.ses.med2 -0.54 0.43 -1.49 0.22
## depblues_mostmug_devx2:rural.ses.med3 0.10 0.31 -0.55 0.72
## depblues_mostmug_devx2:rural.ses.med4 0.04 0.31 -0.65 0.60
## depunfair_mostmug_devx2 -0.06 0.15 -0.36 0.24
## depunfair_mostmug_av12x2 0.05 0.04 -0.03 0.13
## depunfair_mostmug_devx2:rural.ses.med2 -0.02 0.25 -0.52 0.47
## depunfair_mostmug_devx2:rural.ses.med3 0.28 0.36 -0.27 1.17
## depunfair_mostmug_devx2:rural.ses.med4 0.25 0.24 -0.23 0.72
## depmistrt_mostmug_devx2 -0.09 0.18 -0.45 0.25
## depmistrt_mostmug_av12x2 0.11 0.04 0.03 0.20
## depmistrt_mostmug_devx2:rural.ses.med2 0.02 0.26 -0.51 0.56
## depmistrt_mostmug_devx2:rural.ses.med3 -0.23 0.33 -1.00 0.33
## depmistrt_mostmug_devx2:rural.ses.med4 0.43 0.25 -0.04 0.95
## depbetray_mostmug_devx2 -0.03 0.18 -0.42 0.30
## depbetray_mostmug_av12x2 0.12 0.05 0.03 0.22
## depbetray_mostmug_devx2:rural.ses.med2 0.06 0.28 -0.54 0.57
## depbetray_mostmug_devx2:rural.ses.med3 0.38 0.28 -0.15 0.95
## depbetray_mostmug_devx2:rural.ses.med4 0.44 0.26 -0.05 0.98
## Rhat Bulk_ESS Tail_ESS
## depcantgo_Intercept 1.00 1506 2326
## depeffort_Intercept 1.00 2361 2417
## deplonely_Intercept 1.00 2020 2451
## depblues_Intercept 1.00 2355 2530
## depunfair_Intercept 1.00 3014 2895
## depmistrt_Intercept 1.00 2564 2994
## depbetray_Intercept 1.00 2764 2671
## depcantgo_rural.ses.med2 1.00 1206 2781
## depcantgo_rural.ses.med3 1.00 1998 2139
## depcantgo_rural.ses.med4 1.00 2374 2598
## depeffort_rural.ses.med2 1.00 2998 2676
## depeffort_rural.ses.med3 1.00 2768 2516
## depeffort_rural.ses.med4 1.00 2728 3059
## deplonely_rural.ses.med2 1.00 2852 2585
## deplonely_rural.ses.med3 1.00 1989 1498
## deplonely_rural.ses.med4 1.00 2660 2549
## depblues_rural.ses.med2 1.00 3245 2181
## depblues_rural.ses.med3 1.00 3156 2827
## depblues_rural.ses.med4 1.00 3253 3111
## depunfair_rural.ses.med2 1.00 2687 2354
## depunfair_rural.ses.med3 1.00 2843 2384
## depunfair_rural.ses.med4 1.00 2563 2550
## depmistrt_rural.ses.med2 1.00 2877 2711
## depmistrt_rural.ses.med3 1.00 3561 3214
## depmistrt_rural.ses.med4 1.00 3080 2577
## depbetray_rural.ses.med2 1.00 3053 2436
## depbetray_rural.ses.med3 1.00 3339 2752
## depbetray_rural.ses.med4 1.00 3098 2604
## depcantgo_mostmug_devx2 1.00 1345 2207
## depcantgo_mostmug_av12x2 1.00 6065 3055
## depcantgo_mostmug_devx2:rural.ses.med2 1.01 1064 1906
## depcantgo_mostmug_devx2:rural.ses.med3 1.00 1774 1998
## depcantgo_mostmug_devx2:rural.ses.med4 1.00 2134 2746
## depeffort_mostmug_devx2 1.00 2373 2508
## depeffort_mostmug_av12x2 1.00 6867 2719
## depeffort_mostmug_devx2:rural.ses.med2 1.00 2531 2431
## depeffort_mostmug_devx2:rural.ses.med3 1.00 2444 2081
## depeffort_mostmug_devx2:rural.ses.med4 1.00 2279 2028
## deplonely_mostmug_devx2 1.00 1870 2617
## deplonely_mostmug_av12x2 1.00 5943 2881
## deplonely_mostmug_devx2:rural.ses.med2 1.00 2164 2466
## deplonely_mostmug_devx2:rural.ses.med3 1.00 1764 1363
## deplonely_mostmug_devx2:rural.ses.med4 1.00 2418 2482
## depblues_mostmug_devx2 1.00 2979 2647
## depblues_mostmug_av12x2 1.00 5063 3228
## depblues_mostmug_devx2:rural.ses.med2 1.00 2521 2151
## depblues_mostmug_devx2:rural.ses.med3 1.00 2811 2438
## depblues_mostmug_devx2:rural.ses.med4 1.00 2781 2164
## depunfair_mostmug_devx2 1.00 2825 2711
## depunfair_mostmug_av12x2 1.00 6420 3428
## depunfair_mostmug_devx2:rural.ses.med2 1.00 2584 2120
## depunfair_mostmug_devx2:rural.ses.med3 1.00 2388 2056
## depunfair_mostmug_devx2:rural.ses.med4 1.00 2800 2535
## depmistrt_mostmug_devx2 1.00 2638 2674
## depmistrt_mostmug_av12x2 1.00 5885 2849
## depmistrt_mostmug_devx2:rural.ses.med2 1.00 2547 2772
## depmistrt_mostmug_devx2:rural.ses.med3 1.00 2836 2171
## depmistrt_mostmug_devx2:rural.ses.med4 1.00 2948 2626
## depbetray_mostmug_devx2 1.00 2503 2690
## depbetray_mostmug_av12x2 1.00 5952 3166
## depbetray_mostmug_devx2:rural.ses.med2 1.00 2692 2898
## depbetray_mostmug_devx2:rural.ses.med3 1.00 2818 2615
## depbetray_mostmug_devx2:rural.ses.med4 1.00 2877 3015
##
## Monotonic Simplex Parameters:
## Estimate Est.Error l-95% CI u-95% CI
## depcantgo_mostmug_devx21[1] 0.25 0.14 0.04 0.56
## depcantgo_mostmug_devx21[2] 0.28 0.15 0.04 0.60
## depcantgo_mostmug_devx21[3] 0.22 0.13 0.04 0.52
## depcantgo_mostmug_devx21[4] 0.25 0.14 0.04 0.58
## depcantgo_mostmug_av12x21[1] 0.12 0.07 0.02 0.30
## depcantgo_mostmug_av12x21[2] 0.13 0.08 0.02 0.31
## depcantgo_mostmug_av12x21[3] 0.13 0.08 0.02 0.31
## depcantgo_mostmug_av12x21[4] 0.12 0.08 0.02 0.31
## depcantgo_mostmug_av12x21[5] 0.12 0.08 0.02 0.30
## depcantgo_mostmug_av12x21[6] 0.12 0.08 0.02 0.31
## depcantgo_mostmug_av12x21[7] 0.13 0.08 0.02 0.34
## depcantgo_mostmug_av12x21[8] 0.13 0.08 0.02 0.33
## depcantgo_mostmug_devx2:rural.ses.med21[1] 0.20 0.17 0.01 0.63
## depcantgo_mostmug_devx2:rural.ses.med21[2] 0.33 0.26 0.00 0.82
## depcantgo_mostmug_devx2:rural.ses.med21[3] 0.17 0.15 0.01 0.55
## depcantgo_mostmug_devx2:rural.ses.med21[4] 0.30 0.26 0.01 0.88
## depcantgo_mostmug_devx2:rural.ses.med31[1] 0.27 0.20 0.01 0.73
## depcantgo_mostmug_devx2:rural.ses.med31[2] 0.21 0.17 0.01 0.64
## depcantgo_mostmug_devx2:rural.ses.med31[3] 0.21 0.17 0.01 0.64
## depcantgo_mostmug_devx2:rural.ses.med31[4] 0.32 0.22 0.01 0.79
## depcantgo_mostmug_devx2:rural.ses.med41[1] 0.26 0.19 0.01 0.70
## depcantgo_mostmug_devx2:rural.ses.med41[2] 0.23 0.18 0.01 0.66
## depcantgo_mostmug_devx2:rural.ses.med41[3] 0.23 0.18 0.01 0.67
## depcantgo_mostmug_devx2:rural.ses.med41[4] 0.28 0.20 0.01 0.73
## depeffort_mostmug_devx21[1] 0.25 0.14 0.04 0.58
## depeffort_mostmug_devx21[2] 0.24 0.14 0.04 0.55
## depeffort_mostmug_devx21[3] 0.26 0.15 0.04 0.59
## depeffort_mostmug_devx21[4] 0.25 0.15 0.04 0.58
## depeffort_mostmug_av12x21[1] 0.12 0.08 0.02 0.32
## depeffort_mostmug_av12x21[2] 0.12 0.08 0.02 0.31
## depeffort_mostmug_av12x21[3] 0.12 0.08 0.02 0.32
## depeffort_mostmug_av12x21[4] 0.12 0.08 0.01 0.32
## depeffort_mostmug_av12x21[5] 0.12 0.08 0.02 0.32
## depeffort_mostmug_av12x21[6] 0.13 0.08 0.02 0.33
## depeffort_mostmug_av12x21[7] 0.13 0.08 0.02 0.32
## depeffort_mostmug_av12x21[8] 0.13 0.08 0.02 0.33
## depeffort_mostmug_devx2:rural.ses.med21[1] 0.22 0.17 0.01 0.64
## depeffort_mostmug_devx2:rural.ses.med21[2] 0.33 0.21 0.02 0.76
## depeffort_mostmug_devx2:rural.ses.med21[3] 0.21 0.17 0.01 0.61
## depeffort_mostmug_devx2:rural.ses.med21[4] 0.24 0.19 0.01 0.70
## depeffort_mostmug_devx2:rural.ses.med31[1] 0.25 0.19 0.01 0.70
## depeffort_mostmug_devx2:rural.ses.med31[2] 0.27 0.19 0.01 0.70
## depeffort_mostmug_devx2:rural.ses.med31[3] 0.20 0.17 0.01 0.64
## depeffort_mostmug_devx2:rural.ses.med31[4] 0.28 0.21 0.01 0.77
## depeffort_mostmug_devx2:rural.ses.med41[1] 0.24 0.19 0.01 0.69
## depeffort_mostmug_devx2:rural.ses.med41[2] 0.21 0.16 0.01 0.60
## depeffort_mostmug_devx2:rural.ses.med41[3] 0.26 0.20 0.01 0.71
## depeffort_mostmug_devx2:rural.ses.med41[4] 0.29 0.22 0.01 0.80
## deplonely_mostmug_devx21[1] 0.27 0.15 0.04 0.59
## deplonely_mostmug_devx21[2] 0.23 0.13 0.03 0.54
## deplonely_mostmug_devx21[3] 0.24 0.14 0.04 0.56
## deplonely_mostmug_devx21[4] 0.26 0.15 0.04 0.59
## deplonely_mostmug_av12x21[1] 0.12 0.07 0.02 0.29
## deplonely_mostmug_av12x21[2] 0.12 0.07 0.02 0.30
## deplonely_mostmug_av12x21[3] 0.13 0.08 0.02 0.33
## deplonely_mostmug_av12x21[4] 0.12 0.08 0.02 0.32
## deplonely_mostmug_av12x21[5] 0.13 0.08 0.02 0.33
## deplonely_mostmug_av12x21[6] 0.14 0.09 0.02 0.35
## deplonely_mostmug_av12x21[7] 0.12 0.08 0.02 0.32
## deplonely_mostmug_av12x21[8] 0.12 0.08 0.01 0.30
## deplonely_mostmug_devx2:rural.ses.med21[1] 0.18 0.16 0.01 0.59
## deplonely_mostmug_devx2:rural.ses.med21[2] 0.21 0.16 0.01 0.60
## deplonely_mostmug_devx2:rural.ses.med21[3] 0.28 0.18 0.02 0.69
## deplonely_mostmug_devx2:rural.ses.med21[4] 0.33 0.21 0.02 0.76
## deplonely_mostmug_devx2:rural.ses.med31[1] 0.21 0.17 0.01 0.66
## deplonely_mostmug_devx2:rural.ses.med31[2] 0.17 0.15 0.01 0.55
## deplonely_mostmug_devx2:rural.ses.med31[3] 0.36 0.21 0.01 0.77
## deplonely_mostmug_devx2:rural.ses.med31[4] 0.27 0.20 0.01 0.72
## deplonely_mostmug_devx2:rural.ses.med41[1] 0.19 0.16 0.01 0.59
## deplonely_mostmug_devx2:rural.ses.med41[2] 0.20 0.15 0.01 0.58
## deplonely_mostmug_devx2:rural.ses.med41[3] 0.30 0.19 0.02 0.71
## deplonely_mostmug_devx2:rural.ses.med41[4] 0.31 0.20 0.01 0.74
## depblues_mostmug_devx21[1] 0.25 0.14 0.04 0.58
## depblues_mostmug_devx21[2] 0.23 0.13 0.04 0.54
## depblues_mostmug_devx21[3] 0.25 0.14 0.04 0.57
## depblues_mostmug_devx21[4] 0.27 0.15 0.04 0.61
## depblues_mostmug_av12x21[1] 0.12 0.08 0.02 0.32
## depblues_mostmug_av12x21[2] 0.12 0.08 0.02 0.32
## depblues_mostmug_av12x21[3] 0.12 0.08 0.02 0.32
## depblues_mostmug_av12x21[4] 0.12 0.08 0.02 0.31
## depblues_mostmug_av12x21[5] 0.13 0.08 0.02 0.33
## depblues_mostmug_av12x21[6] 0.13 0.08 0.02 0.32
## depblues_mostmug_av12x21[7] 0.13 0.08 0.02 0.33
## depblues_mostmug_av12x21[8] 0.13 0.08 0.02 0.33
## depblues_mostmug_devx2:rural.ses.med21[1] 0.15 0.15 0.00 0.54
## depblues_mostmug_devx2:rural.ses.med21[2] 0.15 0.14 0.00 0.54
## depblues_mostmug_devx2:rural.ses.med21[3] 0.42 0.22 0.03 0.83
## depblues_mostmug_devx2:rural.ses.med21[4] 0.28 0.20 0.01 0.72
## depblues_mostmug_devx2:rural.ses.med31[1] 0.26 0.19 0.01 0.72
## depblues_mostmug_devx2:rural.ses.med31[2] 0.21 0.17 0.01 0.64
## depblues_mostmug_devx2:rural.ses.med31[3] 0.24 0.19 0.01 0.69
## depblues_mostmug_devx2:rural.ses.med31[4] 0.30 0.21 0.01 0.75
## depblues_mostmug_devx2:rural.ses.med41[1] 0.25 0.19 0.01 0.68
## depblues_mostmug_devx2:rural.ses.med41[2] 0.21 0.17 0.01 0.64
## depblues_mostmug_devx2:rural.ses.med41[3] 0.25 0.20 0.01 0.70
## depblues_mostmug_devx2:rural.ses.med41[4] 0.29 0.21 0.01 0.77
## depunfair_mostmug_devx21[1] 0.26 0.14 0.04 0.58
## depunfair_mostmug_devx21[2] 0.25 0.14 0.04 0.57
## depunfair_mostmug_devx21[3] 0.23 0.14 0.03 0.56
## depunfair_mostmug_devx21[4] 0.26 0.15 0.04 0.59
## depunfair_mostmug_av12x21[1] 0.11 0.08 0.01 0.30
## depunfair_mostmug_av12x21[2] 0.10 0.07 0.01 0.27
## depunfair_mostmug_av12x21[3] 0.12 0.08 0.01 0.31
## depunfair_mostmug_av12x21[4] 0.12 0.08 0.02 0.32
## depunfair_mostmug_av12x21[5] 0.13 0.08 0.02 0.34
## depunfair_mostmug_av12x21[6] 0.14 0.08 0.02 0.34
## depunfair_mostmug_av12x21[7] 0.14 0.09 0.02 0.35
## depunfair_mostmug_av12x21[8] 0.13 0.08 0.02 0.33
## depunfair_mostmug_devx2:rural.ses.med21[1] 0.26 0.20 0.01 0.72
## depunfair_mostmug_devx2:rural.ses.med21[2] 0.22 0.18 0.01 0.67
## depunfair_mostmug_devx2:rural.ses.med21[3] 0.23 0.18 0.01 0.68
## depunfair_mostmug_devx2:rural.ses.med21[4] 0.29 0.21 0.01 0.74
## depunfair_mostmug_devx2:rural.ses.med31[1] 0.27 0.20 0.01 0.71
## depunfair_mostmug_devx2:rural.ses.med31[2] 0.19 0.16 0.01 0.61
## depunfair_mostmug_devx2:rural.ses.med31[3] 0.20 0.17 0.01 0.62
## depunfair_mostmug_devx2:rural.ses.med31[4] 0.35 0.23 0.02 0.82
## depunfair_mostmug_devx2:rural.ses.med41[1] 0.29 0.20 0.01 0.73
## depunfair_mostmug_devx2:rural.ses.med41[2] 0.22 0.17 0.01 0.64
## depunfair_mostmug_devx2:rural.ses.med41[3] 0.26 0.19 0.01 0.70
## depunfair_mostmug_devx2:rural.ses.med41[4] 0.23 0.19 0.01 0.69
## depmistrt_mostmug_devx21[1] 0.28 0.15 0.04 0.62
## depmistrt_mostmug_devx21[2] 0.23 0.13 0.04 0.52
## depmistrt_mostmug_devx21[3] 0.23 0.13 0.04 0.53
## depmistrt_mostmug_devx21[4] 0.26 0.15 0.04 0.59
## depmistrt_mostmug_av12x21[1] 0.15 0.09 0.02 0.34
## depmistrt_mostmug_av12x21[2] 0.13 0.08 0.02 0.33
## depmistrt_mostmug_av12x21[3] 0.10 0.06 0.01 0.26
## depmistrt_mostmug_av12x21[4] 0.11 0.07 0.01 0.29
## depmistrt_mostmug_av12x21[5] 0.12 0.07 0.01 0.30
## depmistrt_mostmug_av12x21[6] 0.13 0.08 0.02 0.31
## depmistrt_mostmug_av12x21[7] 0.14 0.09 0.02 0.35
## depmistrt_mostmug_av12x21[8] 0.12 0.08 0.02 0.31
## depmistrt_mostmug_devx2:rural.ses.med21[1] 0.26 0.20 0.01 0.72
## depmistrt_mostmug_devx2:rural.ses.med21[2] 0.22 0.17 0.01 0.65
## depmistrt_mostmug_devx2:rural.ses.med21[3] 0.23 0.18 0.01 0.67
## depmistrt_mostmug_devx2:rural.ses.med21[4] 0.29 0.21 0.01 0.75
## depmistrt_mostmug_devx2:rural.ses.med31[1] 0.25 0.19 0.01 0.68
## depmistrt_mostmug_devx2:rural.ses.med31[2] 0.22 0.17 0.01 0.64
## depmistrt_mostmug_devx2:rural.ses.med31[3] 0.23 0.18 0.01 0.67
## depmistrt_mostmug_devx2:rural.ses.med31[4] 0.31 0.21 0.01 0.77
## depmistrt_mostmug_devx2:rural.ses.med41[1] 0.26 0.18 0.01 0.66
## depmistrt_mostmug_devx2:rural.ses.med41[2] 0.16 0.14 0.01 0.52
## depmistrt_mostmug_devx2:rural.ses.med41[3] 0.38 0.19 0.03 0.76
## depmistrt_mostmug_devx2:rural.ses.med41[4] 0.20 0.16 0.01 0.58
## depbetray_mostmug_devx21[1] 0.27 0.15 0.03 0.60
## depbetray_mostmug_devx21[2] 0.24 0.14 0.03 0.55
## depbetray_mostmug_devx21[3] 0.24 0.14 0.04 0.56
## depbetray_mostmug_devx21[4] 0.26 0.15 0.04 0.59
## depbetray_mostmug_av12x21[1] 0.11 0.07 0.01 0.27
## depbetray_mostmug_av12x21[2] 0.11 0.07 0.01 0.27
## depbetray_mostmug_av12x21[3] 0.09 0.06 0.01 0.24
## depbetray_mostmug_av12x21[4] 0.11 0.07 0.01 0.28
## depbetray_mostmug_av12x21[5] 0.16 0.09 0.02 0.37
## depbetray_mostmug_av12x21[6] 0.17 0.10 0.03 0.41
## depbetray_mostmug_av12x21[7] 0.14 0.09 0.02 0.35
## depbetray_mostmug_av12x21[8] 0.12 0.08 0.02 0.30
## depbetray_mostmug_devx2:rural.ses.med21[1] 0.25 0.20 0.01 0.73
## depbetray_mostmug_devx2:rural.ses.med21[2] 0.25 0.19 0.01 0.71
## depbetray_mostmug_devx2:rural.ses.med21[3] 0.23 0.18 0.01 0.65
## depbetray_mostmug_devx2:rural.ses.med21[4] 0.27 0.20 0.01 0.73
## depbetray_mostmug_devx2:rural.ses.med31[1] 0.23 0.18 0.01 0.66
## depbetray_mostmug_devx2:rural.ses.med31[2] 0.30 0.19 0.02 0.71
## depbetray_mostmug_devx2:rural.ses.med31[3] 0.21 0.16 0.01 0.59
## depbetray_mostmug_devx2:rural.ses.med31[4] 0.26 0.19 0.01 0.70
## depbetray_mostmug_devx2:rural.ses.med41[1] 0.30 0.19 0.02 0.71
## depbetray_mostmug_devx2:rural.ses.med41[2] 0.22 0.16 0.01 0.60
## depbetray_mostmug_devx2:rural.ses.med41[3] 0.29 0.17 0.02 0.67
## depbetray_mostmug_devx2:rural.ses.med41[4] 0.20 0.16 0.01 0.59
## Rhat Bulk_ESS Tail_ESS
## depcantgo_mostmug_devx21[1] 1.00 5810 2690
## depcantgo_mostmug_devx21[2] 1.00 3132 2660
## depcantgo_mostmug_devx21[3] 1.00 5560 3108
## depcantgo_mostmug_devx21[4] 1.00 4465 2739
## depcantgo_mostmug_av12x21[1] 1.00 6186 2558
## depcantgo_mostmug_av12x21[2] 1.00 6117 2197
## depcantgo_mostmug_av12x21[3] 1.00 7576 2524
## depcantgo_mostmug_av12x21[4] 1.00 8494 2998
## depcantgo_mostmug_av12x21[5] 1.00 7347 2165
## depcantgo_mostmug_av12x21[6] 1.00 6423 2548
## depcantgo_mostmug_av12x21[7] 1.00 6695 2815
## depcantgo_mostmug_av12x21[8] 1.00 6510 3016
## depcantgo_mostmug_devx2:rural.ses.med21[1] 1.00 4438 2515
## depcantgo_mostmug_devx2:rural.ses.med21[2] 1.00 1354 1866
## depcantgo_mostmug_devx2:rural.ses.med21[3] 1.00 3881 2400
## depcantgo_mostmug_devx2:rural.ses.med21[4] 1.00 1505 2243
## depcantgo_mostmug_devx2:rural.ses.med31[1] 1.00 5294 2515
## depcantgo_mostmug_devx2:rural.ses.med31[2] 1.00 5488 2683
## depcantgo_mostmug_devx2:rural.ses.med31[3] 1.00 4204 2631
## depcantgo_mostmug_devx2:rural.ses.med31[4] 1.00 4330 3334
## depcantgo_mostmug_devx2:rural.ses.med41[1] 1.00 5029 2167
## depcantgo_mostmug_devx2:rural.ses.med41[2] 1.00 5823 2051
## depcantgo_mostmug_devx2:rural.ses.med41[3] 1.00 5776 2758
## depcantgo_mostmug_devx2:rural.ses.med41[4] 1.00 5716 2575
## depeffort_mostmug_devx21[1] 1.00 6710 2655
## depeffort_mostmug_devx21[2] 1.00 6630 2861
## depeffort_mostmug_devx21[3] 1.00 4443 3052
## depeffort_mostmug_devx21[4] 1.00 5092 2596
## depeffort_mostmug_av12x21[1] 1.00 8067 2207
## depeffort_mostmug_av12x21[2] 1.00 7570 2642
## depeffort_mostmug_av12x21[3] 1.00 7522 2670
## depeffort_mostmug_av12x21[4] 1.00 6477 2591
## depeffort_mostmug_av12x21[5] 1.00 8130 2672
## depeffort_mostmug_av12x21[6] 1.00 7404 2704
## depeffort_mostmug_av12x21[7] 1.00 6501 2613
## depeffort_mostmug_av12x21[8] 1.00 7199 2880
## depeffort_mostmug_devx2:rural.ses.med21[1] 1.00 5577 2309
## depeffort_mostmug_devx2:rural.ses.med21[2] 1.00 3773 1902
## depeffort_mostmug_devx2:rural.ses.med21[3] 1.00 5315 3055
## depeffort_mostmug_devx2:rural.ses.med21[4] 1.00 5863 2944
## depeffort_mostmug_devx2:rural.ses.med31[1] 1.00 5292 2257
## depeffort_mostmug_devx2:rural.ses.med31[2] 1.00 3932 2193
## depeffort_mostmug_devx2:rural.ses.med31[3] 1.00 4554 2783
## depeffort_mostmug_devx2:rural.ses.med31[4] 1.00 4251 2812
## depeffort_mostmug_devx2:rural.ses.med41[1] 1.00 5052 2171
## depeffort_mostmug_devx2:rural.ses.med41[2] 1.00 5642 2162
## depeffort_mostmug_devx2:rural.ses.med41[3] 1.00 3510 2899
## depeffort_mostmug_devx2:rural.ses.med41[4] 1.00 3038 2245
## deplonely_mostmug_devx21[1] 1.00 4839 2765
## deplonely_mostmug_devx21[2] 1.00 6956 2538
## deplonely_mostmug_devx21[3] 1.00 5526 2939
## deplonely_mostmug_devx21[4] 1.00 6363 2593
## deplonely_mostmug_av12x21[1] 1.00 6557 2314
## deplonely_mostmug_av12x21[2] 1.00 6974 2768
## deplonely_mostmug_av12x21[3] 1.00 6562 2000
## deplonely_mostmug_av12x21[4] 1.00 6278 2386
## deplonely_mostmug_av12x21[5] 1.00 8105 2578
## deplonely_mostmug_av12x21[6] 1.00 6869 2675
## deplonely_mostmug_av12x21[7] 1.00 7081 2882
## deplonely_mostmug_av12x21[8] 1.00 7897 2764
## deplonely_mostmug_devx2:rural.ses.med21[1] 1.00 4914 2714
## deplonely_mostmug_devx2:rural.ses.med21[2] 1.00 5483 1926
## deplonely_mostmug_devx2:rural.ses.med21[3] 1.00 3636 2673
## deplonely_mostmug_devx2:rural.ses.med21[4] 1.00 4099 2209
## deplonely_mostmug_devx2:rural.ses.med31[1] 1.00 3965 2664
## deplonely_mostmug_devx2:rural.ses.med31[2] 1.00 5719 2443
## deplonely_mostmug_devx2:rural.ses.med31[3] 1.00 2336 1986
## deplonely_mostmug_devx2:rural.ses.med31[4] 1.00 5093 2382
## deplonely_mostmug_devx2:rural.ses.med41[1] 1.00 4856 2598
## deplonely_mostmug_devx2:rural.ses.med41[2] 1.00 5326 2262
## deplonely_mostmug_devx2:rural.ses.med41[3] 1.00 4304 2454
## deplonely_mostmug_devx2:rural.ses.med41[4] 1.00 5281 2430
## depblues_mostmug_devx21[1] 1.00 5637 2921
## depblues_mostmug_devx21[2] 1.00 6768 2992
## depblues_mostmug_devx21[3] 1.00 6749 2965
## depblues_mostmug_devx21[4] 1.00 6605 2759
## depblues_mostmug_av12x21[1] 1.00 6956 2458
## depblues_mostmug_av12x21[2] 1.00 6165 2296
## depblues_mostmug_av12x21[3] 1.00 7007 2075
## depblues_mostmug_av12x21[4] 1.00 8005 2591
## depblues_mostmug_av12x21[5] 1.00 7086 2531
## depblues_mostmug_av12x21[6] 1.00 6349 2379
## depblues_mostmug_av12x21[7] 1.00 6485 2476
## depblues_mostmug_av12x21[8] 1.00 6507 2944
## depblues_mostmug_devx2:rural.ses.med21[1] 1.00 3395 2352
## depblues_mostmug_devx2:rural.ses.med21[2] 1.00 4871 2584
## depblues_mostmug_devx2:rural.ses.med21[3] 1.00 3609 2353
## depblues_mostmug_devx2:rural.ses.med21[4] 1.00 6183 2728
## depblues_mostmug_devx2:rural.ses.med31[1] 1.00 5296 2137
## depblues_mostmug_devx2:rural.ses.med31[2] 1.00 6640 2747
## depblues_mostmug_devx2:rural.ses.med31[3] 1.00 5652 2756
## depblues_mostmug_devx2:rural.ses.med31[4] 1.00 5429 3221
## depblues_mostmug_devx2:rural.ses.med41[1] 1.00 6463 2367
## depblues_mostmug_devx2:rural.ses.med41[2] 1.00 5935 2912
## depblues_mostmug_devx2:rural.ses.med41[3] 1.00 3693 3038
## depblues_mostmug_devx2:rural.ses.med41[4] 1.00 4234 2680
## depunfair_mostmug_devx21[1] 1.00 6054 2231
## depunfair_mostmug_devx21[2] 1.00 6406 3070
## depunfair_mostmug_devx21[3] 1.00 6050 2891
## depunfair_mostmug_devx21[4] 1.00 6858 2644
## depunfair_mostmug_av12x21[1] 1.00 5949 2365
## depunfair_mostmug_av12x21[2] 1.00 6641 2592
## depunfair_mostmug_av12x21[3] 1.00 6870 2103
## depunfair_mostmug_av12x21[4] 1.00 7035 2116
## depunfair_mostmug_av12x21[5] 1.00 8057 2164
## depunfair_mostmug_av12x21[6] 1.00 7000 2777
## depunfair_mostmug_av12x21[7] 1.00 5925 2647
## depunfair_mostmug_av12x21[8] 1.00 6447 2459
## depunfair_mostmug_devx2:rural.ses.med21[1] 1.00 5824 2648
## depunfair_mostmug_devx2:rural.ses.med21[2] 1.00 5448 2679
## depunfair_mostmug_devx2:rural.ses.med21[3] 1.00 5502 2662
## depunfair_mostmug_devx2:rural.ses.med21[4] 1.00 6225 2892
## depunfair_mostmug_devx2:rural.ses.med31[1] 1.00 4625 2650
## depunfair_mostmug_devx2:rural.ses.med31[2] 1.00 4928 2901
## depunfair_mostmug_devx2:rural.ses.med31[3] 1.00 4112 2052
## depunfair_mostmug_devx2:rural.ses.med31[4] 1.00 3722 2536
## depunfair_mostmug_devx2:rural.ses.med41[1] 1.00 3715 2155
## depunfair_mostmug_devx2:rural.ses.med41[2] 1.00 5101 2185
## depunfair_mostmug_devx2:rural.ses.med41[3] 1.00 5536 2844
## depunfair_mostmug_devx2:rural.ses.med41[4] 1.00 5400 3039
## depmistrt_mostmug_devx21[1] 1.00 5355 2710
## depmistrt_mostmug_devx21[2] 1.00 5907 2370
## depmistrt_mostmug_devx21[3] 1.00 6590 3135
## depmistrt_mostmug_devx21[4] 1.00 7007 2911
## depmistrt_mostmug_av12x21[1] 1.00 8033 2869
## depmistrt_mostmug_av12x21[2] 1.00 6498 2096
## depmistrt_mostmug_av12x21[3] 1.00 6238 2757
## depmistrt_mostmug_av12x21[4] 1.00 6651 2314
## depmistrt_mostmug_av12x21[5] 1.00 5407 2265
## depmistrt_mostmug_av12x21[6] 1.00 7764 2829
## depmistrt_mostmug_av12x21[7] 1.00 5577 2657
## depmistrt_mostmug_av12x21[8] 1.00 7534 2690
## depmistrt_mostmug_devx2:rural.ses.med21[1] 1.00 4829 2686
## depmistrt_mostmug_devx2:rural.ses.med21[2] 1.00 6202 1934
## depmistrt_mostmug_devx2:rural.ses.med21[3] 1.00 5340 2773
## depmistrt_mostmug_devx2:rural.ses.med21[4] 1.00 4880 2548
## depmistrt_mostmug_devx2:rural.ses.med31[1] 1.00 5545 2968
## depmistrt_mostmug_devx2:rural.ses.med31[2] 1.00 4768 2566
## depmistrt_mostmug_devx2:rural.ses.med31[3] 1.00 4312 2799
## depmistrt_mostmug_devx2:rural.ses.med31[4] 1.00 3920 2877
## depmistrt_mostmug_devx2:rural.ses.med41[1] 1.00 4621 2231
## depmistrt_mostmug_devx2:rural.ses.med41[2] 1.00 5134 2838
## depmistrt_mostmug_devx2:rural.ses.med41[3] 1.00 3619 2363
## depmistrt_mostmug_devx2:rural.ses.med41[4] 1.00 5568 2206
## depbetray_mostmug_devx21[1] 1.00 5247 2545
## depbetray_mostmug_devx21[2] 1.00 5269 2905
## depbetray_mostmug_devx21[3] 1.00 5606 2790
## depbetray_mostmug_devx21[4] 1.00 6546 2765
## depbetray_mostmug_av12x21[1] 1.00 6441 2552
## depbetray_mostmug_av12x21[2] 1.00 6894 2747
## depbetray_mostmug_av12x21[3] 1.00 5782 2360
## depbetray_mostmug_av12x21[4] 1.00 7348 2462
## depbetray_mostmug_av12x21[5] 1.00 7166 2533
## depbetray_mostmug_av12x21[6] 1.00 6495 2812
## depbetray_mostmug_av12x21[7] 1.00 6828 2970
## depbetray_mostmug_av12x21[8] 1.00 8165 2314
## depbetray_mostmug_devx2:rural.ses.med21[1] 1.00 4192 2656
## depbetray_mostmug_devx2:rural.ses.med21[2] 1.00 4629 2445
## depbetray_mostmug_devx2:rural.ses.med21[3] 1.00 5344 3012
## depbetray_mostmug_devx2:rural.ses.med21[4] 1.00 5800 2314
## depbetray_mostmug_devx2:rural.ses.med31[1] 1.00 4952 1924
## depbetray_mostmug_devx2:rural.ses.med31[2] 1.00 4208 2049
## depbetray_mostmug_devx2:rural.ses.med31[3] 1.00 4347 2566
## depbetray_mostmug_devx2:rural.ses.med31[4] 1.00 5226 2832
## depbetray_mostmug_devx2:rural.ses.med41[1] 1.00 4364 2665
## depbetray_mostmug_devx2:rural.ses.med41[2] 1.00 5220 2298
## depbetray_mostmug_devx2:rural.ses.med41[3] 1.00 4974 2748
## depbetray_mostmug_devx2:rural.ses.med41[4] 1.00 5274 2578
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).8.1.2.7.4 Prior summary
out.chg.alldepress.stmug.comm.fit[[2]]## prior class coef
## (flat) b
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b
## normal(0, 0.125) b mostmug_av12x2
## normal(0, 0.25) b mostmug_devx2
## normal(0, 1) b mostmug_devx2:rural.ses.med2
## normal(0, 1) b mostmug_devx2:rural.ses.med3
## normal(0, 1) b mostmug_devx2:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## (flat) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd Intercept
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## dirichlet(2, 2, 2, 2, 2, 2, 2, 2) simo mostmug_av12x21
## dirichlet(1) simo mostmug_devx2:rural.ses.med21
## dirichlet(1) simo mostmug_devx2:rural.ses.med31
## dirichlet(1) simo mostmug_devx2:rural.ses.med41
## dirichlet(2, 2, 2, 2) simo mostmug_devx21
## group resp dpar nlpar lb ub source
## default
## depbetray user
## depbetray user
## depbetray user
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depbetray (vectorized)
## depblues user
## depblues user
## depblues user
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depblues (vectorized)
## depcantgo user
## depcantgo user
## depcantgo user
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depcantgo (vectorized)
## depeffort user
## depeffort user
## depeffort user
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## depeffort (vectorized)
## deplonely user
## deplonely user
## deplonely user
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## deplonely (vectorized)
## depmistrt user
## depmistrt user
## depmistrt user
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depmistrt (vectorized)
## depunfair user
## depunfair user
## depunfair user
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## depunfair (vectorized)
## default
## depbetray user
## depblues user
## depcantgo user
## depeffort user
## deplonely user
## depmistrt user
## depunfair user
## depbetray 0 default
## depblues 0 default
## depcantgo 0 default
## depeffort 0 default
## deplonely 0 default
## depmistrt 0 default
## depunfair 0 default
## id depbetray 0 (vectorized)
## id depbetray 0 (vectorized)
## id depblues 0 (vectorized)
## id depblues 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depcantgo 0 (vectorized)
## id depeffort 0 (vectorized)
## id depeffort 0 (vectorized)
## id deplonely 0 (vectorized)
## id deplonely 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depmistrt 0 (vectorized)
## id depunfair 0 (vectorized)
## id depunfair 0 (vectorized)
## depbetray user
## depbetray default
## depbetray default
## depbetray default
## depbetray user
## depblues user
## depblues default
## depblues default
## depblues default
## depblues user
## depcantgo user
## depcantgo default
## depcantgo default
## depcantgo default
## depcantgo user
## depeffort user
## depeffort default
## depeffort default
## depeffort default
## depeffort user
## deplonely user
## deplonely default
## deplonely default
## deplonely default
## deplonely user
## depmistrt user
## depmistrt default
## depmistrt default
## depmistrt default
## depmistrt user
## depunfair user
## depunfair default
## depunfair default
## depunfair default
## depunfair user9 Figure 4: Change correlations by community
(RMD FILE: BDK_2023_Stress_8_Chgcorr_comm_viz)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE9.1 Visualizing change correlations by community
9.1.1 Predictive Marginal Contrasts x Comm (Crim Intent)
load(here("1_Data_Files/Datasets/stress_long.Rdata"))
#load change by comm brms model fits
# criminal intent
chg.prjcrime.stmony.comm.fit <- readRDS(here("Models/chg_prjcrime_stmony_comm_fit.rds"))
chg.prjcrime.sttran.comm.fit <- readRDS(here("Models/chg_prjcrime_sttran_comm_fit.rds"))
chg.prjcrime.stresp.comm.fit <- readRDS(here("Models/chg_prjcrime_stresp_comm_fit.rds"))
chg.prjcrime.stfair.comm.fit <- readRDS(here("Models/chg_prjcrime_stfair_comm_fit.rds"))
chg.prjcrime.stjob.comm.fit <- readRDS(here("Models/chg_prjcrime_stjob_comm_fit.rds"))
chg.prjcrime.stthft.comm.fit <- readRDS(here("Models/chg_prjcrime_stthft_comm_fit.rds"))
chg.prjcrime.stmug.comm.fit <- readRDS(here("Models/chg_prjcrime_stmug_comm_fit.rds"))
chg.anyprjcrime.stmony.comm.fit <- readRDS(here("Models/chg_anyprjcrime_stmony_comm_fit.rds"))
chg.anyprjcrime.sttran.comm.fit <- readRDS(here("Models/chg_anyprjcrime_sttran_comm_fit.rds"))
chg.anyprjcrime.stresp.comm.fit <- readRDS(here("Models/chg_anyprjcrime_stresp_comm_fit.rds"))
chg.anyprjcrime.stfair.comm.fit <- readRDS(here("Models/chg_anyprjcrime_stfair_comm_fit.rds"))
chg.anyprjcrime.stjob.comm.fit <- readRDS(here("Models/chg_anyprjcrime_stjob_comm_fit.rds"))
chg.anyprjcrime.stthft.comm.fit <- readRDS(here("Models/chg_anyprjcrime_stthft_comm_fit.rds"))
chg.anyprjcrime.stmug.comm.fit <- readRDS(here("Models/chg_anyprjcrime_stmug_comm_fit.rds"))
# negative emotions
chg.alldepress.stmony.comm.fit <- readRDS(here("Models/chg_alldepress_stmony_comm_fit.rds"))
chg.alldepress.sttran.comm.fit <- readRDS(here("Models/chg_alldepress_sttran_comm_fit.rds"))
chg.alldepress.stresp.comm.fit <- readRDS(here("Models/chg_alldepress_stresp_comm_fit.rds"))
chg.alldepress.stfair.comm.fit <- readRDS(here("Models/chg_alldepress_stfair_comm_fit.rds"))
chg.alldepress.stjob.comm.fit <- readRDS(here("Models/chg_alldepress_stjob_comm_fit.rds"))
chg.alldepress.stthft.comm.fit <- readRDS(here("Models/chg_alldepress_stthft_comm_fit.rds"))
chg.alldepress.stmug.comm.fit <- readRDS(here("Models/chg_alldepress_stmug_comm_fit.rds"))
#T1 posterior PLME contrasts
# load(here("1_Data_Files/Datasets/twodif_combineddvs3.Rdata"))
# load(here("1_Data_Files/Datasets/twodif_combinedprj3.Rdata"))
# load(here("1_Data_Files/Datasets/twodif_combineddep3.Rdata"))#Manually generate predictive margins data w/following generic structure:
# create new data grid for epred values
# newdata <- mylongdata %>%
# data_grid(xdev, xbar, year)
#
# # generate epred draws (assuming two waves here)
# predmarg_data = epred_draws(mymodelfit,
# newdata = newdata,
# re_formula = NA) %>%
# filter(xdev == -2 & year == 1 |
# xdev == 2 & year == 2) %>%
# group_by(.category, xdev, .draw) %>%
# summarise(`E[y|xdev]` = mean(`.epred`))
#
# # calculate marginal effect contrasts
# margeff_contrast = predmarg_data %>%
# compare_levels(`E[y|xdev]`, by = xdev) %>% # pairwise diffs in `E[y|x]`, by levels of x
# rename(`difference in E[y|change in x (T2-T1)]` = `E[y|xdev]`) # more accurate column name
# create function to generate new data grid for epred values
gen_newdata_comm <- function(mylongdata, xdev, xave){
mylongdata %>%
data_grid({{xdev}}, {{xave}}, rural.ses.med, year)
}
# create function to generate epred draws
# keeping only 2-unit increases from year 1 to year 2 (contrast 2_t2 - -2_t1)
gen_predmarg_data_comm <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
filter({{xdev}} == -2 & year == 1 |
{{xdev}} == 2 & year == 2) %>%
group_by(.category, rural.ses.med, {{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`))
}
# generate epred draws
newdata <- gen_newdata_comm(stress.long, stmony_devx2, stmony_av12x2)
predmarg_stmony_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stmony.comm.fit, stmony_devx2)
newdata <- gen_newdata_comm(stress.long, sttran_devx2, sttran_av12x2)
predmarg_sttran_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.sttran.comm.fit, sttran_devx2)
newdata <- gen_newdata_comm(stress.long, stresp_devx2, stresp_av12x2)
predmarg_stresp_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stresp.comm.fit, stresp_devx2)
newdata <- gen_newdata_comm(stress.long, stfair_devx2, stfair_av12x2)
predmarg_stfair_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stfair.comm.fit, stfair_devx2)
newdata <- gen_newdata_comm(stress.long, stjob_devx2, stjob_av12x2)
predmarg_stjob_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stjob.comm.fit, stjob_devx2)
newdata <- gen_newdata_comm(stress.long, stthft_devx2, stthft_av12x2)
predmarg_stthft_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stthft.comm.fit, stthft_devx2)
newdata <- gen_newdata_comm(stress.long, stmug_devx2, stmug_av12x2)
predmarg_stmug_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stmug.comm.fit, stmug_devx2)
gen_newdata_comm <- function(mylongdata, xdev, xave){
mylongdata %>%
data_grid({{xdev}}, {{xave}}, rural.ses.med, year)
}
# create function to generate epred draws from "any crim intent" models
# keeping only 2-unit increases from year 1 to year 2 (contrast 2_t2 - -2_t1)
gen_predmarg_data_comm <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
filter({{xdev}} == -2 & year == 1 |
{{xdev}} == 2 & year == 2) %>%
group_by(rural.ses.med, {{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
}
# generate "any" epred data & merge with individual outcome item epred data
newdata <- gen_newdata_comm(stress.long, stmony_devx2, stmony_av12x2)
predmarg_stmony_anyprjcrim_chg_comm =
gen_predmarg_data_comm(chg.anyprjcrime.stmony.comm.fit, stmony_devx2)
predmarg_stmony_prjcrim_chg_comm <- bind_rows(predmarg_stmony_prjcrim_chg_comm,
predmarg_stmony_anyprjcrim_chg_comm)
rm(predmarg_stmony_anyprjcrim_chg_comm) #clean environment
newdata <- gen_newdata_comm(stress.long, sttran_devx2, sttran_av12x2)
predmarg_sttran_anyprjcrim_chg_comm =
gen_predmarg_data_comm(chg.anyprjcrime.sttran.comm.fit, sttran_devx2)
predmarg_sttran_prjcrim_chg_comm <- bind_rows(predmarg_sttran_prjcrim_chg_comm,
predmarg_sttran_anyprjcrim_chg_comm)
rm(predmarg_sttran_anyprjcrim_chg_comm) #clean environment
newdata <- gen_newdata_comm(stress.long, stresp_devx2, stresp_av12x2)
predmarg_stresp_anyprjcrim_chg_comm =
gen_predmarg_data_comm(chg.anyprjcrime.stresp.comm.fit, stresp_devx2)
predmarg_stresp_prjcrim_chg_comm <- bind_rows(predmarg_stresp_prjcrim_chg_comm,
predmarg_stresp_anyprjcrim_chg_comm)
rm(predmarg_stresp_anyprjcrim_chg_comm) #clean environment
newdata <- gen_newdata_comm(stress.long, stfair_devx2, stfair_av12x2)
predmarg_stfair_anyprjcrim_chg_comm =
gen_predmarg_data_comm(chg.anyprjcrime.stfair.comm.fit, stfair_devx2)
predmarg_stfair_prjcrim_chg_comm <- bind_rows(predmarg_stfair_prjcrim_chg_comm,
predmarg_stfair_anyprjcrim_chg_comm)
rm(predmarg_stfair_anyprjcrim_chg_comm) #clean environment
newdata <- gen_newdata_comm(stress.long, stjob_devx2, stjob_av12x2)
predmarg_stjob_anyprjcrim_chg_comm =
gen_predmarg_data_comm(chg.anyprjcrime.stjob.comm.fit, stjob_devx2)
predmarg_stjob_prjcrim_chg_comm <- bind_rows(predmarg_stjob_prjcrim_chg_comm,
predmarg_stjob_anyprjcrim_chg_comm)
rm(predmarg_stjob_anyprjcrim_chg_comm) #clean environment
newdata <- gen_newdata_comm(stress.long, stthft_devx2, stthft_av12x2)
predmarg_stthft_anyprjcrim_chg_comm =
gen_predmarg_data_comm(chg.anyprjcrime.stthft.comm.fit, stthft_devx2)
predmarg_stthft_prjcrim_chg_comm <- bind_rows(predmarg_stthft_prjcrim_chg_comm,
predmarg_stthft_anyprjcrim_chg_comm)
rm(predmarg_stthft_anyprjcrim_chg_comm) #clean environment
newdata <- gen_newdata_comm(stress.long, stmug_devx2, stmug_av12x2)
predmarg_stmug_anyprjcrim_chg_comm =
gen_predmarg_data_comm(chg.anyprjcrime.stmug.comm.fit, stmug_devx2)
predmarg_stmug_prjcrim_chg_comm <- bind_rows(predmarg_stmug_prjcrim_chg_comm,
predmarg_stmug_anyprjcrim_chg_comm)
rm(predmarg_stmug_anyprjcrim_chg_comm) #clean environment
#function to calculate mean difference contrasts (marginal effect contrast)
calc_ME_chg_comm <- function(predmarg_data, xdev) {
predmarg_data %>%
compare_levels(`E[y|xdev]`, by = xdev) %>% # pairwise diffs in `E[y|x]`, by levels of x
group_by(rural.ses.med) %>% # generate community-specific marginal contrasts
rename(`PLME2chg` = `E[y|xdev]`) # easy colname reflecting ME 2-unit chg contrast
}
#outputs community-specific predicted differences in E[y]
# associated with 2-Likert category increase in stress (T2-T1)
#marg effect contrasts are marginalized over all person-level avg values of stress (AME)
#generate ME contrasts
PLME2_stmony_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stmony_prjcrim_chg_comm, "stmony_devx2") %>%
mutate(stress_var = "Stress:\nMoney",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmony_devx2)}, file="cache_8_1")
PLME2_sttran_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_sttran_prjcrim_chg_comm, "sttran_devx2") %>%
mutate(stress_var = "Stress:\nTransport",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = sttran_devx2)}, file="cache_8_2")
PLME2_stresp_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stresp_prjcrim_chg_comm, "stresp_devx2") %>%
mutate(stress_var = "Stress:\nRespect",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stresp_devx2)}, file="cache_8_3")
PLME2_stfair_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stfair_prjcrim_chg_comm, "stfair_devx2") %>%
mutate(stress_var = "Stress:\nFair Trtmt",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stfair_devx2)}, file="cache_8_4")
PLME2_stjob_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stjob_prjcrim_chg_comm, "stjob_devx2") %>%
mutate(stress_var = "Stress:\nJob",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stjob_devx2)}, file="cache_8_5")
PLME2_stthft_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stthft_prjcrim_chg_comm, "stthft_devx2") %>%
mutate(stress_var = "Stress:\nTheft Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stthft_devx2)}, file="cache_8_6")
PLME2_stmug_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stmug_prjcrim_chg_comm, "stmug_devx2") %>%
mutate(stress_var = "Stress:\nAssault Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmug_devx2)}, file="cache_8_7")
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
PLME2_combinedprjchg_comm <- bind_rows(
PLME2_stmony_prjcrim_chg_comm,
PLME2_sttran_prjcrim_chg_comm,
PLME2_stresp_prjcrim_chg_comm,
PLME2_stfair_prjcrim_chg_comm,
PLME2_stjob_prjcrim_chg_comm,
PLME2_stthft_prjcrim_chg_comm,
PLME2_stmug_prjcrim_chg_comm) %>%
mutate (stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney")
)
)9.1.2 Predictive Marginal Contrasts x Comm (Negative Emotions)
# create function to generate new data grid for epred values
gen_newdata_comm <- function(mylongdata, xdev, xave){
mylongdata %>%
data_grid({{xdev}}, {{xave}}, rural.ses.med, year)
}
# create function to generate epred draws
# keeping only 2-unit increases from year 1 to year 2 (contrast 2_t2 - -2_t1)
gen_predmarg_data_comm <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
filter({{xdev}} == -2 & year == 1 |
{{xdev}} == 2 & year == 2) %>%
group_by(.category, rural.ses.med, {{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`))
}
# generate epred draws using gen_newdata & gen_predmarg_data function
newdata <- gen_newdata_comm(stress.long, stmony_devx2, stmony_av12x2)
predmarg_stmony_dep_chg_comm = gen_predmarg_data_comm(chg.alldepress.stmony.comm.fit, stmony_devx2)
newdata <- gen_newdata_comm(stress.long, sttran_devx2, sttran_av12x2)
predmarg_sttran_dep_chg_comm = gen_predmarg_data_comm(chg.alldepress.sttran.comm.fit, sttran_devx2)
newdata <- gen_newdata_comm(stress.long, stresp_devx2, stresp_av12x2)
predmarg_stresp_dep_chg_comm = gen_predmarg_data_comm(chg.alldepress.stresp.comm.fit, stresp_devx2)
newdata <- gen_newdata_comm(stress.long, stfair_devx2, stfair_av12x2)
predmarg_stfair_dep_chg_comm = gen_predmarg_data_comm(chg.alldepress.stfair.comm.fit, stfair_devx2)
newdata <- gen_newdata_comm(stress.long, stjob_devx2, stjob_av12x2)
predmarg_stjob_dep_chg_comm = gen_predmarg_data_comm(chg.alldepress.stjob.comm.fit, stjob_devx2)
newdata <- gen_newdata_comm(stress.long, stthft_devx2, stthft_av12x2)
predmarg_stthft_dep_chg_comm = gen_predmarg_data_comm(chg.alldepress.stthft.comm.fit, stthft_devx2)
newdata <- gen_newdata_comm(stress.long, stmug_devx2, stmug_av12x2)
predmarg_stmug_dep_chg_comm = gen_predmarg_data_comm(chg.alldepress.stmug.comm.fit, stmug_devx2)
#outputs predicted difference in E[y] associated with 2-Likert category increase in stress (T2-T1)
#marg effect contrasts are marginalized over all person-level avg values of stress (AME)
#generate ME contrasts using calc_ME_chg function
PLME2_stmony_dep_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stmony_dep_chg_comm, "stmony_devx2") %>%
mutate(stress_var = "Stress:\nMoney",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmony_devx2)}, file="cache_8_8")
PLME2_sttran_dep_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_sttran_dep_chg_comm, "sttran_devx2") %>%
mutate(stress_var = "Stress:\nTransport",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = sttran_devx2)}, file="cache_8_9")
PLME2_stresp_dep_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stresp_dep_chg_comm, "stresp_devx2") %>%
mutate(stress_var = "Stress:\nRespect",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stresp_devx2)}, file="cache_8_10")
PLME2_stfair_dep_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stfair_dep_chg_comm, "stfair_devx2") %>%
mutate(stress_var = "Stress:\nFair Trtmt",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stfair_devx2)}, file="cache_8_11")
PLME2_stjob_dep_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stjob_dep_chg_comm, "stjob_devx2") %>%
mutate(stress_var = "Stress:\nJob",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stjob_devx2)}, file="cache_8_12")
PLME2_stthft_dep_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stthft_dep_chg_comm, "stthft_devx2") %>%
mutate(stress_var = "Stress:\nTheft Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stthft_devx2)}, file="cache_8_13")
PLME2_stmug_dep_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stmug_dep_chg_comm, "stmug_devx2") %>%
mutate(stress_var = "Stress:\nAssault Vctm",
dif_label = "diff in E[y|stress change]") %>%
rename(contrast = stmug_devx2)}, file="cache_8_14")
#Combine data sets w/"bind_rows" command (stacks on top of each other bc all have same vars)
PLME2_combineddepchg_comm <- bind_rows(
PLME2_stmony_dep_chg_comm,
PLME2_sttran_dep_chg_comm,
PLME2_stresp_dep_chg_comm,
PLME2_stfair_dep_chg_comm,
PLME2_stjob_dep_chg_comm,
PLME2_stthft_dep_chg_comm,
PLME2_stmug_dep_chg_comm) %>%
mutate (stress_varf = factor(stress_var, ordered=TRUE,
levels=c("Stress:\nAssault Vctm",
"Stress:\nTheft Vctm",
"Stress:\nJob",
"Stress:\nFair Trtmt",
"Stress:\nRespect",
"Stress:\nTransport",
"Stress:\nMoney")
)
)9.1.3 Fig4 versions: PLME Stress/Outcome Change X Community
#function to find & drop leading zeroes (used for x-axis label)
dropLeadingZero <- function(l){
str_replace(l, '0(?=.)', '')
}
#wrangle PLME prj crime T2-T1 change estimates
PLMEprjchg_comm <- PLME2_combinedprjchg_comm %>%
rename(PLME = PLME2chg) %>%
mutate(
dif_label='diff in E[y|2-cat stress increase]'
) %>%
dplyr::select(-c(contrast)) %>%
group_by(.category, stress_var, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = as.factor(if_else(p_gt0 >= .80, 1, 0))
) %>%
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
#wrangle PLME neg emo T2-T1 change estimates
PLMEdepchg_comm <- PLME2_combineddepchg_comm %>%
rename(PLME = PLME2chg) %>%
mutate(
dif_label='diff in E[y|2-cat stress increase]'
) %>%
dplyr::select(-c(contrast)) %>%
group_by(.category, stress_var, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = as.factor(if_else(p_gt0 >= .80, 1, 0))
)
prjlabs <- c(
"prjthflt5"="Theft <5BAM",
"prjthfgt5"="Theft >5BAM",
"prjthreat"="Threaten",
"prjharm"="Phys. harm",
"prjusedrg"="Use drugs",
"prjhack"="Hack info",
"prjany"="Any crime")
deplabs <- c(
"depcantgo"="Can't go",
"depeffort"="Effort",
"deplonely"="Lonely",
"depblues"="Blues",
"depunfair"="Unfair",
"depmistrt"="Mistreated",
"depbetray"="Betrayed")
stress_varlabs <- c(
"Stress:\nMoney" = "Stress:\nMoney",
"Stress:\nTransport" = "Stress:\nTransport",
"Stress:\nRespect" = "Stress:\nRespect",
"Stress:\nFair Trtmt" = "Stress:\nFair Trtmt",
"Stress:\nJob" = "Stress:\nJob",
"Stress:\nTheft Vctm" = "Stress:\nTheft",
"Stress:\nAssault Vctm" = "Stress:\nAssault")
numcommlabs <- c("1"="Rural/\nLow SES",
"2"="Rural/\nHigh SES",
"3"="Urban/\nLow SES",
"4"="Urban/\nHigh SES")
# Colors from viridis colorblind-friendly palette
# https://waldyrious.net/viridis-palette-generator/
prjcommplot <- ggplot(data = PLMEprjchg_comm,
mapping = aes(x = PLME,
y = reorder(.category, desc(.category)),
color=rural.ses.med)) +
facet_wrap(~reorder(stress_varf, desc(stress_varf)), nrow=1,
labeller = labeller(stress_varf= as_labeller(stress_varlabs))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 1),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .3), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 2),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 3),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 4),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.3), size=.5) +
scale_color_manual(values=c("#bddf26","#7ad151", "#2a788e", "#440154"),
labels=as_labeller(numcommlabs), name=NULL) +
scale_alpha_discrete(range=c(.1,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.31)) +
scale_x_continuous(breaks=c(-.1,0,.1,.2), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=prjlabs) +
theme(axis.title.y = element_blank(),
legend.position = "right",
strip.background = element_blank(),
strip.text.x = element_text(size = 8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# prjcommplot
depcommplot <- ggplot(data = PLMEdepchg_comm,
mapping = aes(x = PLME,
y = reorder(.category, desc(.category)),
color=rural.ses.med)) +
facet_wrap(~reorder(stress_varf, desc(stress_varf)), nrow=1,
labeller = labeller(stress_varf= as_labeller(stress_varlabs))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 1),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .3), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 2),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 3),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 4),
aes(x=PLME, y=reorder(.category, desc(.category)),
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.3), size=.5) +
scale_color_manual(values=c("#bddf26","#7ad151", "#2a788e", "#440154"),
labels=as_labeller(numcommlabs), name=NULL) +
scale_alpha_discrete(range=c(.1,1), guide = "none") +
coord_cartesian(xlim=c(-.51,.71)) +
scale_x_continuous(breaks=c(-.5,0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=deplabs) +
theme(axis.title.y = element_blank(),
legend.position = "right",
strip.background = element_blank(),
strip.text.x = element_blank(),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# depcommplot
design <- "
1113
2223
"
# library(patchwork)
SuppFigure4a <- prjcommplot + depcommplot + guide_area() +
plot_layout(design=design, guides = 'collect', widths = c(4,.6)) +
plot_annotation(
title = 'SUPPLEMENTAL FIGURE 4a\nMarginal Effects of 2-Category Change in Stress on Outcome Probabilities, by Community SES & Urban/Rural Location',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Estimates derived from 14 multivariate (using `brms::mvbind()`) and multilevel between-within Bayesian logistic regression models simultaneously regressing all six specific criminal intent outcomes (7 models) and all seven negative emotion outcomes (7 models) separately on each of the seven stress types, and seven separate models regressing "any criminal intent" on stress. Stress items were separated into a L2 cross-time average (Xbar_i) between-person predictor and a L1 within-person change (X_it - Xbar_i) "fixed effects" estimator. Both L1 and L2 stress variables were specified as monotonic ordinal predictors with a cumulative probit link function. Models also included a factor variable for community and a multiplicative interaction between L1 stress change (fixed effects estimator) and community. Models were estimated in brms with 4 chains and 4000 total post-warmup posterior draws per outcome and per community group. Marginal effect contrast distributions were estimated from the expectation of the posterior predictive distribution for each model as predicted probability difference distributions for 2-category stress increases (T1 to T2 change) by community, averaged over all between-person stress levels. Median posterior density estimates with 95% intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates for the average marginal effect contrast are greater than zero.', width=180)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'right',
legend.key.height = unit(1, 'cm'))
# SuppFigure4a
ggsave("SuppFigure4a.jpeg", width=9, height=6.5, path=here("Output"))
stress_varlabsalt <- c(
"Stress:\nMoney" = "Stress: Money",
"Stress:\nTransport" = "Stress: Transport",
"Stress:\nRespect" = "Stress: Respect",
"Stress:\nFair Trtmt" = "Stress: Fair Trtmt",
"Stress:\nJob" = "Stress: Job",
"Stress:\nTheft Vctm" = "Stress: Theft",
"Stress:\nAssault Vctm" = "Stress: Assault")
prjcommplotalt <- ggplot(data = PLMEprjchg_comm,
mapping = aes(x = PLME,
y = stress_varf,
color=rural.ses.med)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 1),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .3), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 2),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 3),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 4),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.3), size=.5) +
scale_color_manual(values=c("#bddf26","#7ad151", "#2a788e", "#440154"),
labels=as_labeller(numcommlabs), name=NULL) +
scale_alpha_discrete(range=c(.1,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.31)) +
scale_x_continuous(breaks=c(-.1,0,.1,.2), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=stress_varlabsalt) +
theme(axis.title.y = element_blank(),
legend.position = "right",
strip.background = element_blank(),
strip.text.x = element_text(size = 8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# prjcommplotalt
depcommplotalt <- ggplot(data = PLMEdepchg_comm,
mapping = aes(x = PLME,
y = stress_varf,
color=rural.ses.med)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 1),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .3), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 2),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = .1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 3),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.1), size=.5) +
stat_pointinterval(data=function(x) subset(x, rural.ses.med == 4),
aes(x=PLME, y=stress_varf,
alpha=p80_gt0), .width = .95,
position = position_nudge(y = -.3), size=.5) +
scale_color_manual(values=c("#bddf26","#7ad151", "#2a788e", "#440154"),
labels=as_labeller(numcommlabs), name=NULL) +
scale_alpha_discrete(range=c(.1,1), guide = "none") +
coord_cartesian(xlim=c(-.51,.71)) +
scale_x_continuous(breaks=c(-.5,0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=stress_varlabsalt) +
theme(axis.title.y = element_blank(),
legend.position = "right",
strip.background = element_blank(),
strip.text.x = element_text(size = 8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# depcommplotalt
design <- "
1113
2223
"
# library(patchwork)
SuppFigure4b <- prjcommplotalt + depcommplotalt + guide_area() +
plot_layout(design=design, guides = 'collect', widths = c(4,.6)) +
plot_annotation(
title = 'SUPPLEMENTAL FIGURE 4b\nMarginal Effects of 2-Category Change in Stress on Outcome Probabilities, by Community SES & Urban/Rural Location',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Estimates derived from 14 multivariate (using `brms::mvbind()`) and multilevel between-within Bayesian logistic regression models simultaneously regressing all six specific criminal intent outcomes (7 models) and all seven negative emotion outcomes (7 models) separately on each of the seven stress types, and seven separate models regressing "any criminal intent" on stress. Stress items were separated into a L2 cross-time average (Xbar_i) between-person predictor and a L1 within-person change (X_it - Xbar_i) "fixed effects" estimator. Both L1 and L2 stress variables were specified as monotonic ordinal predictors with a cumulative probit link function. Models also included a factor variable for community and a multiplicative interaction between L1 stress change (fixed effects estimator) and community. Models were estimated in brms with 4 chains and 4000 total post-warmup posterior draws per outcome and per community group. Marginal effect contrast distributions were estimated from the expectation of the posterior predictive distribution for each model as predicted probability difference distributions for 2-category stress increases (T1 to T2 change) by community, averaged over all between-person stress levels. Median posterior density estimates with 95% intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates for the average marginal effect contrast are greater than zero.', width=180)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'right',
legend.key.height = unit(1, 'cm'))
# SuppFigure4b
ggsave("SuppFigure4b.jpeg", width=9, height=6.5, path=here("Output"))9.1.3.1 Supp. Fig4a (Alt. groupings)
SuppFigure4a
9.1.3.2 Supp. Fig4b (Alt. groupings)
SuppFigure4b
9.2 FIGURE 4 - Stress Chg on Outcomes by Community
numcommlabs2 <- c("1"="Rural\nLowSES",
"2"="Rural\nHighSES",
"3"="Urban\nLowSES",
"4"="Urban\nHighSES")
prjlabs2 <- c(
"prjthflt5"="Theft\n<5BAM",
"prjthfgt5"="Theft\n>5BAM",
"prjthreat"="Threat",
"prjharm"="Phys.\nharm",
"prjusedrg"="Use\ndrugs",
"prjhack"="Hack\ninfo",
"prjany"="Any crime")
deplabs2 <- c(
"depcantgo"="Can't\ngo",
"depeffort"="Effort",
"deplonely"="Lonely",
"depblues"="Blues",
"depunfair"="Unfair",
"depmistrt"="Mistreated",
"depbetray"="Betrayed")
stress_varlabs <- c(
"Stress:\nMoney" = "Money",
"Stress:\nTransport" = "Transport",
"Stress:\nRespect" = "Respect",
"Stress:\nFair Trtmt" = "Fair Trtmt",
"Stress:\nJob" = "Job",
"Stress:\nTheft Vctm" = "Theft",
"Stress:\nAssault Vctm" = "Assault")
prjcommplotalt2 <- ggplot(data = PLMEprjchg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
group=stress_varf,
color=stress_varf)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLMEprjchg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med),
group_by=stress_varf),
alpha=p80_gt0),
.width = .95, size=.7,
position = position_dodge(width=.7)) +
scale_color_viridis_d(labels=as_labeller(stress_varlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.25,.51)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numcommlabs2) +
labs(subtitle='Criminal Intent\n___________________________________________________________') +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size =8, angle=65),
axis.text.y = element_text(size=10),
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
legend.text = element_text(size = 10),
legend.title=element_text(size=10, face="italic")) +
guides(shape = guide_legend(override.aes = list(size = 0.5)),
color = guide_legend(nrow=1, reverse = TRUE, title="Stress Item:"))
# prjcommplotalt2
depcommplotalt2 <- ggplot(data = PLMEdepchg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
group=stress_varf,
color=stress_varf)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLMEdepchg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med),
group_by=stress_varf),
alpha=p80_gt0),
.width = .95, size=.7,
position = position_dodge(width=.7)) +
scale_color_viridis_d(labels=as_labeller(stress_varlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.25,.71)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numcommlabs2) +
labs(subtitle='Negative Emotions\n_____________________________________________________________') +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size = 8, angle=65),
axis.text.y=element_blank(), #remove y-axis labels
axis.ticks.y=element_blank(), #remove y-axis ticks
axis.line.y=element_blank(), #remove y-axis line
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
legend.text = element_text(size = 10),
legend.title=element_text(size=10, face="italic")) +
guides(shape = guide_legend(override.aes = list(size = 0.5)),
color = guide_legend(nrow=1, reverse = TRUE, title="Stress Item:"))
# depcommplotalt2
design <- "
12
33
"
Figure4 <- prjcommplotalt2 + depcommplotalt2 +
guide_area() +
plot_layout(design=design, guides = 'collect', heights = c(4,.1)) +
plot_annotation(
title = 'FIGURE 4\nMarginal Effect of 2-Category Within-Person Increase in Stress on Change in Outcome Probabilities, by Community',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Each of the 392 intervals displayed represents the estimated marginal effect of a "practically large" 2-category increase in stress on an outcome probability derived from a fixed effects estimator in 98 distinct Bayesian multilevel between/within logistic regression models (98 models*4 estimates per community). Of these, 364 estimates are from 91 multivariate models simultaneously regressing (using `brms::mvbind()`) from multivariate models simultaneously regressing (using `brms:: mvbind()`) the six specific criminal intent outcomes or the seven negative emotions outcomes on each of the seven stress types (13*7=91 models). The other 28 estimates are from seven models regressing "any criminal intent" on each stress type. Stress items were separated into a L2 cross-time average (Xbar_i) between-person predictor and a L1 within-person change (X_it - Xbar_i) "fixed effects" estimator. Both L1 and L2 stress variables were specified as monotonic ordinal predictors with a cumulative probit link function. Models also included a factor variable for community and a multiplicative interaction between L1 stress change (fixed effects estimator) and community. Models were estimated in brms with 4 chains and 4000 total post-warmup posterior draws per outcome and per community group. Marginal effect contrast distributions were estimated from the expectation of the posterior predictive distribution for each model as predicted probability difference distributions for 2-category stress increases (T1 to T2 change) by community, averaged over all between-person stress levels. Median posterior density estimates with 95% intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates for the average marginal effect contrast are greater than zero.', width=195)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'bottom',
legend.key.height = unit(.5, 'cm'),
legend.margin = margin(0,0,0,0),
legend.spacing.y = unit(0, "mm"))
Figure4
ggsave("Figure4.jpeg", width=9, height=6.5, path=here("Output"))As before, let’s pull all these plotted estimates into tables.
Again, we will specifically report some of the bold or “plausibly positive” (i.e., >=80% posterior probability of PLME > 0) in the text, so we make those easier to find by filtering to keep only those bold estimates. We will also report all estimates plotted in Figure 4 in separate tables.
9.2.1 Estimates plotted in Fig4
9.2.1.1 Bold crim intent (chg, >=80% ppd positive)
PLMEprjchg_comm %>%
mutate(
rural.ses.med = if_else(
rural.ses.med == "1", "Rural/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "2", "Rural/HighSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "3", "Urban/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "4", "Urban/HighSES", rural.ses.med)
) %>%
group_by(.category, stress_var, rural.ses.med) %>%
mutate(
n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests
) %>%
dplyr::filter(p_gt0 >= .80) %>%
# summarize PLME estimates across the MCMC draws
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (Chg) by Community PLME Estimates**"),
subtitle = md("Plotted in FIG4 (FIG5 in paper)")
)| Criminal Intent (Chg) by Community PLME Estimates | ||||
| Plotted in FIG4 (FIG5 in paper) | ||||
| rural.ses.med | m | s | ll | ul |
|---|---|---|---|---|
| prjthflt5 - Stress: Assault Vctm | ||||
| Urban/HighSES | 0.034 | 0.096 | -0.0354 | 0.34 |
| prjthflt5 - Stress: Fair Trtmt | ||||
| Urban/LowSES | 0.021 | 0.057 | -0.0113 | 0.17 |
| prjthflt5 - Stress: Job | ||||
| Urban/LowSES | 0.021 | 0.064 | -0.0089 | 0.22 |
| prjthflt5 - Stress: Money | ||||
| Urban/HighSES | 0.089 | 0.159 | 0.0037 | 0.63 |
| Urban/LowSES | 0.028 | 0.093 | -0.0045 | 0.32 |
| prjthflt5 - Stress: Transport | ||||
| Urban/HighSES | 0.069 | 0.176 | -0.0189 | 0.68 |
| prjthfgt5 - Stress: Assault Vctm | ||||
| Urban/HighSES | 0.060 | 0.108 | -0.0255 | 0.38 |
| prjthfgt5 - Stress: Fair Trtmt | ||||
| Urban/HighSES | 0.062 | 0.121 | -0.0223 | 0.46 |
| Urban/LowSES | 0.016 | 0.049 | -0.0251 | 0.15 |
| prjthfgt5 - Stress: Job | ||||
| Urban/LowSES | 0.033 | 0.072 | -0.0088 | 0.26 |
| prjthfgt5 - Stress: Money | ||||
| Urban/HighSES | 0.112 | 0.154 | 0.0088 | 0.61 |
| Urban/LowSES | 0.036 | 0.088 | -0.0050 | 0.30 |
| prjthfgt5 - Stress: Transport | ||||
| Urban/HighSES | 0.065 | 0.127 | -0.0231 | 0.48 |
| prjusedrg - Stress: Assault Vctm | ||||
| Urban/HighSES | 0.024 | 0.069 | -0.0211 | 0.24 |
| prjany - Stress: Assault Vctm | ||||
| Urban/HighSES | 0.182 | 0.198 | -0.1462 | 0.65 |
| prjany - Stress: Fair Trtmt | ||||
| Urban/HighSES | 0.209 | 0.189 | -0.0619 | 0.68 |
| Urban/LowSES | 0.042 | 0.086 | -0.0577 | 0.28 |
| prjany - Stress: Job | ||||
| Urban/LowSES | 0.081 | 0.098 | -0.0239 | 0.36 |
| prjany - Stress: Money | ||||
| Urban/HighSES | 0.320 | 0.219 | -0.0292 | 0.82 |
| Urban/LowSES | 0.105 | 0.140 | -0.0211 | 0.56 |
| prjany - Stress: Transport | ||||
| Urban/HighSES | 0.297 | 0.232 | -0.0586 | 0.84 |
9.2.1.2 Bold neg emotions (chg, >=80% ppd positive)
PLMEdepchg_comm %>%
mutate(
rural.ses.med = if_else(
rural.ses.med == "1", "Rural/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "2", "Rural/HighSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "3", "Urban/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "4", "Urban/HighSES", rural.ses.med)
) %>%
group_by(.category, stress_var, rural.ses.med) %>%
mutate(
n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests
) %>%
dplyr::filter(p_gt0 >= .80) %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var, rural.ses.med) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (Chg) by Community PLME Estimates**"),
subtitle = md("Plotted in FIG4 (FIG5 in paper)")
)| Negative Emotions (Chg) by Community PLME Estimates | ||||
| Plotted in FIG4 (FIG5 in paper) | ||||
| rural.ses.med | m | s | ll | ul |
|---|---|---|---|---|
| depcantgo - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.513 | 0.183 | 0.15217 | 0.85 |
| Rural/LowSES | 0.147 | 0.134 | -0.10261 | 0.43 |
| depcantgo - Stress: Job | ||||
| Rural/HighSES | 0.406 | 0.177 | 0.03778 | 0.73 |
| Rural/LowSES | 0.221 | 0.120 | -0.03806 | 0.44 |
| Urban/HighSES | 0.269 | 0.228 | -0.27323 | 0.63 |
| depcantgo - Stress: Money | ||||
| Rural/HighSES | 0.203 | 0.217 | -0.17191 | 0.70 |
| Rural/LowSES | 0.219 | 0.118 | -0.00383 | 0.46 |
| Urban/HighSES | 0.301 | 0.218 | -0.14792 | 0.74 |
| depcantgo - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.481 | 0.185 | 0.09190 | 0.81 |
| Rural/LowSES | 0.116 | 0.135 | -0.16929 | 0.38 |
| depcantgo - Stress: Transport | ||||
| Rural/HighSES | 0.481 | 0.182 | 0.10531 | 0.81 |
| Rural/LowSES | 0.176 | 0.109 | -0.03484 | 0.38 |
| Urban/HighSES | 0.277 | 0.181 | -0.09016 | 0.65 |
| depeffort - Stress: Assault Vctm | ||||
| Urban/LowSES | 0.132 | 0.171 | -0.18348 | 0.53 |
| depeffort - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.128 | 0.163 | -0.08649 | 0.59 |
| Urban/LowSES | 0.242 | 0.165 | -0.01066 | 0.66 |
| depeffort - Stress: Job | ||||
| Urban/HighSES | 0.181 | 0.207 | -0.13394 | 0.68 |
| depeffort - Stress: Money | ||||
| Urban/HighSES | 0.180 | 0.166 | -0.14306 | 0.55 |
| depeffort - Stress: Respect | ||||
| Rural/HighSES | 0.119 | 0.126 | -0.07022 | 0.45 |
| Urban/LowSES | 0.134 | 0.139 | -0.07376 | 0.48 |
| depeffort - Stress: Theft Vctm | ||||
| Urban/HighSES | 0.133 | 0.169 | -0.14109 | 0.53 |
| Urban/LowSES | 0.117 | 0.127 | -0.14066 | 0.37 |
| deplonely - Stress: Assault Vctm | ||||
| Rural/HighSES | 0.261 | 0.180 | -0.05242 | 0.65 |
| Urban/HighSES | 0.306 | 0.192 | -0.08557 | 0.67 |
| Urban/LowSES | 0.187 | 0.203 | -0.25643 | 0.57 |
| deplonely - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.162 | 0.160 | -0.16396 | 0.49 |
| deplonely - Stress: Job | ||||
| Rural/LowSES | 0.141 | 0.107 | -0.06212 | 0.36 |
| deplonely - Stress: Money | ||||
| Urban/HighSES | 0.428 | 0.205 | 0.05703 | 0.84 |
| deplonely - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.292 | 0.217 | -0.07986 | 0.82 |
| Rural/LowSES | 0.201 | 0.120 | -0.04916 | 0.43 |
| Urban/HighSES | 0.249 | 0.196 | -0.13226 | 0.66 |
| deplonely - Stress: Transport | ||||
| Urban/HighSES | 0.324 | 0.177 | -0.04336 | 0.65 |
| depblues - Stress: Theft Vctm | ||||
| Urban/HighSES | 0.392 | 0.199 | 0.07894 | 0.81 |
| depunfair - Stress: Assault Vctm | ||||
| Urban/HighSES | 0.169 | 0.186 | -0.20791 | 0.51 |
| depunfair - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.316 | 0.172 | 0.00768 | 0.71 |
| Rural/LowSES | 0.116 | 0.086 | -0.03856 | 0.31 |
| Urban/HighSES | 0.308 | 0.192 | -0.09554 | 0.67 |
| Urban/LowSES | 0.157 | 0.158 | -0.12547 | 0.50 |
| depunfair - Stress: Job | ||||
| Rural/HighSES | 0.303 | 0.149 | 0.03667 | 0.63 |
| Urban/HighSES | 0.532 | 0.155 | 0.18758 | 0.79 |
| depunfair - Stress: Money | ||||
| Rural/HighSES | 0.284 | 0.166 | -0.02701 | 0.66 |
| Rural/LowSES | 0.140 | 0.082 | -0.00579 | 0.32 |
| Urban/HighSES | 0.283 | 0.198 | -0.17381 | 0.64 |
| Urban/LowSES | 0.173 | 0.177 | -0.17580 | 0.55 |
| depunfair - Stress: Respect | ||||
| Rural/HighSES | 0.388 | 0.145 | 0.15582 | 0.72 |
| Urban/HighSES | 0.253 | 0.157 | -0.05980 | 0.55 |
| Urban/LowSES | 0.117 | 0.144 | -0.15130 | 0.42 |
| depunfair - Stress: Theft Vctm | ||||
| Urban/HighSES | 0.317 | 0.156 | -0.00011 | 0.61 |
| Urban/LowSES | 0.310 | 0.136 | 0.05076 | 0.60 |
| depunfair - Stress: Transport | ||||
| Rural/HighSES | 0.314 | 0.179 | -0.10265 | 0.67 |
| Rural/LowSES | 0.130 | 0.080 | -0.03927 | 0.28 |
| Urban/HighSES | 0.344 | 0.177 | -0.04847 | 0.67 |
| Urban/LowSES | 0.249 | 0.169 | -0.11300 | 0.60 |
| depmistrt - Stress: Assault Vctm | ||||
| Urban/HighSES | 0.254 | 0.166 | -0.06804 | 0.57 |
| depmistrt - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.161 | 0.180 | -0.10357 | 0.64 |
| Urban/LowSES | 0.159 | 0.124 | -0.09339 | 0.40 |
| depmistrt - Stress: Money | ||||
| Urban/HighSES | 0.262 | 0.230 | -0.05523 | 0.84 |
| depmistrt - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.255 | 0.210 | -0.07162 | 0.79 |
| Urban/HighSES | 0.153 | 0.155 | -0.16415 | 0.45 |
| depmistrt - Stress: Transport | ||||
| Rural/LowSES | 0.058 | 0.076 | -0.09714 | 0.21 |
| depbetray - Stress: Assault Vctm | ||||
| Urban/HighSES | 0.299 | 0.165 | -0.03701 | 0.61 |
| Urban/LowSES | 0.213 | 0.179 | -0.10479 | 0.63 |
| depbetray - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.116 | 0.163 | -0.11443 | 0.56 |
| Urban/LowSES | 0.112 | 0.130 | -0.14435 | 0.37 |
| depbetray - Stress: Job | ||||
| Rural/HighSES | 0.331 | 0.145 | 0.10439 | 0.68 |
| depbetray - Stress: Money | ||||
| Rural/HighSES | 0.125 | 0.133 | -0.11833 | 0.42 |
| Urban/HighSES | 0.154 | 0.182 | -0.19427 | 0.56 |
| depbetray - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.261 | 0.215 | -0.01388 | 0.86 |
| Urban/LowSES | 0.247 | 0.131 | 0.02634 | 0.55 |
9.2.1.3 All crim intent (chg)
PLMEprjchg_comm %>%
mutate(
rural.ses.med = if_else(
rural.ses.med == "1", "Rural/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "2", "Rural/HighSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "3", "Urban/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "4", "Urban/HighSES", rural.ses.med)
) %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var, rural.ses.med) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (Chg) by Community PLME Estimates**"),
subtitle = md("Plotted in FIG4 (FIG5 in paper)")
)| Criminal Intent (Chg) by Community PLME Estimates | ||||
| Plotted in FIG4 (FIG5 in paper) | ||||
| rural.ses.med | m | s | ll | ul |
|---|---|---|---|---|
| prjthflt5 - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.00158094 | 0.0062 | -0.0163 | 0.0011976 |
| Rural/LowSES | -0.00090159 | 0.0043 | -0.0107 | 0.0061629 |
| Urban/HighSES | 0.03443182 | 0.0956 | -0.0354 | 0.3385213 |
| Urban/LowSES | -0.00047146 | 0.0630 | -0.0558 | 0.1197702 |
| prjthflt5 - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00163958 | 0.0072 | -0.0177 | 0.0013682 |
| Rural/LowSES | -0.00047637 | 0.0043 | -0.0094 | 0.0077465 |
| Urban/HighSES | 0.01669615 | 0.0860 | -0.0621 | 0.2634245 |
| Urban/LowSES | 0.02050070 | 0.0565 | -0.0113 | 0.1729251 |
| prjthflt5 - Stress: Job | ||||
| Rural/HighSES | -0.00260125 | 0.0068 | -0.0211 | -0.0001487 |
| Rural/LowSES | -0.00107765 | 0.0036 | -0.0100 | 0.0046447 |
| Urban/HighSES | 0.02572255 | 0.0897 | -0.0554 | 0.2923601 |
| Urban/LowSES | 0.02113944 | 0.0645 | -0.0089 | 0.2223953 |
| prjthflt5 - Stress: Money | ||||
| Rural/HighSES | -0.00130638 | 0.0048 | -0.0134 | 0.0002232 |
| Rural/LowSES | 0.00012116 | 0.0032 | -0.0052 | 0.0074263 |
| Urban/HighSES | 0.08874325 | 0.1595 | 0.0037 | 0.6251763 |
| Urban/LowSES | 0.02827562 | 0.0934 | -0.0045 | 0.3232879 |
| prjthflt5 - Stress: Respect | ||||
| Rural/HighSES | -0.00275302 | 0.0071 | -0.0220 | 0.0000521 |
| Rural/LowSES | -0.00279376 | 0.0054 | -0.0178 | 0.0034116 |
| Urban/HighSES | 0.01346887 | 0.0666 | -0.0674 | 0.1901647 |
| Urban/LowSES | -0.00816063 | 0.0241 | -0.0699 | 0.0257236 |
| prjthflt5 - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.00317805 | 0.0104 | -0.0326 | 0.0000317 |
| Rural/LowSES | -0.00211868 | 0.0055 | -0.0174 | 0.0038382 |
| Urban/HighSES | 0.00528224 | 0.0565 | -0.0858 | 0.1418435 |
| Urban/LowSES | 0.01102381 | 0.0885 | -0.0303 | 0.2743924 |
| prjthflt5 - Stress: Transport | ||||
| Rural/HighSES | -0.00169804 | 0.0059 | -0.0192 | 0.0018803 |
| Rural/LowSES | -0.00076484 | 0.0041 | -0.0095 | 0.0062063 |
| Urban/HighSES | 0.06862394 | 0.1758 | -0.0189 | 0.6756410 |
| Urban/LowSES | 0.00557113 | 0.0683 | -0.0413 | 0.1832424 |
| prjthfgt5 - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.00218675 | 0.0081 | -0.0218 | 0.0010258 |
| Rural/LowSES | -0.00058844 | 0.0070 | -0.0148 | 0.0135791 |
| Urban/HighSES | 0.05956849 | 0.1080 | -0.0255 | 0.3838897 |
| Urban/LowSES | 0.00537067 | 0.0681 | -0.0556 | 0.1827703 |
| prjthfgt5 - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00205920 | 0.0072 | -0.0215 | 0.0013516 |
| Rural/LowSES | 0.00002924 | 0.0055 | -0.0095 | 0.0118691 |
| Urban/HighSES | 0.06176543 | 0.1212 | -0.0223 | 0.4646645 |
| Urban/LowSES | 0.01567591 | 0.0486 | -0.0251 | 0.1479464 |
| prjthfgt5 - Stress: Job | ||||
| Rural/HighSES | -0.00269774 | 0.0093 | -0.0249 | 0.0013454 |
| Rural/LowSES | -0.00109126 | 0.0058 | -0.0140 | 0.0099578 |
| Urban/HighSES | 0.02978770 | 0.0921 | -0.0771 | 0.2885130 |
| Urban/LowSES | 0.03288692 | 0.0719 | -0.0088 | 0.2599347 |
| prjthfgt5 - Stress: Money | ||||
| Rural/HighSES | -0.00146022 | 0.0073 | -0.0155 | 0.0012389 |
| Rural/LowSES | 0.00063480 | 0.0043 | -0.0060 | 0.0121152 |
| Urban/HighSES | 0.11155557 | 0.1537 | 0.0088 | 0.6074268 |
| Urban/LowSES | 0.03578330 | 0.0877 | -0.0050 | 0.2958762 |
| prjthfgt5 - Stress: Respect | ||||
| Rural/HighSES | -0.00322876 | 0.0087 | -0.0292 | 0.0004675 |
| Rural/LowSES | -0.00344126 | 0.0071 | -0.0223 | 0.0055750 |
| Urban/HighSES | 0.01211397 | 0.0724 | -0.0837 | 0.2046446 |
| Urban/LowSES | -0.01052425 | 0.0311 | -0.0894 | 0.0363386 |
| prjthfgt5 - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.00578987 | 0.0161 | -0.0510 | -0.0001861 |
| Rural/LowSES | -0.00289224 | 0.0094 | -0.0268 | 0.0110698 |
| Urban/HighSES | 0.02181668 | 0.0758 | -0.0880 | 0.2108915 |
| Urban/LowSES | 0.01349434 | 0.0917 | -0.0488 | 0.3023956 |
| prjthfgt5 - Stress: Transport | ||||
| Rural/HighSES | -0.00260686 | 0.0094 | -0.0295 | 0.0008443 |
| Rural/LowSES | -0.00072850 | 0.0064 | -0.0144 | 0.0113500 |
| Urban/HighSES | 0.06536789 | 0.1272 | -0.0231 | 0.4825420 |
| Urban/LowSES | -0.00008400 | 0.0643 | -0.0627 | 0.1480932 |
| prjthreat - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.00089617 | 0.0061 | -0.0124 | 0.0037547 |
| Rural/LowSES | -0.00084824 | 0.0033 | -0.0085 | 0.0029524 |
| Urban/HighSES | -0.00573692 | 0.0255 | -0.0657 | 0.0370181 |
| Urban/LowSES | -0.00147528 | 0.0266 | -0.0248 | 0.0285686 |
| prjthreat - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00166048 | 0.0063 | -0.0187 | 0.0010639 |
| Rural/LowSES | -0.00152362 | 0.0035 | -0.0109 | 0.0011499 |
| Urban/HighSES | -0.00037942 | 0.0468 | -0.0477 | 0.1125599 |
| Urban/LowSES | -0.00101389 | 0.0172 | -0.0204 | 0.0238118 |
| prjthreat - Stress: Job | ||||
| Rural/HighSES | -0.00091874 | 0.0115 | -0.0122 | 0.0074216 |
| Rural/LowSES | -0.00055650 | 0.0031 | -0.0079 | 0.0047030 |
| Urban/HighSES | 0.00762679 | 0.0449 | -0.0410 | 0.1261608 |
| Urban/LowSES | 0.00093106 | 0.0153 | -0.0154 | 0.0417247 |
| prjthreat - Stress: Money | ||||
| Rural/HighSES | -0.00106683 | 0.0047 | -0.0127 | 0.0014391 |
| Rural/LowSES | -0.00041019 | 0.0025 | -0.0061 | 0.0035088 |
| Urban/HighSES | 0.00833897 | 0.0722 | -0.0276 | 0.2143597 |
| Urban/LowSES | -0.00007177 | 0.0231 | -0.0175 | 0.0381684 |
| prjthreat - Stress: Respect | ||||
| Rural/HighSES | -0.00145201 | 0.0068 | -0.0177 | 0.0031175 |
| Rural/LowSES | -0.00220404 | 0.0042 | -0.0148 | 0.0005980 |
| Urban/HighSES | -0.00011453 | 0.0390 | -0.0518 | 0.0903556 |
| Urban/LowSES | -0.00476666 | 0.0123 | -0.0402 | 0.0059121 |
| prjthreat - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.00153573 | 0.0211 | -0.0230 | 0.0090816 |
| Rural/LowSES | -0.00181583 | 0.0049 | -0.0155 | 0.0026505 |
| Urban/HighSES | -0.00797705 | 0.0322 | -0.0859 | 0.0423985 |
| Urban/LowSES | -0.00286577 | 0.0190 | -0.0356 | 0.0210013 |
| prjthreat - Stress: Transport | ||||
| Rural/HighSES | -0.00133909 | 0.0059 | -0.0165 | 0.0008593 |
| Rural/LowSES | -0.00086240 | 0.0026 | -0.0081 | 0.0024544 |
| Urban/HighSES | -0.00309687 | 0.0376 | -0.0566 | 0.0708090 |
| Urban/LowSES | -0.00152914 | 0.0272 | -0.0219 | 0.0237147 |
| prjharm - Stress: Assault Vctm | ||||
| Rural/HighSES | 0.00029257 | 0.0287 | -0.0094 | 0.0592233 |
| Rural/LowSES | -0.00045576 | 0.0040 | -0.0087 | 0.0073557 |
| Urban/HighSES | -0.00037847 | 0.0221 | -0.0325 | 0.0479820 |
| Urban/LowSES | -0.00277390 | 0.0185 | -0.0315 | 0.0151167 |
| prjharm - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00161606 | 0.0103 | -0.0191 | 0.0071763 |
| Rural/LowSES | -0.00143353 | 0.0038 | -0.0117 | 0.0023154 |
| Urban/HighSES | -0.00750310 | 0.0233 | -0.0592 | 0.0194579 |
| Urban/LowSES | -0.00033785 | 0.0203 | -0.0158 | 0.0303686 |
| prjharm - Stress: Job | ||||
| Rural/HighSES | -0.00127812 | 0.0077 | -0.0164 | 0.0086614 |
| Rural/LowSES | -0.00011182 | 0.0038 | -0.0076 | 0.0077997 |
| Urban/HighSES | -0.00295323 | 0.0254 | -0.0419 | 0.0409323 |
| Urban/LowSES | -0.00165096 | 0.0106 | -0.0211 | 0.0141202 |
| prjharm - Stress: Money | ||||
| Rural/HighSES | -0.00122517 | 0.0081 | -0.0161 | 0.0082017 |
| Rural/LowSES | -0.00047077 | 0.0032 | -0.0074 | 0.0051510 |
| Urban/HighSES | 0.00292301 | 0.0545 | -0.0249 | 0.1131868 |
| Urban/LowSES | -0.00082693 | 0.0217 | -0.0218 | 0.0296146 |
| prjharm - Stress: Respect | ||||
| Rural/HighSES | -0.00162800 | 0.0080 | -0.0188 | 0.0050562 |
| Rural/LowSES | -0.00134174 | 0.0039 | -0.0122 | 0.0030485 |
| Urban/HighSES | -0.00868173 | 0.0196 | -0.0631 | 0.0064259 |
| Urban/LowSES | -0.00044546 | 0.0109 | -0.0177 | 0.0223627 |
| prjharm - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.00346533 | 0.0145 | -0.0402 | 0.0023111 |
| Rural/LowSES | -0.00262738 | 0.0060 | -0.0216 | 0.0009913 |
| Urban/HighSES | -0.00706313 | 0.0223 | -0.0624 | 0.0192608 |
| Urban/LowSES | -0.00332707 | 0.0130 | -0.0336 | 0.0119253 |
| prjharm - Stress: Transport | ||||
| Rural/HighSES | -0.00387007 | 0.0120 | -0.0387 | 0.0000054 |
| Rural/LowSES | -0.00197644 | 0.0043 | -0.0145 | 0.0018270 |
| Urban/HighSES | -0.01149533 | 0.0238 | -0.0757 | 0.0090732 |
| Urban/LowSES | -0.00055698 | 0.0263 | -0.0196 | 0.0437102 |
| prjusedrg - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.00082361 | 0.0134 | -0.0147 | 0.0109883 |
| Rural/LowSES | -0.00045749 | 0.0055 | -0.0116 | 0.0080974 |
| Urban/HighSES | 0.02448915 | 0.0688 | -0.0211 | 0.2404848 |
| Urban/LowSES | -0.00154845 | 0.0162 | -0.0193 | 0.0074088 |
| prjusedrg - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00282678 | 0.0122 | -0.0266 | 0.0011003 |
| Rural/LowSES | -0.00176450 | 0.0040 | -0.0136 | 0.0016870 |
| Urban/HighSES | 0.00924337 | 0.0648 | -0.0375 | 0.1744457 |
| Urban/LowSES | -0.00195983 | 0.0077 | -0.0206 | 0.0037780 |
| prjusedrg - Stress: Job | ||||
| Rural/HighSES | -0.00094474 | 0.0120 | -0.0169 | 0.0150446 |
| Rural/LowSES | -0.00028824 | 0.0044 | -0.0089 | 0.0092898 |
| Urban/HighSES | 0.00382895 | 0.0520 | -0.0596 | 0.1346532 |
| Urban/LowSES | -0.00174610 | 0.0075 | -0.0195 | 0.0054332 |
| prjusedrg - Stress: Money | ||||
| Rural/HighSES | -0.00091823 | 0.0175 | -0.0177 | 0.0114634 |
| Rural/LowSES | -0.00092759 | 0.0037 | -0.0097 | 0.0036332 |
| Urban/HighSES | 0.00374948 | 0.0726 | -0.0529 | 0.1803764 |
| Urban/LowSES | -0.00165640 | 0.0089 | -0.0184 | 0.0055722 |
| prjusedrg - Stress: Respect | ||||
| Rural/HighSES | -0.00236949 | 0.0085 | -0.0231 | 0.0041981 |
| Rural/LowSES | -0.00287318 | 0.0053 | -0.0171 | 0.0005417 |
| Urban/HighSES | -0.00465803 | 0.0340 | -0.0718 | 0.0620390 |
| Urban/LowSES | -0.00305129 | 0.0080 | -0.0251 | 0.0015024 |
| prjusedrg - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.00092294 | 0.0265 | -0.0151 | 0.0121010 |
| Rural/LowSES | -0.00075291 | 0.0034 | -0.0093 | 0.0048510 |
| Urban/HighSES | 0.00087451 | 0.0316 | -0.0516 | 0.0718913 |
| Urban/LowSES | -0.00163587 | 0.0061 | -0.0175 | 0.0041290 |
| prjusedrg - Stress: Transport | ||||
| Rural/HighSES | -0.00192269 | 0.0095 | -0.0203 | 0.0034057 |
| Rural/LowSES | -0.00128741 | 0.0038 | -0.0106 | 0.0036551 |
| Urban/HighSES | -0.00908540 | 0.0397 | -0.0826 | 0.0538457 |
| Urban/LowSES | -0.00261160 | 0.0081 | -0.0252 | 0.0014759 |
| prjhack - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.00520953 | 0.0246 | -0.0578 | 0.0376390 |
| Rural/LowSES | 0.00000095 | 0.0192 | -0.0335 | 0.0438488 |
| Urban/HighSES | -0.01130634 | 0.0654 | -0.1451 | 0.1290772 |
| Urban/LowSES | 0.00366818 | 0.0726 | -0.0692 | 0.1860251 |
| prjhack - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00805015 | 0.0306 | -0.0775 | 0.0305239 |
| Rural/LowSES | -0.00890961 | 0.0165 | -0.0504 | 0.0146669 |
| Urban/HighSES | -0.03254606 | 0.0630 | -0.1665 | 0.0778911 |
| Urban/LowSES | -0.01579531 | 0.0383 | -0.1002 | 0.0449155 |
| prjhack - Stress: Job | ||||
| Rural/HighSES | -0.00534077 | 0.0242 | -0.0555 | 0.0403825 |
| Rural/LowSES | 0.00327542 | 0.0170 | -0.0261 | 0.0437748 |
| Urban/HighSES | -0.01127410 | 0.0530 | -0.1041 | 0.1105256 |
| Urban/LowSES | -0.00948798 | 0.0283 | -0.0690 | 0.0419329 |
| prjhack - Stress: Money | ||||
| Rural/HighSES | -0.00399020 | 0.0268 | -0.0512 | 0.0400320 |
| Rural/LowSES | 0.00183507 | 0.0153 | -0.0246 | 0.0402821 |
| Urban/HighSES | 0.02836875 | 0.0899 | -0.0735 | 0.2814083 |
| Urban/LowSES | -0.00462542 | 0.0516 | -0.0771 | 0.1063139 |
| prjhack - Stress: Respect | ||||
| Rural/HighSES | -0.00846566 | 0.0251 | -0.0728 | 0.0238427 |
| Rural/LowSES | -0.00976495 | 0.0171 | -0.0569 | 0.0137854 |
| Urban/HighSES | -0.02290337 | 0.0512 | -0.1493 | 0.0655700 |
| Urban/LowSES | -0.02701212 | 0.0413 | -0.1430 | 0.0142090 |
| prjhack - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.01201587 | 0.0384 | -0.1014 | 0.0130728 |
| Rural/LowSES | -0.00835322 | 0.0182 | -0.0527 | 0.0183392 |
| Urban/HighSES | -0.02395402 | 0.0662 | -0.1622 | 0.1141607 |
| Urban/LowSES | -0.00230733 | 0.0432 | -0.0769 | 0.1012141 |
| prjhack - Stress: Transport | ||||
| Rural/HighSES | -0.01182366 | 0.0275 | -0.0892 | 0.0099562 |
| Rural/LowSES | -0.00881672 | 0.0173 | -0.0531 | 0.0166542 |
| Urban/HighSES | -0.05009037 | 0.0671 | -0.2308 | 0.0383661 |
| Urban/LowSES | -0.00966014 | 0.0586 | -0.1067 | 0.0942147 |
| prjany - Stress: Assault Vctm | ||||
| Rural/HighSES | 0.00187526 | 0.0358 | -0.0237 | 0.0870979 |
| Rural/LowSES | -0.00147053 | 0.0190 | -0.0355 | 0.0436702 |
| Urban/HighSES | 0.18166602 | 0.1979 | -0.1462 | 0.6529999 |
| Urban/LowSES | -0.03652088 | 0.0974 | -0.2025 | 0.1739317 |
| prjany - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00935075 | 0.0219 | -0.0699 | 0.0101260 |
| Rural/LowSES | -0.00346379 | 0.0146 | -0.0360 | 0.0237269 |
| Urban/HighSES | 0.20902936 | 0.1893 | -0.0619 | 0.6838385 |
| Urban/LowSES | 0.04205541 | 0.0863 | -0.0577 | 0.2840469 |
| prjany - Stress: Job | ||||
| Rural/HighSES | -0.00835722 | 0.0197 | -0.0584 | 0.0124868 |
| Rural/LowSES | 0.00159856 | 0.0163 | -0.0253 | 0.0416329 |
| Urban/HighSES | 0.09981052 | 0.2022 | -0.2459 | 0.5728979 |
| Urban/LowSES | 0.08056247 | 0.0978 | -0.0239 | 0.3609980 |
| prjany - Stress: Money | ||||
| Rural/HighSES | -0.00408195 | 0.0178 | -0.0399 | 0.0217676 |
| Rural/LowSES | 0.00728413 | 0.0171 | -0.0125 | 0.0548368 |
| Urban/HighSES | 0.32012319 | 0.2186 | -0.0292 | 0.8247629 |
| Urban/LowSES | 0.10544514 | 0.1400 | -0.0211 | 0.5572340 |
| prjany - Stress: Respect | ||||
| Rural/HighSES | -0.01109630 | 0.0215 | -0.0731 | 0.0098086 |
| Rural/LowSES | -0.01401251 | 0.0173 | -0.0601 | 0.0092845 |
| Urban/HighSES | 0.06799576 | 0.1602 | -0.2328 | 0.4343151 |
| Urban/LowSES | -0.03191680 | 0.0634 | -0.1808 | 0.0738670 |
| prjany - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.01545267 | 0.0352 | -0.1096 | 0.0121142 |
| Rural/LowSES | -0.01502300 | 0.0207 | -0.0699 | 0.0109002 |
| Urban/HighSES | 0.06093828 | 0.1698 | -0.2546 | 0.4310546 |
| Urban/LowSES | 0.03213155 | 0.1270 | -0.0990 | 0.4271599 |
| prjany - Stress: Transport | ||||
| Rural/HighSES | -0.01378224 | 0.0284 | -0.1001 | 0.0044937 |
| Rural/LowSES | -0.00153706 | 0.0158 | -0.0321 | 0.0328138 |
| Urban/HighSES | 0.29678607 | 0.2322 | -0.0586 | 0.8380301 |
| Urban/LowSES | 0.04425633 | 0.1262 | -0.0865 | 0.4084651 |
9.2.1.4 All neg emotions (chg)
PLMEdepchg_comm %>%
mutate(
rural.ses.med = if_else(
rural.ses.med == "1", "Rural/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "2", "Rural/HighSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "3", "Urban/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "4", "Urban/HighSES", rural.ses.med)
) %>%
# summarize PLME estimates across the MCMC draws
group_by(.category, stress_var, rural.ses.med) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .025),
ul = quantile(PLME, probs = .975)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (Chg) by Community PLME Estimates**"),
subtitle = md("Plotted in FIG4 (FIG5 in paper)")
)| Negative Emotions (Chg) by Community PLME Estimates | ||||
| Plotted in FIG4 (FIG5 in paper) | ||||
| rural.ses.med | m | s | ll | ul |
|---|---|---|---|---|
| depcantgo - Stress: Assault Vctm | ||||
| Rural/HighSES | 0.15517 | 0.263 | -0.49454 | 0.469 |
| Rural/LowSES | 0.08133 | 0.155 | -0.23944 | 0.366 |
| Urban/HighSES | 0.09498 | 0.196 | -0.29126 | 0.479 |
| Urban/LowSES | 0.13862 | 0.232 | -0.27470 | 0.644 |
| depcantgo - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.51254 | 0.183 | 0.15217 | 0.852 |
| Rural/LowSES | 0.14664 | 0.134 | -0.10261 | 0.426 |
| Urban/HighSES | -0.02861 | 0.207 | -0.43173 | 0.389 |
| Urban/LowSES | 0.09596 | 0.178 | -0.24120 | 0.459 |
| depcantgo - Stress: Job | ||||
| Rural/HighSES | 0.40649 | 0.177 | 0.03778 | 0.725 |
| Rural/LowSES | 0.22088 | 0.120 | -0.03806 | 0.437 |
| Urban/HighSES | 0.26876 | 0.228 | -0.27323 | 0.629 |
| Urban/LowSES | -0.10559 | 0.143 | -0.37845 | 0.169 |
| depcantgo - Stress: Money | ||||
| Rural/HighSES | 0.20286 | 0.217 | -0.17191 | 0.699 |
| Rural/LowSES | 0.21879 | 0.118 | -0.00383 | 0.460 |
| Urban/HighSES | 0.30094 | 0.218 | -0.14792 | 0.740 |
| Urban/LowSES | 0.08431 | 0.195 | -0.31618 | 0.460 |
| depcantgo - Stress: Respect | ||||
| Rural/HighSES | 0.15166 | 0.182 | -0.26863 | 0.456 |
| Rural/LowSES | 0.09172 | 0.127 | -0.16588 | 0.340 |
| Urban/HighSES | -0.04478 | 0.180 | -0.39454 | 0.319 |
| Urban/LowSES | 0.10659 | 0.162 | -0.18675 | 0.455 |
| depcantgo - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.48063 | 0.185 | 0.09190 | 0.810 |
| Rural/LowSES | 0.11646 | 0.135 | -0.16929 | 0.379 |
| Urban/HighSES | 0.12808 | 0.186 | -0.22976 | 0.503 |
| Urban/LowSES | -0.02671 | 0.196 | -0.47259 | 0.294 |
| depcantgo - Stress: Transport | ||||
| Rural/HighSES | 0.48085 | 0.182 | 0.10531 | 0.815 |
| Rural/LowSES | 0.17558 | 0.109 | -0.03484 | 0.383 |
| Urban/HighSES | 0.27680 | 0.181 | -0.09016 | 0.650 |
| Urban/LowSES | -0.09814 | 0.190 | -0.48148 | 0.246 |
| depeffort - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.09778 | 0.113 | -0.33529 | 0.115 |
| Rural/LowSES | 0.03079 | 0.080 | -0.12530 | 0.195 |
| Urban/HighSES | 0.07766 | 0.167 | -0.24811 | 0.405 |
| Urban/LowSES | 0.13220 | 0.171 | -0.18348 | 0.528 |
| depeffort - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.12819 | 0.163 | -0.08649 | 0.595 |
| Rural/LowSES | 0.05641 | 0.074 | -0.08056 | 0.216 |
| Urban/HighSES | 0.05234 | 0.157 | -0.26316 | 0.367 |
| Urban/LowSES | 0.24247 | 0.165 | -0.01066 | 0.658 |
| depeffort - Stress: Job | ||||
| Rural/HighSES | 0.07827 | 0.140 | -0.13428 | 0.453 |
| Rural/LowSES | 0.03149 | 0.065 | -0.09304 | 0.168 |
| Urban/HighSES | 0.18129 | 0.207 | -0.13394 | 0.684 |
| Urban/LowSES | 0.06917 | 0.130 | -0.19891 | 0.313 |
| depeffort - Stress: Money | ||||
| Rural/HighSES | 0.00593 | 0.123 | -0.22862 | 0.253 |
| Rural/LowSES | 0.04208 | 0.069 | -0.09090 | 0.185 |
| Urban/HighSES | 0.18001 | 0.166 | -0.14306 | 0.551 |
| Urban/LowSES | 0.07315 | 0.171 | -0.19019 | 0.515 |
| depeffort - Stress: Respect | ||||
| Rural/HighSES | 0.11890 | 0.126 | -0.07022 | 0.447 |
| Rural/LowSES | 0.02989 | 0.067 | -0.09365 | 0.177 |
| Urban/HighSES | -0.00812 | 0.130 | -0.25650 | 0.254 |
| Urban/LowSES | 0.13392 | 0.139 | -0.07376 | 0.484 |
| depeffort - Stress: Theft Vctm | ||||
| Rural/HighSES | -0.06563 | 0.146 | -0.33657 | 0.239 |
| Rural/LowSES | 0.00452 | 0.070 | -0.12154 | 0.159 |
| Urban/HighSES | 0.13347 | 0.169 | -0.14109 | 0.526 |
| Urban/LowSES | 0.11744 | 0.127 | -0.14066 | 0.366 |
| depeffort - Stress: Transport | ||||
| Rural/HighSES | 0.07507 | 0.152 | -0.22164 | 0.409 |
| Rural/LowSES | 0.01702 | 0.071 | -0.12072 | 0.164 |
| Urban/HighSES | 0.01373 | 0.188 | -0.26800 | 0.529 |
| Urban/LowSES | -0.02394 | 0.150 | -0.30059 | 0.292 |
| deplonely - Stress: Assault Vctm | ||||
| Rural/HighSES | 0.26053 | 0.180 | -0.05242 | 0.651 |
| Rural/LowSES | -0.01032 | 0.149 | -0.31145 | 0.274 |
| Urban/HighSES | 0.30648 | 0.192 | -0.08557 | 0.671 |
| Urban/LowSES | 0.18661 | 0.203 | -0.25643 | 0.565 |
| deplonely - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.16214 | 0.160 | -0.16396 | 0.492 |
| Rural/LowSES | 0.06778 | 0.114 | -0.14359 | 0.315 |
| Urban/HighSES | 0.08941 | 0.210 | -0.31655 | 0.526 |
| Urban/LowSES | 0.02553 | 0.156 | -0.24257 | 0.390 |
| deplonely - Stress: Job | ||||
| Rural/HighSES | -0.04690 | 0.146 | -0.32667 | 0.243 |
| Rural/LowSES | 0.14053 | 0.107 | -0.06212 | 0.362 |
| Urban/HighSES | 0.13160 | 0.201 | -0.27298 | 0.522 |
| Urban/LowSES | -0.04919 | 0.132 | -0.31177 | 0.208 |
| deplonely - Stress: Money | ||||
| Rural/HighSES | -0.00415 | 0.158 | -0.31626 | 0.301 |
| Rural/LowSES | 0.09524 | 0.113 | -0.16293 | 0.282 |
| Urban/HighSES | 0.42793 | 0.205 | 0.05703 | 0.845 |
| Urban/LowSES | -0.07479 | 0.173 | -0.43808 | 0.240 |
| deplonely - Stress: Respect | ||||
| Rural/HighSES | 0.08452 | 0.143 | -0.18912 | 0.386 |
| Rural/LowSES | 0.01172 | 0.110 | -0.20657 | 0.234 |
| Urban/HighSES | 0.11065 | 0.168 | -0.22336 | 0.433 |
| Urban/LowSES | 0.08285 | 0.131 | -0.17403 | 0.355 |
| deplonely - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.29179 | 0.217 | -0.07986 | 0.824 |
| Rural/LowSES | 0.20108 | 0.120 | -0.04916 | 0.433 |
| Urban/HighSES | 0.24941 | 0.196 | -0.13226 | 0.656 |
| Urban/LowSES | -0.05123 | 0.182 | -0.43265 | 0.270 |
| deplonely - Stress: Transport | ||||
| Rural/HighSES | -0.05367 | 0.187 | -0.45358 | 0.293 |
| Rural/LowSES | 0.06195 | 0.120 | -0.19405 | 0.281 |
| Urban/HighSES | 0.32421 | 0.177 | -0.04336 | 0.652 |
| Urban/LowSES | 0.02409 | 0.202 | -0.30674 | 0.586 |
| depblues - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.15539 | 0.097 | -0.34635 | 0.038 |
| Rural/LowSES | -0.01673 | 0.061 | -0.13407 | 0.105 |
| Urban/HighSES | 0.00324 | 0.145 | -0.28598 | 0.304 |
| Urban/LowSES | 0.01729 | 0.144 | -0.22807 | 0.358 |
| depblues - Stress: Fair Trtmt | ||||
| Rural/HighSES | -0.00702 | 0.123 | -0.23369 | 0.254 |
| Rural/LowSES | -0.00514 | 0.057 | -0.12436 | 0.102 |
| Urban/HighSES | 0.00509 | 0.146 | -0.28738 | 0.303 |
| Urban/LowSES | 0.05317 | 0.102 | -0.13393 | 0.280 |
| depblues - Stress: Job | ||||
| Rural/HighSES | -0.00176 | 0.116 | -0.22486 | 0.239 |
| Rural/LowSES | -0.04300 | 0.054 | -0.15862 | 0.061 |
| Urban/HighSES | 0.01646 | 0.235 | -0.29602 | 0.637 |
| Urban/LowSES | -0.02966 | 0.102 | -0.22650 | 0.179 |
| depblues - Stress: Money | ||||
| Rural/HighSES | -0.03989 | 0.161 | -0.33111 | 0.337 |
| Rural/LowSES | -0.05444 | 0.062 | -0.19728 | 0.050 |
| Urban/HighSES | -0.14868 | 0.158 | -0.48413 | 0.158 |
| Urban/LowSES | 0.02934 | 0.132 | -0.19808 | 0.350 |
| depblues - Stress: Respect | ||||
| Rural/HighSES | -0.04802 | 0.122 | -0.29800 | 0.197 |
| Rural/LowSES | -0.03253 | 0.057 | -0.15323 | 0.076 |
| Urban/HighSES | 0.06375 | 0.141 | -0.17069 | 0.399 |
| Urban/LowSES | -0.05085 | 0.097 | -0.24542 | 0.141 |
| depblues - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.01317 | 0.160 | -0.20905 | 0.421 |
| Rural/LowSES | 0.02284 | 0.056 | -0.07962 | 0.150 |
| Urban/HighSES | 0.39201 | 0.199 | 0.07894 | 0.811 |
| Urban/LowSES | 0.03271 | 0.107 | -0.16353 | 0.264 |
| depblues - Stress: Transport | ||||
| Rural/HighSES | -0.09120 | 0.127 | -0.34673 | 0.173 |
| Rural/LowSES | -0.02656 | 0.055 | -0.13313 | 0.089 |
| Urban/HighSES | -0.04098 | 0.146 | -0.26945 | 0.337 |
| Urban/LowSES | -0.05955 | 0.120 | -0.27717 | 0.209 |
| depunfair - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.05113 | 0.155 | -0.34942 | 0.258 |
| Rural/LowSES | -0.03228 | 0.092 | -0.21292 | 0.146 |
| Urban/HighSES | 0.16922 | 0.186 | -0.20791 | 0.515 |
| Urban/LowSES | 0.13916 | 0.250 | -0.24452 | 0.746 |
| depunfair - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.31600 | 0.172 | 0.00768 | 0.709 |
| Rural/LowSES | 0.11566 | 0.086 | -0.03856 | 0.310 |
| Urban/HighSES | 0.30829 | 0.192 | -0.09554 | 0.675 |
| Urban/LowSES | 0.15705 | 0.158 | -0.12547 | 0.503 |
| depunfair - Stress: Job | ||||
| Rural/HighSES | 0.30288 | 0.149 | 0.03667 | 0.631 |
| Rural/LowSES | 0.04858 | 0.087 | -0.11276 | 0.225 |
| Urban/HighSES | 0.53203 | 0.155 | 0.18758 | 0.793 |
| Urban/LowSES | 0.00893 | 0.142 | -0.26217 | 0.312 |
| depunfair - Stress: Money | ||||
| Rural/HighSES | 0.28413 | 0.166 | -0.02701 | 0.658 |
| Rural/LowSES | 0.14043 | 0.082 | -0.00579 | 0.315 |
| Urban/HighSES | 0.28264 | 0.198 | -0.17381 | 0.641 |
| Urban/LowSES | 0.17310 | 0.177 | -0.17580 | 0.547 |
| depunfair - Stress: Respect | ||||
| Rural/HighSES | 0.38780 | 0.145 | 0.15582 | 0.722 |
| Rural/LowSES | 0.05441 | 0.084 | -0.12213 | 0.216 |
| Urban/HighSES | 0.25279 | 0.157 | -0.05980 | 0.554 |
| Urban/LowSES | 0.11687 | 0.144 | -0.15130 | 0.418 |
| depunfair - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.07794 | 0.214 | -0.28526 | 0.615 |
| Rural/LowSES | -0.00423 | 0.092 | -0.18235 | 0.182 |
| Urban/HighSES | 0.31708 | 0.156 | -0.00011 | 0.605 |
| Urban/LowSES | 0.30982 | 0.136 | 0.05076 | 0.601 |
| depunfair - Stress: Transport | ||||
| Rural/HighSES | 0.31401 | 0.179 | -0.10265 | 0.668 |
| Rural/LowSES | 0.12964 | 0.080 | -0.03927 | 0.285 |
| Urban/HighSES | 0.34396 | 0.177 | -0.04847 | 0.674 |
| Urban/LowSES | 0.24949 | 0.169 | -0.11300 | 0.603 |
| depmistrt - Stress: Assault Vctm | ||||
| Rural/HighSES | -0.04336 | 0.167 | -0.37194 | 0.298 |
| Rural/LowSES | -0.03986 | 0.094 | -0.23885 | 0.132 |
| Urban/HighSES | 0.25391 | 0.166 | -0.06804 | 0.568 |
| Urban/LowSES | -0.18619 | 0.169 | -0.52511 | 0.143 |
| depmistrt - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.16097 | 0.180 | -0.10357 | 0.642 |
| Rural/LowSES | 0.00682 | 0.062 | -0.12324 | 0.126 |
| Urban/HighSES | 0.04449 | 0.147 | -0.23503 | 0.343 |
| Urban/LowSES | 0.15910 | 0.124 | -0.09339 | 0.403 |
| depmistrt - Stress: Job | ||||
| Rural/HighSES | 0.08660 | 0.147 | -0.22960 | 0.364 |
| Rural/LowSES | 0.01066 | 0.076 | -0.17374 | 0.139 |
| Urban/HighSES | 0.06661 | 0.177 | -0.28131 | 0.443 |
| Urban/LowSES | -0.12286 | 0.140 | -0.41723 | 0.132 |
| depmistrt - Stress: Money | ||||
| Rural/HighSES | 0.09002 | 0.149 | -0.21266 | 0.401 |
| Rural/LowSES | 0.03029 | 0.066 | -0.08714 | 0.173 |
| Urban/HighSES | 0.26234 | 0.230 | -0.05523 | 0.844 |
| Urban/LowSES | 0.02938 | 0.148 | -0.25081 | 0.317 |
| depmistrt - Stress: Respect | ||||
| Rural/HighSES | 0.16591 | 0.260 | -0.15934 | 0.775 |
| Rural/LowSES | -0.02565 | 0.071 | -0.17473 | 0.113 |
| Urban/HighSES | 0.09587 | 0.129 | -0.15674 | 0.357 |
| Urban/LowSES | 0.05974 | 0.123 | -0.16633 | 0.331 |
| depmistrt - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.25477 | 0.210 | -0.07162 | 0.789 |
| Rural/LowSES | 0.03792 | 0.084 | -0.13402 | 0.201 |
| Urban/HighSES | 0.15291 | 0.155 | -0.16415 | 0.453 |
| Urban/LowSES | -0.07593 | 0.134 | -0.34069 | 0.188 |
| depmistrt - Stress: Transport | ||||
| Rural/HighSES | -0.02369 | 0.164 | -0.37433 | 0.282 |
| Rural/LowSES | 0.05807 | 0.076 | -0.09714 | 0.207 |
| Urban/HighSES | 0.07312 | 0.164 | -0.30021 | 0.360 |
| Urban/LowSES | -0.04784 | 0.150 | -0.37236 | 0.227 |
| depbetray - Stress: Assault Vctm | ||||
| Rural/HighSES | 0.01982 | 0.153 | -0.29884 | 0.328 |
| Rural/LowSES | -0.00783 | 0.078 | -0.17837 | 0.135 |
| Urban/HighSES | 0.29867 | 0.165 | -0.03701 | 0.611 |
| Urban/LowSES | 0.21316 | 0.179 | -0.10479 | 0.625 |
| depbetray - Stress: Fair Trtmt | ||||
| Rural/HighSES | 0.11612 | 0.163 | -0.11443 | 0.562 |
| Rural/LowSES | 0.03138 | 0.057 | -0.06563 | 0.156 |
| Urban/HighSES | 0.12412 | 0.181 | -0.16638 | 0.569 |
| Urban/LowSES | 0.11192 | 0.130 | -0.14435 | 0.375 |
| depbetray - Stress: Job | ||||
| Rural/HighSES | 0.33146 | 0.145 | 0.10439 | 0.680 |
| Rural/LowSES | 0.03964 | 0.056 | -0.06588 | 0.160 |
| Urban/HighSES | 0.03791 | 0.170 | -0.27381 | 0.405 |
| Urban/LowSES | 0.03182 | 0.112 | -0.18242 | 0.266 |
| depbetray - Stress: Money | ||||
| Rural/HighSES | 0.12480 | 0.133 | -0.11833 | 0.424 |
| Rural/LowSES | 0.01972 | 0.058 | -0.09189 | 0.139 |
| Urban/HighSES | 0.15430 | 0.182 | -0.19427 | 0.555 |
| Urban/LowSES | -0.14405 | 0.116 | -0.36802 | 0.092 |
| depbetray - Stress: Respect | ||||
| Rural/HighSES | 0.05355 | 0.117 | -0.17730 | 0.282 |
| Rural/LowSES | 0.02559 | 0.057 | -0.07907 | 0.152 |
| Urban/HighSES | 0.11071 | 0.145 | -0.15930 | 0.416 |
| Urban/LowSES | -0.00397 | 0.120 | -0.24591 | 0.236 |
| depbetray - Stress: Theft Vctm | ||||
| Rural/HighSES | 0.26148 | 0.215 | -0.01388 | 0.857 |
| Rural/LowSES | -0.00019 | 0.068 | -0.14757 | 0.126 |
| Urban/HighSES | 0.11702 | 0.151 | -0.19178 | 0.403 |
| Urban/LowSES | 0.24744 | 0.131 | 0.02634 | 0.551 |
| depbetray - Stress: Transport | ||||
| Rural/HighSES | -0.09004 | 0.146 | -0.39010 | 0.190 |
| Rural/LowSES | 0.01561 | 0.060 | -0.10064 | 0.141 |
| Urban/HighSES | -0.00915 | 0.156 | -0.28970 | 0.326 |
| Urban/LowSES | -0.01476 | 0.148 | -0.27422 | 0.328 |
9.2.2 Median unweighted (meta-analytic avg) estimates from Fig 4
Similar to before, we can also summarize the median unweighted community-specific PLME estimates and 80% posterior interval ranges for these model estimates, this time grouped by community.
9.2.2.1 Crim intent (chg)
PLMEprjchg_comm %>%
ungroup() %>%
mutate(
rural.ses.med = if_else(
rural.ses.med == "1", "Rural/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "2", "Rural/HighSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "3", "Urban/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "4", "Urban/HighSES", rural.ses.med)
) %>%
group_by(rural.ses.med) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .10),
ul = quantile(PLME, probs = .90)) %>%
gt() %>%
tab_header(
title = md("**Criminal Intent (Chg) Median Unadj. PLME Estimate by Comm.**")
)| Criminal Intent (Chg) Median Unadj. PLME Estimate by Comm. | ||||
| rural.ses.med | m | s | ll | ul |
|---|---|---|---|---|
| Rural/HighSES | -0.00244 | 0.019 | -0.018 | 0.00081 |
| Rural/LowSES | -0.00125 | 0.011 | -0.012 | 0.00453 |
| Urban/HighSES | 0.00625 | 0.133 | -0.040 | 0.19256 |
| Urban/LowSES | -0.00062 | 0.068 | -0.025 | 0.06248 |
9.2.2.2 Neg emotions (chg)
PLMEdepchg_comm %>%
ungroup() %>%
mutate(
rural.ses.med = if_else(
rural.ses.med == "1", "Rural/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "2", "Rural/HighSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "3", "Urban/LowSES", rural.ses.med),
rural.ses.med = if_else(
rural.ses.med == "4", "Urban/HighSES", rural.ses.med)
) %>%
group_by(rural.ses.med) %>%
summarise(m = median(PLME),
s = sd(PLME),
ll = quantile(PLME, probs = .10),
ul = quantile(PLME, probs = .90)) %>%
gt() %>%
tab_header(
title = md("**Negative Emotions (Chg) Median Unadj. PLME Estimate by Comm.**")
)| Negative Emotions (Chg) Median Unadj. PLME Estimate by Comm. | ||||
| rural.ses.med | m | s | ll | ul |
|---|---|---|---|---|
| Rural/HighSES | 0.102 | 0.24 | -0.147 | 0.46 |
| Rural/LowSES | 0.034 | 0.11 | -0.077 | 0.20 |
| Urban/HighSES | 0.144 | 0.23 | -0.117 | 0.46 |
| Urban/LowSES | 0.046 | 0.19 | -0.177 | 0.30 |
10 Appendix 1: Composite stress scales
(RMD FILE: BDK_2023_Stress_9_Append1_stress_scales)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE10.1 IRT/SUM STRESS SCALE VERSIONS
Below, we present measurement models for general stress scale with all seven stress items. We create composite stress scales (latent IRT theta scale; sum scale), run models, and plot contrasts. We anticipate a graded response IRT model will be unlikely to fit well and that a sum scale will better capture predictive information contained in stress items due to multidimensionality from different stress sources (e.g., financial; interpersonal) and potentially generate more replicable results (e.g., see here and here). Nonetheless, we will estimate models and plot marginal effects contrasts using IRT and sum scales to assess robustness and present results from both in the Appendix. Later, we will use the stress sum scale in mediation models to estimate direct and indirect effects of stress on criminal intent through criminogenic negative emotions.
load(here("1_Data_Files/Datasets/stress_long.Rdata"))# https://hanhao23.github.io/project/irttutorial/irt-tutorial-in-r-with-mirt-package/
# https://philippmasur.de/2022/05/13/how-to-run-irt-analyses-in-r/
# https://bookdown.org/bean_jerry/using_r_for_social_work_research/item-response-theory.html
# https://journals.sagepub.com/doi/10.1177/09622802211043263
# https://stackoverflow.com/questions/63646722/mirt-package-r-how-to-create-a-independent-model-with-multigroup-function
# https://groups.google.com/g/mirt-package/c/TLv1JFq2tCg
# Estimate graded response IRT model with mirt & save theta scores for stress,
# crim intent, dep symp, & crim emots
# https://rdrr.io/cran/mirt/man/multipleGroup.html
# https://rstudio-pubs-static.s3.amazonaws.com/357155_6674780326ef4afba5f009d17a85d4ae.html
# follow steps here:
# https://bookdown.org/bean_jerry/using_r_for_social_work_research/item-response-theory.html
# mod1 <- (mirt(scale, 1, verbose = FALSE, itemtype = 'graded', SE = TRUE))
#stress - year 1
irtstressy1 <- stress.long %>%
filter(year==1) %>%
dplyr::select(id, stmony, sttran, stresp, stfair, stjob, stthft, stmug)
#graded response IRT model
stscaley1 <- mirt(irtstressy1[-1], 1, verbose = FALSE, itemtype = 'graded')
#save IRT theta (factor) scores
irtstscorey1 <- fscores(stscaley1)
#merge with T1 data
irtstressy1 <- bind_cols(irtstressy1, irtstscorey1) %>%
rename("irtstress" = "F1") %>%
mutate(year="1") %>%
dplyr::select(c(id, year, irtstress))10.1.1 Stress (chg)
10.1.1.1 Model fit
# https://bookdown.org/bean_jerry/using_r_for_social_work_research/item-response-theory.html
# "...we used the R mirt package to fit a graded response model (the recommended model for ordered polytomous response data) using a full-information maximum likelihood fitting function. In addition, we assessed model fit using an index, M2, which is specifically designed to assess the fit of item response models for ordinal data. We used the M2-based root mean square error of approximation as the primary fit index. We also used the standardized root mean square residual (SRMSR) and comparative fit index (CFI) to assess adequacy of model fit."
M2(stscaley1, type = "C2", calcNULL = FALSE) %>% gt()| M2 | df | p | RMSEA | RMSEA_5 | RMSEA_95 | SRMSR | TLI | CFI |
|---|---|---|---|---|---|---|---|---|
| 414 | 14 | 0 | 0.24 | 0.22 | 0.26 | 0.19 | 0.5 | 0.67 |
10.1.1.2 Factor loadings
summary(stscaley1)## F1 h2
## stmony 0.1086 0.01178
## sttran 0.0548 0.00301
## stresp 0.9485 0.89957
## stfair 0.9852 0.97067
## stjob 0.4488 0.20142
## stthft 0.3232 0.10448
## stmug 0.4327 0.18720
##
## SS loadings: 2.4
## Proportion Var: 0.34
##
## Factor correlations:
##
## F1
## F1 110.1.1.3 Item fit
itemfit(stscaley1) ## item S_X2 df.S_X2 RMSEA.S_X2 p.S_X2
## 1 stmony 136 64 0.048 0.000
## 2 sttran 132 63 0.047 0.000
## 3 stresp 339 33 0.138 0.000
## 4 stfair 314 32 0.134 0.000
## 5 stjob 96 57 0.038 0.001
## 6 stthft 85 56 0.033 0.007
## 7 stmug 78 45 0.039 0.00210.1.1.4 Category characteristic curves
plot(stscaley1, type='trace', which.item = c(1,2,3,4,5,6,7), facet_items=T,
as.table = TRUE, auto.key=list(points=F, lines=T, columns=4, space = 'top', cex = .8),
theta_lim = c(-3, 3),
main = "")
10.1.1.5 Item information curves
plot(stscaley1, type='infotrace', which.item = c(1,2,3,4,5,6,7), facet_items=T,
as.table = TRUE, auto.key=list(points=F, lines=T, columns=1, space = 'right', cex = .8),
theta_lim = c(-3, 3),
main="") #interpersonal items driving latent trait - virtually no info from other vars
10.1.1.6 Scale information & conditional SEs
plot(stscaley1, type = 'infoSE', theta_lim = c(-3, 3),
main="") 
10.1.1.7 Conditional reliability
plot(stscaley1, type = 'rxx', theta_lim = c(-3, 3),
main="" ) 
10.1.1.8 Marginal reliability
marginal_rxx(stscaley1) #marginal reliability = .85## [1] 0.8510.1.1.9 Scale characteristic curve
plot(stscaley1, type = 'score', theta_lim = c(-3, 3), main = "") 
#maps latent theta scores to original scale score metric 10.1.2 Stress (T2)
#stress - year 2
irtstressy2 <- stress.long %>%
filter(year==2) %>%
dplyr::select(id, stmony, sttran, stresp, stfair, stjob, stthft, stmug)
#graded response IRT model
stscaley2 <- mirt(irtstressy2[-1], 1, verbose = FALSE, itemtype = 'graded')
#save IRT theta (factor) scores
irtstscorey2 <- fscores(stscaley2)
#merge IRT theta scores with T2 data
irtstressy2 <- bind_cols(irtstressy2, irtstscorey2) %>%
rename("irtstress" = "F1") %>%
mutate(year="2") %>%
dplyr::select(c(id, year, irtstress))
#merge T1 & T2 data with IRT scales
tmpdat <- bind_rows(irtstressy1, irtstressy2)
stress.long <- left_join(stress.long, tmpdat, by=c("id", "year"))
rm(tmpdat)10.1.2.1 Model fit
# https://bookdown.org/bean_jerry/using_r_for_social_work_research/item-response-theory.html
# "...we used the R mirt package to fit a graded response model (the recommended model for ordered polytomous response data) using a full-information maximum likelihood fitting function. In addition, we assessed model fit using an index, M2, which is specifically designed to assess the fit of item response models for ordinal data. We used the M2-based root mean square error of approximation as the primary fit index. We also used the standardized root mean square residual (SRMSR) and comparative fit index (CFI) to assess adequacy of model fit."
M2(stscaley2, type = "C2", calcNULL = FALSE) %>% gt()| M2 | df | p | RMSEA | RMSEA_5 | RMSEA_95 | SRMSR | TLI | CFI |
|---|---|---|---|---|---|---|---|---|
| 257 | 14 | 0 | 0.19 | 0.17 | 0.21 | 0.15 | 0.58 | 0.72 |
10.1.2.2 Factor loadings
summary(stscaley2)## F1 h2
## stmony 0.1612 0.02599
## sttran 0.0808 0.00653
## stresp 0.9254 0.85641
## stfair 0.9540 0.91004
## stjob 0.4760 0.22661
## stthft 0.1936 0.03747
## stmug 0.3088 0.09537
##
## SS loadings: 2.2
## Proportion Var: 0.31
##
## Factor correlations:
##
## F1
## F1 110.1.2.3 Item fit
itemfit(stscaley2) ## item S_X2 df.S_X2 RMSEA.S_X2 p.S_X2
## 1 stmony 114 61 0.042 0.000
## 2 sttran 79 56 0.029 0.022
## 3 stresp 155 34 0.085 0.000
## 4 stfair 130 34 0.076 0.000
## 5 stjob 68 52 0.025 0.067
## 6 stthft 60 56 0.012 0.328
## 7 stmug 57 49 0.018 0.20410.1.2.4 Category characteristic curves
plot(stscaley2, type='trace', which.item = c(1,2,3,4,5,6,7), facet_items=T,
as.table = TRUE, auto.key=list(points=F, lines=T, columns=4, space = 'top', cex = .8),
theta_lim = c(-3, 3),
main = "")
10.1.2.5 Item information curves
plot(stscaley2, type='infotrace', which.item = c(1,2,3,4,5,6,7), facet_items=T,
as.table = TRUE, auto.key=list(points=F, lines=T, columns=1, space = 'right', cex = .8),
theta_lim = c(-3, 3),
main="") #interpersonal items driving latent trait - virtually no info from other vars
10.1.2.6 Scale information & conditional SEs
plot(stscaley2, type = 'infoSE', theta_lim = c(-3, 3),
main="") 
10.1.2.7 Conditional reliability
plot(stscaley2, type = 'rxx', theta_lim = c(-3, 3),
main="" ) 
10.1.2.8 Marginal reliability
marginal_rxx(stscaley2) #marginal reliability = .85## [1] 0.8410.1.2.9 Scale characteristic curve
plot(stscaley2, type = 'score', theta_lim = c(-3, 3), main = "") 
#maps latent theta scores to original scale score metric As anticipated, IRT model suggests the stress items collectively exhibit poor measurement properties as a unidimensional scale; that is, the items do not coalesce well into single stress scale. RMSEA and SRMSR values indicate poor overall fit, while specific item loadings and information criteria indicate interpersonal items almost exclusively drive latent theta scores at each wave. Scale information, conditional SE, and conditional reliability plots indicate ranges where latent theta scores are most precise and exhibit highest scale reliability (T1: [-1.5,1]; T2: [-2,1]). Meanwhile, the overall marginal reliability for the IRT scale is relatively high (T1: 0.85; T2: 0.84).
These graded response IRT models confirm our suspicions that simply lumping stress items together into a single scale at each wave is inadvisable. Most items contributed very little information to the latent stress scale, while latent theta scores largely reflect information contributed by interpersonal stress items. An item-specific approach or perhaps examination of stress subscales by type (e.g., financial; interpersonal) is warranted.
If a general scale is needed (e.g., we will use one later for mediation analysis), then the equal unit weighting assumption underlying simple sum scaling may be better for capturing meaningful outcome-relevant variations in stress; latent scaling may better reflect an underlying unidimensional factor but at the cost of potentially arbitrarily up- or down-weighting component stress items that may vary in their relevance to outcomes of interest. However, either scaling approach purchases data reduction or parsimony at the potentially high cost of precision and interpretive clarity.
Since general scales are common in the field, we repeat the process of estimating between-within models and plotting estimated contrasts representing practically large marginal effects of general stress on outcomes using a latent IRT theta scale and a standardized sum scale.
10.1.3 Descriptives: Composite stress scales
# stdz sum stress scale
# break irtstress & sumstress vars into btw-person means & w/in person changes
stress.long2 <- stress.long %>%
mutate(
sumstress = stmony + sttran + stresp + stfair + stjob + stthft + stmug,
sumstress = (sumstress - mean(sumstress)) / sd(sumstress)
) %>%
group_by(id) %>%
mutate(
irtstressav = mean(irtstress),
sumstressav = mean(sumstress)
) %>%
ungroup() %>%
mutate(
irtstresschg = irtstress - irtstressav,
sumstresschg = sumstress - sumstressav
)10.1.3.1 Summary statistics
#corr btw irt stress scale & standardized sum stress scale
print("Corr: irtstress & sumstress")## [1] "Corr: irtstress & sumstress"cor(stress.long2$irtstress, stress.long2$sumstress)## [1] 0.8cat('\n')#btw-person cross-time average stress items
print("Summary: IRT stress avg")## [1] "Summary: IRT stress avg"summary(stress.long2$irtstressav) ## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -2.10 -0.70 0.14 0.00 0.77 1.73print("Summary: Sum stress avg")## [1] "Summary: Sum stress avg"summary(stress.long2$sumstressav)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -2.47 -0.65 0.11 0.00 0.69 2.22# sd(stress.long2$irtstressav)
# sd(stress.long2$sumstressav)
cat('\n')#w/in-person over time change items
print("Summary: IRT stress chg")## [1] "Summary: IRT stress chg"summary(stress.long2$irtstresschg)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -0.80 -0.15 0.00 0.00 0.15 0.80print("Summary: Sum stress chg")## [1] "Summary: Sum stress chg"summary(stress.long2$sumstresschg)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -0.86 -0.10 0.00 0.00 0.10 0.86 # -.5/.5 change contrast = 1 unit (approx 1SD) change on IRT scale
# table(stress.long2$irtstresschg)10.1.3.2 IRT stress avg (btw)
ggplot(stress.long2, aes(irtstressav)) + geom_histogram(fill="#E99D53")
10.1.3.3 Sum stress avg (btw)
ggplot(stress.long2, aes(sumstressav)) + geom_histogram(fill="#E99D53")
10.1.3.4 IRT stress chg (w/in)
ggplot(stress.long2, aes(irtstresschg)) + geom_histogram(fill="#883E3A")
10.1.3.5 Sum stress chg (w/in)
ggplot(stress.long2, aes(sumstresschg)) + geom_histogram(fill="#883E3A")
10.1.4 Corr X Community: irt stress & prj crime
#list of colnames for projected crime DVs
prjdv_names <- noquote(c("prjthflt5", "prjthfgt5", "prjthreat", "prjharm",
"prjusedrg", "prjhack"))
# Getting a few divergent transitions after warmup
# increased adapt_delta from 0.8 to 0.85
# Could consider priors that better reflect rarity of criminal intent
# E.g., assume probability of each crime intent outcome is 0.1
# & 90% sure that each crime intent mean is less than 0.2
# use LearnBayes package (ProbBayes)
# LearnBayes::beta.select(list(x = 0.1, p = 0.5),
# list(x = 0.2, p = 0.9))
#2.5 20.1
# With these assumptions, could set more informative logit priors:
# set.seed(8675309)
# p_sim <- rbeta(1000, 2.5, 20.1)
# theta_sim <- log(p_sim / (1 - p_sim))
# c(mean(theta_sim), sd(theta_sim))
#Priors: mean = -2.3 SD = .75
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names)
)
chg.prjcrime.stirt.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
irtstresschg + irtstressav + rural.ses.med +
irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id),
data = stress.long2,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
adapt_delta = 0.85,
file = "Models/chg_prjcrime_stirt_comm_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
varlist <- c("^b_prjthflt5_irt","^b_prjthfgt5_irt", "^b_prjthreat_irt",
"^b_prjharm_irt", "^b_usedrg_irt", "^b_prjhack_irt")
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="prjthflt5")
ppcheckdv2 <-pp_check(modelfit, resp="prjthfgt5")
ppcheckdv3 <-pp_check(modelfit, resp="prjthreat")
ppcheckdv4 <-pp_check(modelfit, resp="prjharm")
ppcheckdv5 <-pp_check(modelfit, resp="prjusedrg")
ppcheckdv6 <-pp_check(modelfit, resp="prjhack")
plotcoefs2 <- mcmc_plot(modelfit, variable = varlist, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6, plotcoefs2)
return(allchecks)
}
out.chg.prjcrime.stirt.comm.fit <- ppchecks(chg.prjcrime.stirt.comm.fit)10.1.4.1 Coefficient plot (intervals)
out.chg.prjcrime.stirt.comm.fit[[9]] 
10.1.4.2 PPcheck (density)
p1 <- out.chg.prjcrime.stirt.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stirt.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stirt.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stirt.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stirt.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stirt.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
10.1.4.3 Fit summary
out.chg.prjcrime.stirt.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## prjthreat ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## prjharm ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## prjusedrg ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## prjhack ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## Data: stress.long2 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.79 0.54 2.88 4.95 1.00 1253
## sd(prjthfgt5_Intercept) 3.24 0.47 2.41 4.24 1.00 1104
## sd(prjthreat_Intercept) 3.04 0.51 2.14 4.16 1.00 1130
## sd(prjharm_Intercept) 2.87 0.52 1.93 3.98 1.00 1177
## sd(prjusedrg_Intercept) 2.67 0.51 1.77 3.77 1.00 1155
## sd(prjhack_Intercept) 0.85 0.56 0.03 2.03 1.00 561
## Tail_ESS
## sd(prjthflt5_Intercept) 2190
## sd(prjthfgt5_Intercept) 2203
## sd(prjthreat_Intercept) 2263
## sd(prjharm_Intercept) 1992
## sd(prjusedrg_Intercept) 1733
## sd(prjhack_Intercept) 1228
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_Intercept -5.77 0.69 -7.22 -4.49 1.00
## prjthfgt5_Intercept -5.36 0.65 -6.73 -4.18 1.00
## prjthreat_Intercept -6.59 0.82 -8.36 -5.12 1.00
## prjharm_Intercept -6.29 0.82 -8.06 -4.82 1.00
## prjusedrg_Intercept -6.21 0.80 -7.93 -4.78 1.00
## prjhack_Intercept -4.58 0.62 -6.01 -3.57 1.00
## prjthflt5_irtstresschg -0.12 0.63 -1.34 1.14 1.00
## prjthflt5_irtstressav 0.55 0.47 -0.35 1.52 1.00
## prjthflt5_rural.ses.med2 -1.44 0.67 -2.79 -0.16 1.00
## prjthflt5_rural.ses.med3 1.20 0.57 0.07 2.32 1.00
## prjthflt5_rural.ses.med4 1.95 0.61 0.75 3.11 1.00
## prjthflt5_irtstresschg:rural.ses.med2 -0.33 0.90 -2.10 1.42 1.00
## prjthflt5_irtstresschg:rural.ses.med3 0.29 0.74 -1.16 1.72 1.00
## prjthflt5_irtstresschg:rural.ses.med4 -0.15 0.79 -1.69 1.41 1.00
## prjthflt5_irtstressav:rural.ses.med2 0.43 0.71 -0.92 1.83 1.00
## prjthflt5_irtstressav:rural.ses.med3 0.84 0.61 -0.37 2.07 1.00
## prjthflt5_irtstressav:rural.ses.med4 0.24 0.62 -0.94 1.44 1.00
## prjthfgt5_irtstresschg 0.12 0.62 -1.12 1.37 1.00
## prjthfgt5_irtstressav 0.28 0.45 -0.58 1.19 1.00
## prjthfgt5_rural.ses.med2 -1.53 0.64 -2.77 -0.26 1.00
## prjthfgt5_rural.ses.med3 1.23 0.53 0.19 2.30 1.00
## prjthfgt5_rural.ses.med4 1.86 0.57 0.74 2.98 1.00
## prjthfgt5_irtstresschg:rural.ses.med2 -0.02 0.88 -1.72 1.73 1.00
## prjthfgt5_irtstresschg:rural.ses.med3 -0.09 0.74 -1.51 1.35 1.00
## prjthfgt5_irtstresschg:rural.ses.med4 0.21 0.78 -1.31 1.74 1.00
## prjthfgt5_irtstressav:rural.ses.med2 0.19 0.69 -1.16 1.52 1.00
## prjthfgt5_irtstressav:rural.ses.med3 0.51 0.58 -0.63 1.69 1.00
## prjthfgt5_irtstressav:rural.ses.med4 0.29 0.59 -0.88 1.48 1.00
## prjthreat_irtstresschg -0.44 0.68 -1.76 0.90 1.00
## prjthreat_irtstressav 0.87 0.50 -0.08 1.90 1.00
## prjthreat_rural.ses.med2 -0.84 0.70 -2.23 0.53 1.00
## prjthreat_rural.ses.med3 0.37 0.63 -0.90 1.58 1.00
## prjthreat_rural.ses.med4 1.61 0.60 0.45 2.82 1.00
## prjthreat_irtstresschg:rural.ses.med2 0.12 0.94 -1.75 1.97 1.00
## prjthreat_irtstresschg:rural.ses.med3 0.15 0.80 -1.42 1.75 1.00
## prjthreat_irtstresschg:rural.ses.med4 -0.13 0.82 -1.75 1.44 1.00
## prjthreat_irtstressav:rural.ses.med2 -0.60 0.72 -2.02 0.84 1.00
## prjthreat_irtstressav:rural.ses.med3 0.78 0.66 -0.50 2.10 1.00
## prjthreat_irtstressav:rural.ses.med4 0.29 0.63 -0.96 1.54 1.00
## prjharm_irtstresschg -0.29 0.69 -1.67 1.08 1.00
## prjharm_irtstressav 0.29 0.47 -0.59 1.24 1.00
## prjharm_rural.ses.med2 -0.55 0.67 -1.87 0.76 1.00
## prjharm_rural.ses.med3 0.16 0.62 -1.05 1.36 1.00
## prjharm_rural.ses.med4 1.04 0.62 -0.19 2.27 1.00
## prjharm_irtstresschg:rural.ses.med2 -0.04 0.93 -1.85 1.79 1.00
## prjharm_irtstresschg:rural.ses.med3 0.51 0.80 -1.06 2.05 1.00
## prjharm_irtstresschg:rural.ses.med4 -0.46 0.86 -2.19 1.19 1.00
## prjharm_irtstressav:rural.ses.med2 -0.71 0.64 -1.97 0.53 1.00
## prjharm_irtstressav:rural.ses.med3 0.79 0.64 -0.43 2.10 1.00
## prjharm_irtstressav:rural.ses.med4 0.12 0.64 -1.12 1.45 1.00
## prjusedrg_irtstresschg -0.57 0.68 -1.91 0.75 1.00
## prjusedrg_irtstressav 0.69 0.49 -0.26 1.66 1.00
## prjusedrg_rural.ses.med2 -0.57 0.66 -1.92 0.70 1.00
## prjusedrg_rural.ses.med3 -0.74 0.66 -2.02 0.56 1.00
## prjusedrg_rural.ses.med4 1.63 0.59 0.43 2.78 1.00
## prjusedrg_irtstresschg:rural.ses.med2 -0.24 0.89 -2.00 1.53 1.00
## prjusedrg_irtstresschg:rural.ses.med3 -0.01 0.88 -1.71 1.69 1.00
## prjusedrg_irtstresschg:rural.ses.med4 0.37 0.80 -1.18 1.97 1.00
## prjusedrg_irtstressav:rural.ses.med2 -0.89 0.68 -2.27 0.40 1.00
## prjusedrg_irtstressav:rural.ses.med3 0.50 0.68 -0.81 1.86 1.00
## prjusedrg_irtstressav:rural.ses.med4 0.39 0.63 -0.81 1.60 1.00
## prjhack_irtstresschg -0.43 0.69 -1.78 0.95 1.00
## prjhack_irtstressav 0.56 0.45 -0.31 1.45 1.00
## prjhack_rural.ses.med2 -0.93 0.67 -2.30 0.32 1.00
## prjhack_rural.ses.med3 0.17 0.54 -0.90 1.24 1.00
## prjhack_rural.ses.med4 0.56 0.56 -0.57 1.61 1.00
## prjhack_irtstresschg:rural.ses.med2 0.23 0.93 -1.59 2.03 1.00
## prjhack_irtstresschg:rural.ses.med3 -0.32 0.82 -1.89 1.29 1.00
## prjhack_irtstresschg:rural.ses.med4 -0.15 0.83 -1.77 1.48 1.00
## prjhack_irtstressav:rural.ses.med2 -0.45 0.68 -1.71 0.95 1.00
## prjhack_irtstressav:rural.ses.med3 0.01 0.58 -1.10 1.18 1.00
## prjhack_irtstressav:rural.ses.med4 0.50 0.58 -0.62 1.67 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1570 2320
## prjthfgt5_Intercept 1570 2211
## prjthreat_Intercept 1657 2694
## prjharm_Intercept 1643 2432
## prjusedrg_Intercept 1515 2262
## prjhack_Intercept 1059 1276
## prjthflt5_irtstresschg 4061 3181
## prjthflt5_irtstressav 2500 2831
## prjthflt5_rural.ses.med2 3054 2649
## prjthflt5_rural.ses.med3 2490 2625
## prjthflt5_rural.ses.med4 2632 3167
## prjthflt5_irtstresschg:rural.ses.med2 5818 2903
## prjthflt5_irtstresschg:rural.ses.med3 4397 2803
## prjthflt5_irtstresschg:rural.ses.med4 4633 2727
## prjthflt5_irtstressav:rural.ses.med2 3411 2912
## prjthflt5_irtstressav:rural.ses.med3 2625 2722
## prjthflt5_irtstressav:rural.ses.med4 2794 2969
## prjthfgt5_irtstresschg 3151 2793
## prjthfgt5_irtstressav 2221 2714
## prjthfgt5_rural.ses.med2 3002 2698
## prjthfgt5_rural.ses.med3 2461 2694
## prjthfgt5_rural.ses.med4 2396 2404
## prjthfgt5_irtstresschg:rural.ses.med2 6481 2787
## prjthfgt5_irtstresschg:rural.ses.med3 3716 2599
## prjthfgt5_irtstresschg:rural.ses.med4 4477 2762
## prjthfgt5_irtstressav:rural.ses.med2 3344 2917
## prjthfgt5_irtstressav:rural.ses.med3 2628 2964
## prjthfgt5_irtstressav:rural.ses.med4 2249 2406
## prjthreat_irtstresschg 4232 3030
## prjthreat_irtstressav 2595 2908
## prjthreat_rural.ses.med2 3933 2361
## prjthreat_rural.ses.med3 3370 2810
## prjthreat_rural.ses.med4 2973 2595
## prjthreat_irtstresschg:rural.ses.med2 5066 3018
## prjthreat_irtstresschg:rural.ses.med3 5148 2773
## prjthreat_irtstresschg:rural.ses.med4 5404 3003
## prjthreat_irtstressav:rural.ses.med2 3671 2805
## prjthreat_irtstressav:rural.ses.med3 2951 2898
## prjthreat_irtstressav:rural.ses.med4 2189 2792
## prjharm_irtstresschg 4513 2348
## prjharm_irtstressav 2518 2861
## prjharm_rural.ses.med2 3247 2937
## prjharm_rural.ses.med3 3599 3006
## prjharm_rural.ses.med4 3155 2872
## prjharm_irtstresschg:rural.ses.med2 6386 2312
## prjharm_irtstresschg:rural.ses.med3 4685 2788
## prjharm_irtstresschg:rural.ses.med4 5867 3343
## prjharm_irtstressav:rural.ses.med2 3123 3260
## prjharm_irtstressav:rural.ses.med3 3018 2648
## prjharm_irtstressav:rural.ses.med4 3270 3066
## prjusedrg_irtstresschg 4263 3184
## prjusedrg_irtstressav 3016 2557
## prjusedrg_rural.ses.med2 3890 2925
## prjusedrg_rural.ses.med3 4259 3125
## prjusedrg_rural.ses.med4 3295 3066
## prjusedrg_irtstresschg:rural.ses.med2 5758 3097
## prjusedrg_irtstresschg:rural.ses.med3 5078 2698
## prjusedrg_irtstresschg:rural.ses.med4 4269 3140
## prjusedrg_irtstressav:rural.ses.med2 3222 2685
## prjusedrg_irtstressav:rural.ses.med3 3838 3027
## prjusedrg_irtstressav:rural.ses.med4 2817 2700
## prjhack_irtstresschg 4262 2984
## prjhack_irtstressav 2565 2556
## prjhack_rural.ses.med2 5323 2987
## prjhack_rural.ses.med3 4040 2805
## prjhack_rural.ses.med4 4078 3103
## prjhack_irtstresschg:rural.ses.med2 4823 2855
## prjhack_irtstresschg:rural.ses.med3 4810 3015
## prjhack_irtstresschg:rural.ses.med4 4942 2892
## prjhack_irtstressav:rural.ses.med2 3774 3294
## prjhack_irtstressav:rural.ses.med3 2722 2459
## prjhack_irtstressav:rural.ses.med4 3262 2613
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).10.1.4.4 Prior summary
out.chg.prjcrime.stirt.comm.fit[[2]]## prior class coef group resp
## (flat) b
## normal(0, 1) b prjhack
## normal(0, 1) b irtstressav prjhack
## normal(0, 1) b irtstressav:rural.ses.med2 prjhack
## normal(0, 1) b irtstressav:rural.ses.med3 prjhack
## normal(0, 1) b irtstressav:rural.ses.med4 prjhack
## normal(0, 1) b irtstresschg prjhack
## normal(0, 1) b irtstresschg:rural.ses.med2 prjhack
## normal(0, 1) b irtstresschg:rural.ses.med3 prjhack
## normal(0, 1) b irtstresschg:rural.ses.med4 prjhack
## normal(0, 1) b rural.ses.med2 prjhack
## normal(0, 1) b rural.ses.med3 prjhack
## normal(0, 1) b rural.ses.med4 prjhack
## normal(0, 1) b prjharm
## normal(0, 1) b irtstressav prjharm
## normal(0, 1) b irtstressav:rural.ses.med2 prjharm
## normal(0, 1) b irtstressav:rural.ses.med3 prjharm
## normal(0, 1) b irtstressav:rural.ses.med4 prjharm
## normal(0, 1) b irtstresschg prjharm
## normal(0, 1) b irtstresschg:rural.ses.med2 prjharm
## normal(0, 1) b irtstresschg:rural.ses.med3 prjharm
## normal(0, 1) b irtstresschg:rural.ses.med4 prjharm
## normal(0, 1) b rural.ses.med2 prjharm
## normal(0, 1) b rural.ses.med3 prjharm
## normal(0, 1) b rural.ses.med4 prjharm
## normal(0, 1) b prjthfgt5
## normal(0, 1) b irtstressav prjthfgt5
## normal(0, 1) b irtstressav:rural.ses.med2 prjthfgt5
## normal(0, 1) b irtstressav:rural.ses.med3 prjthfgt5
## normal(0, 1) b irtstressav:rural.ses.med4 prjthfgt5
## normal(0, 1) b irtstresschg prjthfgt5
## normal(0, 1) b irtstresschg:rural.ses.med2 prjthfgt5
## normal(0, 1) b irtstresschg:rural.ses.med3 prjthfgt5
## normal(0, 1) b irtstresschg:rural.ses.med4 prjthfgt5
## normal(0, 1) b rural.ses.med2 prjthfgt5
## normal(0, 1) b rural.ses.med3 prjthfgt5
## normal(0, 1) b rural.ses.med4 prjthfgt5
## normal(0, 1) b prjthflt5
## normal(0, 1) b irtstressav prjthflt5
## normal(0, 1) b irtstressav:rural.ses.med2 prjthflt5
## normal(0, 1) b irtstressav:rural.ses.med3 prjthflt5
## normal(0, 1) b irtstressav:rural.ses.med4 prjthflt5
## normal(0, 1) b irtstresschg prjthflt5
## normal(0, 1) b irtstresschg:rural.ses.med2 prjthflt5
## normal(0, 1) b irtstresschg:rural.ses.med3 prjthflt5
## normal(0, 1) b irtstresschg:rural.ses.med4 prjthflt5
## normal(0, 1) b rural.ses.med2 prjthflt5
## normal(0, 1) b rural.ses.med3 prjthflt5
## normal(0, 1) b rural.ses.med4 prjthflt5
## normal(0, 1) b prjthreat
## normal(0, 1) b irtstressav prjthreat
## normal(0, 1) b irtstressav:rural.ses.med2 prjthreat
## normal(0, 1) b irtstressav:rural.ses.med3 prjthreat
## normal(0, 1) b irtstressav:rural.ses.med4 prjthreat
## normal(0, 1) b irtstresschg prjthreat
## normal(0, 1) b irtstresschg:rural.ses.med2 prjthreat
## normal(0, 1) b irtstresschg:rural.ses.med3 prjthreat
## normal(0, 1) b irtstresschg:rural.ses.med4 prjthreat
## normal(0, 1) b rural.ses.med2 prjthreat
## normal(0, 1) b rural.ses.med3 prjthreat
## normal(0, 1) b rural.ses.med4 prjthreat
## normal(0, 1) b prjusedrg
## normal(0, 1) b irtstressav prjusedrg
## normal(0, 1) b irtstressav:rural.ses.med2 prjusedrg
## normal(0, 1) b irtstressav:rural.ses.med3 prjusedrg
## normal(0, 1) b irtstressav:rural.ses.med4 prjusedrg
## normal(0, 1) b irtstresschg prjusedrg
## normal(0, 1) b irtstresschg:rural.ses.med2 prjusedrg
## normal(0, 1) b irtstresschg:rural.ses.med3 prjusedrg
## normal(0, 1) b irtstresschg:rural.ses.med4 prjusedrg
## normal(0, 1) b rural.ses.med2 prjusedrg
## normal(0, 1) b rural.ses.med3 prjusedrg
## normal(0, 1) b rural.ses.med4 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dpar nlpar lb ub source
## default
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)10.1.5 Corr X Community: irt stress & any prj crime
# "Any crime" outcome & irt stress scale
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept'),
set_prior('normal(0, 1)', class = 'b')
)
chg.anyprjcrime.stirt.comm.fit <- brm(prjany ~ 1 +
irtstresschg + irtstressav + rural.ses.med +
irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id),
data = stress.long2,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_anyprjcrime_stirt_comm_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit)
plotcoefs2 <- mcmc_plot(modelfit, variable = "^b_irt", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, plotcoefs2)
return(allchecks)
}
out.chg.anyprjcrime.stirt.comm.fit <- ppchecks(chg.anyprjcrime.stirt.comm.fit)10.1.5.1 Coefficient plot (intervals)
out.chg.anyprjcrime.stirt.comm.fit[[4]] 
10.1.5.2 PPcheck (density)
out.chg.anyprjcrime.stirt.comm.fit[[3]] + labs(title = "Any Crime Intent (chg)") 
10.1.5.3 Fit summary
out.chg.anyprjcrime.stirt.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## Data: stress.long2 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.02 0.40 2.28 3.86 1.00 1169 1877
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept -4.03 0.49 -5.07 -3.12 1.00 2074
## irtstresschg -0.10 0.56 -1.17 1.01 1.00 4336
## irtstressav 0.51 0.41 -0.28 1.34 1.00 2752
## rural.ses.med2 -1.06 0.54 -2.14 -0.01 1.00 2880
## rural.ses.med3 1.01 0.49 0.06 1.98 1.00 2893
## rural.ses.med4 2.16 0.51 1.18 3.20 1.00 2968
## irtstresschg:rural.ses.med2 -0.61 0.82 -2.21 1.02 1.00 5932
## irtstresschg:rural.ses.med3 0.29 0.68 -1.06 1.61 1.00 4242
## irtstresschg:rural.ses.med4 0.54 0.72 -0.87 1.94 1.00 5401
## irtstressav:rural.ses.med2 -0.48 0.57 -1.57 0.63 1.00 2988
## irtstressav:rural.ses.med3 0.67 0.52 -0.34 1.69 1.00 3167
## irtstressav:rural.ses.med4 0.39 0.55 -0.68 1.46 1.00 3097
## Tail_ESS
## Intercept 1894
## irtstresschg 2925
## irtstressav 2605
## rural.ses.med2 3019
## rural.ses.med3 2861
## rural.ses.med4 3195
## irtstresschg:rural.ses.med2 2908
## irtstresschg:rural.ses.med3 2728
## irtstresschg:rural.ses.med4 3066
## irtstressav:rural.ses.med2 2848
## irtstressav:rural.ses.med3 2695
## irtstressav:rural.ses.med4 2651
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).10.1.5.4 Prior summary
out.chg.anyprjcrime.stirt.comm.fit[[2]]## prior class coef group resp dpar
## normal(0, 1) b
## normal(0, 1) b irtstressav
## normal(0, 1) b irtstressav:rural.ses.med2
## normal(0, 1) b irtstressav:rural.ses.med3
## normal(0, 1) b irtstressav:rural.ses.med4
## normal(0, 1) b irtstresschg
## normal(0, 1) b irtstresschg:rural.ses.med2
## normal(0, 1) b irtstresschg:rural.ses.med3
## normal(0, 1) b irtstresschg:rural.ses.med4
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## nlpar lb ub source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)10.1.6 Corr X Community: irt stress & negative emotions
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names)
)
chg.alldepress.stirt.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + irtstresschg + irtstressav +
rural.ses.med + irtstresschg:rural.ses.med +
irtstressav:rural.ses.med + (1 | id),
data = stress.long2,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stirt_comm_fit",
file_refit = "on_change"
)
##Update function to call all ppchecks for bivar depressive symptom chg models
varlist1 <- c("^b_depcantgo_irt","^b_depeffort_irt", "^b_deplonely_irt",
"^b_depblues_irt")
varlist2 <- c("^b_depunfair_irt", "^b_depmistrt_irt", "^b_depbetray_irt")
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="depcantgo")
ppcheckdv2 <-pp_check(modelfit, resp="depeffort")
ppcheckdv3 <-pp_check(modelfit, resp="deplonely")
ppcheckdv4 <-pp_check(modelfit, resp="depblues")
ppcheckdv5 <-pp_check(modelfit, resp="depunfair")
ppcheckdv6 <-pp_check(modelfit, resp="depmistrt")
ppcheckdv7 <-pp_check(modelfit, resp="depbetray")
plotcoefs1 <- mcmc_plot(modelfit, variable = varlist1, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients\nwith medians, 80%, and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = varlist2, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients\nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2, ppcheckdv3,
ppcheckdv4, ppcheckdv5, ppcheckdv6, ppcheckdv7,
plotcoefs1, plotcoefs2)
return(allchecks)
}
out.chg.alldepress.stirt.comm.fit <- ppchecks(chg.alldepress.stirt.comm.fit)10.1.6.1 Coefficient plot (depressive symptoms)
out.chg.alldepress.stirt.comm.fit[[10]]
10.1.6.2 Coefficient plot (criminogenic emotions)
out.chg.alldepress.stirt.comm.fit[[11]]
10.1.6.3 PPcheck (density)
p1 <- out.chg.alldepress.stirt.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stirt.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stirt.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stirt.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stirt.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stirt.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stirt.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
10.1.6.4 Fit summary
out.chg.alldepress.stirt.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## depeffort ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## deplonely ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## depblues ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## depunfair ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## depmistrt ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## depbetray ~ 1 + irtstresschg + irtstressav + rural.ses.med + irtstresschg:rural.ses.med + irtstressav:rural.ses.med + (1 | id)
## Data: stress.long2 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.33 0.20 0.02 0.72 1.00 705
## sd(depeffort_Intercept) 0.48 0.28 0.03 1.05 1.00 673
## sd(deplonely_Intercept) 0.44 0.24 0.03 0.90 1.01 697
## sd(depblues_Intercept) 0.61 0.31 0.05 1.21 1.01 510
## sd(depunfair_Intercept) 0.23 0.16 0.01 0.59 1.00 1071
## sd(depmistrt_Intercept) 0.33 0.22 0.02 0.83 1.00 779
## sd(depbetray_Intercept) 0.42 0.26 0.02 0.95 1.01 745
## Tail_ESS
## sd(depcantgo_Intercept) 1591
## sd(depeffort_Intercept) 1610
## sd(deplonely_Intercept) 1355
## sd(depblues_Intercept) 931
## sd(depunfair_Intercept) 1719
## sd(depmistrt_Intercept) 976
## sd(depbetray_Intercept) 1333
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_Intercept -0.50 0.14 -0.77 -0.23 1.00
## depeffort_Intercept -2.00 0.22 -2.47 -1.59 1.00
## deplonely_Intercept -1.03 0.17 -1.35 -0.72 1.00
## depblues_Intercept -2.28 0.25 -2.80 -1.81 1.00
## depunfair_Intercept -1.57 0.17 -1.92 -1.23 1.00
## depmistrt_Intercept -2.08 0.21 -2.50 -1.68 1.00
## depbetray_Intercept -2.27 0.23 -2.75 -1.85 1.00
## depcantgo_irtstresschg -0.04 0.44 -0.88 0.82 1.00
## depcantgo_irtstressav -0.16 0.16 -0.48 0.16 1.00
## depcantgo_rural.ses.med2 0.32 0.19 -0.06 0.70 1.00
## depcantgo_rural.ses.med3 0.12 0.19 -0.25 0.49 1.00
## depcantgo_rural.ses.med4 0.09 0.20 -0.30 0.48 1.00
## depcantgo_irtstresschg:rural.ses.med2 0.52 0.59 -0.63 1.67 1.00
## depcantgo_irtstresschg:rural.ses.med3 0.02 0.54 -1.05 1.10 1.00
## depcantgo_irtstresschg:rural.ses.med4 -0.30 0.59 -1.47 0.83 1.00
## depcantgo_irtstressav:rural.ses.med2 0.24 0.21 -0.18 0.65 1.00
## depcantgo_irtstressav:rural.ses.med3 -0.05 0.21 -0.45 0.38 1.00
## depcantgo_irtstressav:rural.ses.med4 0.15 0.23 -0.31 0.61 1.00
## depeffort_irtstresschg 0.14 0.49 -0.82 1.11 1.00
## depeffort_irtstressav 0.03 0.23 -0.41 0.49 1.00
## depeffort_rural.ses.med2 -0.02 0.28 -0.54 0.52 1.00
## depeffort_rural.ses.med3 0.43 0.26 -0.08 0.94 1.00
## depeffort_rural.ses.med4 0.51 0.28 -0.03 1.06 1.00
## depeffort_irtstresschg:rural.ses.med2 0.28 0.71 -1.08 1.67 1.00
## depeffort_irtstresschg:rural.ses.med3 0.23 0.59 -0.93 1.38 1.00
## depeffort_irtstresschg:rural.ses.med4 0.00 0.65 -1.21 1.31 1.00
## depeffort_irtstressav:rural.ses.med2 -0.17 0.30 -0.76 0.42 1.00
## depeffort_irtstressav:rural.ses.med3 0.32 0.29 -0.22 0.88 1.00
## depeffort_irtstressav:rural.ses.med4 0.12 0.31 -0.48 0.71 1.00
## deplonely_irtstresschg -0.24 0.43 -1.08 0.59 1.00
## deplonely_irtstressav -0.12 0.19 -0.48 0.25 1.00
## deplonely_rural.ses.med2 -0.38 0.23 -0.83 0.07 1.00
## deplonely_rural.ses.med3 -0.05 0.21 -0.45 0.37 1.00
## deplonely_rural.ses.med4 0.42 0.23 -0.02 0.87 1.00
## deplonely_irtstresschg:rural.ses.med2 0.33 0.62 -0.90 1.53 1.00
## deplonely_irtstresschg:rural.ses.med3 -0.01 0.54 -1.11 1.03 1.00
## deplonely_irtstresschg:rural.ses.med4 -0.13 0.60 -1.26 1.06 1.00
## deplonely_irtstressav:rural.ses.med2 0.06 0.24 -0.44 0.53 1.00
## deplonely_irtstressav:rural.ses.med3 0.30 0.24 -0.17 0.77 1.00
## deplonely_irtstressav:rural.ses.med4 0.17 0.26 -0.35 0.67 1.00
## depblues_irtstresschg 0.39 0.51 -0.58 1.41 1.00
## depblues_irtstressav 0.04 0.25 -0.45 0.54 1.00
## depblues_rural.ses.med2 0.05 0.30 -0.55 0.63 1.00
## depblues_rural.ses.med3 0.24 0.29 -0.34 0.83 1.00
## depblues_rural.ses.med4 0.47 0.30 -0.11 1.07 1.00
## depblues_irtstresschg:rural.ses.med2 0.42 0.71 -1.00 1.83 1.00
## depblues_irtstresschg:rural.ses.med3 0.23 0.62 -0.96 1.43 1.00
## depblues_irtstresschg:rural.ses.med4 0.21 0.66 -1.07 1.50 1.00
## depblues_irtstressav:rural.ses.med2 -0.68 0.32 -1.31 -0.06 1.00
## depblues_irtstressav:rural.ses.med3 0.55 0.32 -0.09 1.20 1.00
## depblues_irtstressav:rural.ses.med4 0.37 0.34 -0.29 1.02 1.00
## depunfair_irtstresschg 0.84 0.46 -0.07 1.74 1.00
## depunfair_irtstressav 0.27 0.21 -0.14 0.68 1.00
## depunfair_rural.ses.med2 0.24 0.23 -0.21 0.70 1.00
## depunfair_rural.ses.med3 0.69 0.21 0.29 1.12 1.00
## depunfair_rural.ses.med4 0.90 0.23 0.46 1.35 1.00
## depunfair_irtstresschg:rural.ses.med2 0.68 0.64 -0.54 1.93 1.00
## depunfair_irtstresschg:rural.ses.med3 -0.36 0.56 -1.47 0.75 1.00
## depunfair_irtstresschg:rural.ses.med4 -0.18 0.60 -1.36 0.98 1.00
## depunfair_irtstressav:rural.ses.med2 -0.52 0.26 -1.03 -0.00 1.00
## depunfair_irtstressav:rural.ses.med3 0.02 0.25 -0.47 0.52 1.00
## depunfair_irtstressav:rural.ses.med4 0.04 0.27 -0.49 0.58 1.00
## depmistrt_irtstresschg 0.22 0.49 -0.71 1.19 1.00
## depmistrt_irtstressav 0.50 0.25 0.02 1.02 1.00
## depmistrt_rural.ses.med2 0.58 0.26 0.07 1.09 1.00
## depmistrt_rural.ses.med3 0.52 0.26 0.02 1.02 1.00
## depmistrt_rural.ses.med4 0.51 0.28 -0.03 1.07 1.00
## depmistrt_irtstresschg:rural.ses.med2 -0.05 0.66 -1.33 1.26 1.00
## depmistrt_irtstresschg:rural.ses.med3 0.59 0.60 -0.62 1.77 1.00
## depmistrt_irtstresschg:rural.ses.med4 -0.20 0.64 -1.46 1.03 1.00
## depmistrt_irtstressav:rural.ses.med2 -0.22 0.30 -0.82 0.36 1.00
## depmistrt_irtstressav:rural.ses.med3 -0.08 0.30 -0.68 0.50 1.00
## depmistrt_irtstressav:rural.ses.med4 0.19 0.34 -0.47 0.87 1.00
## depbetray_irtstresschg 0.60 0.49 -0.36 1.57 1.00
## depbetray_irtstressav 0.45 0.26 -0.04 0.94 1.00
## depbetray_rural.ses.med2 0.48 0.28 -0.07 1.01 1.00
## depbetray_rural.ses.med3 0.76 0.26 0.26 1.29 1.00
## depbetray_rural.ses.med4 0.77 0.30 0.19 1.36 1.00
## depbetray_irtstresschg:rural.ses.med2 -0.32 0.68 -1.64 1.02 1.00
## depbetray_irtstresschg:rural.ses.med3 -0.35 0.59 -1.52 0.78 1.00
## depbetray_irtstresschg:rural.ses.med4 -0.39 0.63 -1.60 0.88 1.00
## depbetray_irtstressav:rural.ses.med2 -0.33 0.31 -0.96 0.27 1.00
## depbetray_irtstressav:rural.ses.med3 -0.09 0.31 -0.69 0.51 1.00
## depbetray_irtstressav:rural.ses.med4 0.40 0.34 -0.26 1.08 1.00
## Bulk_ESS Tail_ESS
## depcantgo_Intercept 6015 3005
## depeffort_Intercept 2918 2286
## deplonely_Intercept 3490 3016
## depblues_Intercept 1888 2519
## depunfair_Intercept 5223 2977
## depmistrt_Intercept 3945 2729
## depbetray_Intercept 3678 2969
## depcantgo_irtstresschg 4691 3135
## depcantgo_irtstressav 3589 3178
## depcantgo_rural.ses.med2 4971 2922
## depcantgo_rural.ses.med3 5492 3424
## depcantgo_rural.ses.med4 5530 3547
## depcantgo_irtstresschg:rural.ses.med2 5681 2948
## depcantgo_irtstresschg:rural.ses.med3 4910 3033
## depcantgo_irtstresschg:rural.ses.med4 4845 3220
## depcantgo_irtstressav:rural.ses.med2 3801 3433
## depcantgo_irtstressav:rural.ses.med3 3791 3087
## depcantgo_irtstressav:rural.ses.med4 4160 3663
## depeffort_irtstresschg 4736 3329
## depeffort_irtstressav 3432 3032
## depeffort_rural.ses.med2 5146 3097
## depeffort_rural.ses.med3 5301 3167
## depeffort_rural.ses.med4 5018 2892
## depeffort_irtstresschg:rural.ses.med2 6493 3473
## depeffort_irtstresschg:rural.ses.med3 5949 3152
## depeffort_irtstresschg:rural.ses.med4 6043 3525
## depeffort_irtstressav:rural.ses.med2 3958 3241
## depeffort_irtstressav:rural.ses.med3 3421 3072
## depeffort_irtstressav:rural.ses.med4 3716 3403
## deplonely_irtstresschg 4315 3248
## deplonely_irtstressav 3763 3036
## deplonely_rural.ses.med2 4513 3143
## deplonely_rural.ses.med3 5318 3364
## deplonely_rural.ses.med4 4809 2952
## deplonely_irtstresschg:rural.ses.med2 5539 3728
## deplonely_irtstresschg:rural.ses.med3 4676 3339
## deplonely_irtstresschg:rural.ses.med4 5669 3270
## deplonely_irtstressav:rural.ses.med2 4071 2852
## deplonely_irtstressav:rural.ses.med3 3868 3127
## deplonely_irtstressav:rural.ses.med4 4097 3337
## depblues_irtstresschg 5045 2887
## depblues_irtstressav 3797 2848
## depblues_rural.ses.med2 4917 3151
## depblues_rural.ses.med3 5459 2570
## depblues_rural.ses.med4 5009 3075
## depblues_irtstresschg:rural.ses.med2 6633 3198
## depblues_irtstresschg:rural.ses.med3 5545 3520
## depblues_irtstresschg:rural.ses.med4 5991 2915
## depblues_irtstressav:rural.ses.med2 4005 3059
## depblues_irtstressav:rural.ses.med3 4097 3350
## depblues_irtstressav:rural.ses.med4 3806 3514
## depunfair_irtstresschg 4034 3178
## depunfair_irtstressav 3430 2766
## depunfair_rural.ses.med2 5306 3478
## depunfair_rural.ses.med3 5536 3850
## depunfair_rural.ses.med4 5236 3175
## depunfair_irtstresschg:rural.ses.med2 5737 3324
## depunfair_irtstresschg:rural.ses.med3 4749 3271
## depunfair_irtstresschg:rural.ses.med4 4565 3376
## depunfair_irtstressav:rural.ses.med2 3229 3449
## depunfair_irtstressav:rural.ses.med3 3331 3025
## depunfair_irtstressav:rural.ses.med4 3655 3135
## depmistrt_irtstresschg 4414 3276
## depmistrt_irtstressav 3501 2822
## depmistrt_rural.ses.med2 5271 3169
## depmistrt_rural.ses.med3 5064 3047
## depmistrt_rural.ses.med4 4863 3472
## depmistrt_irtstresschg:rural.ses.med2 5961 3147
## depmistrt_irtstresschg:rural.ses.med3 5102 2839
## depmistrt_irtstresschg:rural.ses.med4 6159 3161
## depmistrt_irtstressav:rural.ses.med2 3698 3257
## depmistrt_irtstressav:rural.ses.med3 3899 3172
## depmistrt_irtstressav:rural.ses.med4 3891 3443
## depbetray_irtstresschg 4215 3106
## depbetray_irtstressav 3658 2941
## depbetray_rural.ses.med2 4715 3549
## depbetray_rural.ses.med3 5069 3475
## depbetray_rural.ses.med4 5559 3336
## depbetray_irtstresschg:rural.ses.med2 5815 2945
## depbetray_irtstresschg:rural.ses.med3 5364 3443
## depbetray_irtstresschg:rural.ses.med4 5860 3296
## depbetray_irtstressav:rural.ses.med2 3943 3065
## depbetray_irtstressav:rural.ses.med3 4062 3168
## depbetray_irtstressav:rural.ses.med4 3998 3116
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).10.1.6.5 Prior summary
out.chg.alldepress.stirt.comm.fit[[2]]## prior class coef group resp
## (flat) b
## normal(0, 1) b depbetray
## normal(0, 1) b irtstressav depbetray
## normal(0, 1) b irtstressav:rural.ses.med2 depbetray
## normal(0, 1) b irtstressav:rural.ses.med3 depbetray
## normal(0, 1) b irtstressav:rural.ses.med4 depbetray
## normal(0, 1) b irtstresschg depbetray
## normal(0, 1) b irtstresschg:rural.ses.med2 depbetray
## normal(0, 1) b irtstresschg:rural.ses.med3 depbetray
## normal(0, 1) b irtstresschg:rural.ses.med4 depbetray
## normal(0, 1) b rural.ses.med2 depbetray
## normal(0, 1) b rural.ses.med3 depbetray
## normal(0, 1) b rural.ses.med4 depbetray
## normal(0, 1) b depblues
## normal(0, 1) b irtstressav depblues
## normal(0, 1) b irtstressav:rural.ses.med2 depblues
## normal(0, 1) b irtstressav:rural.ses.med3 depblues
## normal(0, 1) b irtstressav:rural.ses.med4 depblues
## normal(0, 1) b irtstresschg depblues
## normal(0, 1) b irtstresschg:rural.ses.med2 depblues
## normal(0, 1) b irtstresschg:rural.ses.med3 depblues
## normal(0, 1) b irtstresschg:rural.ses.med4 depblues
## normal(0, 1) b rural.ses.med2 depblues
## normal(0, 1) b rural.ses.med3 depblues
## normal(0, 1) b rural.ses.med4 depblues
## normal(0, 1) b depcantgo
## normal(0, 1) b irtstressav depcantgo
## normal(0, 1) b irtstressav:rural.ses.med2 depcantgo
## normal(0, 1) b irtstressav:rural.ses.med3 depcantgo
## normal(0, 1) b irtstressav:rural.ses.med4 depcantgo
## normal(0, 1) b irtstresschg depcantgo
## normal(0, 1) b irtstresschg:rural.ses.med2 depcantgo
## normal(0, 1) b irtstresschg:rural.ses.med3 depcantgo
## normal(0, 1) b irtstresschg:rural.ses.med4 depcantgo
## normal(0, 1) b rural.ses.med2 depcantgo
## normal(0, 1) b rural.ses.med3 depcantgo
## normal(0, 1) b rural.ses.med4 depcantgo
## normal(0, 1) b depeffort
## normal(0, 1) b irtstressav depeffort
## normal(0, 1) b irtstressav:rural.ses.med2 depeffort
## normal(0, 1) b irtstressav:rural.ses.med3 depeffort
## normal(0, 1) b irtstressav:rural.ses.med4 depeffort
## normal(0, 1) b irtstresschg depeffort
## normal(0, 1) b irtstresschg:rural.ses.med2 depeffort
## normal(0, 1) b irtstresschg:rural.ses.med3 depeffort
## normal(0, 1) b irtstresschg:rural.ses.med4 depeffort
## normal(0, 1) b rural.ses.med2 depeffort
## normal(0, 1) b rural.ses.med3 depeffort
## normal(0, 1) b rural.ses.med4 depeffort
## normal(0, 1) b deplonely
## normal(0, 1) b irtstressav deplonely
## normal(0, 1) b irtstressav:rural.ses.med2 deplonely
## normal(0, 1) b irtstressav:rural.ses.med3 deplonely
## normal(0, 1) b irtstressav:rural.ses.med4 deplonely
## normal(0, 1) b irtstresschg deplonely
## normal(0, 1) b irtstresschg:rural.ses.med2 deplonely
## normal(0, 1) b irtstresschg:rural.ses.med3 deplonely
## normal(0, 1) b irtstresschg:rural.ses.med4 deplonely
## normal(0, 1) b rural.ses.med2 deplonely
## normal(0, 1) b rural.ses.med3 deplonely
## normal(0, 1) b rural.ses.med4 deplonely
## normal(0, 1) b depmistrt
## normal(0, 1) b irtstressav depmistrt
## normal(0, 1) b irtstressav:rural.ses.med2 depmistrt
## normal(0, 1) b irtstressav:rural.ses.med3 depmistrt
## normal(0, 1) b irtstressav:rural.ses.med4 depmistrt
## normal(0, 1) b irtstresschg depmistrt
## normal(0, 1) b irtstresschg:rural.ses.med2 depmistrt
## normal(0, 1) b irtstresschg:rural.ses.med3 depmistrt
## normal(0, 1) b irtstresschg:rural.ses.med4 depmistrt
## normal(0, 1) b rural.ses.med2 depmistrt
## normal(0, 1) b rural.ses.med3 depmistrt
## normal(0, 1) b rural.ses.med4 depmistrt
## normal(0, 1) b depunfair
## normal(0, 1) b irtstressav depunfair
## normal(0, 1) b irtstressav:rural.ses.med2 depunfair
## normal(0, 1) b irtstressav:rural.ses.med3 depunfair
## normal(0, 1) b irtstressav:rural.ses.med4 depunfair
## normal(0, 1) b irtstresschg depunfair
## normal(0, 1) b irtstresschg:rural.ses.med2 depunfair
## normal(0, 1) b irtstresschg:rural.ses.med3 depunfair
## normal(0, 1) b irtstresschg:rural.ses.med4 depunfair
## normal(0, 1) b rural.ses.med2 depunfair
## normal(0, 1) b rural.ses.med3 depunfair
## normal(0, 1) b rural.ses.med4 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dpar nlpar lb ub source
## default
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)10.1.7 Corr X Community: stdz sum stress & binary outcomes
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names)
)
chg.prjcrime.stsum.comm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
sumstresschg + sumstressav + rural.ses.med +
sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id),
data = stress.long2,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
adapt_delta = 0.85,
seed = 8675309,
file = "Models/chg_prjcrime_stsum_comm_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
varlist <- c("^b_prjthflt5_sum","^b_prjthfgt5_sum", "^b_prjthreat_sum",
"^b_prjharm_sum", "^b_usedrg_sum", "^b_prjhack_sum")
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="prjthflt5")
ppcheckdv2 <-pp_check(modelfit, resp="prjthfgt5")
ppcheckdv3 <-pp_check(modelfit, resp="prjthreat")
ppcheckdv4 <-pp_check(modelfit, resp="prjharm")
ppcheckdv5 <-pp_check(modelfit, resp="prjusedrg")
ppcheckdv6 <-pp_check(modelfit, resp="prjhack")
plotcoefs2 <- mcmc_plot(modelfit, variable = varlist, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for sum stress scale coefficients\nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6, plotcoefs2)
return(allchecks)
}
out.chg.prjcrime.stsum.comm.fit <- ppchecks(chg.prjcrime.stsum.comm.fit)10.1.7.1 Coefficient plot (intervals)
out.chg.prjcrime.stsum.comm.fit[[9]] 
10.1.7.2 PPcheck (density)
p1 <- out.chg.prjcrime.stsum.comm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stsum.comm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stsum.comm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stsum.comm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stsum.comm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stsum.comm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
10.1.7.3 Fit summary
out.chg.prjcrime.stsum.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## prjthfgt5 ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## prjthreat ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## prjharm ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## prjusedrg ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## prjhack ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## Data: stress.long2 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.76 0.54 2.81 4.94 1.01 953
## sd(prjthfgt5_Intercept) 3.33 0.49 2.45 4.40 1.00 969
## sd(prjthreat_Intercept) 3.15 0.53 2.17 4.26 1.00 1002
## sd(prjharm_Intercept) 3.01 0.54 2.08 4.18 1.00 1019
## sd(prjusedrg_Intercept) 2.76 0.51 1.85 3.84 1.00 1034
## sd(prjhack_Intercept) 0.83 0.55 0.04 1.97 1.00 435
## Tail_ESS
## sd(prjthflt5_Intercept) 1521
## sd(prjthfgt5_Intercept) 1665
## sd(prjthreat_Intercept) 1462
## sd(prjharm_Intercept) 2073
## sd(prjusedrg_Intercept) 1728
## sd(prjhack_Intercept) 789
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_Intercept -5.72 0.70 -7.16 -4.45 1.00
## prjthfgt5_Intercept -5.44 0.68 -6.86 -4.24 1.00
## prjthreat_Intercept -6.66 0.83 -8.43 -5.19 1.00
## prjharm_Intercept -6.49 0.86 -8.29 -4.93 1.00
## prjusedrg_Intercept -6.27 0.83 -8.04 -4.79 1.00
## prjhack_Intercept -4.51 0.61 -5.91 -3.49 1.00
## prjthflt5_sumstresschg 0.52 0.63 -0.75 1.73 1.00
## prjthflt5_sumstressav 0.47 0.46 -0.39 1.41 1.00
## prjthflt5_rural.ses.med2 -1.43 0.67 -2.77 -0.11 1.00
## prjthflt5_rural.ses.med3 1.06 0.60 -0.11 2.21 1.00
## prjthflt5_rural.ses.med4 2.08 0.59 0.95 3.26 1.00
## prjthflt5_sumstresschg:rural.ses.med2 -0.49 0.91 -2.26 1.27 1.00
## prjthflt5_sumstresschg:rural.ses.med3 0.12 0.77 -1.42 1.67 1.00
## prjthflt5_sumstresschg:rural.ses.med4 -0.26 0.80 -1.81 1.33 1.00
## prjthflt5_sumstressav:rural.ses.med2 0.70 0.71 -0.69 2.13 1.00
## prjthflt5_sumstressav:rural.ses.med3 0.69 0.61 -0.50 1.89 1.00
## prjthflt5_sumstressav:rural.ses.med4 -0.28 0.58 -1.45 0.85 1.00
## prjthfgt5_sumstresschg 0.85 0.65 -0.46 2.10 1.00
## prjthfgt5_sumstressav 0.34 0.42 -0.45 1.16 1.00
## prjthfgt5_rural.ses.med2 -1.53 0.65 -2.87 -0.25 1.00
## prjthfgt5_rural.ses.med3 1.04 0.55 -0.03 2.19 1.00
## prjthfgt5_rural.ses.med4 1.89 0.57 0.79 3.03 1.00
## prjthfgt5_sumstresschg:rural.ses.med2 -0.23 0.90 -2.00 1.57 1.00
## prjthfgt5_sumstresschg:rural.ses.med3 -0.12 0.79 -1.67 1.44 1.00
## prjthfgt5_sumstresschg:rural.ses.med4 0.21 0.82 -1.40 1.86 1.00
## prjthfgt5_sumstressav:rural.ses.med2 0.32 0.69 -1.03 1.73 1.00
## prjthfgt5_sumstressav:rural.ses.med3 0.59 0.55 -0.49 1.70 1.00
## prjthfgt5_sumstressav:rural.ses.med4 0.06 0.53 -0.95 1.09 1.00
## prjthreat_sumstresschg -0.44 0.72 -1.85 0.99 1.00
## prjthreat_sumstressav 0.82 0.48 -0.10 1.81 1.00
## prjthreat_rural.ses.med2 -0.88 0.71 -2.31 0.51 1.00
## prjthreat_rural.ses.med3 0.40 0.64 -0.90 1.62 1.00
## prjthreat_rural.ses.med4 1.46 0.63 0.21 2.72 1.00
## prjthreat_sumstresschg:rural.ses.med2 0.39 0.92 -1.42 2.21 1.00
## prjthreat_sumstresschg:rural.ses.med3 -0.16 0.83 -1.87 1.46 1.00
## prjthreat_sumstresschg:rural.ses.med4 -1.00 0.88 -2.71 0.71 1.00
## prjthreat_sumstressav:rural.ses.med2 -0.52 0.74 -1.94 0.94 1.00
## prjthreat_sumstressav:rural.ses.med3 0.23 0.63 -1.01 1.44 1.00
## prjthreat_sumstressav:rural.ses.med4 0.40 0.60 -0.78 1.55 1.00
## prjharm_sumstresschg -0.62 0.71 -2.01 0.77 1.00
## prjharm_sumstressav 0.23 0.45 -0.63 1.14 1.00
## prjharm_rural.ses.med2 -0.57 0.68 -1.89 0.73 1.00
## prjharm_rural.ses.med3 0.21 0.63 -1.05 1.45 1.00
## prjharm_rural.ses.med4 0.92 0.63 -0.34 2.13 1.00
## prjharm_sumstresschg:rural.ses.med2 0.08 0.88 -1.63 1.78 1.00
## prjharm_sumstresschg:rural.ses.med3 0.13 0.87 -1.55 1.81 1.00
## prjharm_sumstresschg:rural.ses.med4 -1.15 0.86 -2.81 0.59 1.00
## prjharm_sumstressav:rural.ses.med2 -0.74 0.66 -2.02 0.59 1.00
## prjharm_sumstressav:rural.ses.med3 0.37 0.63 -0.87 1.58 1.00
## prjharm_sumstressav:rural.ses.med4 0.41 0.61 -0.78 1.61 1.00
## prjusedrg_sumstresschg -0.26 0.75 -1.72 1.24 1.00
## prjusedrg_sumstressav 0.50 0.48 -0.40 1.47 1.00
## prjusedrg_rural.ses.med2 -0.62 0.67 -1.92 0.68 1.00
## prjusedrg_rural.ses.med3 -0.59 0.66 -1.90 0.66 1.00
## prjusedrg_rural.ses.med4 1.44 0.61 0.25 2.62 1.00
## prjusedrg_sumstresschg:rural.ses.med2 0.29 0.91 -1.51 2.07 1.00
## prjusedrg_sumstresschg:rural.ses.med3 -0.09 0.87 -1.80 1.63 1.00
## prjusedrg_sumstresschg:rural.ses.med4 -0.75 0.84 -2.37 0.93 1.00
## prjusedrg_sumstressav:rural.ses.med2 -0.78 0.67 -2.14 0.56 1.00
## prjusedrg_sumstressav:rural.ses.med3 -0.10 0.67 -1.39 1.21 1.00
## prjusedrg_sumstressav:rural.ses.med4 0.72 0.61 -0.45 1.91 1.00
## prjhack_sumstresschg -0.20 0.73 -1.59 1.24 1.00
## prjhack_sumstressav 0.56 0.43 -0.24 1.42 1.00
## prjhack_rural.ses.med2 -0.96 0.67 -2.33 0.30 1.00
## prjhack_rural.ses.med3 0.20 0.53 -0.81 1.22 1.00
## prjhack_rural.ses.med4 0.34 0.58 -0.79 1.46 1.00
## prjhack_sumstresschg:rural.ses.med2 0.45 0.94 -1.38 2.32 1.00
## prjhack_sumstresschg:rural.ses.med3 -0.12 0.84 -1.74 1.55 1.00
## prjhack_sumstresschg:rural.ses.med4 -1.06 0.86 -2.75 0.64 1.00
## prjhack_sumstressav:rural.ses.med2 -0.43 0.67 -1.72 0.87 1.00
## prjhack_sumstressav:rural.ses.med3 -0.36 0.55 -1.41 0.77 1.00
## prjhack_sumstressav:rural.ses.med4 0.55 0.55 -0.52 1.65 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1297 2015
## prjthfgt5_Intercept 1134 2247
## prjthreat_Intercept 1326 2353
## prjharm_Intercept 1249 2060
## prjusedrg_Intercept 1306 2205
## prjhack_Intercept 891 1049
## prjthflt5_sumstresschg 3787 3066
## prjthflt5_sumstressav 1439 2095
## prjthflt5_rural.ses.med2 2403 2859
## prjthflt5_rural.ses.med3 1753 2189
## prjthflt5_rural.ses.med4 1800 2370
## prjthflt5_sumstresschg:rural.ses.med2 4729 2930
## prjthflt5_sumstresschg:rural.ses.med3 3886 3002
## prjthflt5_sumstresschg:rural.ses.med4 4236 2973
## prjthflt5_sumstressav:rural.ses.med2 1613 2122
## prjthflt5_sumstressav:rural.ses.med3 1684 1686
## prjthflt5_sumstressav:rural.ses.med4 1652 2288
## prjthfgt5_sumstresschg 3207 3119
## prjthfgt5_sumstressav 1655 2523
## prjthfgt5_rural.ses.med2 2884 2849
## prjthfgt5_rural.ses.med3 2035 2726
## prjthfgt5_rural.ses.med4 2054 2602
## prjthfgt5_sumstresschg:rural.ses.med2 4606 2974
## prjthfgt5_sumstresschg:rural.ses.med3 3941 2727
## prjthfgt5_sumstresschg:rural.ses.med4 4059 3061
## prjthfgt5_sumstressav:rural.ses.med2 3088 2779
## prjthfgt5_sumstressav:rural.ses.med3 1605 2559
## prjthfgt5_sumstressav:rural.ses.med4 1672 2380
## prjthreat_sumstresschg 3658 2958
## prjthreat_sumstressav 2154 2492
## prjthreat_rural.ses.med2 3403 2913
## prjthreat_rural.ses.med3 2506 2695
## prjthreat_rural.ses.med4 2290 2450
## prjthreat_sumstresschg:rural.ses.med2 4082 2874
## prjthreat_sumstresschg:rural.ses.med3 3886 2788
## prjthreat_sumstresschg:rural.ses.med4 3739 3015
## prjthreat_sumstressav:rural.ses.med2 3262 3187
## prjthreat_sumstressav:rural.ses.med3 2058 2755
## prjthreat_sumstressav:rural.ses.med4 2071 2593
## prjharm_sumstresschg 3249 2811
## prjharm_sumstressav 1740 2279
## prjharm_rural.ses.med2 2534 2500
## prjharm_rural.ses.med3 2347 2701
## prjharm_rural.ses.med4 2387 2764
## prjharm_sumstresschg:rural.ses.med2 4190 3191
## prjharm_sumstresschg:rural.ses.med3 4641 2992
## prjharm_sumstresschg:rural.ses.med4 3611 3079
## prjharm_sumstressav:rural.ses.med2 2584 2868
## prjharm_sumstressav:rural.ses.med3 2545 2986
## prjharm_sumstressav:rural.ses.med4 2414 2808
## prjusedrg_sumstresschg 4567 2994
## prjusedrg_sumstressav 2077 2683
## prjusedrg_rural.ses.med2 3262 2664
## prjusedrg_rural.ses.med3 2710 2876
## prjusedrg_rural.ses.med4 2559 2707
## prjusedrg_sumstresschg:rural.ses.med2 5259 2838
## prjusedrg_sumstresschg:rural.ses.med3 4787 2934
## prjusedrg_sumstresschg:rural.ses.med4 4537 2881
## prjusedrg_sumstressav:rural.ses.med2 2617 2899
## prjusedrg_sumstressav:rural.ses.med3 2849 2907
## prjusedrg_sumstressav:rural.ses.med4 1985 2727
## prjhack_sumstresschg 3883 2969
## prjhack_sumstressav 1596 2533
## prjhack_rural.ses.med2 3970 2736
## prjhack_rural.ses.med3 3179 3126
## prjhack_rural.ses.med4 2718 2612
## prjhack_sumstresschg:rural.ses.med2 4637 3205
## prjhack_sumstresschg:rural.ses.med3 4003 2750
## prjhack_sumstresschg:rural.ses.med4 3796 2729
## prjhack_sumstressav:rural.ses.med2 2991 2913
## prjhack_sumstressav:rural.ses.med3 1994 2258
## prjhack_sumstressav:rural.ses.med4 1601 2091
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).10.1.7.4 Prior summary
out.chg.prjcrime.stsum.comm.fit[[2]]## prior class coef group resp
## (flat) b
## normal(0, 1) b prjhack
## normal(0, 1) b rural.ses.med2 prjhack
## normal(0, 1) b rural.ses.med3 prjhack
## normal(0, 1) b rural.ses.med4 prjhack
## normal(0, 1) b sumstressav prjhack
## normal(0, 1) b sumstressav:rural.ses.med2 prjhack
## normal(0, 1) b sumstressav:rural.ses.med3 prjhack
## normal(0, 1) b sumstressav:rural.ses.med4 prjhack
## normal(0, 1) b sumstresschg prjhack
## normal(0, 1) b sumstresschg:rural.ses.med2 prjhack
## normal(0, 1) b sumstresschg:rural.ses.med3 prjhack
## normal(0, 1) b sumstresschg:rural.ses.med4 prjhack
## normal(0, 1) b prjharm
## normal(0, 1) b rural.ses.med2 prjharm
## normal(0, 1) b rural.ses.med3 prjharm
## normal(0, 1) b rural.ses.med4 prjharm
## normal(0, 1) b sumstressav prjharm
## normal(0, 1) b sumstressav:rural.ses.med2 prjharm
## normal(0, 1) b sumstressav:rural.ses.med3 prjharm
## normal(0, 1) b sumstressav:rural.ses.med4 prjharm
## normal(0, 1) b sumstresschg prjharm
## normal(0, 1) b sumstresschg:rural.ses.med2 prjharm
## normal(0, 1) b sumstresschg:rural.ses.med3 prjharm
## normal(0, 1) b sumstresschg:rural.ses.med4 prjharm
## normal(0, 1) b prjthfgt5
## normal(0, 1) b rural.ses.med2 prjthfgt5
## normal(0, 1) b rural.ses.med3 prjthfgt5
## normal(0, 1) b rural.ses.med4 prjthfgt5
## normal(0, 1) b sumstressav prjthfgt5
## normal(0, 1) b sumstressav:rural.ses.med2 prjthfgt5
## normal(0, 1) b sumstressav:rural.ses.med3 prjthfgt5
## normal(0, 1) b sumstressav:rural.ses.med4 prjthfgt5
## normal(0, 1) b sumstresschg prjthfgt5
## normal(0, 1) b sumstresschg:rural.ses.med2 prjthfgt5
## normal(0, 1) b sumstresschg:rural.ses.med3 prjthfgt5
## normal(0, 1) b sumstresschg:rural.ses.med4 prjthfgt5
## normal(0, 1) b prjthflt5
## normal(0, 1) b rural.ses.med2 prjthflt5
## normal(0, 1) b rural.ses.med3 prjthflt5
## normal(0, 1) b rural.ses.med4 prjthflt5
## normal(0, 1) b sumstressav prjthflt5
## normal(0, 1) b sumstressav:rural.ses.med2 prjthflt5
## normal(0, 1) b sumstressav:rural.ses.med3 prjthflt5
## normal(0, 1) b sumstressav:rural.ses.med4 prjthflt5
## normal(0, 1) b sumstresschg prjthflt5
## normal(0, 1) b sumstresschg:rural.ses.med2 prjthflt5
## normal(0, 1) b sumstresschg:rural.ses.med3 prjthflt5
## normal(0, 1) b sumstresschg:rural.ses.med4 prjthflt5
## normal(0, 1) b prjthreat
## normal(0, 1) b rural.ses.med2 prjthreat
## normal(0, 1) b rural.ses.med3 prjthreat
## normal(0, 1) b rural.ses.med4 prjthreat
## normal(0, 1) b sumstressav prjthreat
## normal(0, 1) b sumstressav:rural.ses.med2 prjthreat
## normal(0, 1) b sumstressav:rural.ses.med3 prjthreat
## normal(0, 1) b sumstressav:rural.ses.med4 prjthreat
## normal(0, 1) b sumstresschg prjthreat
## normal(0, 1) b sumstresschg:rural.ses.med2 prjthreat
## normal(0, 1) b sumstresschg:rural.ses.med3 prjthreat
## normal(0, 1) b sumstresschg:rural.ses.med4 prjthreat
## normal(0, 1) b prjusedrg
## normal(0, 1) b rural.ses.med2 prjusedrg
## normal(0, 1) b rural.ses.med3 prjusedrg
## normal(0, 1) b rural.ses.med4 prjusedrg
## normal(0, 1) b sumstressav prjusedrg
## normal(0, 1) b sumstressav:rural.ses.med2 prjusedrg
## normal(0, 1) b sumstressav:rural.ses.med3 prjusedrg
## normal(0, 1) b sumstressav:rural.ses.med4 prjusedrg
## normal(0, 1) b sumstresschg prjusedrg
## normal(0, 1) b sumstresschg:rural.ses.med2 prjusedrg
## normal(0, 1) b sumstresschg:rural.ses.med3 prjusedrg
## normal(0, 1) b sumstresschg:rural.ses.med4 prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dpar nlpar lb ub source
## default
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)10.1.8 Corr X Community: stdz sum stress & any prj crime
# "Any crime" outcome & sum stress scale
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept'),
set_prior('normal(0, 1)', class = 'b')
)
chg.anyprjcrime.stsum.comm.fit <- brm(prjany ~ 1 +
sumstresschg + sumstressav + rural.ses.med +
sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id),
data = stress.long2,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_anyprjcrime_stsum_comm_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit)
plotcoefs2 <- mcmc_plot(modelfit, variable = "^b_sum", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, plotcoefs2)
return(allchecks)
}
out.chg.anyprjcrime.stsum.comm.fit <- ppchecks(chg.anyprjcrime.stsum.comm.fit)10.1.8.1 Coefficient plot (intervals)
out.chg.anyprjcrime.stsum.comm.fit[[4]] 
10.1.8.2 PPcheck (density)
out.chg.anyprjcrime.stsum.comm.fit[[3]] + labs(title = "Any Crime Intent (chg)") 
10.1.8.3 Fit summary
out.chg.anyprjcrime.stsum.comm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## Data: stress.long2 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.11 0.42 2.36 4.00 1.00 952 1756
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept -4.03 0.49 -5.05 -3.13 1.00 1635
## sumstresschg 0.70 0.57 -0.43 1.83 1.00 3797
## sumstressav 0.46 0.39 -0.29 1.23 1.00 1751
## rural.ses.med2 -1.12 0.55 -2.21 -0.06 1.00 2579
## rural.ses.med3 0.91 0.48 -0.03 1.84 1.00 2247
## rural.ses.med4 2.25 0.52 1.25 3.31 1.00 2142
## sumstresschg:rural.ses.med2 -0.46 0.81 -2.11 1.16 1.00 5110
## sumstresschg:rural.ses.med3 -0.16 0.72 -1.60 1.29 1.00 4227
## sumstresschg:rural.ses.med4 -0.46 0.76 -1.93 1.04 1.00 3981
## sumstressav:rural.ses.med2 -0.31 0.56 -1.39 0.77 1.00 2271
## sumstressav:rural.ses.med3 0.31 0.50 -0.64 1.32 1.00 2016
## sumstressav:rural.ses.med4 0.03 0.49 -0.93 1.01 1.00 2045
## Tail_ESS
## Intercept 2215
## sumstresschg 2810
## sumstressav 2452
## rural.ses.med2 2742
## rural.ses.med3 2777
## rural.ses.med4 2579
## sumstresschg:rural.ses.med2 2371
## sumstresschg:rural.ses.med3 3077
## sumstresschg:rural.ses.med4 2967
## sumstressav:rural.ses.med2 2693
## sumstressav:rural.ses.med3 2713
## sumstressav:rural.ses.med4 2454
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).10.1.8.4 Prior summary
out.chg.anyprjcrime.stsum.comm.fit[[2]]## prior class coef group resp dpar
## normal(0, 1) b
## normal(0, 1) b rural.ses.med2
## normal(0, 1) b rural.ses.med3
## normal(0, 1) b rural.ses.med4
## normal(0, 1) b sumstressav
## normal(0, 1) b sumstressav:rural.ses.med2
## normal(0, 1) b sumstressav:rural.ses.med3
## normal(0, 1) b sumstressav:rural.ses.med4
## normal(0, 1) b sumstresschg
## normal(0, 1) b sumstresschg:rural.ses.med2
## normal(0, 1) b sumstresschg:rural.ses.med3
## normal(0, 1) b sumstresschg:rural.ses.med4
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## nlpar lb ub source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)10.1.9 Corr X Community: stdz sum stress & negative emotions
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names)
)
chg.alldepress.stsum.comm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 + sumstresschg + sumstressav +
rural.ses.med + sumstresschg:rural.ses.med +
sumstressav:rural.ses.med + (1 | id),
data = stress.long2,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stsum_comm_fit",
file_refit = "on_change"
)
##Update function to call all ppchecks for bivar neg emotions chg models
varlist1 <- c("^b_depcantgo_sum","^b_depeffort_sum", "^b_deplonely_sum",
"^b_depblues_sum")
varlist2 <- c("^b_depunfair_sum", "^b_depmistrt_sum", "^b_depbetray_sum")
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="depcantgo")
ppcheckdv2 <-pp_check(modelfit, resp="depeffort")
ppcheckdv3 <-pp_check(modelfit, resp="deplonely")
ppcheckdv4 <-pp_check(modelfit, resp="depblues")
ppcheckdv5 <-pp_check(modelfit, resp="depunfair")
ppcheckdv6 <-pp_check(modelfit, resp="depmistrt")
ppcheckdv7 <-pp_check(modelfit, resp="depbetray")
plotcoefs1 <- mcmc_plot(modelfit, variable = varlist1, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients\nwith medians, 80%, and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = varlist2, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients\nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2, ppcheckdv3,
ppcheckdv4, ppcheckdv5, ppcheckdv6, ppcheckdv7,
plotcoefs1, plotcoefs2)
return(allchecks)
}
out.chg.alldepress.stsum.comm.fit <- ppchecks(chg.alldepress.stsum.comm.fit)10.1.9.1 Coefficient plot (depressive symptoms)
out.chg.alldepress.stsum.comm.fit[[10]]
10.1.9.2 Coefficient plot (criminogenic emotions)
out.chg.alldepress.stsum.comm.fit[[11]]
10.1.9.3 PPcheck (density)
p1 <- out.chg.alldepress.stsum.comm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stsum.comm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stsum.comm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stsum.comm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stsum.comm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stsum.comm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stsum.comm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
10.1.9.4 Fit summary
out.chg.alldepress.stsum.comm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## depeffort ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## deplonely ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## depblues ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## depunfair ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## depmistrt ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## depbetray ~ 1 + sumstresschg + sumstressav + rural.ses.med + sumstresschg:rural.ses.med + sumstressav:rural.ses.med + (1 | id)
## Data: stress.long2 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.38 0.21 0.03 0.79 1.00 538
## sd(depeffort_Intercept) 0.49 0.29 0.02 1.05 1.01 462
## sd(deplonely_Intercept) 0.48 0.26 0.03 0.97 1.02 378
## sd(depblues_Intercept) 0.67 0.32 0.05 1.26 1.01 466
## sd(depunfair_Intercept) 0.26 0.18 0.01 0.67 1.01 659
## sd(depmistrt_Intercept) 0.34 0.21 0.03 0.80 1.00 796
## sd(depbetray_Intercept) 0.41 0.25 0.02 0.94 1.00 650
## Tail_ESS
## sd(depcantgo_Intercept) 1295
## sd(depeffort_Intercept) 778
## sd(deplonely_Intercept) 875
## sd(depblues_Intercept) 882
## sd(depunfair_Intercept) 1550
## sd(depmistrt_Intercept) 1781
## sd(depbetray_Intercept) 1190
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_Intercept -0.54 0.15 -0.84 -0.24 1.00
## depeffort_Intercept -2.08 0.23 -2.56 -1.65 1.00
## deplonely_Intercept -1.06 0.17 -1.41 -0.73 1.00
## depblues_Intercept -2.33 0.27 -2.88 -1.85 1.01
## depunfair_Intercept -1.57 0.18 -1.92 -1.23 1.00
## depmistrt_Intercept -2.02 0.22 -2.46 -1.61 1.00
## depbetray_Intercept -2.24 0.24 -2.73 -1.80 1.00
## depcantgo_sumstresschg 1.42 0.45 0.53 2.31 1.00
## depcantgo_sumstressav -0.14 0.16 -0.45 0.16 1.00
## depcantgo_rural.ses.med2 0.37 0.21 -0.04 0.77 1.00
## depcantgo_rural.ses.med3 0.18 0.20 -0.20 0.56 1.00
## depcantgo_rural.ses.med4 0.09 0.21 -0.34 0.51 1.00
## depcantgo_sumstresschg:rural.ses.med2 0.48 0.61 -0.73 1.69 1.00
## depcantgo_sumstresschg:rural.ses.med3 -1.26 0.56 -2.33 -0.17 1.00
## depcantgo_sumstresschg:rural.ses.med4 -0.21 0.66 -1.55 1.08 1.00
## depcantgo_sumstressav:rural.ses.med2 0.37 0.22 -0.05 0.79 1.00
## depcantgo_sumstressav:rural.ses.med3 -0.02 0.21 -0.44 0.39 1.00
## depcantgo_sumstressav:rural.ses.med4 0.23 0.21 -0.18 0.65 1.00
## depeffort_sumstresschg 0.95 0.50 -0.03 1.94 1.00
## depeffort_sumstressav -0.15 0.20 -0.54 0.25 1.00
## depeffort_rural.ses.med2 0.07 0.29 -0.49 0.64 1.00
## depeffort_rural.ses.med3 0.51 0.27 -0.02 1.06 1.00
## depeffort_rural.ses.med4 0.55 0.28 -0.00 1.13 1.00
## depeffort_sumstresschg:rural.ses.med2 -0.20 0.70 -1.59 1.12 1.00
## depeffort_sumstresschg:rural.ses.med3 -0.14 0.62 -1.36 1.10 1.00
## depeffort_sumstresschg:rural.ses.med4 -0.12 0.71 -1.54 1.22 1.00
## depeffort_sumstressav:rural.ses.med2 0.25 0.29 -0.33 0.82 1.00
## depeffort_sumstressav:rural.ses.med3 0.20 0.27 -0.33 0.72 1.00
## depeffort_sumstressav:rural.ses.med4 0.35 0.27 -0.18 0.88 1.00
## deplonely_sumstresschg 0.94 0.45 0.09 1.84 1.00
## deplonely_sumstressav -0.08 0.16 -0.41 0.22 1.00
## deplonely_rural.ses.med2 -0.39 0.23 -0.86 0.06 1.00
## deplonely_rural.ses.med3 -0.06 0.22 -0.49 0.37 1.00
## deplonely_rural.ses.med4 0.41 0.23 -0.02 0.87 1.00
## deplonely_sumstresschg:rural.ses.med2 0.29 0.64 -0.93 1.60 1.00
## deplonely_sumstresschg:rural.ses.med3 -0.66 0.56 -1.75 0.42 1.00
## deplonely_sumstresschg:rural.ses.med4 0.79 0.65 -0.47 2.07 1.00
## deplonely_sumstressav:rural.ses.med2 0.04 0.24 -0.42 0.52 1.00
## deplonely_sumstressav:rural.ses.med3 0.25 0.23 -0.19 0.71 1.00
## deplonely_sumstressav:rural.ses.med4 0.18 0.22 -0.25 0.62 1.00
## depblues_sumstresschg -0.22 0.53 -1.26 0.84 1.00
## depblues_sumstressav -0.08 0.23 -0.52 0.37 1.00
## depblues_rural.ses.med2 0.10 0.31 -0.51 0.68 1.00
## depblues_rural.ses.med3 0.25 0.30 -0.35 0.84 1.00
## depblues_rural.ses.med4 0.50 0.31 -0.10 1.12 1.00
## depblues_sumstresschg:rural.ses.med2 -0.28 0.71 -1.67 1.10 1.00
## depblues_sumstresschg:rural.ses.med3 0.20 0.65 -1.13 1.45 1.00
## depblues_sumstresschg:rural.ses.med4 0.25 0.73 -1.16 1.68 1.00
## depblues_sumstressav:rural.ses.med2 -0.50 0.32 -1.13 0.12 1.00
## depblues_sumstressav:rural.ses.med3 0.53 0.30 -0.06 1.10 1.00
## depblues_sumstressav:rural.ses.med4 0.51 0.31 -0.10 1.11 1.00
## depunfair_sumstresschg 1.84 0.47 0.94 2.77 1.00
## depunfair_sumstressav 0.20 0.19 -0.18 0.58 1.00
## depunfair_rural.ses.med2 0.26 0.23 -0.19 0.72 1.00
## depunfair_rural.ses.med3 0.62 0.22 0.19 1.06 1.00
## depunfair_rural.ses.med4 0.89 0.23 0.45 1.36 1.00
## depunfair_sumstresschg:rural.ses.med2 0.47 0.66 -0.83 1.74 1.00
## depunfair_sumstresschg:rural.ses.med3 -0.87 0.57 -1.97 0.24 1.00
## depunfair_sumstresschg:rural.ses.med4 0.73 0.69 -0.64 2.15 1.00
## depunfair_sumstressav:rural.ses.med2 -0.08 0.26 -0.58 0.42 1.00
## depunfair_sumstressav:rural.ses.med3 0.13 0.24 -0.36 0.60 1.00
## depunfair_sumstressav:rural.ses.med4 0.07 0.24 -0.39 0.55 1.00
## depmistrt_sumstresschg 0.94 0.52 -0.09 1.96 1.00
## depmistrt_sumstressav 0.38 0.23 -0.05 0.84 1.00
## depmistrt_rural.ses.med2 0.53 0.27 0.01 1.06 1.00
## depmistrt_rural.ses.med3 0.41 0.27 -0.12 0.94 1.00
## depmistrt_rural.ses.med4 0.46 0.30 -0.12 1.04 1.00
## depmistrt_sumstresschg:rural.ses.med2 -0.12 0.67 -1.41 1.18 1.00
## depmistrt_sumstresschg:rural.ses.med3 -0.85 0.64 -2.10 0.41 1.00
## depmistrt_sumstresschg:rural.ses.med4 0.44 0.72 -0.97 1.85 1.00
## depmistrt_sumstressav:rural.ses.med2 0.05 0.30 -0.53 0.62 1.00
## depmistrt_sumstressav:rural.ses.med3 0.00 0.29 -0.58 0.57 1.00
## depmistrt_sumstressav:rural.ses.med4 0.19 0.30 -0.38 0.77 1.00
## depbetray_sumstresschg 1.26 0.52 0.23 2.30 1.00
## depbetray_sumstressav 0.32 0.24 -0.13 0.80 1.00
## depbetray_rural.ses.med2 0.44 0.29 -0.12 1.00 1.00
## depbetray_rural.ses.med3 0.62 0.28 0.06 1.17 1.00
## depbetray_rural.ses.med4 0.68 0.30 0.07 1.27 1.00
## depbetray_sumstresschg:rural.ses.med2 0.18 0.70 -1.22 1.54 1.00
## depbetray_sumstresschg:rural.ses.med3 -0.70 0.63 -1.95 0.56 1.00
## depbetray_sumstresschg:rural.ses.med4 -0.23 0.69 -1.57 1.12 1.00
## depbetray_sumstressav:rural.ses.med2 0.03 0.31 -0.60 0.64 1.00
## depbetray_sumstressav:rural.ses.med3 0.14 0.30 -0.46 0.70 1.00
## depbetray_sumstressav:rural.ses.med4 0.51 0.31 -0.08 1.12 1.00
## Bulk_ESS Tail_ESS
## depcantgo_Intercept 3413 2586
## depeffort_Intercept 2135 2694
## deplonely_Intercept 2425 2948
## depblues_Intercept 1459 2707
## depunfair_Intercept 3099 2843
## depmistrt_Intercept 2694 2748
## depbetray_Intercept 2824 2736
## depcantgo_sumstresschg 3561 3045
## depcantgo_sumstressav 2349 2364
## depcantgo_rural.ses.med2 3742 3180
## depcantgo_rural.ses.med3 3659 2692
## depcantgo_rural.ses.med4 4078 3089
## depcantgo_sumstresschg:rural.ses.med2 4126 2952
## depcantgo_sumstresschg:rural.ses.med3 4258 3664
## depcantgo_sumstresschg:rural.ses.med4 5247 3357
## depcantgo_sumstressav:rural.ses.med2 2692 2950
## depcantgo_sumstressav:rural.ses.med3 2590 2737
## depcantgo_sumstressav:rural.ses.med4 3142 2748
## depeffort_sumstresschg 3564 2755
## depeffort_sumstressav 1925 2028
## depeffort_rural.ses.med2 3611 2657
## depeffort_rural.ses.med3 3775 2905
## depeffort_rural.ses.med4 3585 2941
## depeffort_sumstresschg:rural.ses.med2 5705 2837
## depeffort_sumstresschg:rural.ses.med3 4136 3008
## depeffort_sumstresschg:rural.ses.med4 5588 2668
## depeffort_sumstressav:rural.ses.med2 2800 2965
## depeffort_sumstressav:rural.ses.med3 2169 2684
## depeffort_sumstressav:rural.ses.med4 2082 2420
## deplonely_sumstresschg 3337 3018
## deplonely_sumstressav 2288 2742
## deplonely_rural.ses.med2 3851 3588
## deplonely_rural.ses.med3 3918 2887
## deplonely_rural.ses.med4 4023 3299
## deplonely_sumstresschg:rural.ses.med2 5073 3045
## deplonely_sumstresschg:rural.ses.med3 4146 3237
## deplonely_sumstresschg:rural.ses.med4 4518 3233
## deplonely_sumstressav:rural.ses.med2 3221 3416
## deplonely_sumstressav:rural.ses.med3 2754 2921
## deplonely_sumstressav:rural.ses.med4 2798 2969
## depblues_sumstresschg 3593 2882
## depblues_sumstressav 2505 2886
## depblues_rural.ses.med2 3302 2820
## depblues_rural.ses.med3 3225 3348
## depblues_rural.ses.med4 3196 3051
## depblues_sumstresschg:rural.ses.med2 6058 3143
## depblues_sumstresschg:rural.ses.med3 4774 2986
## depblues_sumstresschg:rural.ses.med4 4978 3068
## depblues_sumstressav:rural.ses.med2 2944 3154
## depblues_sumstressav:rural.ses.med3 2964 2927
## depblues_sumstressav:rural.ses.med4 2964 3206
## depunfair_sumstresschg 3698 3133
## depunfair_sumstressav 2197 2770
## depunfair_rural.ses.med2 4083 3156
## depunfair_rural.ses.med3 3803 3241
## depunfair_rural.ses.med4 3541 3268
## depunfair_sumstresschg:rural.ses.med2 5329 3347
## depunfair_sumstresschg:rural.ses.med3 3549 3157
## depunfair_sumstresschg:rural.ses.med4 5318 3403
## depunfair_sumstressav:rural.ses.med2 2948 3016
## depunfair_sumstressav:rural.ses.med3 2609 3024
## depunfair_sumstressav:rural.ses.med4 2580 2811
## depmistrt_sumstresschg 3414 2816
## depmistrt_sumstressav 3004 3112
## depmistrt_rural.ses.med2 3375 2785
## depmistrt_rural.ses.med3 3091 2739
## depmistrt_rural.ses.med4 3599 2684
## depmistrt_sumstresschg:rural.ses.med2 4784 3045
## depmistrt_sumstresschg:rural.ses.med3 4214 3062
## depmistrt_sumstresschg:rural.ses.med4 5671 3286
## depmistrt_sumstressav:rural.ses.med2 3300 3184
## depmistrt_sumstressav:rural.ses.med3 3480 3580
## depmistrt_sumstressav:rural.ses.med4 3517 3191
## depbetray_sumstresschg 3746 2974
## depbetray_sumstressav 2212 2542
## depbetray_rural.ses.med2 4004 3251
## depbetray_rural.ses.med3 3911 3015
## depbetray_rural.ses.med4 4060 3272
## depbetray_sumstresschg:rural.ses.med2 5151 3107
## depbetray_sumstresschg:rural.ses.med3 4685 3132
## depbetray_sumstresschg:rural.ses.med4 6072 3353
## depbetray_sumstressav:rural.ses.med2 2537 3147
## depbetray_sumstressav:rural.ses.med3 2749 2709
## depbetray_sumstressav:rural.ses.med4 2716 3140
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).10.1.9.5 Prior summary
out.chg.alldepress.stirt.comm.fit[[2]]## prior class coef group resp
## (flat) b
## normal(0, 1) b depbetray
## normal(0, 1) b irtstressav depbetray
## normal(0, 1) b irtstressav:rural.ses.med2 depbetray
## normal(0, 1) b irtstressav:rural.ses.med3 depbetray
## normal(0, 1) b irtstressav:rural.ses.med4 depbetray
## normal(0, 1) b irtstresschg depbetray
## normal(0, 1) b irtstresschg:rural.ses.med2 depbetray
## normal(0, 1) b irtstresschg:rural.ses.med3 depbetray
## normal(0, 1) b irtstresschg:rural.ses.med4 depbetray
## normal(0, 1) b rural.ses.med2 depbetray
## normal(0, 1) b rural.ses.med3 depbetray
## normal(0, 1) b rural.ses.med4 depbetray
## normal(0, 1) b depblues
## normal(0, 1) b irtstressav depblues
## normal(0, 1) b irtstressav:rural.ses.med2 depblues
## normal(0, 1) b irtstressav:rural.ses.med3 depblues
## normal(0, 1) b irtstressav:rural.ses.med4 depblues
## normal(0, 1) b irtstresschg depblues
## normal(0, 1) b irtstresschg:rural.ses.med2 depblues
## normal(0, 1) b irtstresschg:rural.ses.med3 depblues
## normal(0, 1) b irtstresschg:rural.ses.med4 depblues
## normal(0, 1) b rural.ses.med2 depblues
## normal(0, 1) b rural.ses.med3 depblues
## normal(0, 1) b rural.ses.med4 depblues
## normal(0, 1) b depcantgo
## normal(0, 1) b irtstressav depcantgo
## normal(0, 1) b irtstressav:rural.ses.med2 depcantgo
## normal(0, 1) b irtstressav:rural.ses.med3 depcantgo
## normal(0, 1) b irtstressav:rural.ses.med4 depcantgo
## normal(0, 1) b irtstresschg depcantgo
## normal(0, 1) b irtstresschg:rural.ses.med2 depcantgo
## normal(0, 1) b irtstresschg:rural.ses.med3 depcantgo
## normal(0, 1) b irtstresschg:rural.ses.med4 depcantgo
## normal(0, 1) b rural.ses.med2 depcantgo
## normal(0, 1) b rural.ses.med3 depcantgo
## normal(0, 1) b rural.ses.med4 depcantgo
## normal(0, 1) b depeffort
## normal(0, 1) b irtstressav depeffort
## normal(0, 1) b irtstressav:rural.ses.med2 depeffort
## normal(0, 1) b irtstressav:rural.ses.med3 depeffort
## normal(0, 1) b irtstressav:rural.ses.med4 depeffort
## normal(0, 1) b irtstresschg depeffort
## normal(0, 1) b irtstresschg:rural.ses.med2 depeffort
## normal(0, 1) b irtstresschg:rural.ses.med3 depeffort
## normal(0, 1) b irtstresschg:rural.ses.med4 depeffort
## normal(0, 1) b rural.ses.med2 depeffort
## normal(0, 1) b rural.ses.med3 depeffort
## normal(0, 1) b rural.ses.med4 depeffort
## normal(0, 1) b deplonely
## normal(0, 1) b irtstressav deplonely
## normal(0, 1) b irtstressav:rural.ses.med2 deplonely
## normal(0, 1) b irtstressav:rural.ses.med3 deplonely
## normal(0, 1) b irtstressav:rural.ses.med4 deplonely
## normal(0, 1) b irtstresschg deplonely
## normal(0, 1) b irtstresschg:rural.ses.med2 deplonely
## normal(0, 1) b irtstresschg:rural.ses.med3 deplonely
## normal(0, 1) b irtstresschg:rural.ses.med4 deplonely
## normal(0, 1) b rural.ses.med2 deplonely
## normal(0, 1) b rural.ses.med3 deplonely
## normal(0, 1) b rural.ses.med4 deplonely
## normal(0, 1) b depmistrt
## normal(0, 1) b irtstressav depmistrt
## normal(0, 1) b irtstressav:rural.ses.med2 depmistrt
## normal(0, 1) b irtstressav:rural.ses.med3 depmistrt
## normal(0, 1) b irtstressav:rural.ses.med4 depmistrt
## normal(0, 1) b irtstresschg depmistrt
## normal(0, 1) b irtstresschg:rural.ses.med2 depmistrt
## normal(0, 1) b irtstresschg:rural.ses.med3 depmistrt
## normal(0, 1) b irtstresschg:rural.ses.med4 depmistrt
## normal(0, 1) b rural.ses.med2 depmistrt
## normal(0, 1) b rural.ses.med3 depmistrt
## normal(0, 1) b rural.ses.med4 depmistrt
## normal(0, 1) b depunfair
## normal(0, 1) b irtstressav depunfair
## normal(0, 1) b irtstressav:rural.ses.med2 depunfair
## normal(0, 1) b irtstressav:rural.ses.med3 depunfair
## normal(0, 1) b irtstressav:rural.ses.med4 depunfair
## normal(0, 1) b irtstresschg depunfair
## normal(0, 1) b irtstresschg:rural.ses.med2 depunfair
## normal(0, 1) b irtstresschg:rural.ses.med3 depunfair
## normal(0, 1) b irtstresschg:rural.ses.med4 depunfair
## normal(0, 1) b rural.ses.med2 depunfair
## normal(0, 1) b rural.ses.med3 depunfair
## normal(0, 1) b rural.ses.med4 depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dpar nlpar lb ub source
## default
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)10.1.10 Predictive Marginal Contrasts x Comm (IRT & Sum Stress / Crim Intent)
# generate new data grid for specific change contrasts (average over irtstressav)
# keep only 1-unit IRT scale increases from year 1 to year 2 (contrast .5_t2 - -.5_t1)
newdata <- stress.long2 %>%
data_grid(irtstresschg = c(-.5, .5),
irtstressav,
rural.ses.med,
year) %>%
filter(irtstresschg == -0.5 & year == "1" |
irtstresschg == 0.5 & year == "2")
# function to generate epred draws
gen_predmarg_data_comm <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
group_by(.category, rural.ses.med, {{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`))
}
#use function to generate epred draws
predmarg_stirt_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stirt.comm.fit, irtstresschg)
predmarg_stirt_negemots_chg_comm = gen_predmarg_data_comm(chg.alldepress.stirt.comm.fit, irtstresschg)
#generate epred draws for "any crime" outcome & merge w/prjcrm epred draws
predmarg_stirt_anyprjcrim_chg_comm =
epred_draws(chg.prjcrime.stirt.comm.fit,
newdata = newdata,
re_formula = NA) %>%
group_by(rural.ses.med, irtstresschg, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
predmarg_stirt_prjcrim_chg_comm <- bind_rows(predmarg_stirt_prjcrim_chg_comm,
predmarg_stirt_anyprjcrim_chg_comm)
rm(predmarg_stirt_anyprjcrim_chg_comm) #clean environment
#repeat with sum stress scale
newdata <- stress.long2 %>%
data_grid(sumstresschg = c(-.5, .5),
sumstressav,
rural.ses.med,
year) %>%
filter(sumstresschg == -0.5 & year == "1" |
sumstresschg == 0.5 & year == "2")
predmarg_stsum_prjcrim_chg_comm = gen_predmarg_data_comm(chg.prjcrime.stsum.comm.fit, sumstresschg)
predmarg_stsum_negemots_chg_comm = gen_predmarg_data_comm(chg.alldepress.stsum.comm.fit, sumstresschg)
#generate epred draws for "any crime" outcome & merge w/prjcrm epred draws
predmarg_stsum_anyprjcrim_chg_comm =
epred_draws(chg.anyprjcrime.stsum.comm.fit,
newdata = newdata,
re_formula = NA) %>%
group_by(rural.ses.med, sumstresschg, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
predmarg_stsum_prjcrim_chg_comm <- bind_rows(predmarg_stsum_prjcrim_chg_comm,
predmarg_stsum_anyprjcrim_chg_comm)
rm(predmarg_stsum_anyprjcrim_chg_comm) #clean environment
# function to calculate marginal contrasts
calc_ME_chg_comm <- function(predmarg_data, xdev) {
predmarg_data %>%
compare_levels(`E[y|xdev]`, by = xdev) %>% # pairwise diffs in `E[y|x]`, by levels of x
group_by(rural.ses.med) %>% # generate community-specific marginal contrasts
rename(`PLME` = `E[y|xdev]`) # easy colname reflecting ME 2-unit chg contrast
}
#outputs community-specific predicted differences in E[y]
# associated with 1-IRT AVE scale increase in stress (T2-T1, -.5 to .5)
#marg effect contrasts are marginalized over all person-level avg values of stress (AME)
#generate ME contrasts & add alpha indicator
PLME2_stirt_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stirt_prjcrim_chg_comm, "irtstresschg") %>%
mutate(stress_var = "IRT Stress:\nWithin Person\nChange",
dif_label = "diff in E[y|stress diff or increase]",
method="irtchg") %>%
rename(contrast = irtstresschg)}, file="cache_9_1")
PLME2_stirt_prjcrim_chg_comm <- PLME2_stirt_prjcrim_chg_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1"))) %>%
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
PLME2_stirt_negemots_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stirt_negemots_chg_comm, "irtstresschg") %>%
mutate(stress_var = "IRT Stress:\nWithin Person\nChange",
dif_label = "diff in E[y|stress diff or increase]",
method="irtchg") %>%
rename(contrast = irtstresschg)}, file="cache_9_2")
PLME2_stirt_negemots_chg_comm <- PLME2_stirt_negemots_chg_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1")))
#repeat for sum stress scale
PLME2_stsum_prjcrim_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stsum_prjcrim_chg_comm, "sumstresschg") %>%
mutate(stress_var = "Sum Stress:\nWithin Person\nChange",
dif_label = "diff in E[y|stress diff or increase]",
method="sumchg") %>%
rename(contrast = sumstresschg)}, file="cache_9_3")
PLME2_stsum_prjcrim_chg_comm <- PLME2_stsum_prjcrim_chg_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1"))) %>%
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
PLME2_stsum_negemots_chg_comm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stsum_negemots_chg_comm, "sumstresschg") %>%
mutate(stress_var = "Sum Stress:\nWithin Person\nChange",
dif_label = "diff in E[y|stress diff or increase]",
method="sumchg") %>%
rename(contrast = sumstresschg)}, file="cache_9_4")
PLME2_stsum_negemots_chg_comm <- PLME2_stsum_negemots_chg_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1")))
# levels(PLME2_stsum_negemots_chg_comm$p80_gt0)
# NOTE: specify factor level for p80_gt0 indicator
# using as.factor() w/out specifying results in reversed levels in this object
# REPEAT FOR BETWEEN-PERSON DIFF ESTIMATES
# marginalize btw-per 1-0 diff contrast estimates across values of w/in person change & year
# NOTE: base rate probs differ across irtstressav scale, so could collapse representative
# contrasts across scale (-1 - -2, -.5 - -1.5, 0 - -1, .5 - -.5, 1 - 0, 1.5 - .5)
# attempted but too computationally intensive
newdata <- stress.long2 %>%
data_grid(irtstressav = c(0,1),
irtstresschg,
rural.ses.med,
year)
# function to generate epred draws
gen_predmarg_data_comm <- function(mymodelfit, xdif){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
group_by(.category, rural.ses.med, {{xdif}}, .draw) %>%
summarise(`E[y|xdif]` = mean(`.epred`))
}
#use function to generate epred draws
predmarg_stirt_prjcrim_av_comm = gen_predmarg_data_comm(chg.prjcrime.stirt.comm.fit, irtstressav)
predmarg_stirt_negemots_av_comm = gen_predmarg_data_comm(chg.alldepress.stirt.comm.fit, irtstressav)
#generate epred draws for "any crime" outcome & merge w/prjcrm epred draws
predmarg_stirt_anyprjcrim_av_comm =
epred_draws(chg.anyprjcrime.stirt.comm.fit,
newdata = newdata,
re_formula = NA) %>%
group_by(rural.ses.med, irtstressav, .draw) %>%
summarise(`E[y|xdif]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
predmarg_stirt_prjcrim_av_comm <- bind_rows(predmarg_stirt_prjcrim_av_comm,
predmarg_stirt_anyprjcrim_av_comm)
rm(predmarg_stirt_anyprjcrim_av_comm) #clean environment
#repeat with sum scale
newdata <- stress.long2 %>%
data_grid(sumstressav = c(0,1),
sumstresschg,
rural.ses.med,
year)
predmarg_stsum_prjcrim_av_comm = gen_predmarg_data_comm(chg.prjcrime.stsum.comm.fit, sumstressav)
predmarg_stsum_negemots_av_comm = gen_predmarg_data_comm(chg.alldepress.stsum.comm.fit, sumstressav)
#generate epred draws for "any crime" outcome & merge w/prjcrm epred draws
predmarg_stsum_anyprjcrim_av_comm =
epred_draws(chg.anyprjcrime.stsum.comm.fit,
newdata = newdata,
re_formula = NA) %>%
group_by(rural.ses.med, sumstressav, .draw) %>%
summarise(`E[y|xdif]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
predmarg_stsum_prjcrim_av_comm <- bind_rows(predmarg_stsum_prjcrim_av_comm,
predmarg_stsum_anyprjcrim_av_comm)
rm(predmarg_stsum_anyprjcrim_av_comm) #clean environment
calc_ME_diff_comm <- function(predmarg_data, xdif) {
predmarg_data %>%
compare_levels(`E[y|xdif]`, by = xdif) %>% # pairwise diffs in `E[y|x]`, by levels of x
group_by(rural.ses.med) %>% # generate community-specific marginal contrasts
rename(`PLME` = `E[y|xdif]`) # easy colname reflecting ME 2-unit chg contrast
}
#generate ME contrasts & add alpha indicator
PLME2_stirt_prjcrim_av_comm = xfun::cache_rds({calc_ME_diff_comm(predmarg_stirt_prjcrim_av_comm, "irtstressav") %>%
mutate(stress_var = "IRT Stress:\nBetween Person\nDifference",
dif_label = "diff in E[y|stress diff or increase]",
method="irtav") %>%
rename(contrast = irtstressav)}, file="cache_9_5")
PLME2_stirt_prjcrim_av_comm <- PLME2_stirt_prjcrim_av_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1"))) %>%
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
PLME2_stirt_negemots_av_comm = xfun::cache_rds({calc_ME_diff_comm(predmarg_stirt_negemots_av_comm, "irtstressav") %>%
mutate(stress_var = "IRT Stress:\nBetween Person\nDifference",
dif_label = "diff in E[y|stress diff or increase]",
method="irtav") %>%
rename(contrast = irtstressav)}, file="cache_9_6")
PLME2_stirt_negemots_av_comm <- PLME2_stirt_negemots_av_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1")))
#repeat for sum scale
PLME2_stsum_prjcrim_av_comm = xfun::cache_rds({calc_ME_diff_comm(predmarg_stsum_prjcrim_av_comm, "sumstressav") %>%
mutate(stress_var = "Sum Stress:\nBetween Person\nDifference",
dif_label = "diff in E[y|stress diff or increase]",
method="sumav") %>%
rename(contrast = sumstressav)}, file="cache_9_7")
PLME2_stsum_prjcrim_av_comm <- PLME2_stsum_prjcrim_av_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1"))) %>%
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
PLME2_stsum_negemots_av_comm = xfun::cache_rds({calc_ME_diff_comm(predmarg_stsum_negemots_av_comm, "sumstressav") %>%
mutate(stress_var = "Sum Stress:\nBetween Person\nDifference",
dif_label = "diff in E[y|stress diff or increase]",
method="sumav") %>%
rename(contrast = sumstressav)}, file="cache_9_8")
PLME2_stsum_negemots_av_comm <- PLME2_stsum_negemots_av_comm %>%
group_by(.category, rural.ses.med) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1"))) 10.2 APPENDIX 1: PLME Composite Stress Scales/Outcome Change X Community
#function to find & drop leading zeroes (used for x-axis label)
dropLeadingZero <- function(l){
str_replace(l, '0(?=.)', '')
}
numcommlabs2 <- c("1"="Rural\nLowSES",
"2"="Rural\nHighSES",
"3"="Urban\nLowSES",
"4"="Urban\nHighSES")
prjlabs2 <- c(
"prjthflt5"="Theft\n<5BAM",
"prjthfgt5"="Theft\n>5BAM",
"prjthreat"="Threat",
"prjharm"="Phys.\nharm",
"prjusedrg"="Use\ndrugs",
"prjhack"="Hack\ninfo",
"prjany"="Any crime")
deplabs2 <- c(
"depcantgo"="Can't\ngo",
"depeffort"="Effort",
"deplonely"="Lonely",
"depblues"="Blues",
"depunfair"="Unfair",
"depmistrt"="Mistreated",
"depbetray"="Betrayed")
methodlabs <- c(
"irtav"="Between-Person Difference\nEstimator (IRT Scale)",
"irtchg"="Within-Person Change\nEstimator (IRT Scale)",
"sumav"="Between-Person Difference\nEstimator (Sum Scale)",
"sumchg"="Within-Person Change\nEstimator (Sum Scale)")
irtprjcommplot <- ggplot(data = PLME2_stirt_prjcrim_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stirt_prjcrim_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stirt_prjcrim_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#E99D53","#883E3A"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
# coord_cartesian(xlim=c(-.06,.11)) +
# scale_x_continuous(breaks=c(0,.1), labels = dropLeadingZero) +
coord_cartesian(xlim=c(-.055,.355)) + #plot on half scale as neg emotions
scale_x_continuous(breaks=c(0,.25), labels = dropLeadingZero) +
xlab(element_blank()) +
labs(subtitle='Criminal Intent\n___________________________________________________________') +
scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size = 8, angle=65),
axis.line.x = element_blank(), #remove x-axis for top plots
axis.ticks.x = element_blank(), #remove x-axis for top plots
axis.text.x = element_blank(), #remove x-axis for top plots
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# irtprjcommplot
irtdepcommplot <- ggplot(data = PLME2_stirt_negemots_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stirt_negemots_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stirt_negemots_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#E99D53","#883E3A"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.71)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
labs(subtitle='Negative Emotions\n____________________________________________________________') +
# scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size = 8, angle=65),
axis.line.x = element_blank(), #remove x-axis for top plots
axis.ticks.x = element_blank(), #remove x-axis for top plots
axis.text.x = element_blank(), #remove x-axis for top plots
axis.text.y=element_blank(), #remove y-axis labels
axis.ticks.y=element_blank(), #remove y-axis ticks
axis.line.y=element_blank(), #remove y-axis line
plot.title = element_text(size=10, face="bold"),
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# irtdepcommplot
sumprjcommplot <- ggplot(data = PLME2_stsum_prjcrim_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_prjcrim_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stsum_prjcrim_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#7ad151", "#440154"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
# coord_cartesian(xlim=c(-.06,.11)) +
# scale_x_continuous(breaks=c(0,.1), labels = dropLeadingZero) +
coord_cartesian(xlim=c(-.055,.355)) + #plot on half scale as neg emotions
scale_x_continuous(breaks=c(0,.25), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_blank(), #remove facet (outcome) labels
# strip.text.x = element_text(size = 8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumprjcommplot
sumdepcommplot <- ggplot(data = PLME2_stsum_negemots_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_negemots_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stsum_negemots_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#7ad151", "#440154"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.71)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_blank(),
# strip.text.x = element_text(size = 8),
axis.text.y=element_blank(), #remove y-axis labels
axis.ticks.y=element_blank(), #remove y-axis ticks
axis.line.y=element_blank(), #remove y-axis line
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumdepcommplot
design <- "
12
34
55
"
# library(patchwork)
AppendixFig1 <- irtprjcommplot + irtdepcommplot +
sumprjcommplot + sumdepcommplot +
guide_area() +
plot_layout(design=design, guides = 'collect', heights = c(2,2,.1)) +
plot_annotation(
title = 'Appendix 1: Marginal Effects of Stress Scale Increase on Outcome Probabilities, by Estimator, Scaling Method, & Community',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Estimates derived from multivariate (using `brms::mvbind()`) and multilevel between/within Bayesian logistic regression models simultaneously regressing all criminal intent outcomes (6*2=12 models) and all negative emotion outcomes (7*2=14 models) separately on a latent IRT and a standardized sum stress scale, and two separate models regressing "any criminal intent" on each stress scale. Both stress scales were separated into L2 cross-time average (Xbar_i) between-person and L1 within-person change (X_it - Xbar_i) "fixed effects" estimators. Models also included a factor variable for community and multiplicative interactions between community and both L1/L2 stress estimators. Models were estimated in brms with 4 chains and 4000 total post-warmup posterior draws per outcome and per community group. Marginal effect contrast distributions were estimated from the expectation of the posterior predictive distribution for each model as community-specific predicted probability difference distributions averaged over all 1-unit increases on the stress scale (within) or for a 1SD increase from mean (between; "0" vs "1") on initial IRT or standardized latent scale, averaged over the alternative (between or within) stress estimator levels. Median posterior density estimates with 95% intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates for the average marginal effect contrast are greater than zero.', width=195)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'bottom',
legend.key.height = unit(.5, 'cm'),
legend.margin = margin(0,0,0,0),
legend.spacing.y = unit(0, "mm"))
AppendixFig1
ggsave("Appendix1.jpeg", width=9, height=6.5, path=here("Output"))save(stress.long2, file = here("1_Data_Files/Datasets/stress_long2.Rdata"))
saveRDS(PLME2_stsum_prjcrim_chg_comm, file = here("Models/PLME2_stsum_prjcrim_chg_comm.rds"))
saveRDS(PLME2_stsum_prjcrim_av_comm, file = here("Models/PLME2_stsum_prjcrim_av_comm.rds"))
saveRDS(PLME2_stsum_negemots_chg_comm, file = here("Models/PLME2_stsum_negemots_chg_comm.rds"))
saveRDS(PLME2_stsum_negemots_av_comm, file = here("Models/PLME2_stsum_negemots_av_comm.rds"))11 Appendix 2: Robustness of SES Median Split
(RMD FILE: BDK_2023_Stress_10_Append2_median_split)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE11.1 ROBUSTNESS CHECK ON COMMUNITY SES MEDIAN SPLIT
In this section, we re-estimate and plot composite stress scale models to assess robustness of findings to decision to use community-level (L2) SES median split for community identification. Models will include three-way interactions between rural residence, standardized L2 SES, and each stress estimator (between & within). The standardized sum stress scale is used rather than IRT scale because sum scale results more closely summarize those observed in item-specific analysis; as expected, the IRT scale heavily weights one dimension - interpersonal stress - while down-weighting others; thus, results using IRT scale primarily align only with previous findings from item-specific interpersonal stress models (see Appendix 1). Here, the goal is to use a composite measure that adequately captures “overall stress” differences or increases, which the unweighted standardized sum scale appears to do better than the IRT scale.
load(here("1_Data_Files/Datasets/stress_long2.Rdata"))
PLME2_stsum_prjcrim_chg_comm <- readRDS(
file = here("Models/PLME2_stsum_prjcrim_chg_comm.rds"))
PLME2_stsum_prjcrim_av_comm <- readRDS(
file = here("Models/PLME2_stsum_prjcrim_av_comm.rds"))
PLME2_stsum_negemots_chg_comm <- readRDS(
file = here("Models/PLME2_stsum_negemots_chg_comm.rds"))
PLME2_stsum_negemots_av_comm <- readRDS(
file = here("Models/PLME2_stsum_negemots_av_comm.rds"))11.1.1 Standardize L2 SES (T1) variable
stress.long3 <- stress.long2 %>%
group_by(year) %>%
mutate(
L2sesw1z = (L2sesw1 - mean(L2sesw1))/sd(L2sesw1)
) %>%
ungroup()11.1.1.1 L2sesw1 distribution
ggplot(stress.long2, aes(L2sesw1)) + geom_histogram(fill="#E99D53")
11.1.1.2 L2sesw1z distribution
ggplot(stress.long3, aes(L2sesw1z)) + geom_histogram(fill="#E99D53")
11.1.2 Corr X Robust Community: stdz sum stress & prj crime
#list of colnames for projected crime DVs
prjdv_names <- noquote(c("prjthflt5", "prjthfgt5", "prjthreat", "prjharm",
"prjusedrg", "prjhack"))
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = prjdv_names),
set_prior('normal(0, 1)', class = 'b', resp = prjdv_names)
)
chg.prjcrime.stsum.rbstcomm.fit <- brm(
mvbind(prjthflt5, prjthfgt5, prjthreat, prjharm, prjusedrg, prjhack) ~ 1 +
sumstresschg + sumstressav + rural + L2sesw1z +
sumstresschg:rural + sumstressav:rural +
sumstresschg:L2sesw1z + sumstressav:L2sesw1z +
sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z +
(1 | id),
data = stress.long3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
adapt_delta = 0.85,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_prjcrime_stsum_rbstcomm_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
varlist <- c("^b_prjthflt5_sum","^b_prjthfgt5_sum", "^b_prjthreat_sum",
"^b_prjharm_sum", "^b_usedrg_sum", "^b_prjhack_sum")
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="prjthflt5")
ppcheckdv2 <-pp_check(modelfit, resp="prjthfgt5")
ppcheckdv3 <-pp_check(modelfit, resp="prjthreat")
ppcheckdv4 <-pp_check(modelfit, resp="prjharm")
ppcheckdv5 <-pp_check(modelfit, resp="prjusedrg")
ppcheckdv6 <-pp_check(modelfit, resp="prjhack")
plotcoefs2 <- mcmc_plot(modelfit, variable = varlist, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2,
ppcheckdv3, ppcheckdv4, ppcheckdv5, ppcheckdv6, plotcoefs2)
return(allchecks)
}
out.chg.prjcrime.stsum.rbstcomm.fit <- ppchecks(chg.prjcrime.stsum.rbstcomm.fit)11.1.2.1 Coefficient plot (intervals)
out.chg.prjcrime.stsum.rbstcomm.fit[[9]] 
11.1.2.2 PPcheck (density)
p1 <- out.chg.prjcrime.stsum.rbstcomm.fit[[3]] + labs(title = "Theft <5BAM Intent (chg)")
p2 <- out.chg.prjcrime.stsum.rbstcomm.fit[[4]] + labs(title = "Theft >5BAM Intent (chg)")
p3 <- out.chg.prjcrime.stsum.rbstcomm.fit[[5]] + labs(title = "Threat Intent (chg)")
p4 <- out.chg.prjcrime.stsum.rbstcomm.fit[[6]] + labs(title = "Harm Intent (chg)")
p5 <- out.chg.prjcrime.stsum.rbstcomm.fit[[7]] + labs(title = "Use Drugs Intent (chg)")
p6 <- out.chg.prjcrime.stsum.rbstcomm.fit[[8]] + labs(title = "Hack Intent (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6)
11.1.2.3 Fit summary
out.chg.prjcrime.stsum.rbstcomm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: prjthflt5 ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## prjthfgt5 ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## prjthreat ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## prjharm ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## prjusedrg ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## prjhack ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## Data: stress.long3 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(prjthflt5_Intercept) 3.94 0.55 2.95 5.13 1.00 1263
## sd(prjthfgt5_Intercept) 3.36 0.49 2.51 4.41 1.00 1030
## sd(prjthreat_Intercept) 3.25 0.54 2.27 4.40 1.00 939
## sd(prjharm_Intercept) 3.02 0.54 2.05 4.15 1.00 1125
## sd(prjusedrg_Intercept) 2.92 0.52 2.00 3.99 1.00 1242
## sd(prjhack_Intercept) 0.95 0.56 0.05 2.10 1.00 508
## Tail_ESS
## sd(prjthflt5_Intercept) 1909
## sd(prjthfgt5_Intercept) 1220
## sd(prjthreat_Intercept) 2101
## sd(prjharm_Intercept) 1916
## sd(prjusedrg_Intercept) 1915
## sd(prjhack_Intercept) 896
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## prjthflt5_Intercept -4.23 0.62 -5.56 -3.17 1.00
## prjthfgt5_Intercept -3.98 0.55 -5.13 -3.05 1.00
## prjthreat_Intercept -5.78 0.79 -7.43 -4.42 1.00
## prjharm_Intercept -5.89 0.81 -7.60 -4.47 1.00
## prjusedrg_Intercept -6.05 0.83 -7.81 -4.59 1.00
## prjhack_Intercept -4.33 0.59 -5.71 -3.40 1.00
## prjthflt5_sumstresschg 0.41 0.57 -0.71 1.53 1.00
## prjthflt5_sumstressav 0.64 0.36 -0.03 1.35 1.00
## prjthflt5_rural -2.55 0.54 -3.64 -1.49 1.00
## prjthflt5_L2sesw1z 0.27 0.27 -0.26 0.82 1.00
## prjthflt5_sumstresschg:rural 0.37 0.82 -1.26 1.93 1.00
## prjthflt5_sumstressav:rural -0.03 0.55 -1.12 1.07 1.00
## prjthflt5_sumstresschg:L2sesw1z -0.40 0.51 -1.43 0.57 1.00
## prjthflt5_sumstressav:L2sesw1z -0.27 0.29 -0.85 0.31 1.00
## prjthflt5_sumstresschg:rural:L2sesw1z -0.95 0.90 -2.65 0.80 1.00
## prjthflt5_sumstressav:rural:L2sesw1z 0.98 0.66 -0.34 2.26 1.00
## prjthfgt5_sumstresschg 0.77 0.57 -0.35 1.85 1.00
## prjthfgt5_sumstressav 0.64 0.32 0.05 1.27 1.01
## prjthfgt5_rural -2.42 0.49 -3.39 -1.46 1.00
## prjthfgt5_L2sesw1z 0.16 0.24 -0.33 0.64 1.00
## prjthfgt5_sumstresschg:rural 0.48 0.83 -1.14 2.12 1.00
## prjthfgt5_sumstressav:rural -0.34 0.53 -1.38 0.70 1.00
## prjthfgt5_sumstresschg:L2sesw1z -0.03 0.53 -1.07 1.03 1.00
## prjthfgt5_sumstressav:L2sesw1z -0.14 0.27 -0.70 0.40 1.00
## prjthfgt5_sumstresschg:rural:L2sesw1z -1.01 0.88 -2.78 0.74 1.00
## prjthfgt5_sumstressav:rural:L2sesw1z 0.91 0.66 -0.39 2.21 1.00
## prjthreat_sumstresschg -0.68 0.66 -2.00 0.60 1.00
## prjthreat_sumstressav 1.03 0.39 0.30 1.84 1.00
## prjthreat_rural -1.57 0.56 -2.67 -0.49 1.00
## prjthreat_L2sesw1z 0.23 0.31 -0.37 0.84 1.00
## prjthreat_sumstresschg:rural 0.70 0.90 -1.06 2.47 1.00
## prjthreat_sumstressav:rural -0.26 0.62 -1.51 0.95 1.00
## prjthreat_sumstresschg:L2sesw1z -0.76 0.62 -1.99 0.42 1.00
## prjthreat_sumstressav:L2sesw1z 0.17 0.33 -0.50 0.83 1.00
## prjthreat_sumstresschg:rural:L2sesw1z 0.05 0.98 -1.83 1.98 1.00
## prjthreat_sumstressav:rural:L2sesw1z -0.54 0.72 -1.96 0.89 1.00
## prjharm_sumstresschg -0.84 0.69 -2.16 0.52 1.00
## prjharm_sumstressav 0.55 0.37 -0.15 1.29 1.00
## prjharm_rural -0.96 0.57 -2.10 0.13 1.00
## prjharm_L2sesw1z 0.21 0.29 -0.37 0.77 1.00
## prjharm_sumstresschg:rural 0.35 0.86 -1.29 2.09 1.00
## prjharm_sumstressav:rural -0.66 0.57 -1.76 0.45 1.00
## prjharm_sumstresschg:L2sesw1z -0.54 0.65 -1.82 0.72 1.00
## prjharm_sumstressav:L2sesw1z 0.01 0.34 -0.63 0.70 1.00
## prjharm_sumstresschg:rural:L2sesw1z 0.14 0.93 -1.63 1.96 1.00
## prjharm_sumstressav:rural:L2sesw1z -0.27 0.67 -1.58 1.03 1.00
## prjusedrg_sumstresschg -0.42 0.72 -1.78 1.02 1.00
## prjusedrg_sumstressav 0.82 0.40 0.05 1.64 1.00
## prjusedrg_rural -0.84 0.56 -1.99 0.28 1.00
## prjusedrg_L2sesw1z 0.71 0.30 0.14 1.31 1.00
## prjusedrg_sumstresschg:rural 0.58 0.87 -1.13 2.21 1.00
## prjusedrg_sumstressav:rural -0.51 0.59 -1.69 0.64 1.00
## prjusedrg_sumstresschg:L2sesw1z -0.52 0.68 -1.83 0.82 1.00
## prjusedrg_sumstressav:L2sesw1z 0.32 0.34 -0.36 0.99 1.00
## prjusedrg_sumstresschg:rural:L2sesw1z 0.10 0.95 -1.76 1.97 1.00
## prjusedrg_sumstressav:rural:L2sesw1z -0.75 0.65 -2.04 0.52 1.00
## prjhack_sumstresschg -0.56 0.69 -1.89 0.86 1.00
## prjhack_sumstressav 0.72 0.30 0.15 1.33 1.00
## prjhack_rural -0.69 0.50 -1.67 0.26 1.00
## prjhack_L2sesw1z -0.20 0.25 -0.73 0.26 1.00
## prjhack_sumstresschg:rural 0.88 0.90 -0.90 2.63 1.00
## prjhack_sumstressav:rural -0.20 0.52 -1.19 0.82 1.00
## prjhack_sumstresschg:L2sesw1z -0.52 0.61 -1.72 0.67 1.00
## prjhack_sumstressav:L2sesw1z 0.30 0.26 -0.19 0.84 1.00
## prjhack_sumstresschg:rural:L2sesw1z 0.05 0.98 -1.85 1.96 1.00
## prjhack_sumstressav:rural:L2sesw1z -0.49 0.63 -1.74 0.74 1.00
## Bulk_ESS Tail_ESS
## prjthflt5_Intercept 1353 2015
## prjthfgt5_Intercept 1128 1688
## prjthreat_Intercept 1063 2266
## prjharm_Intercept 1340 2294
## prjusedrg_Intercept 1500 2350
## prjhack_Intercept 796 1269
## prjthflt5_sumstresschg 4823 3305
## prjthflt5_sumstressav 1676 1914
## prjthflt5_rural 2355 2766
## prjthflt5_L2sesw1z 1763 2437
## prjthflt5_sumstresschg:rural 5026 2956
## prjthflt5_sumstressav:rural 2133 2839
## prjthflt5_sumstresschg:L2sesw1z 5236 2809
## prjthflt5_sumstressav:L2sesw1z 1823 2330
## prjthflt5_sumstresschg:rural:L2sesw1z 5177 2544
## prjthflt5_sumstressav:rural:L2sesw1z 2522 2141
## prjthfgt5_sumstresschg 5266 2964
## prjthfgt5_sumstressav 1626 2129
## prjthfgt5_rural 3001 2948
## prjthfgt5_L2sesw1z 2273 2890
## prjthfgt5_sumstresschg:rural 5494 2933
## prjthfgt5_sumstressav:rural 2520 2423
## prjthfgt5_sumstresschg:L2sesw1z 4993 2890
## prjthfgt5_sumstressav:L2sesw1z 2109 2227
## prjthfgt5_sumstresschg:rural:L2sesw1z 5767 2748
## prjthfgt5_sumstressav:rural:L2sesw1z 2674 2728
## prjthreat_sumstresschg 5332 3089
## prjthreat_sumstressav 2067 2403
## prjthreat_rural 2923 2599
## prjthreat_L2sesw1z 2506 2556
## prjthreat_sumstresschg:rural 4620 2831
## prjthreat_sumstressav:rural 3065 2882
## prjthreat_sumstresschg:L2sesw1z 5261 3135
## prjthreat_sumstressav:L2sesw1z 2697 2719
## prjthreat_sumstresschg:rural:L2sesw1z 6367 2568
## prjthreat_sumstressav:rural:L2sesw1z 3915 2724
## prjharm_sumstresschg 5575 3334
## prjharm_sumstressav 2599 2918
## prjharm_rural 2644 2490
## prjharm_L2sesw1z 2823 3002
## prjharm_sumstresschg:rural 5924 2592
## prjharm_sumstressav:rural 3144 2826
## prjharm_sumstresschg:L2sesw1z 4076 2689
## prjharm_sumstressav:L2sesw1z 2741 2311
## prjharm_sumstresschg:rural:L2sesw1z 4514 2547
## prjharm_sumstressav:rural:L2sesw1z 2958 2686
## prjusedrg_sumstresschg 4769 2736
## prjusedrg_sumstressav 2327 2353
## prjusedrg_rural 3336 2909
## prjusedrg_L2sesw1z 2649 2493
## prjusedrg_sumstresschg:rural 5583 3158
## prjusedrg_sumstressav:rural 2762 2302
## prjusedrg_sumstresschg:L2sesw1z 5311 2717
## prjusedrg_sumstressav:L2sesw1z 2740 2788
## prjusedrg_sumstresschg:rural:L2sesw1z 6518 2737
## prjusedrg_sumstressav:rural:L2sesw1z 3174 2808
## prjhack_sumstresschg 5550 3195
## prjhack_sumstressav 2328 2040
## prjhack_rural 2976 2820
## prjhack_L2sesw1z 3215 2651
## prjhack_sumstresschg:rural 6377 2885
## prjhack_sumstressav:rural 3073 2854
## prjhack_sumstresschg:L2sesw1z 5375 2662
## prjhack_sumstressav:L2sesw1z 2851 2848
## prjhack_sumstresschg:rural:L2sesw1z 5260 3101
## prjhack_sumstressav:rural:L2sesw1z 3436 2417
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).11.1.2.4 Prior summary
out.chg.prjcrime.stsum.rbstcomm.fit[[2]]## prior class coef group resp
## (flat) b
## normal(0, 1) b prjhack
## normal(0, 1) b L2sesw1z prjhack
## normal(0, 1) b rural prjhack
## normal(0, 1) b sumstressav prjhack
## normal(0, 1) b sumstressav:L2sesw1z prjhack
## normal(0, 1) b sumstressav:rural prjhack
## normal(0, 1) b sumstressav:rural:L2sesw1z prjhack
## normal(0, 1) b sumstresschg prjhack
## normal(0, 1) b sumstresschg:L2sesw1z prjhack
## normal(0, 1) b sumstresschg:rural prjhack
## normal(0, 1) b sumstresschg:rural:L2sesw1z prjhack
## normal(0, 1) b prjharm
## normal(0, 1) b L2sesw1z prjharm
## normal(0, 1) b rural prjharm
## normal(0, 1) b sumstressav prjharm
## normal(0, 1) b sumstressav:L2sesw1z prjharm
## normal(0, 1) b sumstressav:rural prjharm
## normal(0, 1) b sumstressav:rural:L2sesw1z prjharm
## normal(0, 1) b sumstresschg prjharm
## normal(0, 1) b sumstresschg:L2sesw1z prjharm
## normal(0, 1) b sumstresschg:rural prjharm
## normal(0, 1) b sumstresschg:rural:L2sesw1z prjharm
## normal(0, 1) b prjthfgt5
## normal(0, 1) b L2sesw1z prjthfgt5
## normal(0, 1) b rural prjthfgt5
## normal(0, 1) b sumstressav prjthfgt5
## normal(0, 1) b sumstressav:L2sesw1z prjthfgt5
## normal(0, 1) b sumstressav:rural prjthfgt5
## normal(0, 1) b sumstressav:rural:L2sesw1z prjthfgt5
## normal(0, 1) b sumstresschg prjthfgt5
## normal(0, 1) b sumstresschg:L2sesw1z prjthfgt5
## normal(0, 1) b sumstresschg:rural prjthfgt5
## normal(0, 1) b sumstresschg:rural:L2sesw1z prjthfgt5
## normal(0, 1) b prjthflt5
## normal(0, 1) b L2sesw1z prjthflt5
## normal(0, 1) b rural prjthflt5
## normal(0, 1) b sumstressav prjthflt5
## normal(0, 1) b sumstressav:L2sesw1z prjthflt5
## normal(0, 1) b sumstressav:rural prjthflt5
## normal(0, 1) b sumstressav:rural:L2sesw1z prjthflt5
## normal(0, 1) b sumstresschg prjthflt5
## normal(0, 1) b sumstresschg:L2sesw1z prjthflt5
## normal(0, 1) b sumstresschg:rural prjthflt5
## normal(0, 1) b sumstresschg:rural:L2sesw1z prjthflt5
## normal(0, 1) b prjthreat
## normal(0, 1) b L2sesw1z prjthreat
## normal(0, 1) b rural prjthreat
## normal(0, 1) b sumstressav prjthreat
## normal(0, 1) b sumstressav:L2sesw1z prjthreat
## normal(0, 1) b sumstressav:rural prjthreat
## normal(0, 1) b sumstressav:rural:L2sesw1z prjthreat
## normal(0, 1) b sumstresschg prjthreat
## normal(0, 1) b sumstresschg:L2sesw1z prjthreat
## normal(0, 1) b sumstresschg:rural prjthreat
## normal(0, 1) b sumstresschg:rural:L2sesw1z prjthreat
## normal(0, 1) b prjusedrg
## normal(0, 1) b L2sesw1z prjusedrg
## normal(0, 1) b rural prjusedrg
## normal(0, 1) b sumstressav prjusedrg
## normal(0, 1) b sumstressav:L2sesw1z prjusedrg
## normal(0, 1) b sumstressav:rural prjusedrg
## normal(0, 1) b sumstressav:rural:L2sesw1z prjusedrg
## normal(0, 1) b sumstresschg prjusedrg
## normal(0, 1) b sumstresschg:L2sesw1z prjusedrg
## normal(0, 1) b sumstresschg:rural prjusedrg
## normal(0, 1) b sumstresschg:rural:L2sesw1z prjusedrg
## (flat) Intercept
## normal(0, 2) Intercept prjhack
## normal(0, 2) Intercept prjharm
## normal(0, 2) Intercept prjthfgt5
## normal(0, 2) Intercept prjthflt5
## normal(0, 2) Intercept prjthreat
## normal(0, 2) Intercept prjusedrg
## student_t(3, 0, 2.5) sd prjhack
## student_t(3, 0, 2.5) sd prjharm
## student_t(3, 0, 2.5) sd prjthfgt5
## student_t(3, 0, 2.5) sd prjthflt5
## student_t(3, 0, 2.5) sd prjthreat
## student_t(3, 0, 2.5) sd prjusedrg
## student_t(3, 0, 2.5) sd id prjhack
## student_t(3, 0, 2.5) sd Intercept id prjhack
## student_t(3, 0, 2.5) sd id prjharm
## student_t(3, 0, 2.5) sd Intercept id prjharm
## student_t(3, 0, 2.5) sd id prjthfgt5
## student_t(3, 0, 2.5) sd Intercept id prjthfgt5
## student_t(3, 0, 2.5) sd id prjthflt5
## student_t(3, 0, 2.5) sd Intercept id prjthflt5
## student_t(3, 0, 2.5) sd id prjthreat
## student_t(3, 0, 2.5) sd Intercept id prjthreat
## student_t(3, 0, 2.5) sd id prjusedrg
## student_t(3, 0, 2.5) sd Intercept id prjusedrg
## dpar nlpar lb ub source
## default
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)11.1.3 Corr X Robust Community: stdz sum stress & any prj crime
#repeat community robustness model for "any criminal intent" outcome
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept'),
set_prior('normal(0, 1)', class = 'b')
)
chg.anyprjcrime.stsum.rbstcomm.fit <- brm(prjany ~ 1 +
sumstresschg + sumstressav + rural + L2sesw1z +
sumstresschg:rural + sumstressav:rural +
sumstresschg:L2sesw1z + sumstressav:L2sesw1z +
sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z +
(1 | id),
data = stress.long3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
adapt_delta = 0.85,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_anyprjcrime_stsum_rbstcomm_fit",
file_refit = "on_change"
)
#Update function to call all ppchecks for bivar projected crime models
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit)
plotcoefs2 <- mcmc_plot(modelfit, variable = "^b_sum", regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients \nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, plotcoefs2)
return(allchecks)
}
out.chg.anyprjcrime.stsum.rbstcomm.fit <- ppchecks(chg.anyprjcrime.stsum.rbstcomm.fit)11.1.3.1 Coefficient plot (intervals)
out.chg.anyprjcrime.stsum.rbstcomm.fit[[4]] 
11.1.3.2 PPcheck (density)
out.chg.anyprjcrime.stsum.rbstcomm.fit[[3]] + labs(title = "Any Crime Intent (chg)") 
11.1.3.3 Fit summary
out.chg.anyprjcrime.stsum.rbstcomm.fit[[1]]## Family: bernoulli
## Links: mu = logit
## Formula: prjany ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## Data: stress.long3 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 3.25 0.42 2.50 4.16 1.00 1305 2226
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept -2.52 0.39 -3.37 -1.83 1.00 2380
## sumstresschg 0.31 0.51 -0.71 1.31 1.00 7915
## sumstressav 0.59 0.28 0.04 1.14 1.00 3184
## rural -2.35 0.45 -3.25 -1.45 1.00 3970
## L2sesw1z 0.39 0.22 -0.03 0.82 1.00 3273
## sumstresschg:rural 0.75 0.75 -0.73 2.21 1.00 7933
## sumstressav:rural -0.28 0.45 -1.15 0.61 1.00 3865
## sumstresschg:L2sesw1z -0.42 0.48 -1.37 0.50 1.00 8578
## sumstressav:L2sesw1z -0.07 0.24 -0.54 0.40 1.00 3696
## sumstresschg:rural:L2sesw1z -0.98 0.85 -2.63 0.65 1.00 10513
## sumstressav:rural:L2sesw1z -0.01 0.55 -1.07 1.07 1.00 4394
## Tail_ESS
## Intercept 2942
## sumstresschg 3016
## sumstressav 3078
## rural 3133
## L2sesw1z 3051
## sumstresschg:rural 3119
## sumstressav:rural 3310
## sumstresschg:L2sesw1z 3072
## sumstressav:L2sesw1z 2948
## sumstresschg:rural:L2sesw1z 2626
## sumstressav:rural:L2sesw1z 2694
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).11.1.3.4 Prior summary
out.chg.anyprjcrime.stsum.rbstcomm.fit[[2]]## prior class coef group resp dpar
## normal(0, 1) b
## normal(0, 1) b L2sesw1z
## normal(0, 1) b rural
## normal(0, 1) b sumstressav
## normal(0, 1) b sumstressav:L2sesw1z
## normal(0, 1) b sumstressav:rural
## normal(0, 1) b sumstressav:rural:L2sesw1z
## normal(0, 1) b sumstresschg
## normal(0, 1) b sumstresschg:L2sesw1z
## normal(0, 1) b sumstresschg:rural
## normal(0, 1) b sumstresschg:rural:L2sesw1z
## normal(0, 2) Intercept
## student_t(3, 0, 2.5) sd
## student_t(3, 0, 2.5) sd id
## student_t(3, 0, 2.5) sd Intercept id
## nlpar lb ub source
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## 0 default
## 0 (vectorized)
## 0 (vectorized)11.1.4 Corr X Robust Community: stdz sum stress & negative emotions
depdv_names <- noquote(c("depcantgo", "depeffort", "deplonely", "depblues",
"depunfair", "depmistrt", "depbetray"))
#Vectorize priors:
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = depdv_names),
set_prior('normal(0, 1)', class = 'b', resp = depdv_names)
)
chg.alldepress.stsum.rbstcomm.fit <- brm(
mvbind(depcantgo, depeffort, deplonely, depblues, depunfair, depmistrt,
depbetray) ~ 1 +
sumstresschg + sumstressav + rural + L2sesw1z +
sumstresschg:rural + sumstressav:rural +
sumstresschg:L2sesw1z + sumstressav:L2sesw1z +
sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z +
(1 | id),
data = stress.long3,
family = "bernoulli",
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/chg_alldepress_stsum_rbstcomm_fit",
file_refit = "on_change"
)
##Update function to call all ppchecks for bivar neg emotions chg models
varlist1 <- c("^b_depcantgo_sum","^b_depeffort_sum", "^b_deplonely_sum",
"^b_depblues_sum")
varlist2 <- c("^b_depunfair_sum", "^b_depmistrt_sum", "^b_depbetray_sum")
ppchecks <- function(modelfit) {
fitsummary <- summary(modelfit)
priorsummary <- prior_summary(modelfit)
ppcheckdv1 <- pp_check(modelfit, resp="depcantgo")
ppcheckdv2 <-pp_check(modelfit, resp="depeffort")
ppcheckdv3 <-pp_check(modelfit, resp="deplonely")
ppcheckdv4 <-pp_check(modelfit, resp="depblues")
ppcheckdv5 <-pp_check(modelfit, resp="depunfair")
ppcheckdv6 <-pp_check(modelfit, resp="depmistrt")
ppcheckdv7 <-pp_check(modelfit, resp="depbetray")
plotcoefs1 <- mcmc_plot(modelfit, variable = varlist1, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients\nwith medians, 80%, and 95% intervals")
plotcoefs2 <- mcmc_plot(modelfit, variable = varlist2, regex = TRUE,
prob = 0.80, prob_outer = 0.95) +
labs(title = "Coefficient plot",
subtitle = "Posterior intervals for irt stress scale coefficients\nwith medians, 80%, and 95% intervals")
allchecks <- list(fitsummary, priorsummary, ppcheckdv1, ppcheckdv2, ppcheckdv3,
ppcheckdv4, ppcheckdv5, ppcheckdv6, ppcheckdv7,
plotcoefs1, plotcoefs2)
return(allchecks)
}
out.chg.alldepress.stsum.rbstcomm.fit <- ppchecks(chg.alldepress.stsum.rbstcomm.fit)11.1.4.1 Coefficient plot (depressive symptoms)
out.chg.alldepress.stsum.rbstcomm.fit[[10]]
11.1.4.2 Coefficient plot (criminogenic emotions)
out.chg.alldepress.stsum.rbstcomm.fit[[11]]
11.1.4.3 PPcheck (density)
p1 <- out.chg.alldepress.stsum.rbstcomm.fit[[3]] + labs(title = "Can't Get Going (chg)")
p2 <- out.chg.alldepress.stsum.rbstcomm.fit[[4]] + labs(title = "Everything Effort (chg)")
p3 <- out.chg.alldepress.stsum.rbstcomm.fit[[5]] + labs(title = "Lonely (chg)")
p4 <- out.chg.alldepress.stsum.rbstcomm.fit[[6]] + labs(title = "Can't Shake Blues (chg)")
p5 <- out.chg.alldepress.stsum.rbstcomm.fit[[7]] + labs(title = "Felt Life Unfair (chg)")
p6 <- out.chg.alldepress.stsum.rbstcomm.fit[[8]] + labs(title = "Felt Mistreated (chg)")
p7 <- out.chg.alldepress.stsum.rbstcomm.fit[[9]] + labs(title = "Felt Betrayed (chg)")
(p1 + p2) / (p3 + p4) / (p5 + p6) / (p7 + plot_spacer())
11.1.4.4 Fit summary
out.chg.alldepress.stsum.rbstcomm.fit[[1]]## Family: MV(bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli, bernoulli)
## Links: mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## mu = logit
## Formula: depcantgo ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## depeffort ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## deplonely ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## depblues ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## depunfair ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## depmistrt ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## depbetray ~ 1 + sumstresschg + sumstressav + rural + L2sesw1z + sumstresschg:rural + sumstressav:rural + sumstresschg:L2sesw1z + sumstressav:L2sesw1z + sumstresschg:rural:L2sesw1z + sumstressav:rural:L2sesw1z + (1 | id)
## Data: stress.long3 (Number of observations: 978)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Multilevel Hyperparameters:
## ~id (Number of levels: 489)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## sd(depcantgo_Intercept) 0.40 0.22 0.02 0.83 1.00 635
## sd(depeffort_Intercept) 0.49 0.29 0.02 1.08 1.00 595
## sd(deplonely_Intercept) 0.51 0.26 0.03 0.99 1.00 509
## sd(depblues_Intercept) 0.59 0.31 0.04 1.21 1.01 602
## sd(depunfair_Intercept) 0.28 0.19 0.01 0.70 1.00 958
## sd(depmistrt_Intercept) 0.34 0.23 0.01 0.85 1.01 748
## sd(depbetray_Intercept) 0.44 0.27 0.02 0.98 1.01 563
## Tail_ESS
## sd(depcantgo_Intercept) 1293
## sd(depeffort_Intercept) 1526
## sd(deplonely_Intercept) 1097
## sd(depblues_Intercept) 1454
## sd(depunfair_Intercept) 1911
## sd(depmistrt_Intercept) 1821
## sd(depbetray_Intercept) 1650
##
## Regression Coefficients:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## depcantgo_Intercept -0.39 0.10 -0.60 -0.19 1.00
## depeffort_Intercept -1.55 0.15 -1.88 -1.28 1.00
## deplonely_Intercept -0.91 0.12 -1.16 -0.68 1.00
## depblues_Intercept -1.94 0.19 -2.35 -1.60 1.00
## depunfair_Intercept -0.82 0.11 -1.04 -0.61 1.00
## depmistrt_Intercept -1.57 0.14 -1.87 -1.30 1.00
## depbetray_Intercept -1.57 0.15 -1.90 -1.30 1.00
## depcantgo_sumstresschg 0.47 0.34 -0.19 1.13 1.00
## depcantgo_sumstressav -0.04 0.10 -0.24 0.15 1.00
## depcantgo_rural 0.05 0.15 -0.24 0.35 1.00
## depcantgo_L2sesw1z 0.01 0.07 -0.13 0.16 1.00
## depcantgo_sumstresschg:rural 1.56 0.50 0.59 2.54 1.00
## depcantgo_sumstressav:rural 0.10 0.15 -0.19 0.41 1.00
## depcantgo_sumstresschg:L2sesw1z 0.16 0.29 -0.40 0.73 1.00
## depcantgo_sumstressav:L2sesw1z 0.10 0.09 -0.07 0.27 1.00
## depcantgo_sumstresschg:rural:L2sesw1z -0.42 0.61 -1.58 0.79 1.00
## depcantgo_sumstressav:rural:L2sesw1z 0.07 0.20 -0.32 0.47 1.00
## depeffort_sumstresschg 0.72 0.42 -0.11 1.52 1.00
## depeffort_sumstressav 0.14 0.13 -0.12 0.39 1.00
## depeffort_rural -0.50 0.20 -0.90 -0.13 1.00
## depeffort_L2sesw1z -0.06 0.10 -0.25 0.13 1.00
## depeffort_sumstresschg:rural 0.44 0.62 -0.75 1.66 1.00
## depeffort_sumstressav:rural -0.17 0.21 -0.56 0.25 1.00
## depeffort_sumstresschg:L2sesw1z 0.06 0.36 -0.66 0.76 1.00
## depeffort_sumstressav:L2sesw1z -0.12 0.11 -0.34 0.08 1.00
## depeffort_sumstresschg:rural:L2sesw1z -0.81 0.72 -2.19 0.65 1.00
## depeffort_sumstressav:rural:L2sesw1z 0.28 0.26 -0.25 0.79 1.00
## deplonely_sumstresschg 0.87 0.38 0.14 1.63 1.00
## deplonely_sumstressav 0.13 0.11 -0.09 0.35 1.00
## deplonely_rural -0.38 0.16 -0.72 -0.05 1.00
## deplonely_L2sesw1z 0.12 0.08 -0.04 0.29 1.00
## deplonely_sumstresschg:rural 0.32 0.56 -0.76 1.42 1.00
## deplonely_sumstressav:rural -0.23 0.17 -0.58 0.10 1.00
## deplonely_sumstresschg:L2sesw1z 0.69 0.32 0.07 1.32 1.00
## deplonely_sumstressav:L2sesw1z 0.06 0.09 -0.13 0.24 1.00
## deplonely_sumstresschg:rural:L2sesw1z -0.49 0.65 -1.72 0.79 1.00
## deplonely_sumstressav:rural:L2sesw1z 0.20 0.22 -0.23 0.63 1.00
## depblues_sumstresschg -0.01 0.42 -0.83 0.83 1.00
## depblues_sumstressav 0.42 0.15 0.14 0.73 1.00
## depblues_rural -0.36 0.22 -0.79 0.06 1.00
## depblues_L2sesw1z 0.20 0.10 0.00 0.40 1.00
## depblues_sumstresschg:rural -0.50 0.63 -1.72 0.74 1.00
## depblues_sumstressav:rural -0.75 0.22 -1.20 -0.33 1.00
## depblues_sumstresschg:L2sesw1z -0.11 0.38 -0.83 0.65 1.00
## depblues_sumstressav:L2sesw1z 0.01 0.12 -0.22 0.24 1.00
## depblues_sumstresschg:rural:L2sesw1z 0.24 0.72 -1.16 1.65 1.00
## depblues_sumstressav:rural:L2sesw1z -0.50 0.27 -1.03 0.04 1.00
## depunfair_sumstresschg 1.46 0.36 0.75 2.19 1.00
## depunfair_sumstressav 0.30 0.11 0.09 0.51 1.00
## depunfair_rural -0.64 0.17 -0.97 -0.31 1.00
## depunfair_L2sesw1z 0.03 0.08 -0.13 0.18 1.00
## depunfair_sumstresschg:rural 1.03 0.56 -0.06 2.15 1.00
## depunfair_sumstressav:rural -0.12 0.17 -0.46 0.21 1.00
## depunfair_sumstresschg:L2sesw1z 0.33 0.31 -0.27 0.96 1.00
## depunfair_sumstressav:L2sesw1z -0.08 0.09 -0.26 0.10 1.00
## depunfair_sumstresschg:rural:L2sesw1z -0.57 0.64 -1.84 0.69 1.00
## depunfair_sumstressav:rural:L2sesw1z -0.08 0.23 -0.55 0.35 1.00
## depmistrt_sumstresschg 0.56 0.40 -0.23 1.34 1.00
## depmistrt_sumstressav 0.47 0.13 0.22 0.73 1.00
## depmistrt_rural -0.17 0.19 -0.52 0.20 1.00
## depmistrt_L2sesw1z 0.06 0.09 -0.12 0.24 1.00
## depmistrt_sumstresschg:rural 0.64 0.61 -0.55 1.81 1.00
## depmistrt_sumstressav:rural 0.00 0.20 -0.39 0.41 1.00
## depmistrt_sumstresschg:L2sesw1z 0.56 0.35 -0.11 1.25 1.00
## depmistrt_sumstressav:L2sesw1z 0.02 0.11 -0.20 0.24 1.00
## depmistrt_sumstresschg:rural:L2sesw1z -0.78 0.67 -2.10 0.50 1.00
## depmistrt_sumstressav:rural:L2sesw1z -0.20 0.26 -0.68 0.30 1.00
## depbetray_sumstresschg 0.66 0.40 -0.13 1.45 1.00
## depbetray_sumstressav 0.62 0.14 0.36 0.89 1.00
## depbetray_rural -0.48 0.20 -0.87 -0.08 1.00
## depbetray_L2sesw1z 0.05 0.10 -0.15 0.24 1.00
## depbetray_sumstresschg:rural 1.29 0.63 0.04 2.51 1.00
## depbetray_sumstressav:rural -0.24 0.22 -0.67 0.19 1.00
## depbetray_sumstresschg:L2sesw1z 0.13 0.36 -0.58 0.83 1.00
## depbetray_sumstressav:L2sesw1z 0.13 0.12 -0.10 0.36 1.00
## depbetray_sumstresschg:rural:L2sesw1z -1.00 0.73 -2.48 0.42 1.00
## depbetray_sumstressav:rural:L2sesw1z -0.31 0.28 -0.85 0.24 1.00
## Bulk_ESS Tail_ESS
## depcantgo_Intercept 7574 2833
## depeffort_Intercept 1707 2349
## deplonely_Intercept 2408 2739
## depblues_Intercept 1536 2021
## depunfair_Intercept 5153 3028
## depmistrt_Intercept 3543 2728
## depbetray_Intercept 2274 2429
## depcantgo_sumstresschg 7217 3028
## depcantgo_sumstressav 7150 2816
## depcantgo_rural 9284 3137
## depcantgo_L2sesw1z 10574 3058
## depcantgo_sumstresschg:rural 6512 3031
## depcantgo_sumstressav:rural 7149 3353
## depcantgo_sumstresschg:L2sesw1z 7754 2937
## depcantgo_sumstressav:L2sesw1z 6704 3278
## depcantgo_sumstresschg:rural:L2sesw1z 7671 3183
## depcantgo_sumstressav:rural:L2sesw1z 7367 3208
## depeffort_sumstresschg 7002 2892
## depeffort_sumstressav 5921 3449
## depeffort_rural 7674 2905
## depeffort_L2sesw1z 8377 2890
## depeffort_sumstresschg:rural 7336 3421
## depeffort_sumstressav:rural 7198 3385
## depeffort_sumstresschg:L2sesw1z 8136 2871
## depeffort_sumstressav:L2sesw1z 6552 2918
## depeffort_sumstresschg:rural:L2sesw1z 8980 3154
## depeffort_sumstressav:rural:L2sesw1z 8360 3419
## deplonely_sumstresschg 5867 3128
## deplonely_sumstressav 6741 3130
## deplonely_rural 7936 3051
## deplonely_L2sesw1z 7018 2839
## deplonely_sumstresschg:rural 7466 2826
## deplonely_sumstressav:rural 7177 3252
## deplonely_sumstresschg:L2sesw1z 8353 2775
## deplonely_sumstressav:L2sesw1z 6962 3126
## deplonely_sumstresschg:rural:L2sesw1z 8400 3229
## deplonely_sumstressav:rural:L2sesw1z 6245 3475
## depblues_sumstresschg 7318 3155
## depblues_sumstressav 5691 3128
## depblues_rural 8229 3044
## depblues_L2sesw1z 6221 3265
## depblues_sumstresschg:rural 7622 2917
## depblues_sumstressav:rural 5080 2936
## depblues_sumstresschg:L2sesw1z 8217 3339
## depblues_sumstressav:L2sesw1z 6239 2929
## depblues_sumstresschg:rural:L2sesw1z 8440 3396
## depblues_sumstressav:rural:L2sesw1z 6329 3494
## depunfair_sumstresschg 7397 2882
## depunfair_sumstressav 6215 2927
## depunfair_rural 8721 3015
## depunfair_L2sesw1z 8371 3088
## depunfair_sumstresschg:rural 7071 2886
## depunfair_sumstressav:rural 7079 3054
## depunfair_sumstresschg:L2sesw1z 7235 2902
## depunfair_sumstressav:L2sesw1z 8690 3345
## depunfair_sumstresschg:rural:L2sesw1z 8618 3325
## depunfair_sumstressav:rural:L2sesw1z 7619 3179
## depmistrt_sumstresschg 7654 3389
## depmistrt_sumstressav 6114 2525
## depmistrt_rural 8251 3216
## depmistrt_L2sesw1z 9277 3133
## depmistrt_sumstresschg:rural 7430 2756
## depmistrt_sumstressav:rural 7441 3415
## depmistrt_sumstresschg:L2sesw1z 9058 3055
## depmistrt_sumstressav:L2sesw1z 7263 3396
## depmistrt_sumstresschg:rural:L2sesw1z 9467 3806
## depmistrt_sumstressav:rural:L2sesw1z 7687 3175
## depbetray_sumstresschg 7800 2861
## depbetray_sumstressav 6058 2554
## depbetray_rural 7394 3333
## depbetray_L2sesw1z 7190 3210
## depbetray_sumstresschg:rural 8189 3080
## depbetray_sumstressav:rural 7292 3416
## depbetray_sumstresschg:L2sesw1z 9421 3233
## depbetray_sumstressav:L2sesw1z 7061 3249
## depbetray_sumstresschg:rural:L2sesw1z 8430 3102
## depbetray_sumstressav:rural:L2sesw1z 6818 3395
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).11.1.4.5 Prior summary
out.chg.alldepress.stsum.rbstcomm.fit[[2]]## prior class coef group resp
## (flat) b
## normal(0, 1) b depbetray
## normal(0, 1) b L2sesw1z depbetray
## normal(0, 1) b rural depbetray
## normal(0, 1) b sumstressav depbetray
## normal(0, 1) b sumstressav:L2sesw1z depbetray
## normal(0, 1) b sumstressav:rural depbetray
## normal(0, 1) b sumstressav:rural:L2sesw1z depbetray
## normal(0, 1) b sumstresschg depbetray
## normal(0, 1) b sumstresschg:L2sesw1z depbetray
## normal(0, 1) b sumstresschg:rural depbetray
## normal(0, 1) b sumstresschg:rural:L2sesw1z depbetray
## normal(0, 1) b depblues
## normal(0, 1) b L2sesw1z depblues
## normal(0, 1) b rural depblues
## normal(0, 1) b sumstressav depblues
## normal(0, 1) b sumstressav:L2sesw1z depblues
## normal(0, 1) b sumstressav:rural depblues
## normal(0, 1) b sumstressav:rural:L2sesw1z depblues
## normal(0, 1) b sumstresschg depblues
## normal(0, 1) b sumstresschg:L2sesw1z depblues
## normal(0, 1) b sumstresschg:rural depblues
## normal(0, 1) b sumstresschg:rural:L2sesw1z depblues
## normal(0, 1) b depcantgo
## normal(0, 1) b L2sesw1z depcantgo
## normal(0, 1) b rural depcantgo
## normal(0, 1) b sumstressav depcantgo
## normal(0, 1) b sumstressav:L2sesw1z depcantgo
## normal(0, 1) b sumstressav:rural depcantgo
## normal(0, 1) b sumstressav:rural:L2sesw1z depcantgo
## normal(0, 1) b sumstresschg depcantgo
## normal(0, 1) b sumstresschg:L2sesw1z depcantgo
## normal(0, 1) b sumstresschg:rural depcantgo
## normal(0, 1) b sumstresschg:rural:L2sesw1z depcantgo
## normal(0, 1) b depeffort
## normal(0, 1) b L2sesw1z depeffort
## normal(0, 1) b rural depeffort
## normal(0, 1) b sumstressav depeffort
## normal(0, 1) b sumstressav:L2sesw1z depeffort
## normal(0, 1) b sumstressav:rural depeffort
## normal(0, 1) b sumstressav:rural:L2sesw1z depeffort
## normal(0, 1) b sumstresschg depeffort
## normal(0, 1) b sumstresschg:L2sesw1z depeffort
## normal(0, 1) b sumstresschg:rural depeffort
## normal(0, 1) b sumstresschg:rural:L2sesw1z depeffort
## normal(0, 1) b deplonely
## normal(0, 1) b L2sesw1z deplonely
## normal(0, 1) b rural deplonely
## normal(0, 1) b sumstressav deplonely
## normal(0, 1) b sumstressav:L2sesw1z deplonely
## normal(0, 1) b sumstressav:rural deplonely
## normal(0, 1) b sumstressav:rural:L2sesw1z deplonely
## normal(0, 1) b sumstresschg deplonely
## normal(0, 1) b sumstresschg:L2sesw1z deplonely
## normal(0, 1) b sumstresschg:rural deplonely
## normal(0, 1) b sumstresschg:rural:L2sesw1z deplonely
## normal(0, 1) b depmistrt
## normal(0, 1) b L2sesw1z depmistrt
## normal(0, 1) b rural depmistrt
## normal(0, 1) b sumstressav depmistrt
## normal(0, 1) b sumstressav:L2sesw1z depmistrt
## normal(0, 1) b sumstressav:rural depmistrt
## normal(0, 1) b sumstressav:rural:L2sesw1z depmistrt
## normal(0, 1) b sumstresschg depmistrt
## normal(0, 1) b sumstresschg:L2sesw1z depmistrt
## normal(0, 1) b sumstresschg:rural depmistrt
## normal(0, 1) b sumstresschg:rural:L2sesw1z depmistrt
## normal(0, 1) b depunfair
## normal(0, 1) b L2sesw1z depunfair
## normal(0, 1) b rural depunfair
## normal(0, 1) b sumstressav depunfair
## normal(0, 1) b sumstressav:L2sesw1z depunfair
## normal(0, 1) b sumstressav:rural depunfair
## normal(0, 1) b sumstressav:rural:L2sesw1z depunfair
## normal(0, 1) b sumstresschg depunfair
## normal(0, 1) b sumstresschg:L2sesw1z depunfair
## normal(0, 1) b sumstresschg:rural depunfair
## normal(0, 1) b sumstresschg:rural:L2sesw1z depunfair
## (flat) Intercept
## normal(0, 2) Intercept depbetray
## normal(0, 2) Intercept depblues
## normal(0, 2) Intercept depcantgo
## normal(0, 2) Intercept depeffort
## normal(0, 2) Intercept deplonely
## normal(0, 2) Intercept depmistrt
## normal(0, 2) Intercept depunfair
## student_t(3, 0, 2.5) sd depbetray
## student_t(3, 0, 2.5) sd depblues
## student_t(3, 0, 2.5) sd depcantgo
## student_t(3, 0, 2.5) sd depeffort
## student_t(3, 0, 2.5) sd deplonely
## student_t(3, 0, 2.5) sd depmistrt
## student_t(3, 0, 2.5) sd depunfair
## student_t(3, 0, 2.5) sd id depbetray
## student_t(3, 0, 2.5) sd Intercept id depbetray
## student_t(3, 0, 2.5) sd id depblues
## student_t(3, 0, 2.5) sd Intercept id depblues
## student_t(3, 0, 2.5) sd id depcantgo
## student_t(3, 0, 2.5) sd Intercept id depcantgo
## student_t(3, 0, 2.5) sd id depeffort
## student_t(3, 0, 2.5) sd Intercept id depeffort
## student_t(3, 0, 2.5) sd id deplonely
## student_t(3, 0, 2.5) sd Intercept id deplonely
## student_t(3, 0, 2.5) sd id depmistrt
## student_t(3, 0, 2.5) sd Intercept id depmistrt
## student_t(3, 0, 2.5) sd id depunfair
## student_t(3, 0, 2.5) sd Intercept id depunfair
## dpar nlpar lb ub source
## default
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## user
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## user
## user
## user
## user
## user
## user
## user
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 default
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)
## 0 (vectorized)11.1.5 Predictive Marginal Contrasts x Robust Comm (Stdz Sum Stress & Crim Intent)
Let’s compare marginal effect contrasts from models with the standardized sum stress scale and four discretely measured communities above to an alternative model that instead includes a rural/urban binary indicator and a continuous standardized community level SES variable in three-way interactions with the same stress variable.
# generate new data grid for specific change contrasts (average over sumstressav)
# keep only 1-unit Stdz sum scale increases from year 1 to year 2 (contrast .5_t2 - -.5_t1)
# generate data for rural/urban & for -1/1 SD L2sesw1z
newdata <- stress.long3 %>%
data_grid(sumstresschg = c(-.5, .5),
sumstressav,
rural,
L2sesw1z = c(-1,1),
year) %>%
filter(sumstresschg == -0.5 & year == "1" |
sumstresschg == 0.5 & year == "2")
# function to generate epred draws
gen_predmarg_data_comm <- function(mymodelfit, xdev){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
group_by(.category, rural, L2sesw1z, {{xdev}}, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`))
}
#generate epred draws
predmarg_stsum_prjcrim_chg_rbstcomm = gen_predmarg_data_comm(chg.prjcrime.stsum.rbstcomm.fit, sumstresschg)
predmarg_stsum_negemots_chg_rbstcomm = gen_predmarg_data_comm(chg.alldepress.stsum.rbstcomm.fit, sumstresschg)
#generate epred draws for "any crim intent" & merge with prjcrime
predmarg_stsum_anyprjcrim_chg_rbstcomm =
epred_draws(chg.anyprjcrime.stsum.rbstcomm.fit,
newdata = newdata,
re_formula = NA) %>%
group_by(rural, L2sesw1z, sumstresschg, .draw) %>%
summarise(`E[y|xdev]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
predmarg_stsum_prjcrim_chg_rbstcomm <- bind_rows(predmarg_stsum_prjcrim_chg_rbstcomm,
predmarg_stsum_anyprjcrim_chg_rbstcomm)
rm(predmarg_stsum_anyprjcrim_chg_rbstcomm) #clean environment
# function to calculate marginal contrasts
calc_ME_chg_comm <- function(predmarg_data, xdev) {
predmarg_data %>%
compare_levels(`E[y|xdev]`, by = xdev) %>% # pairwise diffs in `E[y|x]`, by levels of x
group_by(rural, L2sesw1z) %>% # generate community-specific marginal contrasts
rename(`PLME` = `E[y|xdev]`) # easy colname reflecting ME 2-unit chg contrast
}
# outputs community-specific predicted differences in E[y]
# associated with 1-Sum AVE scale increase in stress (T2-T1, -.5 to .5)
# marg effect contrasts are marginalized over all person-level avg values of stress (AME)
#generate ME contrasts & add alpha indicator
PLME2_stsum_prjcrim_chg_rbstcomm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stsum_prjcrim_chg_rbstcomm, "sumstresschg") %>%
mutate(stress_var = "Sum Stress:\nWithin Person\nChange",
dif_label = "diff in E[y|stress diff or increase]",
method="rbstcommchg") %>%
rename(contrast = sumstresschg)}, file="cache_10_1")
PLME2_stsum_prjcrim_chg_rbstcomm <- PLME2_stsum_prjcrim_chg_rbstcomm %>%
group_by(.category, rural, L2sesw1z) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1")),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==-1, 1, 0),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==1, 2, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==-1, 3, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==1, 4, rural.ses.rbst),
rural.ses.rbst = factor(rural.ses.rbst, levels = c("1", "2", "3", "4"))
) %>%
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
PLME2_stsum_negemots_chg_rbstcomm = xfun::cache_rds({calc_ME_chg_comm(predmarg_stsum_negemots_chg_rbstcomm, "sumstresschg") %>%
mutate(stress_var = "Sum Stress:\nWithin Person\nChange",
dif_label = "diff in E[y|stress diff or increase]",
method="rbstcommchg") %>%
rename(contrast = sumstresschg)}, file="cache_10_2")
PLME2_stsum_negemots_chg_rbstcomm <- PLME2_stsum_negemots_chg_rbstcomm %>%
group_by(.category, rural, L2sesw1z) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1")),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==-1, 1, 0),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==1, 2, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==-1, 3, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==1, 4, rural.ses.rbst),
rural.ses.rbst = factor(rural.ses.rbst, levels = c("1", "2", "3", "4"))
)
# REPEAT FOR BETWEEN-PERSON DIFF ESTIMATES
# marginalize btw-per 1-0 diff contrast estimates across values of w/in person change & year
# NOTE: base rate probs differ across sumstressav scale, so could collapse representative
# contrasts across scale (-1 - -2, -.5 - -1.5, 0 - -1, .5 - -.5, 1 - 0, 1.5 - .5)
# attempted but too computationally intensive
newdata <- stress.long3 %>%
data_grid(sumstressav = c(0,1),
sumstresschg,
rural,
L2sesw1z = c(-1,1),
year)
# function to generate epred draws
gen_predmarg_data_comm <- function(mymodelfit, xdif){
epred_draws(mymodelfit,
newdata = newdata,
re_formula = NA) %>%
group_by(.category, rural, L2sesw1z, {{xdif}}, .draw) %>%
summarise(`E[y|xdif]` = mean(`.epred`))
}
#use function to generate epred draws
predmarg_stsum_prjcrim_av_rbstcomm = gen_predmarg_data_comm(chg.prjcrime.stsum.rbstcomm.fit, sumstressav)
predmarg_stsum_negemots_av_rbstcomm = gen_predmarg_data_comm(chg.alldepress.stsum.rbstcomm.fit, sumstressav)
#generate epred draws for "any crim intent" & merge with prjcrime
predmarg_stsum_anyprjcrim_av_rbstcomm =
epred_draws(chg.anyprjcrime.stsum.rbstcomm.fit,
newdata = newdata,
re_formula = NA) %>%
group_by(rural, L2sesw1z, sumstressav, .draw) %>%
summarise(`E[y|xdif]` = mean(`.epred`)) %>%
ungroup() %>%
mutate(.category="prjany")
predmarg_stsum_prjcrim_av_rbstcomm <- bind_rows(predmarg_stsum_prjcrim_av_rbstcomm,
predmarg_stsum_anyprjcrim_av_rbstcomm)
rm(predmarg_stsum_anyprjcrim_av_rbstcomm) #clean environment
calc_ME_diff_comm <- function(predmarg_data, xdif) {
predmarg_data %>%
compare_levels(`E[y|xdif]`, by = xdif) %>% # pairwise diffs in `E[y|x]`, by levels of x
group_by(rural, L2sesw1z) %>% # generate community-specific marginal contrasts
rename(`PLME` = `E[y|xdif]`) # easy colname reflecting ME 2-unit chg contrast
}
#generate ME contrasts & add alpha indicator & rural.ses.rbst
PLME2_stsum_prjcrim_av_rbstcomm = xfun::cache_rds({calc_ME_diff_comm(predmarg_stsum_prjcrim_av_rbstcomm, "sumstressav") %>%
mutate(stress_var = "Sum Stress:\nBetween Person\nDifference",
dif_label = "diff in E[y|stress diff or increase]",
method="rbstcommav") %>%
rename(contrast = sumstressav)}, file="cache_10_3")
PLME2_stsum_prjcrim_av_rbstcomm <- PLME2_stsum_prjcrim_av_rbstcomm %>%
group_by(.category, rural, L2sesw1z) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1")),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==-1, 1, 0),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==1, 2, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==-1, 3, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==1, 4, rural.ses.rbst),
rural.ses.rbst = factor(rural.ses.rbst, levels = c("1", "2", "3", "4"))
) %>%
ungroup() %>%
mutate(
.category = factor(.category,
levels=c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack",
"prjany"))
)
PLME2_stsum_negemots_av_rbstcomm = xfun::cache_rds({calc_ME_diff_comm(predmarg_stsum_negemots_av_rbstcomm, "sumstressav") %>%
mutate(stress_var = "Sum Stress:\nBetween Person\nDifference",
dif_label = "diff in E[y|stress diff or increase]",
method="rbstcommav") %>%
rename(contrast = sumstressav)}, file="cache_10_4")
PLME2_stsum_negemots_av_rbstcomm <- PLME2_stsum_negemots_av_rbstcomm %>%
group_by(.category, rural, L2sesw1z) %>%
mutate(n_ests = n(),
n_gt0 = sum(PLME>0),
p_gt0 = n_gt0 / n_ests,
p80_gt0 = if_else(p_gt0 >= .80, 1, 0),
p80_gt0 = factor(p80_gt0, levels=c("0","1")),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==-1, 1, 0),
rural.ses.rbst = if_else(rural==1 & L2sesw1z==1, 2, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==-1, 3, rural.ses.rbst),
rural.ses.rbst = if_else(rural==0 & L2sesw1z==1, 4, rural.ses.rbst),
rural.ses.rbst = factor(rural.ses.rbst, levels = c("1", "2", "3", "4"))
) 11.2 APPENDIX 2: PLME Stress/Outcome Change X Community
#function to find & drop leading zeroes (used for x-axis label)
dropLeadingZero <- function(l){
str_replace(l, '0(?=.)', '')
}
numcommlabs2 <- c("1"="Rural\nLowSES",
"2"="Rural\nHighSES",
"3"="Urban\nLowSES",
"4"="Urban\nHighSES")
numrbstcommlabs2 <- c("1"="Rural\nLowSES",
"2"="Rural\nHighSES",
"3"="Urban\nLowSES",
"4"="Urban\nHighSES")
prjlabs2 <- c(
"prjthflt5"="Theft\n<5BAM",
"prjthfgt5"="Theft\n>5BAM",
"prjthreat"="Threat",
"prjharm"="Phys.\nharm",
"prjusedrg"="Use\ndrugs",
"prjhack"="Hack\ninfo",
"prjany"="Any crime")
deplabs2 <- c(
"depcantgo"="Can't\ngo",
"depeffort"="Effort",
"deplonely"="Lonely",
"depblues"="Blues",
"depunfair"="Unfair",
"depmistrt"="Mistreated",
"depbetray"="Betrayed")
methodlabs <- c(
"sumav"="Between-Person Difference\n(Discrete Communities)",
"sumchg"="Within-Person Change\n(Discrete Communities)",
"rbstcommav"="Between-Person Difference\n(Continuous Communities)",
"rbstcommchg"="Within-Person Change\n(Continuous Communities)")
sumprjcommplot2 <- ggplot(data = PLME2_stsum_prjcrim_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_prjcrim_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stsum_prjcrim_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#7ad151", "#440154"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
# coord_cartesian(xlim=c(-.06,.11)) +
# scale_x_continuous(breaks=c(0,.1), labels = dropLeadingZero) +
coord_cartesian(xlim=c(-.055,.355)) + #plot on half scale as neg emotions
scale_x_continuous(breaks=c(0,.25), labels = dropLeadingZero) +
xlab(element_blank()) +
labs(subtitle='Criminal Intent\n___________________________________________________________') +
scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size = 8, angle=65),
axis.line.x = element_blank(), #remove x-axis for top plots
axis.ticks.x = element_blank(), #remove x-axis for top plots
axis.text.x = element_blank(), #remove x-axis for top plots
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumprjcommplot2
sumdepcommplot2 <- ggplot(data = PLME2_stsum_negemots_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_negemots_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stsum_negemots_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#7ad151", "#440154"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.71)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
labs(subtitle='Negative Emotions\n____________________________________________________________') +
# scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size = 8, angle=65),
axis.line.x = element_blank(), #remove x-axis for top plots
axis.ticks.x = element_blank(), #remove x-axis for top plots
axis.text.x = element_blank(), #remove x-axis for top plots
axis.text.y=element_blank(), #remove y-axis labels
axis.ticks.y=element_blank(), #remove y-axis ticks
axis.line.y=element_blank(), #remove y-axis line
plot.title = element_text(size=10, face="bold"),
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumdepcommplot2
sumprjrbstcommplot <- ggplot(data = PLME2_stsum_prjcrim_chg_rbstcomm,
mapping = aes(x = PLME,
y = reorder(rural.ses.rbst, desc(rural.ses.rbst)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_prjcrim_chg_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stsum_prjcrim_av_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#fcae12", "#a92e5e"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
# coord_cartesian(xlim=c(-.06,.11)) +
# scale_x_continuous(breaks=c(0,.1), labels = dropLeadingZero) +
coord_cartesian(xlim=c(-.055,.355)) + #plot on half scale as neg emotions
scale_x_continuous(breaks=c(0,.25), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numrbstcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_blank(), #remove facet (outcome) labels
# strip.text.x = element_text(size = 8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumprjrbstcommplot
sumdeprbstcommplot <- ggplot(data = PLME2_stsum_negemots_chg_rbstcomm,
mapping = aes(x = PLME,
y = reorder(rural.ses.rbst, desc(rural.ses.rbst)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_negemots_chg_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
stat_pointinterval(data=PLME2_stsum_negemots_av_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
scale_color_manual(values=c("#fcae12", "#a92e5e"),
labels=as_labeller(methodlabs), name=NULL) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.71)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numrbstcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_blank(),
# strip.text.x = element_text(size = 8),
axis.text.y=element_blank(), #remove y-axis labels
axis.ticks.y=element_blank(), #remove y-axis ticks
axis.line.y=element_blank(), #remove y-axis line
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumdeprbstcommplot
design <- "
12
34
55
"
# library(patchwork)
SuppAppendix2a <- sumprjcommplot2 + sumdepcommplot2 +
sumprjrbstcommplot + sumdeprbstcommplot +
guide_area() +
plot_layout(design=design, guides = 'collect', heights = c(2,2,.1)) +
plot_annotation(
title = 'SUPPLEMENTAL APPENDIX 2a\nMarginal Effects of 1-Unit Stress Scale Increase on Outcome Probabilities, Comparing Estimators by Discrete or Continuous Community',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Estimates derived from 4 multivariate (using `brms::mvbind()`) and multilevel between-within Bayesian logistic regression models simultaneously regressing all criminal intent outcomes (2 models) and all negative emotion outcomes (2 models) on a standardized sum stress scale separated into L2 cross-time average (Xbar_i) between-person and L1 within-person change (X_it - Xbar_i) "fixed effects" estimators. Models summarized in top panel included a factor variable for community and multiplicative interactions between community and both L1/L2 stress estimators. Models summarized in bottom panel included three-way interactions between stress, a rural/urban binary indicator, and a continuous standardized community level SES variable. Models were estimated in brms with 4 chains and 4000 total post-warmup posterior draws per outcome and per community group. Marginal effect contrast distributions were estimated from the expectation of the posterior predictive distribution for each model as community-specific predicted probability difference distributions averaged over all 1-unit increases on the stress scale (within) or for a 1SD increase from mean (between; "0" vs "1") on initial IRT or standardized latent scale, averaged over the alternative (between or within) stress estimator levels. In bottom panel, predicted contrasts were estimated for rural/urban communities with -1SD (low) and +1SD (high) continuous L2 SES levels. Median posterior density estimates with 95% intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates for the average marginal effect contrast are greater than zero.', width=190)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'bottom',
legend.key.height = unit(1.5, 'cm'))
SuppAppendix2a
ggsave("SuppAppendix2a.jpeg", width=9, height=6.5, path=here("Output"))numcommlabs2 <- c("1"="Rural\nLowSES",
"2"="Rural\nHighSES",
"3"="Urban\nLowSES",
"4"="Urban\nHighSES")
numrbstcommlabs2 <- c("1"="Rural\nLowSES",
"2"="Rural\nHighSES",
"3"="Urban\nLowSES",
"4"="Urban\nHighSES")
prjlabs2 <- c(
"prjthflt5"="Theft\n<5BAM",
"prjthfgt5"="Theft\n>5BAM",
"prjthreat"="Threat",
"prjharm"="Phys.\nharm",
"prjusedrg"="Use\ndrugs",
"prjhack"="Hack\ninfo",
"prjany"="Any crime")
deplabs2 <- c(
"depcantgo"="Can't\ngo",
"depeffort"="Effort",
"deplonely"="Lonely",
"depblues"="Blues",
"depunfair"="Unfair",
"depmistrt"="Mistreated",
"depbetray"="Betrayed")
methodlabs <- c(
"sumav"="Between-Person Difference\n(Discrete Communities)",
"rbstcommav"="Between-Person Difference\n(Continuous Communities)",
"sumchg"="Within-Person Change\n(Discrete Communities)",
"rbstcommchg"="Within-Person Change\n(Continuous Communities)")
sumprjwfecommplot2 <- ggplot(data = PLME2_stsum_prjcrim_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_prjcrim_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
stat_pointinterval(data=PLME2_stsum_prjcrim_chg_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
scale_color_manual(values=c("#7ad151", "#440154"),
labels=as_labeller(methodlabs), name=NULL,
breaks=c('sumchg', 'rbstcommchg')) + #change order of legend items
scale_alpha_discrete(range=c(.2,1), guide = "none") +
# coord_cartesian(xlim=c(-.06,.11)) +
# scale_x_continuous(breaks=c(0,.1), labels = dropLeadingZero) +
coord_cartesian(xlim=c(-.055,.355)) + #plot on half scale as neg emotions
scale_x_continuous(breaks=c(0,.25), labels = dropLeadingZero) +
xlab(element_blank()) +
labs(subtitle='Criminal Intent\n___________________________________________________________') +
scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size = 8, angle=65),
axis.line.x = element_blank(), #remove x-axis for top plots
axis.ticks.x = element_blank(), #remove x-axis for top plots
axis.text.x = element_blank(), #remove x-axis for top plots
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumprjwfecommplot2
sumdepwfecommplot2 <- ggplot(data = PLME2_stsum_negemots_chg_comm,
mapping = aes(x = PLME,
y = reorder(rural.ses.med, desc(rural.ses.med)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_negemots_chg_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
stat_pointinterval(data=PLME2_stsum_negemots_chg_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
scale_color_manual(values=c("#7ad151", "#440154"),
labels=as_labeller(methodlabs), name=NULL,
breaks=c('sumchg', 'rbstcommchg')) + #change order of legend items
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.71)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
labs(subtitle='Negative Emotions\n____________________________________________________________') +
# scale_y_discrete(labels=numcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_text(size = 8, angle=65),
axis.line.x = element_blank(), #remove x-axis for top plots
axis.ticks.x = element_blank(), #remove x-axis for top plots
axis.text.x = element_blank(), #remove x-axis for top plots
axis.text.y=element_blank(), #remove y-axis labels
axis.ticks.y=element_blank(), #remove y-axis ticks
axis.line.y=element_blank(), #remove y-axis line
plot.title = element_text(size=10, face="bold"),
plot.subtitle=element_text(size=8, hjust=0.5, face="italic"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumdepwfecommplot2
sumprjbtwcommplot <- ggplot(data = PLME2_stsum_prjcrim_chg_rbstcomm,
mapping = aes(x = PLME,
y = reorder(rural.ses.rbst, desc(rural.ses.rbst)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(prjlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_prjcrim_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
stat_pointinterval(data=PLME2_stsum_prjcrim_av_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
scale_color_manual(values=c("#fcae12", "#a92e5e"),
labels=as_labeller(methodlabs), name=NULL,
breaks=c('sumav', 'rbstcommav')) + #change order of legend items
scale_alpha_discrete(range=c(.2,1), guide = "none") +
# coord_cartesian(xlim=c(-.06,.11)) +
# scale_x_continuous(breaks=c(0,.1), labels = dropLeadingZero) +
coord_cartesian(xlim=c(-.055,.355)) + #plot on half scale as neg emotions
scale_x_continuous(breaks=c(0,.25), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numrbstcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_blank(), #remove facet (outcome) labels
# strip.text.x = element_text(size = 8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumprjbtwcommplot
sumdepbtwcommplot <- ggplot(data = PLME2_stsum_negemots_chg_rbstcomm,
mapping = aes(x = PLME,
y = reorder(rural.ses.rbst, desc(rural.ses.rbst)),
color=method)) +
facet_wrap(~.category, nrow=1,
labeller = labeller(.category = as_labeller(deplabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
stat_pointinterval(data=PLME2_stsum_negemots_av_comm,
aes(x=PLME, y=reorder(rural.ses.med, desc(rural.ses.med)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = .15)) +
stat_pointinterval(data=PLME2_stsum_negemots_av_rbstcomm,
aes(x=PLME, y=reorder(rural.ses.rbst, desc(rural.ses.rbst)),
alpha=p80_gt0),
.width = .95, size = .7,
position = position_nudge(y = -.15)) +
scale_color_manual(values=c("#fcae12", "#a92e5e"),
labels=as_labeller(methodlabs), name=NULL,
breaks=c('sumav', 'rbstcommav')) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
coord_cartesian(xlim=c(-.11,.71)) +
scale_x_continuous(breaks=c(0,.5), labels = dropLeadingZero) +
xlab(element_blank()) +
scale_y_discrete(labels=numrbstcommlabs2) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
# legend.direction="vertical",
strip.background = element_blank(),
strip.text.x = element_blank(),
# strip.text.x = element_text(size = 8),
axis.text.y=element_blank(), #remove y-axis labels
axis.ticks.y=element_blank(), #remove y-axis ticks
axis.line.y=element_blank(), #remove y-axis line
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)))
# sumdepbtwcommplot
design <- "
12
34
55
"
# library(patchwork)
Appendix2 <- sumprjwfecommplot2 + sumdepwfecommplot2 +
sumprjbtwcommplot + sumdepbtwcommplot +
guide_area() +
plot_layout(design=design, guides = 'collect', heights = c(2,2,.1)) +
plot_annotation(
title = 'APPENDIX 2\nMarginal Effects of Stress Scale Increase on Outcome Probabilities, by Estimator & Discrete or Continuous Community Measure',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Estimates derived from multivariate (using `brms::mvbind()`) and multilevel between-within Bayesian logistic regression models simultaneously regressing all criminal intent outcomes (6*2=12 models) and all negative emotion outcomes (7*2=14 models) on a standardized sum stress scale separated into L2 cross-time average (Xbar_i) between-person and L1 within-person change (X_it - Xbar_i) "fixed effects" estimators. Two separate models also regressed "any criminal intent" on stress. "Discrete" community models included a factor variable for community and multiplicative interactions between community and both L1/L2 stress estimators. "Continuous" community models included three- way interactions between stress, a rural/urban binary indicator, and a continuous standardized community level SES variable to assess robustness of results across community measurement. Models were estimated in brms with 4 chains and 4000 total post-warmup posterior draws per outcome and per community group. Marginal effect contrast distributions were estimated from the expectation of the posterior predictive distribution for each model as community-specific predicted probability difference distributions averaged over all 1-unit increases on the stress scale (within) or for a 1SD increase from mean (between; "0" vs "1") on initial IRT or standardized latent scale, averaged over the alternative (between or within) stress estimator levels. In bottom panel, predicted contrasts were estimated for rural/urban communities with -1SD (low) and +1SD (high) continuous L2 SES levels. Median posterior density estimates with 95% intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates for the average marginal effect contrast are greater than zero.', width=190)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'bottom',
legend.key.height = unit(.5, 'cm'),
legend.margin = margin(0,0,0,0),
legend.spacing.y = unit(0, "mm"))
Appendix2
ggsave("Appendix2.jpeg", width=9, height=6.5, path=here("Output"))save(stress.long3, file = here("1_Data_Files/Datasets/stress_long3.Rdata"))12 Figure 5: Mediation models
(RMD FILE: BDK_2023_Stress_11_Fig5_mediate)
## [1] "T/F: Root 'here()' folder contains subfolder 'Models'"## [1] TRUE12.1 Mediation Modeling
General strain theory posits that stress causes negative emotions that, under certain conditions, cause criminal coping reactions, while certain stressors and resulting negative emotions are expected to be more “criminogenic” than others. For instance, theory and research suggests that individuals who experience “internalizing emotions” such as depressive symptoms in response to stressors are less likely to engage in aggressive or criminal coping behaviors. In contrast, stressors that invoke perceptions of unfair treatment, mistreatment, or betrayal presumably are more likely to result in “externalizing” responses such as criminal coping. Hence, we explore whether multivariate correlational patterns are consistent with the prediction that “criminogenic” emotions mediate stress-criminal intent associations.
Unfortunately, estimating mediation models with every multivariate combination of stress item, criminogenic emotion item, and criminal intent outcome item compounds the number of potential estimates, making summary and visualization even more untenable. Thus, for this exploratory analysis, we will reduce the number of potential estimates by, first, relying on our standardized sum stress composite scale and, second, collapsing our negative emotions items into a variety index.
The sum stress scale will permit estimation of the posited effects of between-person differences and within-person changes on criminogenic emotions and stress. Alternative scaling methods, such as factor-based or IRT scaling, typically enforce strong assumptions about the unidimensionality of an underlying latent construct and, by weighting items according to underlying theta or factor loadings, may overrepresent effect estimates corresponding to upweighted component items while suppressing those corresponding to down-weighted items. In this case, while a general “stressed” construct might be partly be driving reporting on each component stress item, multidimensionality cannot be ruled out - that is, one can imagine experiencing high financial stress, low relational stress, high job-related stress, and low victimization stress - or any other possible combination of stress levels. In such instances, weighted scales should be avoided altogether absent compelling theoretical reasoning and, if scaling is necessary, it should be conducted with caution and unweighted (e.g., simple sum) scales should be more effective in capturing and summarizing any effects of underlying components (see Appendix 1).
For similar reasons, we employ a simple variety index of the reported number of negative criminogenic emotions experienced, which permits us to estimate the change in the number of potentially criminogenic emotional “symptoms” associated with an increase in overall (sum) stress levels as well as the change in the probability of criminal intent associated with a one-symptom increase in criminogenic negative emotions. For robustness purposes, we also estimate (and present results below) mediation models using a sum scale for criminogenic emotions.
As for modeling, at this time, there is a lack of accessible options
for multilevel mediation modeling - particularly if one wishes to
generate between- and within-unit estimates or to specify a cumulative
probit/logit function for ordinal predictors. Ultimately, we chose to
estimate between/within models in brms similar to those
estimated earlier, then we use easystats::mediate()
function to estimate mediation pathway coefficients (e.g., direct and
indirect effects). Since this package generates average causal mediation
estimates from posterior distributions, the hope was that it would be
flexible enough to apply to our models with ordinal predictors.
Unfortunately, that did not work (see github issue). So, we instead
settled for simpler mediation models by specifying X (stress) and M
(emotions) as metric continuous variables (linear M models) and Y
(criminal intent) as binary (logistic Y models). We supplemented
estimates from these models with within-person X->M and X->Y
effect plots generated using bmlm package, which is
designed specifically to estimate within-unit mediation processes.
Ultimately, we present exploratory causal mediation estimates generated from these complementary approaches to assess whether these data plausibly might have been generated by the posited causal process - that is, whether evidence supports (among other plausible alternatives) the existence of indirect linear causal effects of differences or changes in stress (sum scale) on criminal intent (binary items) through differences or changes in “criminogenic” negative emotions items.
12.1.1 Data wrangling & summaries
load(here("1_Data_Files/Datasets/stress_wide4.Rdata"))
load(here("1_Data_Files/Datasets/stress_long3.Rdata"))# summarytools::freq
# freq(stress.wide4$depunfairordw1)
# freq(stress.wide4$depunfairordw2)
# freq(stress.wide4$depmistrtordw1)
# freq(stress.wide4$depmistrtordw2)
# freq(stress.wide4$depbetrayordw1)
# freq(stress.wide4$depbetrayordw2)
# missing (NA) on one id for depunfairw1
# freq(stress.long3$depunfair)
# freq(stress.long3$depmistrt)
# freq(stress.long3$depbetray)
stress.long4 <- stress.long3 %>%
mutate(
indcrmemo = (as.numeric(depunfair) + as.numeric(depmistrt) + as.numeric(depbetray))-3,
sumcrmemo = depunfairord + depmistrtord + depbetrayord,
prjthflt5i = as.integer(recode(prjthflt5,
"0" = 0, "1" = 1)), # need binary y as integer for bmlm
prjthfgt5i = as.integer(recode(prjthfgt5,
"0" = 0, "1" = 1)),
prjthreati = as.integer(recode(prjthreat,
"0" = 0, "1" = 1)),
prjharmi = as.integer(recode(prjharm,
"0" = 0, "1" = 1)),
prjusedrgi = as.integer(recode(prjusedrg,
"0" = 0, "1" = 1)),
prjhacki = as.integer(recode(prjhack,
"0" = 0, "1" = 1)),
prjanyi = as.integer(recode(prjany,
"0" = 0, "1" = 1))
) %>%
drop_na(sumcrmemo) %>% # row for id w/missing unfairw1 obs dropped
group_by(id) %>%
mutate(
nobs = n() #count num observations by id to identify id w/missing row
) %>%
ungroup() %>%
filter(nobs ==2) %>%
mutate(
sumcrmemoz = (sumcrmemo - mean(sumcrmemo))/sd(sumcrmemo), #stdz sum crim emo var
indcrmemoc = indcrmemo - mean(indcrmemo)
) %>%
group_by(id) %>% #w2 row for id w/missing unfairw1 obs dropped
mutate(
sumcrmemozav = mean(sumcrmemoz), #create btw-person version of crim emo
indcrmemoav = mean(indcrmemo),
indcrmemocav = mean(indcrmemoc)
) %>%
ungroup() %>%
mutate(
sumcrmemozchg = sumcrmemoz - sumcrmemozav, #create w/in-person version of crim emo
indcrmemocchg = indcrmemoc - indcrmemocav, #create w/in-person version of crim emo
indcrmemocchgi = as.integer(indcrmemocchg*2), #factor & integer versions (x2) for ordinal models
indcrmemocchgf = factor(indcrmemocchgi, ordered=TRUE, levels = c(-3, -2, -1, 0, 1, 2, 3)),
indcrmemoavi = as.integer(indcrmemoav*2),
indcrmemoavf = factor(indcrmemoavi, ordered=TRUE, levels = c(0, 1, 2, 3, 4, 5, 6)),
) %>%
dplyr::select(-nobs)
# 976 obs from 488 ids (one id dropped)12.1.1.1 Crim. Emotions (Sum)
ggplot(stress.long4, aes(sumcrmemo)) + geom_histogram(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Sum Score)")
12.1.1.2 Crim. Emotions (Stdz Sum)
ggplot(stress.long4, aes(sumcrmemoz)) + geom_histogram(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Std Sum Score)")
12.1.1.3 Crim. Emotions (Stdz Sum, Between)
ggplot(stress.long4, aes(sumcrmemozav)) + geom_histogram(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Std Sum Score; btw-id cross-time avg)")
12.1.1.4 Crim. Emotions (Stdz Sum, Within)
ggplot(stress.long4, aes(sumcrmemozchg)) + geom_histogram(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Sum Score; w/in-id chg)")
12.1.1.5 Crim. Emotions (Variety)
ggplot(stress.long4, aes(indcrmemo)) + geom_bar(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Variety Index)")
12.1.1.6 Crim. Emotions (Centered Variety)
ggplot(stress.long4, aes(indcrmemoc)) + geom_bar(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Centered Variety Index)")
12.1.1.7 Crim. Emotions (Centered Variety, Between)
ggplot(stress.long4, aes(indcrmemocav)) + geom_bar(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Centered Variety Index; btw-id avg)")
12.1.1.8 Crim. Emotions (Centered Variety, Within)
ggplot(stress.long4, aes(indcrmemocchg)) + geom_bar(fill="#E99D53") + xlab("sumcrmemo (Crim Emotions Centered Variety Index; w/in-id chg)")
12.1.1.9 Corr: Stdz Sum & Variety Crim. Emotions (Between)
cor(stress.long4$sumcrmemozav, stress.long4$indcrmemocav)## [1] 0.8112.1.1.10 Corr: Stdz Sum & Variety Crim. Emotions (Within)
cor(stress.long4$sumcrmemozchg, stress.long4$indcrmemocchg)## [1] 0.8112.2 B/W Mediation Models (brms; mediate)
12.2.1 X=stdz sum stress, M=crim emotions (variety index)
#Attempted to specify emotion mediators as ordinal
# models estimate fine but mediation() throws errors
# raised issue here: https://github.com/easystats/bayestestR/issues/576
# f1.1o <- bf(prjthflt5 ~ 1 + sumstresschg + sumstressav +
# mo(indcrmemoavf) + mo(indcrmemocchgf) + (1 | id),
# family = "bernoulli")
# f1.2o <- bf(indcrmemocchgf ~ 1 + sumstressav + sumstresschg +
# indcrmemocav + (1 | id),
# family = cumulative("probit"))
# m1wo <- brm(f1.1o + f1.2o, set_rescor(FALSE),
# data = stress.long4,
# # prior = prior1,
# cores = nCoresphys,
# chains = 4,
# backend = "cmdstanr",
# seed = 8675309)
# summary(m1wo)
#
# medy1wo <- mediation(m1wo,
# treatment="sumstresschg",
# mediator="indcrmemocchgf",
# response=c(m="indcrmemocchgf", y="prjthflt5"),
# ci=.95
# )
# medy1wo
# "simple" mediation models treating X & M as metric & Y as binary (logistic)
# default priors for M models, same Int & beta priors for Y models as before
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = "prjthflt5"),
set_prior('normal(0, 1)', class = 'b', resp = "prjthflt5")
)
fy1w <- bf(prjthflt5 ~ 1 + sumstresschg + sumstressav +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
fmw <- bf(indcrmemocchg ~ 1 + sumstressav + sumstresschg +
indcrmemocav + (1 | id),
family = "gaussian")
m1w <- brm(fy1w + fmw, set_rescor(FALSE),
data = stress.long4,
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m1w",
file_refit = "on_change"
)
fy1b <- bf(prjthflt5 ~ 1 + sumstressav + sumstresschg +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
fmb <- bf(indcrmemocav ~ 1 + sumstressav + sumstresschg +
indcrmemocchg + (1 | id),
family = "gaussian")
m1b <- brm(fy1b + fmb, set_rescor(FALSE),
data = stress.long4,
prior = prior1,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m1b",
file_refit = "on_change"
)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = "prjthfgt5"),
set_prior('normal(0, 1)', class = 'b', resp = "prjthfgt5")
)
fy2w <- bf(prjthfgt5 ~ 1 + sumstresschg + sumstressav +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m2w <- brm(fy2w + fmw, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m2w",
file_refit = "on_change"
)
fy2b <- bf(prjthfgt5 ~ 1 + sumstressav + sumstresschg +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m2b <- brm(fy2b + fmb, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m2b",
file_refit = "on_change"
)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = "prjthreat"),
set_prior('normal(0, 1)', class = 'b', resp = "prjthreat")
)
fy3w <- bf(prjthreat ~ 1 + sumstresschg + sumstressav +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m3w <- brm(fy3w + fmw, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m3w",
file_refit = "on_change"
)
fy3b <- bf(prjthreat ~ 1 + sumstressav + sumstresschg +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m3b <- brm(fy3b + fmb, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m3b",
file_refit = "on_change"
)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = "prjharm"),
set_prior('normal(0, 1)', class = 'b', resp = "prjharm")
)
fy4w <- bf(prjharm ~ 1 + sumstresschg + sumstressav +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m4w <- brm(fy4w + fmw, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m4w",
file_refit = "on_change"
)
fy4b <- bf(prjharm ~ 1 + sumstressav + sumstresschg +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m4b <- brm(fy4b + fmb, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m4b",
file_refit = "on_change"
)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = "prjusedrg"),
set_prior('normal(0, 1)', class = 'b', resp = "prjusedrg")
)
fy5w <- bf(prjusedrg ~ 1 + sumstresschg + sumstressav +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m5w <- brm(fy5w + fmw, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m5w",
file_refit = "on_change"
)
fy5b <- bf(prjusedrg ~ 1 + sumstressav + sumstresschg +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m5b <- brm(fy5b + fmb, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m5b",
file_refit = "on_change"
)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = "prjhack"),
set_prior('normal(0, 1)', class = 'b', resp = "prjhack")
)
fy6w <- bf(prjhack ~ 1 + sumstresschg + sumstressav +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m6w <- brm(fy6w + fmw, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m6w",
file_refit = "on_change"
)
fy6b <- bf(prjhack ~ 1 + sumstressav + sumstresschg +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m6b <- brm(fy6b + fmb, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m6b",
file_refit = "on_change"
)
prior1 <- c(
set_prior('normal(0, 2)', class = 'Intercept', resp = "prjany"),
set_prior('normal(0, 1)', class = 'b', resp = "prjany")
)
fy7w <- bf(prjany ~ 1 + sumstresschg + sumstressav +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m7w <- brm(fy7w + fmw, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m7w",
file_refit = "on_change"
)
fy7b <- bf(prjany ~ 1 + sumstressav + sumstresschg +
indcrmemocav + indcrmemocchg + (1 | id),
family = "bernoulli")
m7b <- brm(fy7b + fmb, set_rescor(FALSE),
data = stress.long4,
cores = nCoresphys,
chains = 4,
backend = "cmdstanr",
seed = 8675309,
file = "Models/m7b",
file_refit = "on_change"
)12.2.1.1 Within-person mediation (x=chg, m=chg)
12.2.1.1.1 Theft <5BAM
medy1w <- mediation(m1w,
treatment="sumstresschg",
mediator="indcrmemocchg",
ci=.95)
medy1w## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstresschg
## Mediator : indcrmemocchg
## Response : prjthflt5
##
## Effect | Estimate | 95% ETI
## ---------------------------------------------------
## Direct Effect (ADE) | 0.516 | [-0.687, 1.686]
## Indirect Effect (ACME) | -0.044 | [-0.330, 0.233]
## Mediator Effect | -0.078 | [-0.554, 0.401]
## Total Effect | 0.474 | [-0.688, 1.602]
##
## Proportion mediated: -9.21% [-248.17%, 229.74%]12.2.1.1.2 Theft >5BAM
medy2w <- mediation(m2w,
treatment="sumstresschg",
mediator="indcrmemocchg",
ci=.95)
medy2w## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstresschg
## Mediator : indcrmemocchg
## Response : prjthfgt5
##
## Effect | Estimate | 95% ETI
## ---------------------------------------------------
## Direct Effect (ADE) | 1.586 | [ 0.018, 3.262]
## Indirect Effect (ACME) | -0.083 | [-0.408, 0.217]
## Mediator Effect | -0.144 | [-0.681, 0.380]
## Total Effect | 1.504 | [ 0.010, 3.094]
##
## Proportion mediated: -5.51% [-48.81%, 37.79%]12.2.1.1.3 Threat
medy3w <- mediation(m3w,
treatment="sumstresschg",
mediator="indcrmemocchg",
ci=.95)
medy3w## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstresschg
## Mediator : indcrmemocchg
## Response : prjthreat
##
## Effect | Estimate | 95% ETI
## ---------------------------------------------------
## Direct Effect (ADE) | -1.745 | [-4.178, 0.305]
## Indirect Effect (ACME) | 0.105 | [-0.276, 0.508]
## Mediator Effect | 0.186 | [-0.478, 0.855]
## Total Effect | -1.646 | [-3.906, 0.302]
##
## Proportion mediated: -6.39% [-75.98%, 63.20%]12.2.1.1.4 Phys. Harm
medy4w <- mediation(m4w,
treatment="sumstresschg",
mediator="indcrmemocchg",
ci=.95)
medy4w## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstresschg
## Mediator : indcrmemocchg
## Response : prjharm
##
## Effect | Estimate | 95% ETI
## ----------------------------------------------------
## Direct Effect (ADE) | -2.262 | [-4.772, -0.187]
## Indirect Effect (ACME) | 0.338 | [-0.050, 0.842]
## Mediator Effect | 0.593 | [-0.090, 1.434]
## Total Effect | -1.943 | [-4.257, 0.015]
##
## Proportion mediated: -17.39% [-82.36%, 47.59%]12.2.1.1.5 Use drugs
medy5w <- mediation(m5w,
treatment="sumstresschg",
mediator="indcrmemocchg",
ci=.95)
medy5w## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstresschg
## Mediator : indcrmemocchg
## Response : prjusedrg
##
## Effect | Estimate | 95% ETI
## ---------------------------------------------------
## Direct Effect (ADE) | -1.333 | [-3.807, 0.956]
## Indirect Effect (ACME) | 0.133 | [-0.234, 0.548]
## Mediator Effect | 0.233 | [-0.416, 0.908]
## Total Effect | -1.207 | [-3.601, 1.008]
##
## Proportion mediated: -11.05% [-150.09%, 127.98%]12.2.1.1.6 Hack info.
medy6w <- mediation(m6w,
treatment="sumstresschg",
mediator="indcrmemocchg",
ci=.95)
medy6w## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstresschg
## Mediator : indcrmemocchg
## Response : prjhack
##
## Effect | Estimate | 95% ETI
## ---------------------------------------------------
## Direct Effect (ADE) | -0.911 | [-2.943, 1.042]
## Indirect Effect (ACME) | 0.131 | [-0.243, 0.530]
## Mediator Effect | 0.231 | [-0.416, 0.876]
## Total Effect | -0.784 | [-2.690, 1.092]
##
## Proportion mediated: -16.74% [-252.25%, 218.76%]12.2.1.1.7 Any crime intent
medy7w <- mediation(m1w,
treatment="sumstresschg",
mediator="indcrmemocchg",
ci=.95)
medy7w## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstresschg
## Mediator : indcrmemocchg
## Response : prjthflt5
##
## Effect | Estimate | 95% ETI
## ---------------------------------------------------
## Direct Effect (ADE) | 0.516 | [-0.687, 1.686]
## Indirect Effect (ACME) | -0.044 | [-0.330, 0.233]
## Mediator Effect | -0.078 | [-0.554, 0.401]
## Total Effect | 0.474 | [-0.688, 1.602]
##
## Proportion mediated: -9.21% [-248.17%, 229.74%]12.2.1.2 Between-person mediation (x=av, m=av)
12.2.1.2.1 Theft <5BAM
medy1b <- mediation(m1b,
treatment="sumstressav",
mediator="indcrmemocav",
ci=.95)
medy1b## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstressav
## Mediator : indcrmemocav
## Response : prjthflt5
##
## Effect | Estimate | 95% ETI
## --------------------------------------------------
## Direct Effect (ADE) | 0.684 | [0.194, 1.233]
## Indirect Effect (ACME) | 0.202 | [0.150, 0.264]
## Mediator Effect | 1.179 | [0.769, 1.518]
## Total Effect | 0.839 | [0.438, 1.496]
##
## Proportion mediated: 24.02% [3.64%, 44.40%]12.2.1.2.2 Theft >5BAM
medy2b <- mediation(m2b,
treatment="sumstressav",
mediator="indcrmemocav",
ci=.95)
medy2b## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstressav
## Mediator : indcrmemocav
## Response : prjthfgt5
##
## Effect | Estimate | 95% ETI
## --------------------------------------------------
## Direct Effect (ADE) | 0.956 | [0.287, 1.414]
## Indirect Effect (ACME) | 0.139 | [0.067, 0.312]
## Mediator Effect | 0.909 | [0.471, 1.648]
## Total Effect | 1.145 | [0.370, 1.608]
##
## Proportion mediated: 12.11% [5.08%, 19.15%]12.2.1.2.3 Threat
medy3b <- mediation(m3b,
treatment="sumstressav",
mediator="indcrmemocav",
ci=.95)
medy3b## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstressav
## Mediator : indcrmemocav
## Response : prjthreat
##
## Effect | Estimate | 95% ETI
## --------------------------------------------------
## Direct Effect (ADE) | 1.173 | [1.013, 1.267]
## Indirect Effect (ACME) | 0.222 | [0.095, 0.485]
## Mediator Effect | 1.194 | [0.578, 2.580]
## Total Effect | 1.349 | [1.199, 1.752]
##
## Proportion mediated: 16.47% [6.32%, 26.62%]12.2.1.2.4 Phys. Harm
medy4b <- mediation(m4b,
treatment="sumstressav",
mediator="indcrmemocav",
ci=.95)
medy4b## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstressav
## Mediator : indcrmemocav
## Response : prjharm
##
## Effect | Estimate | 95% ETI
## ---------------------------------------------------
## Direct Effect (ADE) | 0.113 | [-0.162, 0.875]
## Indirect Effect (ACME) | 0.205 | [ 0.152, 0.334]
## Mediator Effect | 1.013 | [ 0.741, 1.862]
## Total Effect | 0.356 | [ 0.022, 1.100]
##
## Proportion mediated: 57.58% [-348.42%, 463.58%]12.2.1.2.5 Use drugs
medy5b <- mediation(m5b,
treatment="sumstressav",
mediator="indcrmemocav",
ci=.95)
medy5b## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstressav
## Mediator : indcrmemocav
## Response : prjusedrg
##
## Effect | Estimate | 95% ETI
## --------------------------------------------------
## Direct Effect (ADE) | 0.460 | [0.415, 0.894]
## Indirect Effect (ACME) | 0.210 | [0.152, 0.308]
## Mediator Effect | 1.241 | [0.677, 1.634]
## Total Effect | 0.670 | [0.567, 1.203]
##
## Proportion mediated: 31.40% [27.00%, 35.80%]12.2.1.2.6 Hack info.
medy6b <- mediation(m6b,
treatment="sumstressav",
mediator="indcrmemocav",
ci=.95)
medy6b## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstressav
## Mediator : indcrmemocav
## Response : prjhack
##
## Effect | Estimate | 95% ETI
## --------------------------------------------------
## Direct Effect (ADE) | 0.731 | [0.373, 1.483]
## Indirect Effect (ACME) | 0.160 | [0.023, 0.224]
## Mediator Effect | 0.801 | [0.103, 1.157]
## Total Effect | 0.823 | [0.532, 1.707]
##
## Proportion mediated: 19.41% [5.58%, 33.23%]12.2.1.2.7 Any crime intent
medy7b <- mediation(m1b,
treatment="sumstressav",
mediator="indcrmemocav",
ci=.95)
medy7b## # Causal Mediation Analysis for Stan Model
##
## Treatment: sumstressav
## Mediator : indcrmemocav
## Response : prjthflt5
##
## Effect | Estimate | 95% ETI
## --------------------------------------------------
## Direct Effect (ADE) | 0.684 | [0.194, 1.233]
## Indirect Effect (ACME) | 0.202 | [0.150, 0.264]
## Mediator Effect | 1.179 | [0.769, 1.518]
## Total Effect | 0.839 | [0.438, 1.496]
##
## Proportion mediated: 24.02% [3.64%, 44.40%]12.3 Alt. Within-person Mediation Models (bmlm)
The models below present an alternative to between/within models for
estimating within-person mediation effects using the bmlm
package’s within-subjects mediation modeling. See here
and here for more
details.
12.3.1 X=stdz sum stress change, M=crim emotions (variety index)
# Within-subjects bayesian mediation models (bmlm)
# X: stress change (sum scale)
# M: criminogenic emotions (variety index)
# Y: binary criminal intent items
# no built-in file save option, knitr & xfun cache options not working as desired during knitting
# instead, manual saving & loading bmlm models
if (file.exists(here("Models","fit_med_1"))) {
fit_med.1 <- readRDS("Models/fit_med_1")
} else {
# mediation analysis - prjthflt5
fit_med.1 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "indcrmemocchg",
y = "prjthflt5i",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_med.1, file = "Models/fit_med_1")
}
# mediation model - prjthfgt5
if (file.exists(here("Models","fit_med_2"))) {
fit_med.2 <- readRDS("Models/fit_med_2")
} else {
fit_med.2 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "indcrmemocchg",
y = "prjthfgt5i",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_med.2, file = "Models/fit_med_2")
}
# mediation model - prjthreati
if (file.exists(here("Models","fit_med_3"))) {
fit_med.3 <- readRDS("Models/fit_med_3")
} else {
fit_med.3 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "indcrmemocchg",
y = "prjthreati",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_med.3, file = "Models/fit_med_3")
}
# mediation model - prjharmi
if (file.exists(here("Models","fit_med_4"))) {
fit_med.4 <- readRDS("Models/fit_med_4")
} else {
fit_med.4 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "indcrmemocchg",
y = "prjharmi",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_med.4, file = "Models/fit_med_4")
}
# mediation model - prjusedrgi
if (file.exists(here("Models","fit_med_5"))) {
fit_med.5 <- readRDS("Models/fit_med_5")
} else {
fit_med.5 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "indcrmemocchg",
y = "prjusedrgi",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_med.5, file = "Models/fit_med_5")
}
# mediation model - prjhacki
if (file.exists(here("Models","fit_med_6"))) {
fit_med.6 <- readRDS("Models/fit_med_6")
} else {
fit_med.6 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "indcrmemocchg",
y = "prjhacki",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_med.6, file = "Models/fit_med_6")
}
# mediation model - prjanyi
if (file.exists(here("Models","fit_med_7"))) {
fit_med.7 <- readRDS("Models/fit_med_7")
} else {
fit_med.7 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "indcrmemocchg",
y = "prjanyi",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_med.7, file = "Models/fit_med_7")
}
pars <- c("a", "b", "cp", "corrab")
# Diagnostic messages
# check_hmc_diagnostics(fit_med.1)
# check_hmc_diagnostics(fit_med.2)
# check_hmc_diagnostics(fit_med.3)
# check_hmc_diagnostics(fit_med.4)
# check_hmc_diagnostics(fit_med.5)
# check_hmc_diagnostics(fit_med.6)
# check_hmc_diagnostics(fit_med.7)12.3.1.1 Theft <5BAM
12.3.1.1.1 Plot parameters (violin)
mlm_pars_plot(fit_med.1,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5))
12.3.1.1.2 Plot effects (pop-avg fitted values)
medplot1 <- mlm_spaghetti_plot(
mod = fit_med.1,
d = stress.long4,
x = "sumstresschg", m = "indcrmemocchg", y = "prjthflt5i", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
medplot1[[1]] + labs(title="Path a (X -> M)"),
medplot1[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.1.1.3 Path diagram
mlm_path_plot(fit_med.1,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(variety)",
ylab = "Theft\n<5BAM")
# grid.echo() #grab grid element (path plot)
# path1 <- grid.grab() #save as grid object12.3.1.1.4 Fit summary
#mediation model summary - prjthftlt5
mlm_summary(fit_med.1)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.78 0.19 0.78 0.41 1.15 7890 1
## 2 b -0.16 0.47 -0.16 -1.11 0.81 442 1
## 3 cp 1.01 1.60 0.98 -2.25 4.37 262 1
## 4 me -0.10 0.61 -0.09 -1.37 1.18 394 1
## 5 c 0.92 1.62 0.88 -2.35 4.27 245 1
## 6 pme -0.15 15.90 0.00 -4.19 3.91 19639 112.3.1.1.5 Trace plots
mcmc_trace(as.data.frame(fit_med.1), pars = pars)
12.3.1.2 Theft >5BAM
12.3.1.2.1 Plot parameters (violin)
mlm_pars_plot(fit_med.2,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5))
12.3.1.2.2 Plot effects (pop-avg fitted values)
medplot2 <- mlm_spaghetti_plot(
mod = fit_med.2,
d = stress.long4,
x = "sumstresschg", m = "indcrmemocchg", y = "prjthfgt5i", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
medplot2[[1]] + labs(title="Path a (X -> M)"),
medplot2[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.1.2.3 Path diagram
mlm_path_plot(fit_med.2,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(variety)",
ylab = "Theft\n>5BAM")
# grid.echo() #grab grid element (path plot)
# path2 <- grid.grab() #save as grid object12.3.1.2.4 Fit summary
#mediation model summary - prjthftgt5
mlm_summary(fit_med.2)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.78 0.19 0.78 0.41 1.16 8857 1
## 2 b -0.23 0.42 -0.22 -1.08 0.60 366 1
## 3 cp 1.43 1.50 1.46 -1.81 4.44 382 1
## 4 me -0.26 0.54 -0.23 -1.46 0.79 311 1
## 5 c 1.17 1.49 1.23 -2.12 4.08 355 1
## 6 pme 0.35 48.87 -0.08 -3.56 3.12 19958 112.3.1.2.5 Trace plots
mcmc_trace(as.data.frame(fit_med.2), pars = pars)
12.3.1.3 Threat
12.3.1.3.1 Plot parameters (violin)
mlm_pars_plot(fit_med.3,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5))
12.3.1.3.2 Plot effects (pop-avg fitted values)
medplot3 <- mlm_spaghetti_plot(
mod = fit_med.3,
d = stress.long4,
x = "sumstresschg", m = "indcrmemocchg", y = "prjthreati", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
medplot3[[1]] + labs(title="Path a (X -> M)"),
medplot3[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.1.3.3 Path diagram
mlm_path_plot(fit_med.3,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(variety)",
ylab = "Threat")
# grid.echo() #grab grid element (path plot)
# path3 <- grid.grab() #save as grid object12.3.1.3.4 Fit summary
#mediation model summary - prjthreat
mlm_summary(fit_med.3)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.77 0.19 0.77 0.41 1.15 10000 1
## 2 b 0.12 0.53 0.12 -0.94 1.19 940 1
## 3 cp -1.10 1.85 -1.18 -4.62 2.91 830 1
## 4 me -0.31 0.66 -0.23 -1.82 0.83 602 1
## 5 c -1.41 1.86 -1.46 -5.08 2.61 795 1
## 6 pme 2.14 208.00 0.10 -2.78 2.79 20008 112.3.1.3.5 Trace plots
mcmc_trace(as.data.frame(fit_med.3), pars = pars)
12.3.1.4 Phys. Harm
12.3.1.4.1 Plot parameters (violin)
mlm_pars_plot(fit_med.4,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.1.4.2 Plot effects (pop-avg fitted values)
medplot4 <- mlm_spaghetti_plot(
mod = fit_med.4,
d = stress.long4,
x = "sumstresschg", m = "indcrmemocchg", y = "prjharmi", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
medplot4[[1]] + labs(title="Path a (X -> M)"),
medplot4[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.1.4.3 Path diagram
mlm_path_plot(fit_med.4,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(variety)",
ylab = "Harm")
# grid.echo() #grab grid element (path plot)
# path4 <- grid.grab() #save as grid object12.3.1.4.4 Fit summary
#mediation model summary - prjharm
mlm_summary(fit_med.4)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.78 0.19 0.78 0.40 1.1 13521 1
## 2 b 0.47 0.55 0.49 -0.68 1.6 1053 1
## 3 cp -1.85 2.11 -1.91 -6.05 2.7 1078 1
## 4 me 0.80 0.70 0.71 -0.38 2.4 698 1
## 5 c -1.05 2.15 -1.14 -5.21 3.6 991 1
## 6 pme 0.28 74.62 -0.19 -6.17 5.5 19888 112.3.1.4.5 Trace plots
mcmc_trace(as.data.frame(fit_med.4), pars = pars)
12.3.1.5 Use drugs
12.3.1.5.1 Plot parameters (violin)
mlm_pars_plot(fit_med.5,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.1.5.2 Plot effects (pop-avg fitted values)
medplot5 <- mlm_spaghetti_plot(
mod = fit_med.5,
d = stress.long4,
x = "sumstresschg", m = "indcrmemocchg", y = "prjusedrgi", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
medplot5[[1]] + labs(title="Path a (X -> M)"),
medplot5[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.1.5.3 Path diagram
mlm_path_plot(fit_med.5,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(variety)",
ylab = "Use\nDrugs")
# grid.echo() #grab grid element (path plot)
# path5 <- grid.grab() #save as grid object12.3.1.5.4 Fit summary
#mediation model summary - prjusedrg
mlm_summary(fit_med.5)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.81 0.19 0.81 0.44 1.19 11849 1
## 2 b 0.40 0.59 0.39 -0.76 1.60 905 1
## 3 cp -0.73 1.87 -0.81 -4.30 3.27 863 1
## 4 me -1.01 0.94 -0.92 -3.09 0.52 936 1
## 5 c -1.75 1.93 -1.74 -5.59 2.17 1065 1
## 6 pme 0.00 39.35 0.37 -4.08 4.82 20101 112.3.1.5.5 Trace plots
mcmc_trace(as.data.frame(fit_med.5), pars = pars)
12.3.1.6 Hack info.
12.3.1.6.1 Plot parameters (violin)
mlm_pars_plot(fit_med.6,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.1.6.2 Plot effects (pop-avg fitted values)
medplot6 <- mlm_spaghetti_plot(
mod = fit_med.6,
d = stress.long4,
x = "sumstresschg", m = "indcrmemocchg", y = "prjhacki", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
medplot6[[1]] + labs(title="Path a (X -> M)"),
medplot6[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.1.6.3 Path diagram
mlm_path_plot(fit_med.6,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(variety)",
ylab = "Hack\nInfo")
# grid.echo() #grab grid element (path plot)
# path6 <- grid.grab() #save as grid object12.3.1.6.4 Fit summary
#mediation model summary - prjhack
mlm_summary(fit_med.6)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.77 0.19 0.77 0.40 1.1 9072 1
## 2 b 0.21 0.43 0.20 -0.64 1.1 1988 1
## 3 cp -0.89 1.12 -0.88 -3.10 1.3 4840 1
## 4 me 0.34 0.67 0.30 -0.97 1.8 446 1
## 5 c -0.55 1.17 -0.56 -2.85 1.8 3420 1
## 6 pme -0.28 81.13 0.00 -7.61 7.4 19993 112.3.1.6.5 Trace plots
mcmc_trace(as.data.frame(fit_med.6), pars = pars)
12.3.1.7 Any crime intent
12.3.1.7.1 Plot parameters (violin)
mlm_pars_plot(fit_med.7,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.1.7.2 Plot effects (pop-avg fitted values)
medplot7 <- mlm_spaghetti_plot(
mod = fit_med.7,
d = stress.long4,
x = "sumstresschg", m = "indcrmemocchg", y = "prjanyi", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
medplot7[[1]] + labs(title="Path a (X -> M)"),
medplot7[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.1.7.3 Path diagram
mlm_path_plot(fit_med.7,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(variety)",
ylab = "Any\nCrime")
# grid.echo() #grab grid element (path plot)
# path7 <- grid.grab() #save as grid object12.3.1.7.4 Fit summary
#mediation model summary - prjany
mlm_summary(fit_med.7)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.79 0.19 0.79 0.42 1.16 8727 1
## 2 b 0.16 0.40 0.16 -0.67 0.96 418 1
## 3 cp 0.63 1.06 0.60 -1.48 2.84 512 1
## 4 me 0.16 0.54 0.15 -0.94 1.30 453 1
## 5 c 0.78 1.09 0.76 -1.38 3.02 335 1
## 6 pme -0.33 74.40 0.17 -4.40 4.89 19996 112.3.1.7.5 Trace plots
mcmc_trace(as.data.frame(fit_med.7), pars = pars)
12.3.1.8 Combine bmlm plots (variety M)
# Combine path plots
# grid.arrange(path1, path2, path3, path4, path5, path6, path7)
# text does not scale properly
#Generate custom X->M & M->Y plots for Fig5
#function to find & drop leading zeroes (used for x-axis label)
dropLeadingZero <- function(l){
str_replace(l, '0(?=.)', '')
}
#Function to generate custom X-M spaghetti plots
# NOTE: same x->m plot for all models
xmplot <- function(){
testdat %>%
ggplot(
aes(y=m_fitted_mean, x=sumstresschg,
ymin=m_fitted_lower, ymax=m_fitted_upper)
) +
geom_line(color="#883E3A") +
geom_ribbon(fill="#883E3A", alpha=.2) +
scale_x_continuous(limits=c(-.51,.51),breaks=c(-.5,.5),
labels = dropLeadingZero) +
scale_y_continuous(limits=c(-.6,.6),breaks=c(-.5,.5),
labels = dropLeadingZero) +
ylab("\U25B3M") +
xlab("\U25B3X") +
theme(axis.title = element_blank(),
axis.text = element_blank())
}
#Function to generate custom M-Y spaghetti plots
medyplot <- function(){
testdat %>%
ggplot(
aes(y=y_fitted_mean, x=indcrmemocchg,
ymin=y_fitted_lower, ymax=y_fitted_upper)
) +
geom_line(color="#883E3A") +
geom_ribbon(fill="#883E3A", alpha=.2) +
scale_x_continuous(limits=c(-.6,.6),breaks=c(-.5,.5),
labels = dropLeadingZero) +
scale_y_continuous(limits=c(-.01,.21),breaks=c(0,.2),
labels = dropLeadingZero) +
ylab("\U25B3Y") +
xlab("\U25B3M") +
theme(axis.title = element_blank(),
axis.text = element_blank())
}
#Generate custom spaghetti plots
#prjthftlt5
testdat <- medplot1[[1]]$layer[[1]]$data
med1xmplot <- xmplot()
testdat <- medplot1[[2]]$layer[[2]]$data
med1myplot <- medyplot()
#prjthftgt5
testdat <- medplot2[[1]]$layer[[1]]$data
med2xmplot <- xmplot()
testdat <- medplot2[[2]]$layer[[2]]$data
med2myplot <- medyplot()
#prjthreat
testdat <- medplot3[[1]]$layer[[1]]$data
med3xmplot <- xmplot()
testdat <- medplot3[[2]]$layer[[2]]$data
med3myplot <- medyplot()
#prjharm
testdat <- medplot4[[1]]$layer[[1]]$data
med4xmplot <- xmplot()
testdat <- medplot4[[2]]$layer[[2]]$data
med4myplot <- medyplot()
#prjusedrg
testdat <- medplot5[[1]]$layer[[1]]$data
med5xmplot <- xmplot()
testdat <- medplot5[[2]]$layer[[2]]$data
med5myplot <- medyplot()
#prjhack
testdat <- medplot6[[1]]$layer[[1]]$data
med6xmplot <- xmplot()
testdat <- medplot6[[2]]$layer[[2]]$data
med6myplot <- medyplot()
#prjany
testdat <- medplot7[[1]]$layer[[1]]$data
med7xmplot <- xmplot()
testdat <- medplot7[[2]]$layer[[2]]$data
med7myplot <- medyplot()
# new approach = merge & facet_grid
medplot1xmdat <- medplot1[[1]]$layer[[1]]$data
medplot1xmdat <- medplot1xmdat %>% mutate(yvar="prjthftlt5", plotvar="xm")
medplot1mydat <- medplot1[[2]]$layer[[2]]$data
medplot1mydat <- medplot1mydat %>% mutate(yvar="prjthftlt5", plotvar="my")
medplot2xmdat <- medplot2[[1]]$layer[[1]]$data
medplot2xmdat <- medplot2xmdat %>% mutate(yvar="prjthftgt5", plotvar="xm")
medplot2mydat <- medplot2[[2]]$layer[[2]]$data
medplot2mydat <- medplot2mydat %>% mutate(yvar="prjthftgt5", plotvar="my")
medplot3xmdat <- medplot3[[1]]$layer[[1]]$data
medplot3xmdat <- medplot3xmdat %>% mutate(yvar="prjthreat", plotvar="xm")
medplot3mydat <- medplot3[[2]]$layer[[2]]$data
medplot3mydat <- medplot3mydat %>% mutate(yvar="prjthreat", plotvar="my")
medplot4xmdat <- medplot4[[1]]$layer[[1]]$data
medplot4xmdat <- medplot4xmdat %>% mutate(yvar="prjharm", plotvar="xm")
medplot4mydat <- medplot4[[2]]$layer[[2]]$data
medplot4mydat <- medplot4mydat %>% mutate(yvar="prjharm", plotvar="my")
medplot5xmdat <- medplot5[[1]]$layer[[1]]$data
medplot5xmdat <- medplot5xmdat %>% mutate(yvar="prjusedrg", plotvar="xm")
medplot5mydat <- medplot5[[2]]$layer[[2]]$data
medplot5mydat <- medplot5mydat %>% mutate(yvar="prjusedrg", plotvar="my")
medplot6xmdat <- medplot6[[1]]$layer[[1]]$data
medplot6xmdat <- medplot6xmdat %>% mutate(yvar="prjhack", plotvar="xm")
medplot6mydat <- medplot6[[2]]$layer[[2]]$data
medplot6mydat <- medplot6mydat %>% mutate(yvar="prjhack", plotvar="my")
medplot7xmdat <- medplot7[[1]]$layer[[1]]$data
medplot7xmdat <- medplot7xmdat %>% mutate(yvar="prjany", plotvar="xm")
medplot7mydat <- medplot7[[2]]$layer[[2]]$data
medplot7mydat <- medplot7mydat %>% mutate(yvar="prjany", plotvar="my")
medplotxmdat <- bind_rows(medplot1xmdat, medplot2xmdat, medplot3xmdat,
medplot4xmdat, medplot5xmdat, medplot6xmdat,
medplot7xmdat) %>%
mutate(yvar=factor(yvar, levels=c("prjthftlt5", "prjthftgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack", "prjany")))
medplotmydat <- bind_rows(medplot1mydat, medplot2mydat, medplot3mydat,
medplot4mydat, medplot5mydat, medplot6mydat,
medplot7mydat) %>%
mutate(yvar=factor(yvar, levels=c("prjthftlt5", "prjthftgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack", "prjany")))12.3.2 Alt Med (X=stdz sum stress change, M=crim emotions sum scale)
As a robustness check, we re-estimated the bmlm within-person medition models above using a standardized sum scale instead of a variety “symptom” index for our mediating criminogenic negative emotions measure. As above, these models generate null within-person mediation effect estimates, due to near-zero associations between negative emotions and criminal intent (i.e., null “b” path or M-Y estimates).
# Within-subjects bayesian mediation models (bmlm)
# X: stress change (sum scale)
# M: criminogenic emotions (sum scale)
# Y: binary criminal intent items
# no built-in file save option, knitr & xfun cache options not working as desired during knitting
# instead, manual saving & loading bmlm models
if (file.exists(here("Models","fit_altmed_1"))) {
fit_altmed.1 <- readRDS("Models/fit_altmed_1")
} else {
# mediation analysis - prjthflt5
fit_altmed.1 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "sumcrmemozchg",
y = "prjthflt5i",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_altmed.1, file = "Models/fit_altmed_1")
}
# mediation model - prjthfgt5
if (file.exists(here("Models","fit_altmed_2"))) {
fit_altmed.2 <- readRDS("Models/fit_altmed_2")
} else {
fit_altmed.2 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "sumcrmemozchg",
y = "prjthfgt5i",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_altmed.2, file = "Models/fit_altmed_2")
}
# mediation model - prjthreati
if (file.exists(here("Models","fit_altmed_3"))) {
fit_altmed.3 <- readRDS("Models/fit_altmed_3")
} else {
fit_altmed.3 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "sumcrmemozchg",
y = "prjthreati",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_altmed.3, file = "Models/fit_altmed_3")
}
# mediation model - prjharmi
if (file.exists(here("Models","fit_altmed_4"))) {
fit_altmed.4 <- readRDS("Models/fit_altmed_4")
} else {
fit_altmed.4 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "sumcrmemozchg",
y = "prjharmi",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_altmed.4, file = "Models/fit_altmed_4")
}
# mediation model - prjusedrgi
if (file.exists(here("Models","fit_altmed_5"))) {
fit_altmed.5 <- readRDS("Models/fit_altmed_5")
} else {
fit_altmed.5 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "sumcrmemozchg",
y = "prjusedrgi",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_altmed.5, file = "Models/fit_altmed_5")
}
# mediation model - prjhacki
if (file.exists(here("Models","fit_altmed_6"))) {
fit_altmed.6 <- readRDS("Models/fit_altmed_6")
} else {
fit_altmed.6 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "sumcrmemozchg",
y = "prjhacki",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_altmed.6, file = "Models/fit_altmed_6")
}
# mediation model - prjanyi
if (file.exists(here("Models","fit_altmed_7"))) {
fit_altmed.7 <- readRDS("Models/fit_altmed_7")
} else {
fit_altmed.7 <- mlm(d = stress.long4,
id = "id",
x = "sumstresschg",
m = "sumcrmemozchg",
y = "prjanyi",
binary_y = TRUE,
priors = list(dy = 2, dm = 2, b = 1),
cores = 4,
iter = 10000,
seed = 8675309)
saveRDS(fit_altmed.7, file = "Models/fit_altmed_7")
}
pars <- c("a", "b", "cp", "corrab")
# Diagnostic messages
# check_hmc_diagnostics(fit_altmed.1)
# check_hmc_diagnostics(fit_altmed.2)
# check_hmc_diagnostics(fit_altmed.3)
# check_hmc_diagnostics(fit_altmed.4)
# check_hmc_diagnostics(fit_altmed.5)
# check_hmc_diagnostics(fit_altmed.6)
# check_hmc_diagnostics(fit_altmed.7)12.3.2.1 Theft <5BAM
12.3.2.1.1 Plot parameters (violin)
mlm_pars_plot(fit_altmed.1,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5))
12.3.2.1.2 Plot effects (pop-avg fitted values)
altmedplot1 <- mlm_spaghetti_plot(
mod = fit_altmed.1,
d = stress.long4,
x = "sumstresschg", m = "sumcrmemozchg", y = "prjthflt5i", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
altmedplot1[[1]] + labs(title="Path a (X -> M)"),
altmedplot1[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.2.1.3 Path diagram
mlm_path_plot(fit_altmed.1,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(sum scale)",
ylab = "Theft\n<5BAM")
# grid.echo() #grab grid element (path plot)
# altpath1 <- grid.grab() #save as grid object12.3.2.1.4 Fit summary
#mediation model summary - prjthftlt5
mlm_summary(fit_altmed.1)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.94 0.19 0.94 0.57 1.32 16282 1
## 2 b -0.10 0.41 -0.09 -0.95 0.69 838 1
## 3 cp 0.73 1.65 0.70 -2.63 4.29 530 1
## 4 me 0.11 0.62 0.07 -1.06 1.47 615 1
## 5 c 0.84 1.66 0.83 -2.56 4.39 526 1
## 6 pme 0.10 27.60 0.08 -3.82 3.92 20010 112.3.2.1.5 Trace plots
mcmc_trace(as.data.frame(fit_altmed.1), pars = pars)
12.3.2.2 Theft >5BAM
12.3.2.2.1 Plot parameters (violin)
mlm_pars_plot(fit_altmed.2,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5))
12.3.2.2.2 Plot effects (pop-avg fitted values)
altmedplot2 <- mlm_spaghetti_plot(
mod = fit_altmed.2,
d = stress.long4,
x = "sumstresschg", m = "sumcrmemozchg", y = "prjthfgt5i", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
altmedplot2[[1]] + labs(title="Path a (X -> M)"),
altmedplot2[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.2.2.3 Path diagram
mlm_path_plot(fit_altmed.2,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(sum scale)",
ylab = "Theft\n>5BAM")
# grid.echo() #grab grid element (path plot)
# altpath2 <- grid.grab() #save as grid object12.3.2.2.4 Fit summary
#mediation model summary - prjthftgt5
mlm_summary(fit_altmed.2)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.95 0.19 0.95 0.56 1.32 17771 1
## 2 b -0.16 0.42 -0.15 -1.05 0.65 646 1
## 3 cp 1.40 1.51 1.40 -1.75 4.49 493 1
## 4 me 0.12 0.63 0.05 -0.98 1.58 500 1
## 5 c 1.52 1.53 1.53 -1.73 4.62 473 1
## 6 pme 0.45 44.49 0.05 -2.15 2.22 19962 112.3.2.2.5 Trace plots
mcmc_trace(as.data.frame(fit_altmed.2), pars = pars)
12.3.2.3 Threat
12.3.2.3.1 Plot parameters (violin)
mlm_pars_plot(fit_altmed.3,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5))
12.3.2.3.2 Plot effects (pop-avg fitted values)
altmedplot3 <- mlm_spaghetti_plot(
mod = fit_altmed.3,
d = stress.long4,
x = "sumstresschg", m = "sumcrmemozchg", y = "prjthreati", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
altmedplot3[[1]] + labs(title="Path a (X -> M)"),
altmedplot3[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.2.3.3 Path diagram
mlm_path_plot(fit_altmed.3,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(sum scale)",
ylab = "Threat")
# grid.echo() #grab grid element (path plot)
# altpath3 <- grid.grab() #save as grid object12.3.2.3.4 Fit summary
#mediation model summary - prjthreat
mlm_summary(fit_altmed.3)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.94 0.19 0.95 0.57 1.3 13804 1
## 2 b 0.03 0.53 0.04 -1.08 1.1 1032 1
## 3 cp -0.97 1.95 -1.09 -4.53 3.4 875 1
## 4 me -0.19 0.79 -0.13 -1.94 1.3 553 1
## 5 c -1.15 1.95 -1.23 -4.85 3.2 916 1
## 6 pme -0.20 46.37 0.07 -3.75 3.6 19961 112.3.2.3.5 Trace plots
mcmc_trace(as.data.frame(fit_altmed.3), pars = pars)
12.3.2.4 Phys. Harm
12.3.2.4.1 Plot parameters (violin)
mlm_pars_plot(fit_altmed.4,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.2.4.2 Plot effects (pop-avg fitted values)
altmedplot4 <- mlm_spaghetti_plot(
mod = fit_altmed.4,
d = stress.long4,
x = "sumstresschg", m = "sumcrmemozchg", y = "prjharmi", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
altmedplot4[[1]] + labs(title="Path a (X -> M)"),
altmedplot4[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.2.4.3 Path diagram
mlm_path_plot(fit_altmed.4,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(sum scale)",
ylab = "Harm")
# grid.echo() #grab grid element (path plot)
# altpath4 <- grid.grab() #save as grid object12.3.2.4.4 Fit summary
#mediation model summary - prjharm
mlm_summary(fit_altmed.4)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.94 0.19 0.94 0.57 1.3 13056 1
## 2 b 0.19 0.61 0.18 -1.02 1.4 1512 1
## 3 cp -1.72 2.09 -1.78 -5.78 2.8 1013 1
## 4 me 0.88 1.06 0.76 -0.98 3.3 439 1
## 5 c -0.83 2.17 -0.94 -4.99 3.9 1021 1
## 6 pme 0.19 81.29 -0.10 -7.48 7.7 19993 112.3.2.4.5 Trace plots
mcmc_trace(as.data.frame(fit_altmed.4), pars = pars)
12.3.2.5 Use drugs
12.3.2.5.1 Plot parameters (violin)
mlm_pars_plot(fit_altmed.5,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.2.5.2 Plot effects (pop-avg fitted values)
altmedplot5 <- mlm_spaghetti_plot(
mod = fit_altmed.5,
d = stress.long4,
x = "sumstresschg", m = "sumcrmemozchg", y = "prjusedrgi", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
altmedplot5[[1]] + labs(title="Path a (X -> M)"),
altmedplot5[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.2.5.3 Path diagram
mlm_path_plot(fit_altmed.5,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(sum scale)",
ylab = "Use\nDrugs")
# grid.echo() #grab grid element (path plot)
# altpath5 <- grid.grab() #save as grid object12.3.2.5.4 Fit summary
#mediation model summary - prjusedrg
mlm_summary(fit_altmed.5)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.95 0.19 0.95 0.58 1.3 15503 1
## 2 b 0.27 0.50 0.27 -0.73 1.3 1476 1
## 3 cp -0.71 1.76 -0.81 -3.99 3.1 1644 1
## 4 me -0.16 0.80 -0.06 -2.02 1.2 849 1
## 5 c -0.88 1.80 -0.92 -4.35 2.9 1473 1
## 6 pme -4.96 743.73 0.08 -3.88 4.2 20008 112.3.2.5.5 Trace plots
mcmc_trace(as.data.frame(fit_altmed.5), pars = pars)
12.3.2.6 Hack info.
12.3.2.6.1 Plot parameters (violin)
mlm_pars_plot(fit_altmed.6,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.2.6.2 Plot effects (pop-avg fitted values)
altmedplot6 <- mlm_spaghetti_plot(
mod = fit_altmed.6,
d = stress.long4,
x = "sumstresschg", m = "sumcrmemozchg", y = "prjhacki", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
altmedplot6[[1]] + labs(title="Path a (X -> M)"),
altmedplot6[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.2.6.3 Path diagram
mlm_path_plot(fit_altmed.6,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(sum scale)",
ylab = "Hack\nInfo")
# grid.echo() #grab grid element (path plot)
# altpath6 <- grid.grab() #save as grid object12.3.2.6.4 Fit summary
#altmediation model summary - prjhack
mlm_summary(fit_altmed.6)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.93 0.19 0.93 0.56 1.31 10304 1
## 2 b 0.14 0.42 0.14 -0.72 0.98 2489 1
## 3 cp -1.08 1.16 -1.08 -3.36 1.20 3699 1
## 4 me 0.89 0.89 0.80 -0.62 2.82 314 1
## 5 c -0.19 1.22 -0.23 -2.49 2.33 1689 1
## 6 pme -0.19 150.21 0.03 -12.99 13.50 19767 112.3.2.6.5 Trace plots
mcmc_trace(as.data.frame(fit_altmed.6), pars = pars)
12.3.2.7 Any crime intent
12.3.2.7.1 Plot parameters (violin)
mlm_pars_plot(fit_altmed.7,
type = "violin",
pars = c("a", "b", "cp", "c", "me")) +
scale_y_continuous(breaks = seq(-50, 10, 5)) 
12.3.2.7.2 Plot effects (pop-avg fitted values)
altmedplot7 <- mlm_spaghetti_plot(
mod = fit_altmed.7,
d = stress.long4,
x = "sumstresschg", m = "sumcrmemozchg", y = "prjanyi", id = "id",
fixed = TRUE, random = FALSE, binary_y = TRUE, n = 20)
grid.arrange(
altmedplot7[[1]] + labs(title="Path a (X -> M)"),
altmedplot7[[2]] + labs(title="Path b (M -> Y)") +
coord_cartesian(ylim=c(0,.05)),
nrow=1)
12.3.2.7.3 Path diagram
mlm_path_plot(fit_altmed.7,
xlab = "Stress\nChange",
mlab = "Criminogenic\nEmotions\n(sum scale)",
ylab = "Any\nCrime")
# grid.echo() #grab grid element (path plot)
# altpath7 <- grid.grab() #save as grid object12.3.2.7.4 Fit summary
#mediation model summary - prjany
mlm_summary(fit_altmed.7)## Parameter Mean SE Median 2.5% 97.5% n_eff Rhat
## 1 a 0.95 0.19 0.95 0.57 1.32 14832 1
## 2 b 0.05 0.37 0.06 -0.74 0.76 563 1
## 3 cp 0.46 1.17 0.41 -1.75 3.06 228 1
## 4 me 0.65 0.67 0.56 -0.46 2.14 771 1
## 5 c 1.10 1.19 1.05 -1.14 3.71 225 1
## 6 pme 5.12 684.97 0.40 -4.09 5.17 20007 112.3.2.7.5 Trace plots
mcmc_trace(as.data.frame(fit_altmed.7), pars = pars)
12.4 FIGURE 5: Mediation
#wrangle mediation output - add yvar for merging & faceting
gen_meddata <- function(modname, yvarname){
tibble(modname %>%
mutate(
yvar = yvarname
)
)}
#create a dummy indicator ==1 if lower 80% interval > 0
ci80gt0 <- function(modname, xvar, mvar){
tibble(mediation(modname,
treatment=xvar,
mediator=mvar,
ci=.80)) %>%
mutate(
ci80gt0 = if_else(CI_low > 0, 1, 0)
) %>%
dplyr::select(Effect, ci80gt0)
}
medy1w <- gen_meddata(medy1w, "prjthflt5")
tempdat <- ci80gt0(m1w, "sumstresschg", "indcrmemocchg")
medy1w <- left_join(medy1w, tempdat)
medy2w <- gen_meddata(medy2w, "prjthfgt5")
tempdat <- ci80gt0(m2w, "sumstresschg", "indcrmemocchg")
medy2w <- left_join(medy2w, tempdat)
medy3w <- gen_meddata(medy3w, "prjthreat")
tempdat <- ci80gt0(m3w, "sumstresschg", "indcrmemocchg")
medy3w <- left_join(medy3w, tempdat)
medy4w <- gen_meddata(medy4w, "prjharm")
tempdat <- ci80gt0(m4w, "sumstresschg", "indcrmemocchg")
medy4w <- left_join(medy4w, tempdat)
medy5w <- gen_meddata(medy5w, "prjusedrg")
tempdat <- ci80gt0(m5w, "sumstresschg", "indcrmemocchg")
medy5w <- left_join(medy5w, tempdat)
medy6w <- gen_meddata(medy6w, "prjhack")
tempdat <- ci80gt0(m6w, "sumstresschg", "indcrmemocchg")
medy6w <- left_join(medy6w, tempdat)
medy7w <- gen_meddata(medy7w, "prjany")
tempdat <- ci80gt0(m6w, "sumstresschg", "indcrmemocchg")
medy7w <- left_join(medy7w, tempdat)
medy1b <- gen_meddata(medy1b, "prjthflt5")
tempdat <- ci80gt0(m1b, "sumstressav", "indcrmemocav")
medy1b <- left_join(medy1b, tempdat)
medy2b <- gen_meddata(medy2b, "prjthfgt5")
tempdat <- ci80gt0(m2b, "sumstressav", "indcrmemocav")
medy2b <- left_join(medy2b, tempdat)
medy3b <- gen_meddata(medy3b, "prjthreat")
tempdat <- ci80gt0(m3b, "sumstressav", "indcrmemocav")
medy3b <- left_join(medy3b, tempdat)
medy4b <- gen_meddata(medy4b, "prjharm")
tempdat <- ci80gt0(m4b, "sumstressav", "indcrmemocav")
medy4b <- left_join(medy4b, tempdat)
medy5b <- gen_meddata(medy5b, "prjusedrg")
tempdat <- ci80gt0(m5b, "sumstressav", "indcrmemocav")
medy5b <- left_join(medy5b, tempdat)
medy6b <- gen_meddata(medy6b, "prjhack")
tempdat <- ci80gt0(m6b, "sumstressav", "indcrmemocav")
medy6b <- left_join(medy6b, tempdat)
medy7b <- gen_meddata(medy7b, "prjany")
tempdat <- ci80gt0(m6b, "sumstressav", "indcrmemocav")
medy7b <- left_join(medy7b, tempdat)
meddataw <- bind_rows(medy1w, medy2w, medy3w, medy4w, medy5w, medy6w, medy7w) %>%
mutate(
yvar = factor(yvar, ordered=TRUE,
levels = c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack", "prjany")),
ci80gt0 = factor(ci80gt0, levels=c(0,1)),
Effect = factor(Effect, levels=c("Direct Effect (ADE)", "Mediator Effect",
"Indirect Effect (ACME)", "Total Effect",
"Proportion Mediated")),
method = "within"
)
meddatab <- bind_rows(medy1b, medy2b, medy3b, medy4b, medy5b, medy6b, medy7b) %>%
mutate(
yvar = factor(yvar, ordered=TRUE,
levels = c("prjthflt5", "prjthfgt5", "prjthreat",
"prjharm", "prjusedrg", "prjhack", "prjany")),
ci80gt0 = factor(ci80gt0, levels=c(0,1)),
Effect = factor(Effect, levels=c("Direct Effect (ADE)", "Mediator Effect",
"Indirect Effect (ACME)", "Total Effect",
"Proportion Mediated")),
method = "between"
)
meddata <- bind_rows(meddataw, meddatab)
prjlabs3 <- c(
"prjthflt5"="Theft\n<5BAM",
"prjthfgt5"="Theft\n>5BAM",
"prjthreat"="Threaten",
"prjharm"="Phys.\nharm",
"prjusedrg"="Use\ndrugs",
"prjhack"="Hack\ninfo",
"prjany"="Any\ncrime")
effectlabs <- c(
"Direct Effect (ADE)"="Direct\nEffect\n(ADE)\nX\U2192Y",
"Mediator Effect"=" \nMediator\nEffect\nM\U2192Y",
"Indirect Effect (ACME)"="Indirect\nEffect\n(ACME)\nX\U2192M\U2192Y",
"Total Effect"="Total Effect\nX\U2192Y\n+\nX\U2192M\U2192Y",
"Proportion Mediated"=" \nProportion\nMediated\n(Ind/Total)")
methodlabs2 <- c(
"between"="Between-Person Difference (Cross-Time Mean) Estimator",
"within"="Within-Person Change (T2-T1) \"Fixed Effects\" Estimator")
# Effect types have different scales, so changed to plot faceted by Effect
# ggplot(meddataw,
# aes(y=reorder(Effect, desc(Effect)), x=Estimate)) +
# facet_wrap(~yvar, nrow=1,
# scales = "free_x",
# labeller = labeller(yvar = as_labeller(prjlabs2))) +
# geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
# geom_pointrange(data=meddataw,
# aes(x=Estimate, xmin = CI_low, xmax = CI_high, alpha=ci80gt0),
# color="#883E3A", fill="#883E3A",
# position = position_nudge(y=-.1)) +
# geom_pointrange(data=meddatab,
# aes(x=Estimate, xmin = CI_low, xmax = CI_high), color="#E99D53", fill="#E99D53",
# position = position_nudge(y=.1)) +
# scale_y_discrete(labels=effectlabs) +
# scale_alpha_discrete(range=c(.2,1), guide = "none")
Fig5v1 <- ggplot(meddataw,
aes(y=reorder(yvar, desc(yvar)), x=Estimate, color=method)) +
facet_wrap(~Effect, nrow=1,
scales = "free_x",
labeller = labeller(Effect = as_labeller(effectlabs))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
geom_pointrangeh(data=meddataw, # use ggstance::geom_pointrangeh() for horiz lines in legend
aes(x=Estimate, xmin = CI_low, xmax = CI_high, alpha=ci80gt0),
# color="#883E3A", fill="#883E3A",
position = position_nudge(y=-.1)) +
geom_pointrangeh(data=meddatab,
aes(x=Estimate, xmin = CI_low, xmax = CI_high, alpha=ci80gt0),
# color="#E99D53", fill="#E99D53",
position = position_nudge(y=.1)) +
scale_y_discrete(labels=prjlabs3) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
scale_color_manual(values=c("#E99D53","#883E3A"),
labels=as_labeller(methodlabs2), name=NULL) +
xlab(element_blank()) +
facetted_pos_scales(
x = list(
Effect == "Proportion Mediated" ~
scale_x_continuous(breaks=c(-1,-.5,0,.5,1),
limits = c(-1.1,1.1), labels = dropLeadingZero)
# Effect %in% c("depsymw1", "negemow1") ~
# scale_x_continuous(breaks=c(-.05,0,.05,.1, .15),
# limits = c(-.06,.16), labels = dropLeadingZero)
) ) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
strip.background = element_blank(),
strip.text.x = element_text(size=8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)),
color = guide_legend(nrow=1, reverse = FALSE))
design <- "
11
22
"
SuppFigure5 <- Fig5v1 + guide_area() +
plot_layout(design=design, guides = 'collect', heights=c(30,.1)) +
plot_annotation(
title = 'SUPPLEMENTAL FIGURE 5\nEstimated Total, Direct, & Indirect Effects of Stress on Criminal Intent via Criminogenic Emotions, by Estimator',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. Estimates derived from 14 multivariate and multilevel between-within Bayesian regression models simultaneously regressing (using `brms`) each binary criminal intent outcome (six binary logistic models) and either a between-person or a within-person mediator model (two Gaussian models) predicting differences or changes in a variety index of the number of criminogenic emotions reported. All models included a standardized sum stress scale separated into L2 cross-time average (Xbar_i) between-person and L1 within-person change (X_it - Xbar_i) /"fixed effects"/ estimators. The outcome (criminal intent) models also included both L2 (between) and L1 (change) measures of criminogenic emotions. Mediation estimates reflect untransformed beta parameter estimates derived using `bayestestR::mediation()` R package. Median posterior density estimates with 95% equal-tailed (ETI) credible intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates are greater than zero.', width=195)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'bottom',
legend.key.height = unit(.5, 'cm'),
legend.margin = margin(0,0,0,0),
legend.spacing.y = unit(0, "mm"))
SuppFigure5
# New version with BMLM plots
meddatawsub <- meddataw %>%
filter(Effect != "Proportion Mediated") %>%
droplevels()
meddatabsub <- meddatab %>%
filter(Effect != "Proportion Mediated") %>%
droplevels()
effectlabs2 <- c(
"Direct Effect (ADE)"="mediation()\nDirect Effect\n\"ADE\"\n(X\U2192Y)",
"Mediator Effect"=" mediation()\nMediator\nEffect\n(M\U2192Y)",
"Indirect Effect (ACME)"="mediation()\nIndirect Effect\n\"ACME\"\n(X\U2192M\U2192Y)",
"Total Effect"="mediation()\nTotal Effect\n(X\U2192Y) +\n(X\U2192M\U2192Y)")
Fig5 <- ggplot(meddatawsub,
aes(y=reorder(yvar, desc(yvar)), x=Estimate, color=method)) +
facet_wrap(~Effect, nrow=1,
scales = "free_x",
labeller = labeller(Effect = as_labeller(effectlabs2))) +
geom_vline(xintercept = 0, linetype = "dashed", size=.5, alpha=.4) +
geom_pointrangeh(data=meddatawsub, # use ggstance::geom_pointrangeh() for horiz lines in legend
aes(x=Estimate, xmin = CI_low, xmax = CI_high, alpha=ci80gt0),
# color="#883E3A", fill="#883E3A",
position = position_nudge(y=-.1)) +
geom_pointrangeh(data=meddatabsub,
aes(x=Estimate, xmin = CI_low, xmax = CI_high, alpha=ci80gt0),
# color="#E99D53", fill="#E99D53",
position = position_nudge(y=.1)) +
scale_y_discrete(labels=prjlabs3) +
scale_alpha_discrete(range=c(.2,1), guide = "none") +
scale_color_manual(values=c("#E99D53","#883E3A"),
labels=as_labeller(methodlabs2), name=NULL) +
xlab(element_blank()) +
theme(axis.title.y = element_blank(),
legend.position = "bottom",
strip.background = element_blank(),
strip.text.x = element_text(size=8),
axis.text.y = element_text(size=8),
plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0),
legend.text = element_text(size = 8)) +
guides(shape = guide_legend(override.aes = list(size = 0.5)),
color = guide_legend(nrow=1, reverse = FALSE))
plotvarlabs<- c(
"xm"="bmlm()\nWithin-\nPerson\n(\U25B3X\U2192\U25B3M)",
"my"="bmlm()\nWithin-\nPerson\n(\U25B3M\U2192\U25B3Y)")
#Custom X-M spaghetti plots
Fig5.2 <- medplotxmdat %>%
ggplot(
aes(y=m_fitted_mean, x=sumstresschg,
ymin=m_fitted_lower, ymax=m_fitted_upper)
) +
facet_grid(yvar~plotvar,
labeller = labeller(plotvar = as_labeller(plotvarlabs))) +
geom_line(color="#883E3A") +
geom_ribbon(fill="#883E3A", alpha=.3) +
scale_x_continuous(breaks=c(-.5,.5),
labels = dropLeadingZero) +
scale_y_continuous(breaks=c(-.5,.5),
labels = dropLeadingZero) +
# ylab("\U25B3M") +
# xlab("\U25B3X") +
theme(
legend.position = "none",
strip.text.y = element_blank(),
strip.text.x = element_text(size=8),
axis.title = element_blank(),
axis.text = element_text(size=8))
#Custom M-Y spaghetti plots
Fig5.3 <- medplotmydat %>%
ggplot(
aes(y=y_fitted_mean, x=indcrmemocchg,
ymin=y_fitted_lower, ymax=y_fitted_upper)
) +
facet_grid(yvar~plotvar, scales="free_y",
labeller = labeller(plotvar = as_labeller(plotvarlabs))) +
geom_line(color="#883E3A") +
geom_ribbon(fill="#883E3A", alpha=.3) +
scale_x_continuous(breaks=c(-.5,.5),
labels = dropLeadingZero) +
# scale_y_continuous(breaks=c(0,.02),
# labels = dropLeadingZero) +
# coord_cartesian(xlim=c(-.71,.71), expand=FALSE) +
facetted_pos_scales(
y = list(
yvar == "prjany" ~
scale_y_continuous(breaks=c(0,.04),
limits = c(-.001,.05), labels = dropLeadingZero),
yvar != "prjany" ~
scale_y_continuous(breaks=c(0,.02),
limits = c(-.001,.035), labels = dropLeadingZero)
) ) +
# ylab("\U25B3Y") +
# xlab("\U25B3M") +
theme(
legend.position = "none",
strip.text.y = element_blank(),
strip.text.x = element_text(size=8),
axis.title = element_blank(),
axis.text = element_text(size=8))
design <- "
123
444
"
Figure5 <- Fig5 + Fig5.2 + Fig5.3 + guide_area() +
plot_layout(design=design, guides="collect",
widths=c(4,.5,1),
heights=c(40,.1)) +
plot_annotation(
title = 'FIGURE 5\nEstimated Total, Direct, & Indirect Effects of Stress on Criminal Intent via Criminogenic Emotions, by Estimator',
#subtitle = 'Subtitle here',
caption = str_wrap('Note: N=489 respondents participating at both survey waves. COLUMNS 1-4: Multilevel b/w Bayesian models predicted binary criminal intent outcomes (7 logistic "Y" models) and a between- or a within-person mediator (2 Gaussian "M" models). All models included a standardized sum stress scale separated into L2 between-person (Xbar_i) and L1 within-person change (X_it - Xbar_i) /"fixed effects"/ estimators ("X"). "Y" models also included both L2 (between) and L1 (change) measures of differences/changes in the number of criminogenic emotions reported. Untransformed posterior mediation estimates generated from fitted models using `bayestestR::mediation()` R package. Median posterior density estimates with 95% equal-tailed (ETI) credible intervals displayed. Bold point-intervals indicate at least 80% of posterior estimates are greater than zero. COLUMNS 5-6: Comparable within-person mediation models were fit using `bmlm::mlm()` R package to generate model-implied posterior effect estimates averaged over random effects for plotting in original item metrics. Column 5 (X\U2192M) displays estimated effect of changes in X (X=-.5 to X=.5 equates to 1 SD unit increase in stress) on changes in M (M=-.5 to M=.5 = equates to increase of 1 additional criminogenic emotion reported). Column 6 (M\U2192Y) displays estimated effect of changes in M on changes in the probability of Y (e.g., increase from Y=0 at M=-.5 to Y=0.02 at M=.5 would imply an increase of one reported emotion causes a 2-percentage-point increase in the probability of criminal intent). These estimates also reflect any indirect effect of X on Y through M, while x-axis range displays model-implied degree of change in M (emotions) caused by change in X (stress).', width=190)) &
theme(plot.title = element_text(size=10, face="bold"),
plot.caption = element_text(size=8, hjust = 0), #move caption to left of plot
legend.position = 'bottom',
legend.key.height = unit(.5, 'cm'),
legend.margin = margin(0,0,0,0),
legend.spacing.y = unit(0, "mm"))
Figure5
ggsave("Figure5.jpeg", width=9, height=6.5, path=here("Output"))13 Directed Acyclic Graphs (DAGS)
(RMD FILE: BDK_2023_Stress_12_DAG)
Our manuscript includes DAGs that communicate the simplistic causal assumptions underlying our analysis. You can think of them as communicating these as necessary beliefs if one wishes to make causal inferences from results that appear to support theoretical expectations. To be clear, I do not think these DAGs truly represent the complex causal processes generating our data. Rather, they represent a starting point for analysis and interpretation. We can improve future research by interrogating both our causal assumptions - these DAGs - and our results, identifying what we deem to be potential serious issues with the assumptions or design, and then develop more convincing DAGs and subsequent research designs intended to better identify (if possible) posited causal effects of interest.
13.1 Load Packages
13.2 DAGs
# DAG 1 (RQ2A)
DAG1 <- dagify(
CrimIntent ~ Stress + Indiv + C,
Stress ~ Indiv + C,
D ~ Stress + CrimIntent ,
exposure = "Stress",
outcome = "CrimIntent",
coords=list(
x=c(Stress=1, Indiv=1, CrimIntent=2, C=2, D=2),
y=c(Stress=0, Indiv=1, CrimIntent=0, C=-1, D=1)
)) %>% tidy_dagitty() %>%
dplyr::mutate(confound = if_else(name == "D",
"grey", "darkslategrey"),
confound = if_else(name %in% c("Indiv", "C"),
"maroon", confound)
)
#function to shorten arrows - set percentage to shorten
DAG1p <- shorten_dag_arrows(DAG1, 0.2)
#create factor variable to isolate edge of interest, permits specifying edge color
testdat <- DAG1p %>% dplyr::mutate(
myedge1 = if_else(DAG1p$data$name == "Indiv" |
DAG1p$data$name == "C",
"unblocked", "focal"),
myedge1 = if_else(DAG1p$data$to == "D",
"blocked", myedge1),
modlinetype = ifelse(myedge1 == "unblocked", "solid", "dashed")
)
DAG1p3 <- testdat %>% ggplot(aes(x=x, y=y, xend=xend, yend=yend)) +
geom_dag_edges(aes(x = xstart, y = ystart,
edge_color=myedge1,
edge_linetype = modlinetype), show.legend = FALSE) +
geom_dag_text(label=c("Time-varying\nconfounders",
"Criminal Intent\n(T1)",
"Colliders\n(common effects)",
"Time-stable\nconfounders",
"Stress\n(T1)"),
aes(color = confound), size=4) +
theme_dag() +
guides(fill = 'none', color = 'none', edge_color = 'none') +
scale_color_manual(values = c("darkslategrey", "grey", "maroon")) +
scale_edge_colour_manual(values=c("grey", "darkslategrey","maroon")) +
scale_x_continuous(expand=expansion(mult=c(0.3,0.3))) +
scale_y_continuous(expand=expansion(mult=c(0.3,0.3))) +
# ggtitle("DAG 1 (RQ2A: Between-person)") +
theme(plot.title = element_text(size = 10))
################
# DAG 2 (RQ2B)
DAG2 <- dagify(
CrimIntent ~ Stress + Indiv + C,
Stress ~ Indiv + C,
D ~ Stress + CrimIntent,
exposure = "Stress",
outcome = "CrimIntent",
coords=list(
x=c(Stress=1, Indiv=1, CrimIntent=2, C=2, D=2),
y=c(Stress=0, Indiv=1, CrimIntent=0, C=-1, D=1)
)) %>% tidy_dagitty() %>%
dplyr::mutate(confound = if_else(name %in% c("Indiv","D"),
"grey", "darkslategrey"),
confound = if_else(name == "C",
"maroon", confound)
)
#function to shorten arrows - set percentage to shorten
DAG2p <- shorten_dag_arrows(DAG2, 0.2)
#create factor variable to isolate edge of interest, permits specifying edge color
testdat <- DAG2p %>% dplyr::mutate(
myedge1 = if_else(DAG2p$data$name == "Indiv" |
DAG1p$data$to == "D",
"blocked", "focal"),
myedge1 = if_else(DAG2p$data$name == "C", "unblocked", myedge1),
modlinetype = ifelse(myedge1 == "unblocked", "solid", "dashed")
)
DAG2p3 <- testdat %>% ggplot(aes(x=x, y=y, xend=xend, yend=yend)) +
geom_dag_edges(aes(x = xstart, y = ystart,
edge_color=myedge1,
edge_linetype = modlinetype), show.legend = FALSE) +
geom_dag_text(label=c("Time-varying\nconfounders",
"Criminal Intent\n(change)",
"Colliders\n(time-varying\ncommon effects)",
"Individual\n(time-stable\nconfounders)",
"Stress\n(change)"),
aes(color = confound), size=4) +
theme_dag() +
guides(fill = 'none', color = 'none', edge_color = 'none') +
scale_color_manual(values = c("darkslategrey", "grey", "maroon")) +
scale_edge_colour_manual(values=c("grey", "darkslategrey", "maroon")) +
scale_x_continuous(expand=expansion(mult=c(0.3,0.3))) +
scale_y_continuous(expand=expansion(mult=c(0.3,0.3))) +
# ggtitle("DAG 2 (RQ2B: Within-person)") +
theme(plot.title = element_text(size = 10))
################
# DAG 3 (RQ3)
DAG3 <- dagify(
CrimIntent ~ Stress + Indiv + Community + C,
Stress ~ Indiv + Community + C,
D ~ Stress + CrimIntent,
exposure = "Stress",
outcome = "CrimIntent",
coords=list(
x=c(Stress=1, Community=1, Indiv=1, CrimIntent=2, C=2, D=2),
y=c(Stress=0, Community=-1, Indiv=1, CrimIntent=0, C=-1, D=1)
)) %>% tidy_dagitty() %>%
dplyr::mutate(confound = if_else(name %in% c("Indiv","Community", "D"),
"grey", "darkslategrey"),
confound = if_else(name == "C",
"maroon", confound)
)
#function to shorten arrows - set percentage to shorten
DAG3p <- shorten_dag_arrows(DAG3, 0.2)
#create factor variable to isolate edge of interest, permits specifying edge color
testdat <- DAG3p %>% dplyr::mutate(
myedge1 = if_else(DAG3p$data$name %in% c("Indiv","Community") |
DAG3p$data$to == "D" ,
"blocked", "focal"),
myedge1 = if_else(DAG3p$data$name == "C", "unblocked", myedge1),
modlinetype = ifelse(myedge1 == "unblocked", "solid", "dashed")
)
DAG3p3 <- testdat %>% ggplot(aes(x=x, y=y, xend=xend, yend=yend)) +
geom_dag_edges(aes(x = xstart, y = ystart,
edge_color=myedge1,
edge_linetype = modlinetype), show.legend = FALSE) +
geom_dag_text(label=c("Time-varying\nconfounders",
"Community\n(Rurality/SES)",
"Criminal Intent\n(change)",
"Colliders\n(time-varying\ncommon effects)",
"Individual\n(time-stable\nconfounders)",
"Stress\n(change)"),
aes(color=confound), size=4) +
theme_dag() +
guides(fill = 'none', color = 'none', edge_color = 'none') +
scale_color_manual(values = c("darkslategrey","grey", "maroon")) +
scale_edge_colour_manual(values=c("grey", "darkslategrey", "maroon")) +
scale_x_continuous(expand=expansion(mult=c(0.3,0.3))) +
scale_y_continuous(expand=expansion(mult=c(0.3,0.3))) +
# ggtitle("DAG 3 (RQ3: Effect heterogeneity)") +
theme(plot.title = element_text(size = 10))
################
# DAG 4 (RQ4)
DAG4 <- dagify(
CrimIntent ~ Stress + Indiv + CrimEmots + C,
CrimEmots ~ Stress + C,
Stress ~ Indiv + C,
D ~ Stress + CrimEmots + CrimIntent,
exposure = "Stress",
outcome = "CrimIntent",
coords=list(
x=c(Stress=1, CrimEmots=1, Indiv=1, CrimIntent=2, C=2, D=2),
y=c(Stress=0, CrimEmots=-1, Indiv=1, CrimIntent=0, C=-1, D=1)
)) %>% tidy_dagitty() %>%
dplyr::mutate(confound = if_else(name %in% c("Indiv", "D"),
"grey", "darkslategrey"),
confound = if_else(name == "C",
"maroon", confound)
)
#function to shorten arrows - set percentage to shorten
DAG4p <- shorten_dag_arrows(DAG4, 0.2)
#create factor variable to isolate edge of interest, permits specifying edge color
testdat <- DAG4p %>% dplyr::mutate(
myedge1 = if_else(DAG4p$data$name == "Indiv" |
DAG4p$data$to == "D" ,
"blocked", "focal"),
myedge1 = if_else(DAG4p$data$name == "C", "unblocked", myedge1),
modlinetype = ifelse(myedge1 == "unblocked", "solid", "dashed")
)
DAG4p3 <- testdat %>% ggplot(aes(x=x, y=y, xend=xend, yend=yend)) +
geom_dag_edges(aes(x = xstart, y = ystart,
edge_color=myedge1,
edge_linetype = modlinetype), show.legend = FALSE) +
geom_dag_text(label=c("Time-varying\nconfounders",
"Criminogenic\nNeg. Emotions\n(change)",
"Criminal Intent\n(change)",
"Colliders\n(time-varying\ncommon effects)",
"Individual\n(time-stable\nconfounders)",
"Stress\n(change)"),
aes(color=confound), size=4) +
theme_dag() +
guides(fill = 'none', color = 'none', edge_color = 'none') +
scale_color_manual(values = c("darkslategrey","grey", "maroon")) +
scale_edge_colour_manual(values=c("grey", "darkslategrey", "maroon")) +
scale_x_continuous(expand=expansion(mult=c(0.3,0.3))) +
scale_y_continuous(expand=expansion(mult=c(0.3,0.3))) +
# ggtitle("DAG 4 (RQ4: Mediation)") +
theme(plot.title = element_text(size = 10))# DAG1p3
#
# DAG2p3
#
# DAG3p3
#
# DAG4p3
# ggsave("DAG1p3.jpg", DAG1p3)
# ggsave("DAG2p3.jpg", DAG2p3)
# ggsave("DAG3p3.jpg", DAG3p3)
# ggsave("DAG4p3.jpg", DAG4p3)(DAG1p3 + DAG2p3) / (DAG3p3 + DAG4p3)
In these figures, the solid dark slate paths represent focal effects of interest. Maroon dashed arrows represent unblocked backdoor paths through which potentially biasing information might affect our estimates. The solid light grey arrows represent paths blocked by design or analysis. For instance, the within-person “fixed effects” estimate from “between-within” multilevel modeling designs adjust for time-stable confounders. I included colliders to note that they are adjusted by default without explicit stratification, whereas inclusion of covariates (“control variables”) without strong causal understanding of the underlying data-generating process (or selection in design/sampling phases) could open these backdoor paths and introduce collider bias.
The top row of DAGs communicate the simplistic causal assumptions underlying the between-person (top left) and within-person change (top right) estimates from our earlier models (e.g., see Figure 3 in supplement). In particular, comparison of these top two DAGs communicates the within-person change estimates’ adjustment by design for all “Individual” - measured or unmeasured - sources of time-stable, between-person confounding. Shout-out to Huntington-Klein here and here for inspiring this way of representing fixed effects in DAGs. From these DAGs, it should be clearer that the unadjusted between-person estimates are at greater risk of transmitting biasing information from these varied individual-specific sources of confounding. In the bottom row, explicit stratification on community blocks confounding (between-person; within-person already blocked) and permits assessment of effect heterogeneity across communities (RQ3), and “criminogenic” emotions are a posited mediator of the stress-crime relationship (RQ4).