2024USElectionStudy_Trust+Democracy_emotions
2025-08-07
Prep
Libraries & Data
library(tidyverse)
library(sandwich)
library(gtsummary)
library(flextable)
library(tables)
library(psych)
library(ggplot2)
library(openxlsx)
library(lmerTest)
library(lme4)
library(knitr)
library(hrbrthemes)
library(interactions)
library(sjPlot)
library(cowplot)
d <- read.csv("/Users/af13/Documents/Research/24_ElectionStudy/data + analyses/2024ElectionStudy_combined_cleancases_full.csv", na.strings = c("","NA"), header = T, fileEncoding="latin1")mcSummary() function
Variable Construction
Single-measures
# AI Risk
# numeracy
d <- d %>%
group_by(pid) %>%
fill(numeracy, .direction = "downup") %>%
ungroup()
addmargins(table(d$numeracy,d$wave, exclude = F))
# trustGovt
d[d$wave == 1 | d$wave == 2,] <- d[d$wave == 1 | d$wave == 2,] %>%
group_by(pid) %>%
fill(trustGovt1, .direction = "downup") %>%
ungroup()
addmargins(table(d$trustGovt1,d$wave, exclude = F))
d[d$wave == 1 | d$wave == 2,] <- d[d$wave == 1 | d$wave == 2,] %>%
group_by(pid) %>%
fill(trustGovt2, .direction = "downup") %>%
ungroup()
addmargins(table(d$trustGovt2,d$wave, exclude = F))
d$trustGovt <- rowMeans(d[,c("trustGovt1","trustGovt2")], na.rm = T)
d$trustGovt.c <- d$trustGovt - mean(d$trustGovt, na.rm = T)
# trustSci
psych::alpha(d[ ,c("trustSci_1","trustSci_2","trustSci_3","trustSci_4")])
d <- d %>%
group_by(pid) %>%
fill(trustSci_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$trustSci_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(trustSci_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$trustSci_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(trustSci_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$trustSci_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(trustSci_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$trustSci_4,d$wave, exclude = F))
d$trustSci <- rowMeans(d[, c("trustSci_1", "trustSci_2", "trustSci_3", "trustSci_4")], na.rm = T)
table(d$trustSci, d$wave)
# affProj_Trump
d[d$wave == 2 | d$wave == 3,] <- d[d$wave == 2 | d$wave == 3,] %>%
group_by(pid) %>%
fill(affProj_Trump_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$affProj_Trump_1,d$wave, exclude = F))
d[d$wave == 2 | d$wave == 3,] <- d[d$wave == 2 | d$wave == 3,] %>%
group_by(pid) %>%
fill(affProj_Trump_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$affProj_Trump_2,d$wave, exclude = F))
d[d$wave == 2 | d$wave == 3,] <- d[d$wave == 2 | d$wave == 3,] %>%
group_by(pid) %>%
fill(affProj_Trump_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$affProj_Trump_3,d$wave, exclude = F))
d[d$wave == 2 | d$wave == 3,] <- d[d$wave == 2 | d$wave == 3,] %>%
group_by(pid) %>%
fill(affProj_Trump_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$affProj_Trump_4,d$wave, exclude = F))
# punitive policies
d <- d %>%
group_by(pid) %>%
fill(deathpenalty, .direction = "downup") %>%
ungroup()
addmargins(table(d$deathpenalty,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(abortionban, .direction = "downup") %>%
ungroup()
addmargins(table(d$abortionban,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(ICEraids, .direction = "downup") %>%
ungroup()
addmargins(table(d$ICEraids,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(gunctrl, .direction = "downup") %>%
ungroup()
addmargins(table(d$gunctrl,d$wave, exclude = F))
addmargins(table(d$gunctrl,d$wave))
d <- d %>%
group_by(pid) %>%
fill(vaxxpass, .direction = "downup") %>%
ungroup()
addmargins(table(d$vaxxpass,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(hatespeech, .direction = "downup") %>%
ungroup()
addmargins(table(d$hatespeech,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(jailCEOs, .direction = "downup") %>%
ungroup()
addmargins(table(d$jailCEOs,d$wave, exclude = F))
# ideological position based on punitive policies
## deathpenalty, abortionban, ICEraids, gunctrl => negative = more liberal, positive = more conservative
## vaxxpass, hatespeech, jailCEOs => positive = more liberal, negative = more conservative
d$vaxxpass.r <- d$vaxxpass * -1
d$hatespeech.r <- d$hatespeech * -1
d$jailCEOs.r <- d$jailCEOs * -1
describe(d[,c('deathpenalty', 'abortionban', 'ICEraids', 'gunctrl','vaxxpass.r', 'hatespeech.r', 'jailCEOs.r')])
d$issue_ideology <- rowMeans(d[,c('deathpenalty', 'abortionban', 'ICEraids', 'gunctrl','vaxxpass.r', 'hatespeech.r', 'jailCEOs.r')], na.rm = T)
psych::alpha(d[,c('deathpenalty', 'abortionban', 'ICEraids', 'gunctrl','vaxxpass.r', 'hatespeech.r', 'jailCEOs.r')]) # a bit low!
# hist(d$issue_ideology)
psych::describe(d$issue_ideology) # that is a normal distribution
d <- d %>%
group_by(pid) %>%
fill(issue_ideology, .direction = "downup") %>%
ungroup()
addmargins(table(d$issue_ideology,d$wave, exclude = F))
# climate change belief
d$climbel_1 <- d$climatechange_belief_1
d$climbel_2 <- d$climatechange_belief_2
d$climbel_3 <- d$climatechange_belief_3
d$climbel_4 <- d$climatechange_belief_4
d <- d %>%
group_by(pid) %>%
fill(climatechange_belief_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$climatechange_belief_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(climatechange_belief_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$climatechange_belief_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(climatechange_belief_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$climatechange_belief_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(climatechange_belief_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$climatechange_belief_4,d$wave, exclude = F))
# GCB
d <- d %>%
group_by(pid) %>%
fill(GCB_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$GCB_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(GCB_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$GCB_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(GCB_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$GCB_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(GCB_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$GCB_4,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(GCB_5, .direction = "downup") %>%
ungroup()
addmargins(table(d$GCB_5,d$wave, exclude = F))
psych::alpha(d[,c("GCB_1", "GCB_2", "GCB_3", "GCB_4", "GCB_5")])
d <- d %>%
group_by(pid) %>%
fill(GCB, .direction = "downup") %>%
ungroup()
addmargins(table(d$GCB,d$wave, exclude = F))
# AI risk
d <- d %>%
group_by(pid) %>%
fill(AIrisk_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$AIrisk_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(AIrisk_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$AIrisk_2,d$wave, exclude = F))
d$AIrisk <- rowMeans(d[,c("AIrisk_1","AIrisk_2")], na.rm = T)
d$AIrisk.c <- d$AIrisk - mean(d$AIrisk, na.rm = T)
# Group FT
d <- d %>%
group_by(pid) %>%
fill(FT_groups_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$FT_groups_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(FT_groups_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$FT_groups_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(FT_groups_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$FT_groups_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(FT_groups_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$FT_groups_4,d$wave, exclude = F))
# Religion
d <- d %>%
group_by(pid) %>%
fill(religion, .direction = "downup") %>%
ungroup()
addmargins(table(d$religion,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(Evangelical, .direction = "downup") %>%
ungroup()
addmargins(table(d$Evangelical,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(rel_importance, .direction = "downup") %>%
ungroup()
addmargins(table(d$rel_importance,d$wave, exclude = F))
# ZIP code
d <- d %>%
group_by(pid) %>%
fill(ZIPcode, .direction = "downup") %>%
ungroup()
addmargins(table(d$ZIPcode,d$wave, exclude = F))
# State
d <- d %>%
group_by(pid) %>%
fill(state, .direction = "downup") %>%
ungroup()
addmargins(table(d$state,d$wave, exclude = F))
d$state <- as.factor(d$state)
d$state[d$state == "invalid"] <- NA
d$state[d$state == "Invalid"] <- NA
d$state <- droplevels(d$state)
table(d$state, exclude = F)
# per_gop
d <- d %>%
group_by(pid) %>%
fill(per_gop, .direction = "downup") %>%
ungroup()
addmargins(table(d$per_gop,d$wave, exclude = F))
d$per_gop <- as.numeric(d$per_gop)
# per_dem
d <- d %>%
group_by(pid) %>%
fill(per_dem, .direction = "downup") %>%
ungroup()
addmargins(table(d$per_dem,d$wave, exclude = F))
d$per_dem <- as.numeric(d$per_dem)
# vote_MC
d <- d %>%
group_by(pid) %>%
fill(vote_MC_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$vote_MC_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(vote_MC_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$vote_MC_2,d$wave, exclude = F))
# Israel
table(d$israel)
d <- d %>%
group_by(pid) %>%
fill(israel, .direction = "downup") %>%
ungroup()
addmargins(table(d$israel,d$wave, exclude = F))
table(d$israel_usRole)
d <- d %>%
group_by(pid) %>%
fill(israel_usRole, .direction = "downup") %>%
ungroup()
addmargins(table(d$israel_usRole,d$wave, exclude = F))
d$israel_usRole <- as.numeric(as.character(d$israel_usRole))
# border
table(d$border_Harris)
d <- d %>%
group_by(pid) %>%
fill(border_Harris, .direction = "downup") %>%
ungroup()
addmargins(table(d$border_Harris,d$wave, exclude = F))
table(d$border_Trump)
d <- d %>%
group_by(pid) %>%
fill(border_Trump, .direction = "downup") %>%
ungroup()
addmargins(table(d$border_Trump,d$wave, exclude = F))
# polling
d <- d %>%
group_by(pid) %>%
fill(pollFreq, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollFreq,d$wave, exclude = F))
d$pollFreq <- dplyr::recode(d$pollFreq,
`1` = 7,
`2` = 6,
`3` = 5,
`4` = 4,
`5` = 3,
`6` = 2,
`7` = 1)
d[d$wave == 1 | d$wave ==2,] <- d[d$wave == 1 | d$wave ==2,] %>%
group_by(pid) %>%
fill(pollAcc, .direction = "downup") %>%
ungroup()
d[d$wave == 3 | d$wave == 4,] <- d[d$wave == 3 | d$wave == 4,] %>%
group_by(pid) %>%
fill(pollAcc, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollAcc,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollPred, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollPred,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollsource, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollsource,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollsource_7_TEXT, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollsource_7_TEXT,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollrel1, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollrel1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollbias1, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollbias1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollrel2, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollrel2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollbias2, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollbias2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollrel3, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollrel3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollbias3, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollbias3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollrel4, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollrel4,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(pollbias4, .direction = "downup") %>%
ungroup()
addmargins(table(d$pollbias4,d$wave, exclude = F))
# NFC
d <- d %>%
group_by(pid) %>%
fill(NFC6_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$NFC6_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(NFC6_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$NFC6_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(NFC6_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$NFC6_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(NFC6_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$NFC6_4,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(NFC6_5, .direction = "downup") %>%
ungroup()
addmargins(table(d$NFC6_5,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(NFC6_6, .direction = "downup") %>%
ungroup()
addmargins(table(d$NFC6_6,d$wave, exclude = F))
d$NFC6_3.r <- d$NFC6_3 * -1
d$NFC6_4.r <- d$NFC6_4 * -1
d$NFC <- rowMeans(d[,c("NFC6_1", "NFC6_2", "NFC6_3.r", "NFC6_4.r", "NFC6_5", "NFC6_6")], na.rm = T)
psych::alpha(d[,c("NFC6_1", "NFC6_2", "NFC6_3.r", "NFC6_4.r", "NFC6_5", "NFC6_6")])
# Executive order measures
d <- d %>%
group_by(pid) %>%
fill(EO_impact_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_impact_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_impact_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_impact_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_4,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_valence_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_valence_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_valence_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_valence_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_4,d$wave, exclude = F))
# gov't process measures
d <- d %>%
group_by(pid) %>%
fill(govt_process, .direction = "downup") %>%
ungroup()
addmargins(table(d$govt_process,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(representation_Dem, .direction = "downup") %>%
ungroup()
addmargins(table(d$representation_Dem,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(representation_Rep, .direction = "downup") %>%
ungroup()
addmargins(table(d$representation_Rep,d$wave, exclude = F))
# ideology
d <- d %>%
group_by(pid) %>%
fill(ideology, .direction = "downup") %>%
ungroup()
addmargins(table(d$ideology,d$wave, exclude = F))
# EO valence & impact
d <- d %>%
group_by(pid) %>%
fill(EO_valence_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_valence_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_valence_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_valence_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_valence_4,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_impact_1, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_1,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_impact_2, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_2,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_impact_3, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_3,d$wave, exclude = F))
d <- d %>%
group_by(pid) %>%
fill(EO_impact_4, .direction = "downup") %>%
ungroup()
addmargins(table(d$EO_impact_4,d$wave, exclude = F))
d$EO_valence <- rowMeans(d[,c("EO_valence_1","EO_valence_2","EO_valence_3","EO_valence_4")],na.rm = T)
d$EO_impact <- rowMeans(d[,c("EO_impact_1","EO_impact_2","EO_impact_3","EO_impact_4")],na.rm = T)repeated measures
# emotions
d$Neg_Emo <- rowMeans(d[,c("Emotions_1", "Emotions_2", "Emotions_5")], na.rm = T)
psych::alpha(d[,c("Emotions_1", "Emotions_2", "Emotions_5")])
cor.test(d$Emotions_1[d$wave == 1], d$Emotions_2[d$wave == 1])
cor.test(d$Emotions_1[d$wave == 2], d$Emotions_2[d$wave == 2])
psych::alpha(d[d$wave == 3,c("Emotions_1", "Emotions_2", "Emotions_5")])
psych::alpha(d[d$wave == 4,c("Emotions_1", "Emotions_2", "Emotions_5")])
d$Neg_Emo.c <- d$Neg_Emo - mean(d$Neg_Emo, na.rm = T)
d$Pos_Emo <- rowMeans(d[,c("Emotions_3", "Emotions_4")], na.rm = T)
cor.test(d$Emotions_3, d$Emotions_4)
cor.test(d$Emotions_3[d$wave == 1], d$Emotions_4[d$wave == 1])
cor.test(d$Emotions_3[d$wave == 2], d$Emotions_4[d$wave == 2])
cor.test(d$Emotions_3[d$wave == 3], d$Emotions_4[d$wave == 3])
cor.test(d$Emotions_3[d$wave == 4], d$Emotions_4[d$wave == 4])
d$Pos_Emo.c <- d$Pos_Emo - mean(d$Pos_Emo, na.rm = T)
# vote confidence
d$voteconfidence <- rowMeans(d[,c('voteconf_natl','voteconf_self')], na.rm = T)
cor.test(d$voteconf_natl, d$voteconf_self)
cor.test(d$voteconf_natl[d$wave == 1], d$voteconf_self[d$wave == 1])
cor.test(d$voteconf_natl[d$wave == 2], d$voteconf_self[d$wave == 2])
cor.test(d$voteconf_natl[d$wave == 3], d$voteconf_self[d$wave == 3])
cor.test(d$voteconf_natl[d$wave == 4], d$voteconf_self[d$wave == 4])
# party ID
## Creating continuous measure of party from branching ANES question
d$partyCont <- NA
d$partyCont[d$demStrength == 1] <- -3
d$partyCont[d$demStrength == 2] <- -2
d$partyCont[d$partyClose == 1] <- -1
d$partyCont[d$partyClose == 3] <- 0
d$partyCont[d$partyClose == 2] <- 1
d$partyCont[d$repStrength == 2] <- 2
d$partyCont[d$repStrength == 1] <- 3
## Creating factor measure of party from continuous measure
d$party_factor <- NA
d$party_factor[d$partyCont == -3 | d$partyCont == -2 | d$partyCont == -1] <- "Democrat"
d$party_factor[d$partyCont == 3 | d$partyCont == 2 | d$partyCont == 1] <- "Republican"
d$party_factor[d$partyCont == 0] <- "Independent"
table(d$party_factor)
# affective polarization
## Parties
d$FT_Outgroup <- ifelse(d$party_factor == "Democrat", d$FT_parties_2,
ifelse(d$party_factor == "Republican", d$FT_parties_1, NA))
d$FT_diff <- NA
d$FT_diff <- d$FT_parties_1 - d$FT_parties_2
d$affPol <- ifelse(d$party_factor == "Democrat", d$FT_parties_1 - d$FT_parties_2,
ifelse(d$party_factor == "Republican", d$FT_parties_2 - d$FT_parties_1,
ifelse(d$party_factor == "Independent", NA, NA)))
## Politicians
d$FT_Outcandidate <- ifelse(d$party_factor == "Democrat", d$FT_pols_2,
ifelse(d$party_factor == "Republican", d$FT_pols_1, NA))
d$FT_CandDiff <- NA
d$FT_CandDiff <- d$FT_pols_1 - d$FT_pols_2
describeBy(d$FT_CandDiff, d$party_factor)
d$affPol_cand <- ifelse(d$party_factor == "Democrat", d$FT_pols_1 - d$FT_pols_2,
ifelse(d$party_factor == "Republican", d$FT_pols_2 - d$FT_pols_1,
ifelse(d$party_factor == "Independent", NA, NA)))
# connectedness
d$connectedness_6 <- as.numeric(d$connectedness_6)
d$connectedness <- rowMeans(d[,c("connectedness_1","connectedness_2","connectedness_3","connectedness_4","connectedness_5","connectedness_6","connectedness_7","connectedness_8")], na.rm = T)
psych::alpha(d[ , c("connectedness_1","connectedness_2","connectedness_3","connectedness_4","connectedness_5","connectedness_6","connectedness_7","connectedness_8")])
# demNorms
d$antidemNorms <- rowMeans(d[,c("demNorms_Reps_1", "demNorms_Reps_2", "demNorms_Reps_3", "demNorms_Reps_4", "demNorms_Dems_1", "demNorms_Dems_2", "demNorms_Dems_3", "demNorms_Dems_4")], na.rm = T)
psych::alpha(d[,c("demNorms_Reps_1", "demNorms_Reps_2", "demNorms_Reps_3", "demNorms_Reps_4")])
psych::alpha(d[d$wave == 1,c("demNorms_Reps_1", "demNorms_Reps_2", "demNorms_Reps_3", "demNorms_Reps_4")])
psych::alpha(d[d$wave == 2,c("demNorms_Reps_1", "demNorms_Reps_2", "demNorms_Reps_3", "demNorms_Reps_4")])
psych::alpha(d[d$wave == 3,c("demNorms_Reps_1", "demNorms_Reps_2", "demNorms_Reps_3", "demNorms_Reps_4")])
psych::alpha(d[d$wave == 4,c("demNorms_Reps_1", "demNorms_Reps_2", "demNorms_Reps_3", "demNorms_Reps_4")])
psych::alpha(d[,c("demNorms_Dems_1", "demNorms_Dems_2", "demNorms_Dems_3", "demNorms_Dems_4")])
psych::alpha(d[d$wave == 1,c("demNorms_Dems_1", "demNorms_Dems_2", "demNorms_Dems_3", "demNorms_Dems_4")])
psych::alpha(d[d$wave == 2,c("demNorms_Dems_1", "demNorms_Dems_2", "demNorms_Dems_3", "demNorms_Dems_4")])
psych::alpha(d[d$wave == 3,c("demNorms_Dems_1", "demNorms_Dems_2", "demNorms_Dems_3", "demNorms_Dems_4")])
psych::alpha(d[d$wave == 4,c("demNorms_Dems_1", "demNorms_Dems_2", "demNorms_Dems_3", "demNorms_Dems_4")])
d$demNorms <- (d$antidemNorms*-1)
psych::describe(d$demNorms)
psych::describe(d$antidemNorms)
# anti-Asian animus
d$AAA <- rowMeans(d[c("AAA_1", "AAA_2", "AAA_3")], na.rm = T)
psych::alpha(d[,c("demNorms_Dems_1", "demNorms_Dems_2", "demNorms_Dems_3", "demNorms_Dems_4")])
# gov't trust
cor.test(d$trustGovt1, d$trustGovt2) # yeah
cor.test(d$trustGovt1[d$wave == 1], d$trustGovt2[d$wave == 1])
cor.test(d$trustGovt1[d$wave == 2], d$trustGovt2[d$wave == 2])
cor.test(d$trustGovt1[d$wave == 3], d$trustGovt2[d$wave == 3])
cor.test(d$trustGovt1[d$wave == 4], d$trustGovt2[d$wave == 4])
d$trustGovt <- rowMeans(d[c("trustGovt1", "trustGovt2")], na.rm = T)
# science trust
d$trustSci <- rowMeans(d[c("trustSci_1", "trustSci_2","trustSci_3","trustSci_4")], na.rm = T)
psych::alpha(d[c("trustSci_1", "trustSci_2","trustSci_3","trustSci_4")])
# Vote Choice
table(d$voteChoice_final, d$wave)
d$voteID <- NA
d$voteID[d$voteChoice_final == 1] <- "Trump"
d$voteID[d$voteChoice_final == 2] <- "Harris"
d$voteID[d$VoteChoice == 1] <- "Trump"
d$voteID[d$VoteChoice == 2] <- "Harris"
d$voteID[d$VoteLean == 1] <- "Trump"
d$voteID[d$VoteLean == 2] <- "Harris"
d$voteID[d$VoteLean == 3] <- "Other"
d$voteID[d$voteChoice_final == 4] <- "Did Not Vote"
d$voteID[d$voteChoice_final == 3] <- "Other" # Hmmm
table(d$voteID, exclude = F)
# counter-proj
## ID
d$proj_ID <- NA
d$proj_ID[d$voteID == "Harris"] <- "Left"
d$proj_ID[d$voteID == "Trump"] <- "Right"
d$proj_ID[d$voteID == "Other" & d$party_factor == "Democrat"] <- "Left"
d$proj_ID[d$voteID == "Other" & d$party_factor == "Republican"] <- "Right"
## affective polarization
d$affPol_proj_Harris <- d$FT_proj_Harris_2 - d$FT_proj_Harris_1
d$affPol_proj_Trump <- d$FT_proj_Trump_1 - d$FT_proj_Trump_2
d$affPol_proj_Dem <- d$FT_proj_Dem_2 - d$FT_proj_Dem_1
d$affPol_proj_Rep <- d$FT_proj_Rep_1 - d$FT_proj_Rep_2
## dem norms
d$demNorms_proj_harris <- rowMeans(d[,c("demNorms_proj_harris_1", "demNorms_proj_harris_2", "demNorms_proj_harris_3", "demNorms_proj_harris_4")], na.rm = T)
d$demNorms_proj_Trump <- rowMeans(d[,c("demNorms_proj_Trump_1", "demNorms_proj_Trump_2", "demNorms_proj_Trump_3", "demNorms_proj_Trump_4")], na.rm = T)
d$demNorms_proj_Dem <- rowMeans(d[,c("demNorms_proj_Dem_1", "demNorms_proj_Dem_2", "demNorms_proj_Dem_3", "demNorms_proj_Dem_4")], na.rm = T)
d$demNorms_proj_Rep <- rowMeans(d[,c("demNorms_proj_Rep_1", "demNorms_proj_Rep_2", "demNorms_proj_Rep_3", "demNorms_proj_Rep_4")], na.rm = T)
# Affective Polarization
#d$affPol_proj <- ifelse(d$voteID == "Harris", d$affPol_proj_Harris,
# ifelse(d$voteID == "Trump", d$affPol_proj_Trump,
# ifelse(d$voteID == "Other" & d$party_factor == "Democrat", #d$affPol_proj_Dem,
# ifelse(d$voteID == "Other" & d$party_factor == "Republican", #d$affPol_proj_Rep, NA))))
d$affPol_proj <- rowMeans(d[,c("affPol_proj_Harris","affPol_proj_Trump","affPol_proj_Dem","affPol_proj_Rep")], na.rm = T)
describeBy(d$affPol_proj, list(d$proj_ID, d$wave))
## Negative Polarization
d$negPol_proj <- ifelse(d$voteID == "Harris", d$FT_proj_Harris_1,
ifelse(d$voteID == "Trump", d$FT_proj_Trump_2,
ifelse(d$voteID == "Other" & d$party_factor == "Democrat", d$FT_proj_Dem_1,
ifelse(d$voteID == "Other" & d$party_factor == "Republican", d$FT_proj_Rep_2, NA))))
## Democratic Norms
d$demNorms_proj <- ifelse(d$voteID == "Harris", d$demNorms_proj_harris,
ifelse(d$voteID == "Trump", d$demNorms_proj_Trump,
ifelse(d$voteID == "Other" & d$party_factor == "Democrat", d$demNorms_proj_Dem,
ifelse(d$voteID == "Other" & d$party_factor == "Republican", d$demNorms_proj_Rep, NA))))
# climate attitudes
d$climBel_proj <- ifelse(d$voteID == "Harris", d$climBel_proj_Harris,
ifelse(d$voteID == "Trump", d$climbel_proj_Trump,
ifelse(d$voteID == "Other" & d$party_factor == "Democrat", d$climBel_proj_Dem,
ifelse(d$voteID == "Other" & d$party_factor == "Republican", d$climBel_proj_Rep, NA))))
# ID centrality variables
## ID formation 1
d$IDform_1 <- rowMeans(d[, c("IDform_Dem_1", "IDform_Rep_1", "IDform_ind_1", "IDform_lib_1", "IDform_leftist_1", "IDform_prog_1", "IDform_con_1", "IDform_right_1", "IDform_mod_1", "IDform_socialist_1", "IDform_natlist_1", "IDform_libertarian_1", "IDform_feminist_1", "IDform_prolife_1", "IDform_prochoice_1", "IDform_MAGA_1", "IDform_enviro_1", "IDform_antiTrump_1", "IDform_pop_1", "IDform_other_1")], na.rm = T)
## ID formation 2
d$IDform_2 <- rowMeans(d[, c("IDform_Dem_2", "IDform_Rep_2", "IDform_ind_2", "IDform_lib_2", "IDform_leftist_2", "IDform_prog_2", "IDform_con_2", "IDform_right_2", "IDform_mod_2", "IDform_socialist_2", "IDform_natlist_2", "IDform_libertarian_2", "IDform_feminist_2", "IDform_prolife_2", "IDform_prochoice_2", "IDform_MAGA_2", "IDform_enviro_2", "IDform_antiTrump_2", "IDform_pop_2", "IDform_other_2")], na.rm = T)
## ID formation 3
d$IDform_3 <- rowMeans(d[, c("IDform_Dem_3", "IDform_Rep_3", "IDform_ind_3", "IDform_lib_3", "IDform_leftist_3", "IDform_prog_3", "IDform_con_3", "IDform_right_3", "IDform_mod_3", "IDform_socialist_3", "IDform_natlist_3", "IDform_libertarian_3", "IDform_feminist_3", "IDform_prolife_3", "IDform_prochoice_3", "IDform_MAGA_3", "IDform_enviro_3", "IDform_antiTrump_3", "IDform_pop_3", "IDform_other_3")], na.rm = T)
## ID formation 4
d$IDform_4 <- rowMeans(d[, c("IDform_Dem_4", "IDform_Rep_4", "IDform_ind_4", "IDform_lib_4", "IDform_leftist_4", "IDform_prog_4", "IDform_con_4", "IDform_right_4", "IDform_mod_4", "IDform_socialist_4", "IDform_natlist_4", "IDform_libertarian_4", "IDform_feminist_4", "IDform_prolife_4", "IDform_prochoice_4", "IDform_MAGA_4", "IDform_enviro_4", "IDform_antiTrump_4", "IDform_pop_4", "IDform_other_4")], na.rm = T)
## ID formation 5
d$IDform_5 <- rowMeans(d[, c("IDform_Dem_5", "IDform_Rep_5", "IDform_ind_5", "IDform_lib_5", "IDform_leftist_5", "IDform_prog_5", "IDform_con_5", "IDform_right_5", "IDform_mod_5", "IDform_socialist_5", "IDform_natlist_5", "IDform_libertarian_5", "IDform_feminist_5", "IDform_prolife_5", "IDform_prochoice_5", "IDform_MAGA_5", "IDform_enviro_5", "IDform_antiTrump_5", "IDform_pop_5", "IDform_other_5")], na.rm = T)
## ID formation 6
d$IDform_6 <- rowMeans(d[, c("IDform_Dem_6", "IDform_Rep_6", "IDform_ind_6", "IDform_lib_6", "IDform_leftist_6", "IDform_prog_6", "IDform_con_6", "IDform_right_6", "IDform_mod_6", "IDform_socialist_6", "IDform_natlist_6", "IDform_libertarian_6", "IDform_feminist_6", "IDform_prolife_6", "IDform_prochoice_6", "IDform_MAGA_6", "IDform_enviro_6", "IDform_antiTrump_6", "IDform_pop_6", "IDform_other_6")], na.rm = T)
## composite identity centrality
d$IDform <- rowMeans(d[,c("IDform_1", "IDform_2", "IDform_3", "IDform_4", "IDform_5", "IDform_6")], na.rm = T)
psych::alpha(d[,c("IDform_1", "IDform_2", "IDform_3", "IDform_4", "IDform_5", "IDform_6")])
# Identity Variables
d$numID_selected <- NA
d$numID_selected <- rowSums(d[,c("polID_Liberal_select", "polID_Leftist_select", "polID_Progressive_select", "polID_Conservative_select", "polID_RightWing_select", "polID_Moderate_select", "polID_Socialist_select", "polID_Nationalist_select", "polID_Libertarian_select", "polID_Feminist_select", "polID_Prolife_select", "polID_Prochoice_select", "polID_MAGA_select", "polID_AntiTrump_select", "polID_Environmentalist_select", "polID_Populist_select", "polID_Other_select")],na.rm = T)
table(d$numID_selected, d$wave)
## Calculate & join ID sds for each participant
ID_sds <- d %>%
dplyr::group_by(pid) %>%
dplyr::summarize(sdID_selected = sd(numID_selected, na.rm = TRUE))
d <- d %>%
left_join(ID_sds, by = "pid")Results
Figures:
Figures are typically sized at 90mm (single column) or 180mm (double column), with a maximum height of 170mm, and font sizes should be 5-7pt at this size
Prep: Analytic codes & centering
### partisan identity vars
# Contrast Codes
## dems vs. reps
d$pDem_Rep <- NA
d$pDem_Rep[d$party_factor == "Democrat"] <- -1/2
d$pDem_Rep[d$party_factor == "Independent"] <- 0
d$pDem_Rep[d$party_factor == "Republican"] <- 1/2
## partisans vs. inds
d$pParty_Ind <- NA
d$pParty_Ind[d$party_factor == "Democrat"] <- -1/3
d$pParty_Ind[d$party_factor == "Independent"] <- 2/3
d$pParty_Ind[d$party_factor == "Republican"] <- -1/3
# Dummy Codes
## Democrats
### dems vs. R
d$pDem_R <- NA
d$pDem_R[d$party_factor == "Democrat"] <- 0
d$pDem_R[d$party_factor == "Independent"] <- 0
d$pDem_R[d$party_factor == "Republican"] <- 1
### dems vs. I
d$pDem_I <- NA
d$pDem_I[d$party_factor == "Democrat"] <- 0
d$pDem_I[d$party_factor == "Independent"] <- 1
d$pDem_I[d$party_factor == "Republican"] <- 0
## Republicans
### reps vs. D
d$pRep_D <- NA
d$pRep_D[d$party_factor == "Democrat"] <- 1
d$pRep_D[d$party_factor == "Independent"] <- 0
d$pRep_D[d$party_factor == "Republican"] <- 0
### reps vs. I
d$pRep_I <- NA
d$pRep_I[d$party_factor == "Democrat"] <- 0
d$pRep_I[d$party_factor == "Independent"] <- 1
d$pRep_I[d$party_factor == "Republican"] <- 0
## Independents
### inds vs. D
d$pInd_D <- NA
d$pInd_D[d$party_factor == "Democrat"] <- 1
d$pInd_D[d$party_factor == "Independent"] <- 0
d$pInd_D[d$party_factor == "Republican"] <- 0
### inds vs. R
d$pInd_R <- NA
d$pInd_R[d$party_factor == "Democrat"] <- 0
d$pInd_R[d$party_factor == "Independent"] <- 0
d$pInd_R[d$party_factor == "Republican"] <- 1
# party switch
d$party_switch.d <- ifelse(d$party_switch == "TRUE", 1, 0)
### wave
d$wave <- as.factor(d$wave)
# contrast codes
d$wave.lin <- NA
d$wave.lin[d$wave == 1] <- -1/2
d$wave.lin[d$wave == 2] <- -1/4
d$wave.lin[d$wave == 3] <- 1/4
d$wave.lin[d$wave == 4] <- 1/2
d$wave.quad <- NA
d$wave.quad[d$wave == 1] <- -1/2
d$wave.quad[d$wave == 2] <- 1/2
d$wave.quad[d$wave == 3] <- 1/2
d$wave.quad[d$wave == 4] <- -1/2
d$wave.cub <- NA
d$wave.cub[d$wave == 1] <- 1/4
d$wave.cub[d$wave == 2] <- -1/2
d$wave.cub[d$wave == 3] <- 1/2
d$wave.cub[d$wave == 4] <- -1/4
# dummy codes
## wave 1
d$wave1_2 <- NA
d$wave1_2[d$wave == 1] <- 0
d$wave1_2[d$wave == 2] <- 1
d$wave1_2[d$wave == 3] <- 0
d$wave1_2[d$wave == 4] <- 0
d$wave1_3 <- NA
d$wave1_3[d$wave == 1] <- 0
d$wave1_3[d$wave == 2] <- 0
d$wave1_3[d$wave == 3] <- 1
d$wave1_3[d$wave == 4] <- 0
d$wave1_4 <- NA
d$wave1_4[d$wave == 1] <- 0
d$wave1_4[d$wave == 2] <- 0
d$wave1_4[d$wave == 3] <- 0
d$wave1_4[d$wave == 4] <- 1
## wave 2
d$wave2_1 <- NA
d$wave2_1[d$wave == 1] <- 1
d$wave2_1[d$wave == 2] <- 0
d$wave2_1[d$wave == 3] <- 0
d$wave2_1[d$wave == 4] <- 0
d$wave2_3 <- NA
d$wave2_3[d$wave == 1] <- 0
d$wave2_3[d$wave == 2] <- 0
d$wave2_3[d$wave == 3] <- 1
d$wave2_3[d$wave == 4] <- 0
d$wave2_4 <- NA
d$wave2_4[d$wave == 1] <- 0
d$wave2_4[d$wave == 2] <- 0
d$wave2_4[d$wave == 3] <- 0
d$wave2_4[d$wave == 4] <- 1
## wave 3
d$wave3_1 <- NA
d$wave3_1[d$wave == 1] <- 1
d$wave3_1[d$wave == 2] <- 0
d$wave3_1[d$wave == 3] <- 0
d$wave3_1[d$wave == 4] <- 0
d$wave3_2 <- NA
d$wave3_2[d$wave == 1] <- 0
d$wave3_2[d$wave == 2] <- 1
d$wave3_2[d$wave == 3] <- 0
d$wave3_2[d$wave == 4] <- 0
d$wave3_4 <- NA
d$wave3_4[d$wave == 1] <- 0
d$wave3_4[d$wave == 2] <- 0
d$wave3_4[d$wave == 3] <- 0
d$wave3_4[d$wave == 4] <- 1
## wave 4
d$wave4_1 <- NA
d$wave4_1[d$wave == 1] <- 1
d$wave4_1[d$wave == 2] <- 0
d$wave4_1[d$wave == 3] <- 0
d$wave4_1[d$wave == 4] <- 0
d$wave4_2 <- NA
d$wave4_2[d$wave == 1] <- 0
d$wave4_2[d$wave == 2] <- 1
d$wave4_2[d$wave == 3] <- 0
d$wave4_2[d$wave == 4] <- 0
d$wave4_3 <- NA
d$wave4_3[d$wave == 1] <- 0
d$wave4_3[d$wave == 2] <- 0
d$wave4_3[d$wave == 3] <- 1
d$wave4_3[d$wave == 4] <- 0
# Polling Variable
d$pollFreq.c <- d$pollFreq - mean(d$pollFreq, na.rm = T)
# Race
d$vs_race <- as.factor(as.character(d$vs_race))
d$vs_race <- relevel(d$vs_race, ref = "White")
# Gender
d$vs_gender <- as.factor(as.character(d$vs_gender))
d$male_female <- NA
d$male_female[d$vs_gender == "Male"] <- -1/2
d$male_female[d$vs_gender == "Female"] <- 1/2
d$male_female[d$vs_gender == "Other"] <- 0
d$nonbinary_mf <- NA
d$nonbinary_mf[d$vs_gender == "Male"] <- 1/3
d$nonbinary_mf[d$vs_gender == "Female"] <- 1/3
d$nonbinary_mf[d$vs_gender == "Other"] <- -2/3
# generic conspiracist beliefs
d$GCB.c <- d$GCB - mean(d$GCB, na.rm = T)
# affPol (scaled)
d$affPol.100 <- d$affPol/100
# poll reliability & poll ideology bias
## averaging across all media sources
d$pollrel <- rowMeans(d[,c("pollrel1", "pollrel2", "pollrel3", "pollrel4")], na.rm = T)
d$pollbias <- rowMeans(d[,c("pollbias1", "pollbias2", "pollbias3", "pollbias4")], na.rm = T)
## scaled variables
d$pollrel.z <- scale(d$pollrel)
d$pollbias.z <- 2 * (d$pollbias - min(d$pollbias, na.rm = TRUE)) / (max(d$pollbias, na.rm = TRUE) - min(d$pollbias, na.rm = TRUE)) - 1
## magnitude and tilt separate
d$pollbias_abs <- abs(d$pollbias)
d$pollbias_tilt <- ifelse(d$pollbias > 0, "conservative",
ifelse(d$pollbias < 0, "liberal", NA))
table(d$pollbias_tilt)##
## conservative liberal
## 2014 3764# Vote share measures
d$per_sum <- rowSums(d[,c("per_dem","per_gop")], na.rm = T)
d$per_sum[d$per_sum == 0] <- NA
d$per_inparty <- ifelse(d$party_factor == "Democrat", d$per_dem,
ifelse(d$party_factor == "Republican", d$per_gop,
ifelse(d$party_factor == "Independent", NA, NA)))
# plot labels
wave_label <- c(`1` = "Wave 1", `2` = "Wave 2", `3` = "Wave 3", `4` = "Wave 4")
wave_posemo_label.b <- c(`1` = "Wave 1", `2` = "Wave 2", `3` = "Wave 3", `4` = "Wave 4", "-1 SD" = "Low Pos. Emo.", "Mean Positive Emotions" = "Mean Pos. Emo.", "+1 SD" = "High Pos. Emo.")
wave_negemo_label.b <- c(`1` = "Wave 1", `2` = "Wave 2", `3` = "Wave 3", `4` = "Wave 4", "-1 SD" = "Low Neg. Emo.", "Mean Negative Emotions" = "Mean Neg. Emo.", "+1 SD" = "High Neg. Emo.")
wave_posemo_label.t <- c(`1` = "Wave 1", `2` = "Wave 2", `3` = "Wave 3", `4` = "Wave 4", "1st tercile" = "Low Pos. Emo.", "2nd tercile" = "Mid Pos. Emo.", "3rd tercile" = "High Pos. Emo.")
wave_negemo_label.t <- c(`1` = "Wave 1", `2` = "Wave 2", `3` = "Wave 3", `4` = "Wave 4", "1st tercile" = "Low Neg. Emo.", "2nd tercile" = "Mid Neg. Emo.", "3rd tercile" = "High Neg. Emo.")
# terciles & bins
## emotions - pos & neg
d$Pos_Emo_terciles <- ntile(d$Pos_Emo, 3)
d$Pos_Emo_terciles <- as.factor(d$Pos_Emo_terciles)
d$Pos_Emo_terciles <- recode_factor(d$Pos_Emo_terciles,
"1" = "1st tercile",
"2" = "2nd tercile",
"3" = "3rd tercile")
d$Pos_Emo_bins <- NA
d$Pos_Emo_bins[d$Pos_Emo <= (mean(d$Pos_Emo, na.rm = T) - sd(d$Pos_Emo, na.rm = T))] <- "-1 SD"
d$Pos_Emo_bins[d$Pos_Emo > (mean(d$Pos_Emo, na.rm = T) - sd(d$Pos_Emo, na.rm = T)) & d$Pos_Emo < (mean(d$Pos_Emo, na.rm = T) + sd(d$Pos_Emo, na.rm = T))] <- "Mean Positive Emotions"
d$Pos_Emo_bins[d$Pos_Emo >= (mean(d$Pos_Emo, na.rm = T) + sd(d$Pos_Emo, na.rm = T))] <- "+1 SD"
d$Pos_Emo_bins <- factor(d$Pos_Emo_bins,
levels = c("-1 SD",
"Mean Positive Emotions",
"+1 SD"))
d$Neg_Emo_terciles <- ntile(d$Neg_Emo, 3)
d$Neg_Emo_terciles <- as.factor(d$Neg_Emo_terciles)
d$Neg_Emo_terciles <- recode_factor(d$Neg_Emo_terciles,
"1" = "1st tercile",
"2" = "2nd tercile",
"3" = "3rd tercile")
d$Neg_Emo_bins <- NA
d$Neg_Emo_bins[d$Neg_Emo <= (mean(d$Neg_Emo, na.rm = T) - sd(d$Neg_Emo, na.rm = T))] <- "-1 SD"
d$Neg_Emo_bins[d$Neg_Emo > (mean(d$Neg_Emo, na.rm = T) - sd(d$Neg_Emo, na.rm = T)) & d$Neg_Emo < (mean(d$Neg_Emo, na.rm = T) + sd(d$Neg_Emo, na.rm = T))] <- "Mean Negative Emotions"
d$Neg_Emo_bins[d$Neg_Emo >= (mean(d$Neg_Emo, na.rm = T) + sd(d$Neg_Emo, na.rm = T))] <- "+1 SD"
d$Neg_Emo_bins <- factor(d$Neg_Emo_bins,
levels = c("-1 SD",
"Mean Negative Emotions",
"+1 SD"))
d$per_gop.100 <- d$per_gop/100
d$per_dem.100 <- d$per_dem/1001. Perceived Legitimacy by Social Norms and Emotions
a. Media Norms
# Density of Partisan Media Bias
ggplot(d, aes(x=pollbias.z)) +
geom_density(fill="#69b3a2", color="#e9ecef", alpha=0.9) +
geom_vline(xintercept = 0, linetype="dashed",
color = "black", size=.5) +
theme_bw() +
labs(x = "Partisan Media Bias (Normalized)")
# Density by party
ggplot(d[!is.na(d$party_factor),], aes(x=pollbias.z, fill = party_factor)) +
geom_density(color="black", alpha=0.5, position = 'identity') +
theme_bw() +
geom_vline(xintercept = 0, linetype="dashed",
color = "black", size=.5) +
scale_fill_manual("Participant Party ID",
values = c("#1696d2","grey","#db2b27")) +
labs(x = "Partisan Media Bias (Normalized)")
# Just bias
ggplot(d[!is.na(d$party_factor),],
aes(x = pollbias.z,
y = voteconfidence,
fill = party_factor,
color = party_factor)) +
geom_smooth(method = "lm",
fullrange = T) +
facet_grid(~wave, labeller = as_labeller(wave_label)) +
labs(x = "Partisan Media Bias (Normalized)",
y = "Perceived Election Legitimacy")+
scale_fill_manual("Participant
Partisan Identity",
values = c("#1696d2","grey","#db2b27")) +
scale_color_manual("Participant
Partisan Identity",
values = c("#1696d2","grey","#db2b27")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
# bias x positive emotions - tercs
ggplot(d[!is.na(d$party_factor),],
aes(x = pollbias.z,
y = voteconfidence,
fill = party_factor,
color = party_factor)) +
geom_jitter(height = .3, width = .3, alpha = .2, size = .2) +
geom_smooth(method = "lm",
fullrange = T, se = F) +
facet_grid(Pos_Emo_terciles~wave, labeller = as_labeller(wave_posemo_label.t)) +
labs(x = "Partisan Media Bias (Normalized)",
y = "Perceived Election Legitimacy")+
scale_fill_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
scale_color_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
## Negative Emotions
# bias x negative emotions - tercs
ggplot(d[!is.na(d$party_factor),],
aes(x = pollbias.z,
y = voteconfidence,
fill = party_factor,
color = party_factor)) +
geom_jitter(height = .3, width = .3, alpha = .2, size = .2) +
geom_smooth(method = "lm",
fullrange = T, se = F) +
facet_grid(Neg_Emo_terciles~wave, labeller = as_labeller(wave_negemo_label.t)) +
labs(x = "Partisan Media Bias (Normalized)",
y = "Perceived Election Legitimacy") +
scale_fill_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
scale_color_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
Models
Main model
m1 <- lmer(voteconfidence ~ (pDem_Rep + pParty_Ind) * pollbias.z * (wave.lin + wave.quad + wave.cub) * (Pos_Emo.c + Neg_Emo.c)
+ pollAcc + pollFreq.c
+ vs_race + vs_age + male_female + nonbinary_mf
+ (Pos_Emo.c + Neg_Emo.c|pid),
data = d)
tab_model(m1, show.stat = T, show.df = T, df.method = "satterthwaite")| voteconfidence | |||||
|---|---|---|---|---|---|
| Predictors | Estimates | CI | Statistic | p | df |
| (Intercept) | 3.01 | 2.76 – 3.26 | 23.60 | <0.001 | 1347.92 |
| pDem Rep | -0.37 | -0.47 – -0.27 | -7.39 | <0.001 | 3381.85 |
| pParty Ind | -0.19 | -0.31 – -0.08 | -3.24 | 0.001 | 4722.83 |
| pollbias z | -0.54 | -0.71 – -0.38 | -6.58 | <0.001 | 2657.99 |
| wave lin | 0.27 | 0.19 – 0.35 | 6.71 | <0.001 | 4086.07 |
| wave quad | 0.10 | 0.04 – 0.16 | 3.39 | 0.001 | 3920.85 |
| wave cub | -0.05 | -0.13 – 0.02 | -1.43 | 0.154 | 3884.77 |
| Pos Emo c | 0.21 | 0.17 – 0.24 | 11.16 | <0.001 | 2202.86 |
| Neg Emo c | -0.07 | -0.11 – -0.03 | -3.63 | <0.001 | 1752.79 |
| pollAcc | 0.06 | 0.04 – 0.08 | 6.19 | <0.001 | 5231.73 |
| pollFreq c | -0.01 | -0.03 – 0.00 | -1.57 | 0.117 | 1615.15 |
| vs race [Black] | -0.30 | -0.43 – -0.18 | -4.95 | <0.001 | 1744.81 |
| vs race [Hispanic] | -0.12 | -0.25 – 0.01 | -1.86 | 0.063 | 1620.80 |
| vs race [Other] | -0.10 | -0.26 – 0.06 | -1.25 | 0.211 | 1679.00 |
| vs age | 0.00 | 0.00 – 0.01 | 3.02 | 0.003 | 1667.00 |
| male female | -0.30 | -0.38 – -0.22 | -7.51 | <0.001 | 1572.66 |
| nonbinary mf | 0.78 | 0.14 – 1.42 | 2.40 | 0.017 | 991.43 |
| pDem Rep × pollbias z | 0.20 | -0.08 – 0.48 | 1.38 | 0.166 | 3325.19 |
| pParty Ind × pollbias z | -0.09 | -0.45 – 0.26 | -0.52 | 0.603 | 4934.75 |
| pDem Rep × wave lin | 0.90 | 0.75 – 1.05 | 12.01 | <0.001 | 4266.81 |
| pDem Rep × wave quad | 0.16 | 0.05 – 0.27 | 2.80 | 0.005 | 4086.26 |
| pDem Rep × wave cub | -0.28 | -0.41 – -0.14 | -4.02 | <0.001 | 3892.97 |
| pParty Ind × wave lin | 0.08 | -0.12 – 0.27 | 0.74 | 0.460 | 4020.50 |
| pParty Ind × wave quad | 0.02 | -0.13 – 0.17 | 0.30 | 0.764 | 3909.47 |
| pParty Ind × wave cub | 0.07 | -0.12 – 0.25 | 0.68 | 0.497 | 3912.71 |
| pollbias z × wave lin | 0.28 | 0.04 – 0.52 | 2.31 | 0.021 | 4074.09 |
| pollbias z × wave quad | 0.14 | -0.04 – 0.32 | 1.53 | 0.127 | 3916.11 |
| pollbias z × wave cub | 0.03 | -0.19 – 0.25 | 0.29 | 0.769 | 3897.92 |
| pDem Rep × Pos Emo c | 0.05 | -0.01 – 0.10 | 1.51 | 0.131 | 1674.66 |
| pDem Rep × Neg Emo c | -0.04 | -0.10 – 0.03 | -1.19 | 0.233 | 1654.50 |
| pParty Ind × Pos Emo c | 0.02 | -0.07 – 0.11 | 0.40 | 0.693 | 3529.36 |
| pParty Ind × Neg Emo c | 0.07 | -0.02 – 0.16 | 1.62 | 0.105 | 2746.16 |
| pollbias z × Pos Emo c | 0.10 | -0.01 – 0.21 | 1.82 | 0.069 | 2228.38 |
| pollbias z × Neg Emo c | 0.00 | -0.11 – 0.11 | 0.03 | 0.980 | 1878.87 |
| wave lin × Pos Emo c | 0.03 | -0.05 – 0.10 | 0.72 | 0.470 | 4043.95 |
| wave lin × Neg Emo c | 0.00 | -0.07 – 0.07 | 0.04 | 0.971 | 4164.67 |
| wave quad × Pos Emo c | 0.01 | -0.05 – 0.06 | 0.19 | 0.849 | 4044.11 |
| wave quad × Neg Emo c | -0.05 | -0.10 – 0.00 | -1.78 | 0.076 | 3920.58 |
| wave cub × Pos Emo c | -0.02 | -0.09 – 0.04 | -0.62 | 0.535 | 4053.58 |
| wave cub × Neg Emo c | 0.03 | -0.03 – 0.09 | 0.98 | 0.327 | 3884.47 |
|
(pDem Rep × pollbias z) × wave lin |
0.17 | -0.26 – 0.60 | 0.77 | 0.442 | 4262.24 |
|
(pDem Rep × pollbias z) × wave quad |
0.24 | -0.09 – 0.56 | 1.42 | 0.156 | 4090.10 |
|
(pDem Rep × pollbias z) × wave cub |
0.03 | -0.36 – 0.43 | 0.17 | 0.868 | 3958.40 |
|
(pParty Ind × pollbias z) × wave lin |
0.21 | -0.40 – 0.83 | 0.68 | 0.497 | 4024.84 |
|
(pParty Ind × pollbias z) × wave quad |
-0.17 | -0.63 – 0.29 | -0.73 | 0.467 | 3919.51 |
|
(pParty Ind × pollbias z) × wave cub |
0.01 | -0.56 – 0.58 | 0.03 | 0.975 | 3918.10 |
|
(pDem Rep × pollbias z) × Pos Emo c |
-0.13 | -0.30 – 0.03 | -1.57 | 0.117 | 1650.13 |
|
(pDem Rep × pollbias z) × Neg Emo c |
-0.02 | -0.20 – 0.16 | -0.22 | 0.824 | 1652.42 |
|
(pParty Ind × pollbias z) × Pos Emo c |
0.03 | -0.25 – 0.31 | 0.23 | 0.819 | 3271.31 |
|
(pParty Ind × pollbias z) × Neg Emo c |
-0.01 | -0.28 – 0.27 | -0.04 | 0.971 | 2888.85 |
|
(pDem Rep × wave lin) × Pos Emo c |
-0.02 | -0.14 – 0.09 | -0.37 | 0.709 | 4305.32 |
|
(pDem Rep × wave lin) × Neg Emo c |
0.14 | 0.01 – 0.26 | 2.12 | 0.034 | 4292.73 |
|
(pDem Rep × wave quad) × Pos Emo c |
0.06 | -0.02 – 0.15 | 1.43 | 0.153 | 4134.31 |
|
(pDem Rep × wave quad) × Neg Emo c |
-0.09 | -0.18 – 0.01 | -1.79 | 0.073 | 4142.47 |
|
(pDem Rep × wave cub) × Pos Emo c |
0.09 | -0.01 – 0.19 | 1.75 | 0.080 | 4019.73 |
|
(pDem Rep × wave cub) × Neg Emo c |
0.04 | -0.07 – 0.15 | 0.72 | 0.469 | 3999.96 |
|
(pParty Ind × wave lin) × Pos Emo c |
0.31 | 0.12 – 0.50 | 3.15 | 0.002 | 3957.59 |
|
(pParty Ind × wave lin) × Neg Emo c |
-0.04 | -0.21 – 0.14 | -0.43 | 0.670 | 4122.92 |
|
(pParty Ind × wave quad) × Pos Emo c |
-0.00 | -0.15 – 0.15 | -0.01 | 0.989 | 4048.99 |
|
(pParty Ind × wave quad) × Neg Emo c |
-0.06 | -0.19 – 0.07 | -0.90 | 0.369 | 3876.99 |
|
(pParty Ind × wave cub) × Pos Emo c |
-0.03 | -0.21 – 0.15 | -0.33 | 0.745 | 4103.72 |
|
(pParty Ind × wave cub) × Neg Emo c |
0.00 | -0.16 – 0.16 | 0.00 | 0.997 | 3883.74 |
|
(pollbias z × wave lin) × Pos Emo c |
-0.02 | -0.24 – 0.20 | -0.19 | 0.848 | 3921.25 |
|
(pollbias z × wave lin) × Neg Emo c |
0.10 | -0.11 – 0.30 | 0.92 | 0.358 | 4101.56 |
|
(pollbias z × wave quad) × Pos Emo c |
0.06 | -0.10 – 0.23 | 0.75 | 0.454 | 3965.45 |
|
(pollbias z × wave quad) × Neg Emo c |
-0.05 | -0.20 – 0.11 | -0.61 | 0.545 | 3899.42 |
|
(pollbias z × wave cub) × Pos Emo c |
-0.13 | -0.33 – 0.08 | -1.23 | 0.220 | 4123.47 |
|
(pollbias z × wave cub) × Neg Emo c |
-0.00 | -0.19 – 0.19 | -0.04 | 0.970 | 3893.74 |
|
(pDem Rep × pollbias z × wave lin) × Pos Emo c |
-0.19 | -0.51 – 0.14 | -1.11 | 0.266 | 4256.46 |
|
(pDem Rep × pollbias z × wave lin) × Neg Emo c |
-0.12 | -0.48 – 0.24 | -0.68 | 0.499 | 4249.18 |
|
(pDem Rep × pollbias z × wave quad) × Pos Emo c |
0.06 | -0.18 – 0.30 | 0.48 | 0.632 | 4121.60 |
|
(pDem Rep × pollbias z × wave quad) × Neg Emo c |
0.20 | -0.07 – 0.46 | 1.47 | 0.143 | 4127.88 |
|
(pDem Rep × pollbias z × wave cub) × Pos Emo c |
0.12 | -0.16 – 0.41 | 0.84 | 0.402 | 4029.54 |
|
(pDem Rep × pollbias z × wave cub) × Neg Emo c |
0.30 | -0.02 – 0.61 | 1.84 | 0.066 | 3992.63 |
|
(pParty Ind × pollbias z × wave lin) × Pos Emo c |
-0.31 | -0.91 – 0.29 | -1.01 | 0.314 | 3820.50 |
|
(pParty Ind × pollbias z × wave lin) × Neg Emo c |
0.07 | -0.47 – 0.61 | 0.25 | 0.805 | 4025.92 |
|
(pParty Ind × pollbias z × wave quad) × Pos Emo c |
0.36 | -0.10 – 0.81 | 1.54 | 0.124 | 3940.65 |
|
(pParty Ind × pollbias z × wave quad) × Neg Emo c |
-0.30 | -0.71 – 0.10 | -1.47 | 0.141 | 3872.32 |
|
(pParty Ind × pollbias z × wave cub) × Pos Emo c |
-0.03 | -0.59 – 0.53 | -0.11 | 0.916 | 4149.23 |
|
(pParty Ind × pollbias z × wave cub) × Neg Emo c |
-0.14 | -0.64 – 0.37 | -0.52 | 0.603 | 3894.89 |
| Random Effects | |||||
| σ2 | 0.41 | ||||
| τ00 pid | 0.49 | ||||
| τ11 pid.Pos_Emo.c | 0.04 | ||||
| τ11 pid.Neg_Emo.c | 0.04 | ||||
| ρ01 | -0.45 | ||||
| 0.14 | |||||
| ICC | 0.60 | ||||
| N pid | 1674 | ||||
| Observations | 5753 | ||||
| Marginal R2 / Conditional R2 | 0.206 / 0.683 | ||||
Significant Effects
Main Effects
Dem_Rep: b = -0.37, t(3381.85) = -7.39, p < .001
pollbias: b = -0.54, t(2657.99) = -6.58, p < .001
wave lin: b = 0.27, t(4086.07) = 6.71, p < .001
Pos Emo: b = 0.21, t(2202.86) = 11.16, p < .001
Neg Emo: b = -0.07, t(1752.79) = -3.63, p < .001
2-way Interactions
Dem_Rep × wave lin: b = 0.90, t(4266.81) = 12.01, p < 0.001
pollbias × wave lin: b = 0.28, t(4074.09) = 2.31, p = 0.021
3-way Interactions
- Dem_Rep × wave lin x Neg Emo: b = 0.14, t(4292.73) = 2.12, p = 0.034
b. Geographic Norms
# Density plot of Trump vote share
ggplot(d, aes(x=per_gop)) +
geom_density(fill="#69b3a2", color="#e9ecef", alpha=0.9) +
theme_bw() +
labs(x = "Trump Vote Share by County")
# Density by party
ggplot(d[!is.na(d$party_factor),], aes(x=per_gop, fill = party_factor)) +
geom_density(color="black", alpha=0.5, position = 'identity') +
theme_bw() +
scale_fill_manual("Participant Party ID",
values = c("#1696d2","grey","#db2b27")) +
labs(x = "Trump Vote Share by County")
# Just vote share
ggplot(d[!is.na(d$party_factor),],
aes(x = per_gop,
y = voteconfidence,
fill = party_factor,
color = party_factor)) +
geom_smooth(method = "lm",
fullrange = T) +
facet_grid(~wave, labeller = as_labeller(wave_label)) +
labs(x = "Trump Vote Share by County",
y = "Perceived Election Legitimacy")+
scale_fill_manual("Participant
Partisan Identity",
values = c("#1696d2","grey","#db2b27")) +
scale_color_manual("Participant
Partisan Identity",
values = c("#1696d2","grey","#db2b27")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
# bias x positive emotions - tercs
ggplot(d[!is.na(d$party_factor),],
aes(x = per_gop,
y = voteconfidence,
fill = party_factor,
color = party_factor)) +
geom_jitter(height = .3, width = .3, alpha = .2, size = .2) +
geom_smooth(method = "lm",
fullrange = T, se = F) +
facet_grid(Pos_Emo_terciles~wave, labeller = as_labeller(wave_posemo_label.t)) +
labs(x = "Trump Vote Share by County",
y = "Perceived Election Legitimacy")+
scale_fill_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
scale_color_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
# bias x negative emotions - tercs
ggplot(d[!is.na(d$party_factor),],
aes(x = per_gop,
y = voteconfidence,
fill = party_factor,
color = party_factor)) +
geom_jitter(height = .3, width = .3, alpha = .2, size = .2) +
geom_smooth(method = "lm",
fullrange = T, se = F) +
facet_grid(Neg_Emo_terciles~wave, labeller = as_labeller(wave_negemo_label.t)) +
labs(x = "Trump Vote Share by County",
y = "Perceived Election Legitimacy") +
scale_fill_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
scale_color_manual("Participant
Partisan Identity",
values = c("#1696d2","black","#db2b27")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
#paletteer::paletteer_d("fishualize::Aluterus_scriptus")
# Party x Negative emotions x Trump vote share
ggplot(d[!is.na(d$party_factor),],
aes(x = per_gop,
y = voteconfidence,
fill = Neg_Emo_terciles,
color = Neg_Emo_terciles)) +
geom_smooth(method = "lm",
fullrange = T) +
labs(x = "Trump Vote Share by County",
y = "Perceived Election Legitimacy") +
facet_wrap(~party_factor) +
scale_fill_manual("Negative Emotions",
values = c("#05DBF2FF","#035AA6FF","#011C40FF")) +
scale_color_manual("Negative Emotions",
values = c("#05DBF2FF","#035AA6FF","#011C40FF")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
#paletteer::paletteer_d("MoMAColors::Exter")
# Just positive emotions x Trump vote share
ggplot(d[!is.na(d$party_factor),],
aes(x = per_gop,
y = voteconfidence,
fill = Pos_Emo_terciles,
color = Pos_Emo_terciles)) +
geom_smooth(method = "lm",
fullrange = T) +
labs(x = "Trump Vote Share by County",
y = "Perceived Election Legitimacy") +
scale_fill_manual("Positive Emotions",
values = c("#FAC881FF","#E87444FF","#BF2729FF")) +
scale_color_manual("Positive Emotions",
values = c("#FAC881FF","#E87444FF","#BF2729FF")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
# Party x Positive emotions x Trump vote share
ggplot(d[!is.na(d$party_factor),],
aes(x = per_gop,
y = voteconfidence,
fill = Pos_Emo_terciles,
color = Pos_Emo_terciles)) +
geom_smooth(method = "lm",
fullrange = T) +
facet_wrap(~party_factor) +
labs(x = "Trump Vote Share by County",
y = "Perceived Election Legitimacy") +
scale_fill_manual("Positive Emotions",
values = c("#FAC881FF","#E87444FF","#BF2729FF")) +
scale_color_manual("Positive Emotions",
values = c("#FAC881FF","#E87444FF","#BF2729FF")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) 
Models
Main Model
m2 <- lmer(voteconfidence ~ (pDem_Rep + pParty_Ind) * per_gop.100 * (wave.lin + wave.quad + wave.cub) * (Pos_Emo.c + Neg_Emo.c)
+ vs_age
+ vs_race
+ nonbinary_mf
+ male_female
+ (Pos_Emo.c + Neg_Emo.c|pid),
data = d)
tab_model(m2, show.stat = T, show.df = T, df.method = "satterthwaite")| voteconfidence | |||||
|---|---|---|---|---|---|
| Predictors | Estimates | CI | Statistic | p | df |
| (Intercept) | 3.29 | 3.02 – 3.56 | 24.01 | <0.001 | 2788.35 |
| pDem Rep | -0.73 | -1.06 – -0.40 | -4.36 | <0.001 | 4887.31 |
| pParty Ind | -0.34 | -0.71 – 0.03 | -1.82 | 0.068 | 6065.77 |
| per gop 100 | -0.34 | -0.69 – 0.00 | -1.95 | 0.052 | 3485.12 |
| wave lin | 0.19 | -0.07 – 0.44 | 1.41 | 0.160 | 5356.44 |
| wave quad | -0.08 | -0.27 – 0.12 | -0.76 | 0.448 | 5210.29 |
| wave cub | -0.01 | -0.25 – 0.23 | -0.07 | 0.942 | 5105.85 |
| Pos Emo c | -0.01 | -0.12 – 0.10 | -0.17 | 0.868 | 2761.93 |
| Neg Emo c | -0.03 | -0.14 – 0.08 | -0.51 | 0.611 | 2082.69 |
| vs age | 0.00 | 0.00 – 0.01 | 3.27 | 0.001 | 2355.89 |
| vs race [Black] | -0.31 | -0.42 – -0.20 | -5.47 | <0.001 | 2451.73 |
| vs race [Hispanic] | -0.07 | -0.19 – 0.04 | -1.20 | 0.230 | 2333.40 |
| vs race [Other] | -0.11 | -0.25 – 0.03 | -1.53 | 0.125 | 2338.90 |
| nonbinary mf | 0.33 | -0.17 – 0.83 | 1.28 | 0.201 | 1576.36 |
| male female | -0.31 | -0.38 – -0.24 | -8.52 | <0.001 | 2248.67 |
| pDem Rep × per gop 100 | 0.50 | -0.15 – 1.15 | 1.51 | 0.132 | 4728.28 |
| pParty Ind × per gop 100 | 0.21 | -0.53 – 0.95 | 0.56 | 0.575 | 5989.16 |
| pDem Rep × wave lin | 0.65 | 0.13 – 1.18 | 2.43 | 0.015 | 5526.69 |
| pDem Rep × wave quad | -0.13 | -0.53 – 0.27 | -0.64 | 0.521 | 5336.10 |
| pDem Rep × wave cub | 0.11 | -0.38 – 0.60 | 0.45 | 0.651 | 5050.86 |
| pParty Ind × wave lin | 0.09 | -0.54 – 0.72 | 0.29 | 0.771 | 5158.91 |
| pParty Ind × wave quad | 0.25 | -0.23 – 0.74 | 1.02 | 0.308 | 5276.99 |
| pParty Ind × wave cub | 0.02 | -0.56 – 0.61 | 0.08 | 0.936 | 5155.69 |
| per gop 100 × wave lin | 0.21 | -0.30 – 0.73 | 0.81 | 0.416 | 5342.61 |
| per gop 100 × wave quad | 0.46 | 0.07 – 0.85 | 2.32 | 0.020 | 5191.42 |
| per gop 100 × wave cub | -0.15 | -0.61 – 0.32 | -0.62 | 0.538 | 5099.98 |
| pDem Rep × Pos Emo c | 0.33 | 0.14 – 0.53 | 3.32 | 0.001 | 2448.10 |
| pDem Rep × Neg Emo c | 0.05 | -0.17 – 0.26 | 0.45 | 0.655 | 2398.70 |
| pParty Ind × Pos Emo c | -0.04 | -0.31 – 0.23 | -0.30 | 0.767 | 4333.42 |
| pParty Ind × Neg Emo c | 0.19 | -0.07 – 0.46 | 1.42 | 0.155 | 2727.51 |
| per gop 100 × Pos Emo c | 0.41 | 0.19 – 0.64 | 3.64 | <0.001 | 2714.81 |
| per gop 100 × Neg Emo c | -0.10 | -0.32 – 0.12 | -0.87 | 0.383 | 2105.38 |
| wave lin × Pos Emo c | -0.23 | -0.45 – -0.01 | -2.07 | 0.039 | 5222.22 |
| wave lin × Neg Emo c | -0.06 | -0.28 – 0.16 | -0.55 | 0.581 | 5361.55 |
| wave quad × Pos Emo c | 0.01 | -0.15 – 0.18 | 0.13 | 0.894 | 5288.57 |
| wave quad × Neg Emo c | 0.10 | -0.07 – 0.26 | 1.15 | 0.251 | 5117.40 |
| wave cub × Pos Emo c | -0.03 | -0.23 – 0.17 | -0.31 | 0.760 | 5325.31 |
| wave cub × Neg Emo c | 0.04 | -0.15 – 0.24 | 0.44 | 0.657 | 4879.18 |
|
(pDem Rep × per gop 100) × wave lin |
0.67 | -0.36 – 1.71 | 1.28 | 0.202 | 5540.32 |
|
(pDem Rep × per gop 100) × wave quad |
0.70 | -0.08 – 1.48 | 1.76 | 0.078 | 5322.75 |
|
(pDem Rep × per gop 100) × wave cub |
-0.80 | -1.75 – 0.15 | -1.65 | 0.099 | 5037.40 |
|
(pParty Ind × per gop 100) × wave lin |
-0.00 | -1.26 – 1.26 | -0.00 | 0.997 | 5145.05 |
|
(pParty Ind × per gop 100) × wave quad |
-0.31 | -1.28 – 0.65 | -0.64 | 0.521 | 5250.88 |
|
(pParty Ind × per gop 100) × wave cub |
-0.01 | -1.16 – 1.14 | -0.02 | 0.985 | 5148.05 |
|
(pDem Rep × per gop 100) × Pos Emo c |
-0.63 | -1.02 – -0.24 | -3.17 | 0.002 | 2430.43 |
|
(pDem Rep × per gop 100) × Neg Emo c |
-0.20 | -0.63 – 0.22 | -0.93 | 0.351 | 2351.15 |
|
(pParty Ind × per gop 100) × Pos Emo c |
0.13 | -0.41 – 0.66 | 0.46 | 0.645 | 4206.57 |
|
(pParty Ind × per gop 100) × Neg Emo c |
-0.27 | -0.80 – 0.26 | -0.99 | 0.324 | 2785.07 |
|
(pDem Rep × wave lin) × Pos Emo c |
0.34 | -0.04 – 0.72 | 1.76 | 0.079 | 5560.13 |
|
(pDem Rep × wave lin) × Neg Emo c |
0.24 | -0.20 – 0.67 | 1.07 | 0.285 | 5440.35 |
|
(pDem Rep × wave quad) × Pos Emo c |
0.03 | -0.25 – 0.31 | 0.18 | 0.856 | 5400.60 |
|
(pDem Rep × wave quad) × Neg Emo c |
-0.43 | -0.74 – -0.11 | -2.66 | 0.008 | 5237.32 |
|
(pDem Rep × wave cub) × Pos Emo c |
-0.16 | -0.49 – 0.17 | -0.94 | 0.345 | 5205.82 |
|
(pDem Rep × wave cub) × Neg Emo c |
0.02 | -0.35 – 0.40 | 0.12 | 0.907 | 5030.03 |
|
(pParty Ind × wave lin) × Pos Emo c |
0.06 | -0.52 – 0.64 | 0.20 | 0.840 | 5133.73 |
|
(pParty Ind × wave lin) × Neg Emo c |
0.19 | -0.35 – 0.73 | 0.68 | 0.496 | 5324.69 |
|
(pParty Ind × wave quad) × Pos Emo c |
-0.26 | -0.69 – 0.17 | -1.17 | 0.242 | 5237.83 |
|
(pParty Ind × wave quad) × Neg Emo c |
0.22 | -0.19 – 0.63 | 1.06 | 0.289 | 5090.25 |
|
(pParty Ind × wave cub) × Pos Emo c |
0.25 | -0.27 – 0.78 | 0.95 | 0.343 | 5430.00 |
|
(pParty Ind × wave cub) × Neg Emo c |
0.11 | -0.39 – 0.61 | 0.43 | 0.665 | 4868.24 |
|
(per gop 100 × wave lin) × Pos Emo c |
0.52 | 0.08 – 0.96 | 2.34 | 0.020 | 5211.30 |
|
(per gop 100 × wave lin) × Neg Emo c |
0.07 | -0.37 – 0.51 | 0.29 | 0.772 | 5374.47 |
|
(per gop 100 × wave quad) × Pos Emo c |
0.05 | -0.27 – 0.38 | 0.31 | 0.758 | 5269.03 |
|
(per gop 100 × wave quad) × Neg Emo c |
-0.24 | -0.57 – 0.09 | -1.43 | 0.152 | 5114.83 |
|
(per gop 100 × wave cub) × Pos Emo c |
-0.01 | -0.39 – 0.38 | -0.04 | 0.971 | 5304.76 |
|
(per gop 100 × wave cub) × Neg Emo c |
-0.00 | -0.39 – 0.39 | -0.01 | 0.995 | 4880.21 |
|
(pDem Rep × per gop 100 × wave lin) × Pos Emo c |
-0.78 | -1.54 – -0.03 | -2.04 | 0.041 | 5520.23 |
|
(pDem Rep × per gop 100 × wave lin) × Neg Emo c |
-0.05 | -0.91 – 0.81 | -0.11 | 0.912 | 5468.28 |
|
(pDem Rep × per gop 100 × wave quad) × Pos Emo c |
0.06 | -0.49 – 0.62 | 0.22 | 0.827 | 5370.51 |
|
(pDem Rep × per gop 100 × wave quad) × Neg Emo c |
0.72 | 0.09 – 1.35 | 2.25 | 0.025 | 5261.50 |
|
(pDem Rep × per gop 100 × wave cub) × Pos Emo c |
0.40 | -0.25 – 1.05 | 1.20 | 0.232 | 5180.06 |
|
(pDem Rep × per gop 100 × wave cub) × Neg Emo c |
0.03 | -0.71 – 0.78 | 0.08 | 0.935 | 5030.77 |
|
(pParty Ind × per gop 100 × wave lin) × Pos Emo c |
0.43 | -0.72 – 1.58 | 0.73 | 0.462 | 5130.77 |
|
(pParty Ind × per gop 100 × wave lin) × Neg Emo c |
-0.60 | -1.70 – 0.50 | -1.06 | 0.288 | 5339.04 |
|
(pParty Ind × per gop 100 × wave quad) × Pos Emo c |
0.59 | -0.26 – 1.44 | 1.36 | 0.174 | 5218.02 |
|
(pParty Ind × per gop 100 × wave quad) × Neg Emo c |
-0.44 | -1.26 – 0.38 | -1.04 | 0.298 | 5060.91 |
|
(pParty Ind × per gop 100 × wave cub) × Pos Emo c |
-0.50 | -1.54 – 0.53 | -0.96 | 0.338 | 5411.42 |
|
(pParty Ind × per gop 100 × wave cub) × Neg Emo c |
-0.26 | -1.25 – 0.72 | -0.53 | 0.599 | 4856.27 |
| Random Effects | |||||
| σ2 | 0.42 | ||||
| τ00 pid | 0.57 | ||||
| τ11 pid.Pos_Emo.c | 0.05 | ||||
| τ11 pid.Neg_Emo.c | 0.03 | ||||
| ρ01 | -0.32 | ||||
| 0.20 | |||||
| ICC | 0.63 | ||||
| N pid | 2456 | ||||
| Observations | 7616 | ||||
| Marginal R2 / Conditional R2 | 0.190 / 0.697 | ||||
Significant Effects
Main Effects
- Dem_Rep: b = -0.73, t(4887.31) = -4.36, p < .001
2-way Interactions
Dem_Rep × wave lin: b = 0.65, t(5526.69) = 2.43, p = .015
per gop × wave quad: b = 0.46, t(5191.42) = 2.32, p = .020
Dem_Rep × Pos Emo: b = 0.33, t(2448.10) = 3.32, p = .001
per gop.100 × Pos Emo.c: b = 0.41, t(2714.81) = 3.46, p < .001
wave lin × Pos Emo c: b = -0.23, t(5222.22) = -2.07, p = .039
3-way Interactions
Dem_Rep × per gop 100 × Pos Emo c: b = -0.63, t(2430.43) = -3.17, p = .002
Dem_Rep × wave quad × Neg Emo c: b = -0.43, t(5237.32) = -2.66, p = .008
per gop 100 × wave lin × Pos Emo c: b = 0.52, t(5211.30) = 2.34, p = .020
4-way Interactions
Dem_Rep × per gop 100 × wave lin × Pos Emo c: b = -0.78, t(5520.23) = -2.04, p = .041
Dem_Rep × per gop 100 × wave quad × Neg Emo c: b = 0.72, t(5261.50) = 2.25, p = .025
i. Emotions as mediator?
m2.med1 <- lmer(voteconfidence ~ (pDem_Rep + pParty_Ind) * (wave.lin + wave.quad + wave.cub) + per_gop.100
+ vs_age
+ vs_race
+ nonbinary_mf
+ male_female
+ (1|pid),
data = d)
summary(m2.med1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: voteconfidence ~ (pDem_Rep + pParty_Ind) * (wave.lin + wave.quad +
## wave.cub) + per_gop.100 + vs_age + vs_race + nonbinary_mf +
## male_female + (1 | pid)
## Data: d
##
## REML criterion at convergence: 20733.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.7722 -0.5091 0.0599 0.5329 3.8359
##
## Random effects:
## Groups Name Variance Std.Dev.
## pid (Intercept) 0.6552 0.8095
## Residual 0.5429 0.7368
## Number of obs: 7616, groups: pid, 2456
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 3.184e+00 1.244e-01 2.489e+03 25.599 < 2e-16 ***
## pDem_Rep -1.715e-01 3.908e-02 3.715e+03 -4.388 1.17e-05 ***
## pParty_Ind -3.911e-01 4.831e-02 6.240e+03 -8.094 6.87e-16 ***
## wave.lin 1.796e-01 2.930e-02 5.618e+03 6.129 9.42e-10 ***
## wave.quad 2.365e-01 2.208e-02 5.402e+03 10.714 < 2e-16 ***
## wave.cub -1.043e-01 2.714e-02 5.311e+03 -3.841 0.000124 ***
## per_gop.100 -5.534e-01 1.493e-01 2.405e+03 -3.707 0.000214 ***
## vs_age 6.943e-03 1.336e-03 2.413e+03 5.198 2.18e-07 ***
## vs_raceBlack -2.330e-01 5.939e-02 2.484e+03 -3.923 8.98e-05 ***
## vs_raceHispanic -3.995e-02 6.169e-02 2.442e+03 -0.648 0.517368
## vs_raceOther -1.165e-01 7.343e-02 2.385e+03 -1.586 0.112844
## nonbinary_mf 5.295e-01 2.324e-01 2.546e+03 2.279 0.022761 *
## male_female -3.749e-01 3.862e-02 2.392e+03 -9.708 < 2e-16 ***
## pDem_Rep:wave.lin 1.204e+00 5.171e-02 5.581e+03 23.283 < 2e-16 ***
## pDem_Rep:wave.quad 5.337e-01 3.850e-02 5.380e+03 13.862 < 2e-16 ***
## pDem_Rep:wave.cub -6.393e-01 4.651e-02 5.294e+03 -13.746 < 2e-16 ***
## pParty_Ind:wave.lin 3.295e-02 7.639e-02 5.687e+03 0.431 0.666230
## pParty_Ind:wave.quad 1.441e-02 5.803e-02 5.486e+03 0.248 0.803927
## pParty_Ind:wave.cub 8.453e-02 7.167e-02 5.384e+03 1.179 0.238281
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1m2.med2 <- lmer(voteconfidence ~ (pDem_Rep + pParty_Ind) * (wave.lin + wave.quad + wave.cub) + Pos_Emo.c + Neg_Emo.c
+ vs_age
+ vs_race
+ nonbinary_mf
+ male_female
+ (1|pid),
data = d)
summary(m2.med2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: voteconfidence ~ (pDem_Rep + pParty_Ind) * (wave.lin + wave.quad +
## wave.cub) + Pos_Emo.c + Neg_Emo.c + vs_age + vs_race + nonbinary_mf +
## male_female + (1 | pid)
## Data: d
##
## REML criterion at convergence: 21517.8
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.9487 -0.5074 0.0437 0.5430 4.1273
##
## Random effects:
## Groups Name Variance Std.Dev.
## pid (Intercept) 0.6382 0.7988
## Residual 0.5055 0.7110
## Number of obs: 8079, groups: pid, 2606
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 3.145e+00 9.653e-02 2.671e+03 32.584 < 2e-16 ***
## pDem_Rep -4.907e-01 3.933e-02 4.795e+03 -12.478 < 2e-16 ***
## pParty_Ind -3.173e-01 4.561e-02 6.767e+03 -6.957 3.81e-12 ***
## wave.lin 2.332e-01 2.763e-02 5.967e+03 8.442 < 2e-16 ***
## wave.quad 2.227e-01 2.065e-02 5.715e+03 10.782 < 2e-16 ***
## wave.cub -1.120e-01 2.534e-02 5.624e+03 -4.421 1.00e-05 ***
## Pos_Emo.c 1.965e-01 1.018e-02 7.966e+03 19.305 < 2e-16 ***
## Neg_Emo.c -8.143e-02 1.045e-02 8.006e+03 -7.794 7.31e-15 ***
## vs_age 4.379e-03 1.279e-03 2.578e+03 3.423 0.000628 ***
## vs_raceBlack -3.562e-01 5.703e-02 2.678e+03 -6.246 4.89e-10 ***
## vs_raceHispanic -1.070e-01 5.915e-02 2.615e+03 -1.808 0.070679 .
## vs_raceOther -9.140e-02 7.089e-02 2.527e+03 -1.289 0.197424
## nonbinary_mf 2.763e-01 2.210e-01 2.668e+03 1.250 0.211293
## male_female -3.598e-01 3.682e-02 2.541e+03 -9.772 < 2e-16 ***
## pDem_Rep:wave.lin 8.654e-01 5.106e-02 6.124e+03 16.949 < 2e-16 ***
## pDem_Rep:wave.quad 3.079e-01 3.760e-02 5.892e+03 8.191 3.15e-16 ***
## pDem_Rep:wave.cub -4.228e-01 4.482e-02 5.744e+03 -9.435 < 2e-16 ***
## pParty_Ind:wave.lin 4.353e-02 7.172e-02 6.026e+03 0.607 0.543899
## pParty_Ind:wave.quad 1.641e-03 5.418e-02 5.798e+03 0.030 0.975843
## pParty_Ind:wave.cub 8.134e-02 6.669e-02 5.691e+03 1.220 0.222631
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1m2.med3 <- lmer(voteconfidence ~ (pDem_Rep + pParty_Ind) * (wave.lin + wave.quad + wave.cub) + per_gop.100 + Pos_Emo.c + Neg_Emo.c
+ vs_age
+ vs_race
+ nonbinary_mf
+ male_female
+ (1|pid),
data = d)
summary(m2.med3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: voteconfidence ~ (pDem_Rep + pParty_Ind) * (wave.lin + wave.quad +
## wave.cub) + per_gop.100 + Pos_Emo.c + Neg_Emo.c + vs_age +
## vs_race + nonbinary_mf + male_female + (1 | pid)
## Data: d
##
## REML criterion at convergence: 20241.7
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.9994 -0.5124 0.0451 0.5427 4.1405
##
## Random effects:
## Groups Name Variance Std.Dev.
## pid (Intercept) 0.6348 0.7967
## Residual 0.5023 0.7087
## Number of obs: 7616, groups: pid, 2456
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 3.448e+00 1.223e-01 2.511e+03 28.183 < 2e-16 ***
## pDem_Rep -4.599e-01 4.062e-02 4.540e+03 -11.322 < 2e-16 ***
## pParty_Ind -3.135e-01 4.711e-02 6.390e+03 -6.656 3.06e-11 ***
## wave.lin 2.430e-01 2.835e-02 5.612e+03 8.571 < 2e-16 ***
## wave.quad 2.225e-01 2.127e-02 5.390e+03 10.459 < 2e-16 ***
## wave.cub -1.197e-01 2.615e-02 5.301e+03 -4.575 4.87e-06 ***
## per_gop.100 -6.178e-01 1.462e-01 2.399e+03 -4.225 2.48e-05 ***
## Pos_Emo.c 1.963e-01 1.041e-02 7.496e+03 18.851 < 2e-16 ***
## Neg_Emo.c -8.058e-02 1.076e-02 7.541e+03 -7.492 7.54e-14 ***
## vs_age 4.407e-03 1.313e-03 2.429e+03 3.358 0.000798 ***
## vs_raceBlack -3.723e-01 5.847e-02 2.529e+03 -6.366 2.29e-10 ***
## vs_raceHispanic -1.215e-01 6.053e-02 2.448e+03 -2.008 0.044806 *
## vs_raceOther -1.179e-01 7.192e-02 2.378e+03 -1.639 0.101367
## nonbinary_mf 3.181e-01 2.277e-01 2.541e+03 1.397 0.162419
## male_female -3.536e-01 3.786e-02 2.391e+03 -9.339 < 2e-16 ***
## pDem_Rep:wave.lin 8.683e-01 5.231e-02 5.765e+03 16.599 < 2e-16 ***
## pDem_Rep:wave.quad 2.953e-01 3.853e-02 5.550e+03 7.663 2.14e-14 ***
## pDem_Rep:wave.cub -4.216e-01 4.592e-02 5.411e+03 -9.182 < 2e-16 ***
## pParty_Ind:wave.lin 7.038e-02 7.355e-02 5.661e+03 0.957 0.338666
## pParty_Ind:wave.quad 1.479e-03 5.586e-02 5.465e+03 0.026 0.978874
## pParty_Ind:wave.cub 6.001e-02 6.897e-02 5.365e+03 0.870 0.384286
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1No evidence of mediation (with emotions entered as mediator for effect of geographic partisan norms).
c. Media vs. Geography Comparison
PEL.m.normplot <- lmer(voteconfidence ~ party_factor * pollbias.z * wave * per_gop
+ pollAcc
+ pollFreq.c
+ vs_age
+ nonbinary_mf
+ male_female
+ vs_race
+ (1|pid),
data = d)
ggplot(d[!is.na(d$party_factor),],
aes(x = per_gop,
y = voteconfidence,
fill = party_factor,
color = party_factor)) +
geom_smooth(method = "lm",
fullrange = T) +
facet_grid(~wave, labeller = as_labeller(wave_label)) +
labs(x = "Percentage of County voting for Donald Trump",
y = "Perceived Election Legitimacy") +
scale_fill_manual("Participant
Partisan Identity",
values = c("#1696d2","grey","#db2b27")) +
scale_color_manual("Participant
Partisan Identity",
values = c("#1696d2","grey","#db2b27")) +
theme_bw() +
coord_cartesian(ylim = c(1,5)) +
scale_x_continuous(breaks = seq(from = 0, to = 90, length.out = 7))
PEL.m.normplot <- lmer(voteconfidence ~ party_factor * pollbias.z * wave * per_gop +
pollAcc + pollFreq.c + vs_age + nonbinary_mf + male_female +
vs_race + (1 | pid),
data = d)
interactions::interact_plot(PEL.m.normplot,
pred = per_gop,
modx = pollbias.z,
mod2 = party_factor,
x.label = "% of County Voted for Trump",
y.label = "Perceived Election Legitimacy",
legend.main = "Polling Bias",
modx.labels = c("Left-Lean",
"Neutral",
"Right-Lean"),
mod2.labels = c("Democrats",
"Independents",
"Republicans"))
Summary
With both media bias and geographic social norms entered as moderators in a model predicting perceived election legitimacy (PEL) WITH emotions entered as a moderator too–Trump vote share retained significance, but media bias lost significance. Further, media bias was not involved in any higher level interactions.
Trump vote share was a significant negative predictor of PEL (b = -0.52, 95% CI = [-0.97, -0.08], t(2496.92) = -2.30, p = .021). Trump vote share was involved in several higher order interactions, including a 3-way interaction with partisan identity (Democratic vs. Republican; b = -0.78, 95% CI = [-0.97, -0.08], t(2410.96) = -3.15, p = .002), and two 4-way interactions. One 4-way interaction was with positive emotions, linear time, and partisan identity (Democratic vs. Republican; b = -1.22, 95% CI = [-2.23, -0.20], t(4196.73) = -2.35, p = .019) and one was with negative emotions, quadratic time, and partisan identity (Democratic vs. Republican; b = 0.98, 95% CI = [0.18, 1.77], t(4082.22) = 2.41, p = .016).
Models
Main Model
m3 <- lmer(voteconfidence ~ (pDem_Rep + pParty_Ind) * pollbias.z * per_gop.100 * (wave.lin + wave.quad + wave.cub) * (Pos_Emo.c + Neg_Emo.c)
+ pollAcc
+ pollFreq.c
+ vs_age
+ nonbinary_mf
+ male_female
+ vs_race
+ (Pos_Emo.c | pid),
data = d)
tab_model(m3, show.stat = T, show.df = T, df.method = "satterthwaite")| voteconfidence | |||||
|---|---|---|---|---|---|
| Predictors | Estimates | CI | Statistic | p | df |
| (Intercept) | 3.26 | 2.92 – 3.60 | 18.78 | <0.001 | 1790.08 |
| pDem Rep | -0.63 | -1.02 – -0.23 | -3.08 | 0.002 | 3028.52 |
| pParty Ind | -0.07 | -0.57 – 0.42 | -0.29 | 0.772 | 4399.59 |
| pollbias z | -0.61 | -1.31 – 0.08 | -1.73 | 0.084 | 2811.28 |
| per gop 100 | -0.52 | -0.97 – -0.08 | -2.30 | 0.021 | 2496.92 |
| wave lin | 0.10 | -0.23 – 0.44 | 0.61 | 0.541 | 3962.31 |
| wave quad | -0.10 | -0.36 – 0.15 | -0.79 | 0.431 | 3789.14 |
| wave cub | 0.18 | -0.14 – 0.49 | 1.09 | 0.278 | 3886.80 |
| Pos Emo c | 0.08 | -0.07 – 0.23 | 1.06 | 0.287 | 3212.86 |
| Neg Emo c | 0.06 | -0.09 – 0.22 | 0.81 | 0.417 | 5232.15 |
| pollAcc | 0.06 | 0.04 – 0.08 | 5.89 | <0.001 | 4895.08 |
| pollFreq c | -0.01 | -0.03 – 0.01 | -1.44 | 0.149 | 1525.56 |
| vs age | 0.00 | 0.00 – 0.01 | 2.75 | 0.006 | 1586.50 |
| nonbinary mf | 0.88 | 0.21 – 1.55 | 2.59 | 0.010 | 1018.14 |
| male female | -0.29 | -0.38 – -0.21 | -7.03 | <0.001 | 1495.15 |
| vs race [Black] | -0.32 | -0.44 – -0.19 | -4.97 | <0.001 | 1631.72 |
| vs race [Hispanic] | -0.10 | -0.23 – 0.03 | -1.51 | 0.130 | 1520.02 |
| vs race [Other] | -0.15 | -0.31 – 0.01 | -1.81 | 0.071 | 1574.67 |
| pDem Rep × pollbias z | -0.32 | -1.49 – 0.86 | -0.53 | 0.594 | 3068.86 |
| pParty Ind × pollbias z | -0.07 | -1.65 – 1.50 | -0.09 | 0.928 | 4769.05 |
| pDem Rep × per gop 100 | 0.58 | -0.21 – 1.37 | 1.45 | 0.148 | 2909.38 |
| pParty Ind × per gop 100 | -0.23 | -1.23 – 0.76 | -0.46 | 0.645 | 4373.04 |
| pollbias z × per gop 100 | 0.09 | -1.29 – 1.48 | 0.13 | 0.896 | 2824.99 |
| pDem Rep × wave lin | 0.38 | -0.26 – 1.01 | 1.16 | 0.244 | 4072.35 |
| pDem Rep × wave quad | -0.24 | -0.72 – 0.24 | -0.99 | 0.321 | 3939.16 |
| pDem Rep × wave cub | 0.06 | -0.52 – 0.64 | 0.20 | 0.845 | 3785.87 |
| pParty Ind × wave lin | 0.29 | -0.56 – 1.14 | 0.67 | 0.504 | 3888.46 |
| pParty Ind × wave quad | 0.32 | -0.33 – 0.96 | 0.96 | 0.339 | 3819.53 |
| pParty Ind × wave cub | 0.40 | -0.43 – 1.22 | 0.95 | 0.345 | 3956.57 |
| pollbias z × wave lin | 0.24 | -0.87 – 1.35 | 0.42 | 0.675 | 4037.85 |
| pollbias z × wave quad | 0.15 | -0.67 – 0.97 | 0.37 | 0.713 | 3831.47 |
| pollbias z × wave cub | -0.28 | -1.29 – 0.74 | -0.54 | 0.591 | 3886.69 |
| per gop 100 × wave lin | 0.34 | -0.33 – 1.02 | 0.99 | 0.320 | 3935.89 |
| per gop 100 × wave quad | 0.39 | -0.12 – 0.90 | 1.50 | 0.133 | 3779.56 |
| per gop 100 × wave cub | -0.46 | -1.09 – 0.17 | -1.43 | 0.152 | 3862.71 |
| pDem Rep × Pos Emo c | 0.40 | 0.16 – 0.65 | 3.22 | 0.001 | 2429.39 |
| pDem Rep × Neg Emo c | 0.07 | -0.18 – 0.31 | 0.54 | 0.591 | 5077.47 |
| pParty Ind × Pos Emo c | 0.20 | -0.17 – 0.57 | 1.05 | 0.292 | 4349.92 |
| pParty Ind × Neg Emo c | 0.45 | 0.05 – 0.84 | 2.20 | 0.028 | 5174.77 |
| pollbias z × Pos Emo c | -0.20 | -0.68 – 0.27 | -0.84 | 0.403 | 3497.14 |
| pollbias z × Neg Emo c | -0.09 | -0.59 – 0.41 | -0.35 | 0.724 | 5240.46 |
| per gop 100 × Pos Emo c | 0.30 | -0.00 – 0.59 | 1.93 | 0.054 | 3217.42 |
| per gop 100 × Neg Emo c | -0.26 | -0.57 – 0.05 | -1.63 | 0.102 | 5208.92 |
| wave lin × Pos Emo c | -0.21 | -0.53 – 0.10 | -1.33 | 0.185 | 4128.49 |
| wave lin × Neg Emo c | 0.16 | -0.16 – 0.48 | 0.96 | 0.338 | 4117.93 |
| wave quad × Pos Emo c | 0.11 | -0.12 – 0.34 | 0.94 | 0.348 | 4047.22 |
| wave quad × Neg Emo c | 0.05 | -0.19 – 0.28 | 0.39 | 0.698 | 3894.29 |
| wave cub × Pos Emo c | -0.06 | -0.33 – 0.22 | -0.39 | 0.694 | 4068.19 |
| wave cub × Neg Emo c | -0.01 | -0.30 – 0.27 | -0.10 | 0.919 | 3810.44 |
|
(pDem Rep × pollbias z) × per gop 100 |
1.04 | -1.29 – 3.38 | 0.88 | 0.380 | 2979.50 |
|
(pParty Ind × pollbias z) × per gop 100 |
-0.36 | -3.54 – 2.82 | -0.22 | 0.826 | 4749.22 |
|
(pDem Rep × pollbias z) × wave lin |
0.34 | -1.57 – 2.25 | 0.34 | 0.730 | 4095.35 |
|
(pDem Rep × pollbias z) × wave quad |
0.31 | -1.13 – 1.75 | 0.42 | 0.674 | 3967.66 |
|
(pDem Rep × pollbias z) × wave cub |
-1.02 | -2.75 – 0.71 | -1.16 | 0.247 | 3837.99 |
|
(pParty Ind × pollbias z) × wave lin |
-0.16 | -3.05 – 2.73 | -0.11 | 0.915 | 4004.90 |
|
(pParty Ind × pollbias z) × wave quad |
-1.51 | -3.67 – 0.64 | -1.37 | 0.170 | 3867.15 |
|
(pParty Ind × pollbias z) × wave cub |
0.12 | -2.58 – 2.83 | 0.09 | 0.930 | 3961.53 |
|
(pDem Rep × per gop 100) × wave lin |
1.08 | -0.16 – 2.31 | 1.71 | 0.087 | 4065.35 |
|
(pDem Rep × per gop 100) × wave quad |
0.80 | -0.13 – 1.73 | 1.68 | 0.093 | 3919.15 |
|
(pDem Rep × per gop 100) × wave cub |
-0.71 | -1.83 – 0.42 | -1.23 | 0.218 | 3759.65 |
|
(pParty Ind × per gop 100) × wave lin |
-0.38 | -2.11 – 1.36 | -0.42 | 0.672 | 3853.85 |
|
(pParty Ind × per gop 100) × wave quad |
-0.63 | -1.94 – 0.68 | -0.95 | 0.345 | 3806.34 |
|
(pParty Ind × per gop 100) × wave cub |
-0.61 | -2.25 – 1.03 | -0.73 | 0.468 | 3935.10 |
|
(pollbias z × per gop 100) × wave lin |
0.05 | -2.21 – 2.31 | 0.04 | 0.966 | 3999.39 |
|
(pollbias z × per gop 100) × wave quad |
0.02 | -1.63 – 1.67 | 0.03 | 0.979 | 3794.74 |
|
(pollbias z × per gop 100) × wave cub |
0.66 | -1.35 – 2.67 | 0.64 | 0.522 | 3862.75 |
|
(pDem Rep × pollbias z) × Pos Emo c |
-0.21 | -0.92 – 0.51 | -0.57 | 0.568 | 2433.69 |
|
(pDem Rep × pollbias z) × Neg Emo c |
-0.07 | -0.79 – 0.65 | -0.18 | 0.853 | 5085.06 |
|
(pParty Ind × pollbias z) × Pos Emo c |
0.03 | -1.19 – 1.26 | 0.06 | 0.956 | 4303.91 |
|
(pParty Ind × pollbias z) × Neg Emo c |
0.35 | -0.99 – 1.69 | 0.52 | 0.605 | 5226.75 |
|
(pDem Rep × per gop 100) × Pos Emo c |
-0.78 | -1.27 – -0.29 | -3.15 | 0.002 | 2410.96 |
|
(pDem Rep × per gop 100) × Neg Emo c |
-0.24 | -0.73 – 0.25 | -0.97 | 0.330 | 5114.77 |
|
(pParty Ind × per gop 100) × Pos Emo c |
-0.34 | -1.08 – 0.40 | -0.89 | 0.373 | 4290.31 |
|
(pParty Ind × per gop 100) × Neg Emo c |
-0.79 | -1.61 – 0.02 | -1.91 | 0.057 | 5139.02 |
|
(pollbias z × per gop 100) × Pos Emo c |
0.62 | -0.35 – 1.59 | 1.25 | 0.211 | 3572.53 |
|
(pollbias z × per gop 100) × Neg Emo c |
0.14 | -0.89 – 1.17 | 0.27 | 0.787 | 5235.34 |
|
(pDem Rep × wave lin) × Pos Emo c |
0.55 | 0.04 – 1.06 | 2.11 | 0.035 | 4201.83 |
|
(pDem Rep × wave lin) × Neg Emo c |
0.02 | -0.52 – 0.57 | 0.09 | 0.929 | 4230.89 |
|
(pDem Rep × wave quad) × Pos Emo c |
0.09 | -0.29 – 0.47 | 0.47 | 0.637 | 4167.13 |
|
(pDem Rep × wave quad) × Neg Emo c |
-0.52 | -0.92 – -0.12 | -2.53 | 0.011 | 4069.86 |
|
(pDem Rep × wave cub) × Pos Emo c |
-0.09 | -0.52 – 0.35 | -0.38 | 0.702 | 3979.08 |
|
(pDem Rep × wave cub) × Neg Emo c |
0.00 | -0.47 – 0.47 | 0.00 | 0.998 | 3955.14 |
|
(pParty Ind × wave lin) × Pos Emo c |
0.21 | -0.62 – 1.05 | 0.50 | 0.618 | 4147.38 |
|
(pParty Ind × wave lin) × Neg Emo c |
1.01 | 0.17 – 1.85 | 2.35 | 0.019 | 4077.51 |
|
(pParty Ind × wave quad) × Pos Emo c |
-0.06 | -0.67 – 0.54 | -0.20 | 0.840 | 4004.29 |
|
(pParty Ind × wave quad) × Neg Emo c |
0.10 | -0.52 – 0.72 | 0.32 | 0.749 | 3852.43 |
|
(pParty Ind × wave cub) × Pos Emo c |
0.05 | -0.69 – 0.80 | 0.14 | 0.890 | 4194.54 |
|
(pParty Ind × wave cub) × Neg Emo c |
-0.09 | -0.85 – 0.66 | -0.24 | 0.811 | 3830.39 |
|
(pollbias z × wave lin) × Pos Emo c |
-0.40 | -1.49 – 0.69 | -0.72 | 0.474 | 4199.62 |
|
(pollbias z × wave lin) × Neg Emo c |
0.07 | -0.96 – 1.10 | 0.13 | 0.895 | 4138.47 |
|
(pollbias z × wave quad) × Pos Emo c |
0.48 | -0.29 – 1.24 | 1.23 | 0.220 | 4080.99 |
|
(pollbias z × wave quad) × Neg Emo c |
0.66 | -0.09 – 1.42 | 1.72 | 0.085 | 3908.38 |
|
(pollbias z × wave cub) × Pos Emo c |
-0.47 | -1.33 – 0.40 | -1.06 | 0.291 | 4129.45 |
|
(pollbias z × wave cub) × Neg Emo c |
0.46 | -0.45 – 1.36 | 0.99 | 0.321 | 3806.22 |
|
(per gop 100 × wave lin) × Pos Emo c |
0.50 | -0.14 – 1.15 | 1.53 | 0.125 | 4099.45 |
|
(per gop 100 × wave lin) × Neg Emo c |
-0.36 | -1.02 – 0.30 | -1.07 | 0.284 | 4114.44 |
|
(per gop 100 × wave quad) × Pos Emo c |
-0.17 | -0.64 – 0.29 | -0.73 | 0.466 | 4021.12 |
|
(per gop 100 × wave quad) × Neg Emo c |
-0.18 | -0.66 – 0.30 | -0.72 | 0.469 | 3873.86 |
|
(per gop 100 × wave cub) × Pos Emo c |
0.04 | -0.50 – 0.58 | 0.14 | 0.885 | 4052.77 |
|
(per gop 100 × wave cub) × Neg Emo c |
0.05 | -0.52 – 0.62 | 0.16 | 0.870 | 3799.82 |
|
(pDem Rep × pollbias z × per gop 100) × wave lin |
-0.28 | -4.07 – 3.51 | -0.14 | 0.886 | 4090.61 |
|
(pDem Rep × pollbias z × per gop 100) × wave quad |
-0.05 | -2.88 – 2.77 | -0.04 | 0.972 | 3944.92 |
|
(pDem Rep × pollbias z × per gop 100) × wave cub |
2.06 | -1.33 – 5.44 | 1.19 | 0.233 | 3823.22 |
|
(pParty Ind × pollbias z × per gop 100) × wave lin |
0.67 | -5.27 – 6.60 | 0.22 | 0.825 | 3953.93 |
|
(pParty Ind × pollbias z × per gop 100) × wave quad |
3.13 | -1.22 – 7.48 | 1.41 | 0.158 | 3822.65 |
|
(pParty Ind × pollbias z × per gop 100) × wave cub |
-0.04 | -5.41 – 5.34 | -0.01 | 0.989 | 3937.04 |
|
(pDem Rep × pollbias z × per gop 100) × Pos Emo c |
0.07 | -1.36 – 1.49 | 0.09 | 0.928 | 2428.90 |
|
(pDem Rep × pollbias z × per gop 100) × Neg Emo c |
0.03 | -1.41 – 1.48 | 0.05 | 0.963 | 5103.01 |
|
(pParty Ind × pollbias z × per gop 100) × Pos Emo c |
-0.03 | -2.55 – 2.48 | -0.03 | 0.978 | 4333.33 |
|
(pParty Ind × pollbias z × per gop 100) × Neg Emo c |
-0.95 | -3.71 – 1.82 | -0.67 | 0.502 | 5199.35 |
|
(pDem Rep × pollbias z × wave lin) × Pos Emo c |
0.18 | -1.35 – 1.70 | 0.23 | 0.819 | 4187.26 |
|
(pDem Rep × pollbias z × wave lin) × Neg Emo c |
-0.10 | -1.73 – 1.53 | -0.12 | 0.903 | 4213.22 |
|
(pDem Rep × pollbias z × wave quad) × Pos Emo c |
0.18 | -0.94 – 1.30 | 0.32 | 0.750 | 4197.21 |
|
(pDem Rep × pollbias z × wave quad) × Neg Emo c |
0.69 | -0.50 – 1.89 | 1.14 | 0.256 | 4147.84 |
|
(pDem Rep × pollbias z × wave cub) × Pos Emo c |
0.06 | -1.21 – 1.34 | 0.10 | 0.921 | 4009.74 |
|
(pDem Rep × pollbias z × wave cub) × Neg Emo c |
-0.00 | -1.39 – 1.38 | -0.01 | 0.996 | 4041.22 |
|
(pParty Ind × pollbias z × wave lin) × Pos Emo c |
0.11 | -2.91 – 3.13 | 0.07 | 0.944 | 4221.90 |
|
(pParty Ind × pollbias z × wave lin) × Neg Emo c |
0.42 | -2.34 – 3.19 | 0.30 | 0.763 | 4101.13 |
|
(pParty Ind × pollbias z × wave quad) × Pos Emo c |
0.49 | -1.58 – 2.56 | 0.46 | 0.644 | 4042.94 |
|
(pParty Ind × pollbias z × wave quad) × Neg Emo c |
0.70 | -1.32 – 2.72 | 0.68 | 0.496 | 3866.24 |
|
(pParty Ind × pollbias z × wave cub) × Pos Emo c |
-1.29 | -3.69 – 1.11 | -1.06 | 0.291 | 4220.14 |
|
(pParty Ind × pollbias z × wave cub) × Neg Emo c |
1.71 | -0.74 – 4.17 | 1.37 | 0.171 | 3806.09 |
|
(pDem Rep × per gop 100 × wave lin) × Pos Emo c |
-1.22 | -2.23 – -0.20 | -2.35 | 0.019 | 4196.73 |
|
(pDem Rep × per gop 100 × wave lin) × Neg Emo c |
0.23 | -0.86 – 1.31 | 0.41 | 0.681 | 4228.13 |
|
(pDem Rep × per gop 100 × wave quad) × Pos Emo c |
0.00 | -0.75 – 0.75 | 0.00 | 0.999 | 4144.55 |
|
(pDem Rep × per gop 100 × wave quad) × Neg Emo c |
0.98 | 0.18 – 1.77 | 2.41 | 0.016 | 4082.22 |
|
(pDem Rep × per gop 100 × wave cub) × Pos Emo c |
0.36 | -0.50 – 1.22 | 0.82 | 0.411 | 3951.93 |
|
(pDem Rep × per gop 100 × wave cub) × Neg Emo c |
0.03 | -0.90 – 0.95 | 0.05 | 0.957 | 3932.71 |
|
(pParty Ind × per gop 100 × wave lin) × Pos Emo c |
0.12 | -1.62 – 1.85 | 0.13 | 0.895 | 4115.37 |
|
(pParty Ind × per gop 100 × wave lin) × Neg Emo c |
-2.29 | -4.05 – -0.54 | -2.56 | 0.010 | 4081.65 |
|
(pParty Ind × per gop 100 × wave quad) × Pos Emo c |
0.12 | -1.12 – 1.35 | 0.18 | 0.854 | 3988.58 |
|
(pParty Ind × per gop 100 × wave quad) × Neg Emo c |
-0.36 | -1.63 – 0.90 | -0.56 | 0.576 | 3825.99 |
|
(pParty Ind × per gop 100 × wave cub) × Pos Emo c |
-0.16 | -1.64 – 1.33 | -0.21 | 0.837 | 4166.66 |
|
(pParty Ind × per gop 100 × wave cub) × Neg Emo c |
0.16 | -1.38 – 1.69 | 0.20 | 0.840 | 3816.65 |
|
(pollbias z × per gop 100 × wave lin) × Pos Emo c |
0.79 | -1.48 – 3.06 | 0.68 | 0.494 | 4183.06 |
|
(pollbias z × per gop 100 × wave lin) × Neg Emo c |
0.01 | -2.15 – 2.17 | 0.01 | 0.993 | 4120.59 |
|
(pollbias z × per gop 100 × wave quad) × Pos Emo c |
-0.81 | -2.38 – 0.75 | -1.02 | 0.309 | 4049.99 |
|
(pollbias z × per gop 100 × wave quad) × Neg Emo c |
-1.50 | -3.05 – 0.06 | -1.89 | 0.059 | 3879.46 |
|
(pollbias z × per gop 100 × wave cub) × Pos Emo c |
0.69 | -1.06 – 2.43 | 0.77 | 0.441 | 4104.22 |
|
(pollbias z × per gop 100 × wave cub) × Neg Emo c |
-1.02 | -2.87 – 0.82 | -1.09 | 0.277 | 3786.00 |
|
(pDem Rep × pollbias z × per gop 100 × wave lin) × Pos Emo c |
-0.96 | -4.01 – 2.10 | -0.61 | 0.539 | 4188.43 |
|
(pDem Rep × pollbias z × per gop 100 × wave lin) × Neg Emo c |
-0.14 | -3.43 – 3.15 | -0.08 | 0.933 | 4204.44 |
|
(pDem Rep × pollbias z × per gop 100 × wave quad) × Pos Emo c |
-0.33 | -2.56 – 1.90 | -0.29 | 0.771 | 4174.77 |
|
(pDem Rep × pollbias z × per gop 100 × wave quad) × Neg Emo c |
-1.11 | -3.50 – 1.29 | -0.90 | 0.366 | 4144.06 |
|
(pDem Rep × pollbias z × per gop 100 × wave cub) × Pos Emo c |
0.13 | -2.39 – 2.65 | 0.10 | 0.917 | 3988.97 |
|
(pDem Rep × pollbias z × per gop 100 × wave cub) × Neg Emo c |
0.63 | -2.11 – 3.38 | 0.45 | 0.652 | 4032.75 |
|
(pParty Ind × pollbias z × per gop 100 × wave lin) × Pos Emo c |
-1.02 | -7.33 – 5.28 | -0.32 | 0.750 | 4196.68 |
|
(pParty Ind × pollbias z × per gop 100 × wave lin) × Neg Emo c |
-0.85 | -6.68 – 4.97 | -0.29 | 0.774 | 4082.56 |
|
(pParty Ind × pollbias z × per gop 100 × wave quad) × Pos Emo c |
-0.11 | -4.39 – 4.16 | -0.05 | 0.958 | 4023.23 |
|
(pParty Ind × pollbias z × per gop 100 × wave quad) × Neg Emo c |
-1.90 | -6.09 – 2.28 | -0.89 | 0.373 | 3833.12 |
|
(pParty Ind × pollbias z × per gop 100 × wave cub) × Pos Emo c |
2.30 | -2.54 – 7.15 | 0.93 | 0.351 | 4181.93 |
|
(pParty Ind × pollbias z × per gop 100 × wave cub) × Neg Emo c |
-4.16 | -9.21 – 0.88 | -1.62 | 0.106 | 3779.25 |
| Random Effects | |||||
| σ2 | 0.43 | ||||
| τ00 pid | 0.51 | ||||
| τ11 pid.Pos_Emo.c | 0.04 | ||||
| ρ01 pid | -0.51 | ||||
| ICC | 0.58 | ||||
| N pid | 1578 | ||||
| Observations | 5428 | ||||
| Marginal R2 / Conditional R2 | 0.225 / 0.671 | ||||
Significant Effects
Main Effects
pDem Rep: b = -0.63, t(3028.52) = -3.08, p = .002
per gop: b = -0.52, t(2496.92) = -2.30, p = .021
2-way interactions
- pDem Rep × Pos Emo c: b = 0.40, t(2429.39) = 3.22, p = .001
3-way interactions
pDem Rep × per gop 100 × Pos Emo c: b = -0.78, t(2410.96) = -3.15, p = .002
pDem Rep × wave lin × Pos Emo c: b = 0.55, t(4201.83) = 2.11, p = .035
pDem Rep × wave quad × Neg Emo c: b = -0.52, t(4069.86) = -2.53, p = .011
pDem Rep × wave quad × Neg Emo c: b = -0.52, t(4069.86) = -2.53, p = .011
4-way interactions
pDem Rep × per gop 100 × wave lin × Pos Emo c: b = -1.22, t(4196.73) = -2.35, p = .019
pDem Rep × per gop 100 × wave quad × Neg Emo c: b = 0.98, t(4082.22) = 2.41, p = .016