Institutional Signaling Pilot 1

1 Setup

Libraries and functions

knitr::opts_chunk$set(warning = FALSE, message = FALSE) 

Mypackages <-
  c("lme4","tidyverse","effects","ggplot2","psych",
    "MASS","Rmisc","lmerTest","ggthemes", "knitr",
    "lsmeans","pastecs","sjstats","car","ordinal",
    "Rcpp","corrplot", "ggpubr", "EnvStats",
    "easyStats", "cowplot","see","datawizard", "ggcorrplot",
    "corrplot", "effects"
    )

# install.packages(Mypackages) #you must remove the # in this comment if you need to install the packages! 
lapply(Mypackages,
       require,
       character.only = TRUE)

options(knitr.kable.NA = '—')
set.seed(1)  

1.1 Load Data

# read in data files
gjg <-read.csv("/Users/mtrenfield17/Desktop/Research/Boston College Research/Institutional Signaling/Institutional Virtue Signaling Pilot 1.csv")

1.2 Functions

plot_cooker <- function(data, iv, dv, title) {
  part1 <- ggplot(data, aes(x = {{iv}}, y = {{dv}}, fill = {{iv}})) +
    geom_violin(alpha = 0.3, scale = "count") + 
  stat_summary(fun = "mean", geom = "point", size = 3, color = "black") +
    stat_summary(fun.data = mean_cl_normal, geom = "errorbar", width = 0.2,
                 #change to make a data set from allEffects with mean, low CI, high CI
                 size = 1.5, color = "black") +
    theme_classic() +
    xlab("") +
    ylab("") +
    ggtitle(title)
  ggpar(part1, legend = "none")
}

wrapped_plot_cooker <- function(data, iv, dv, title, facet_var) {
  part1 <- ggplot(data, aes(x = {{iv}}, y = {{dv}}, fill = {{iv}})) +
    geom_violin(alpha = 0.3, scale = "count") + 
  stat_summary(fun = "mean", geom = "point", size = 3, color = "black") +
    stat_summary(fun.data = mean_cl_normal, geom = "errorbar", width = 0.2, 
                 size = 1.5, color = "black") +
    theme_classic() +
    xlab("") +
    ylab("") +
    ggtitle(title) +
    facet_wrap(vars({{facet_var}}))
  ggpar(part1, legend = "none")
}

1.3 Reshaping data

#### filtering people who failed the attn check ####
filtered_gjg <- gjg %>% filter(attentionCheck == "Companies' Social Media")

#### make dataset long ####
gjg_long<-filtered_gjg %>% gather(stim, resp, "B_S_Chic_SGen":"C_N_Tar_Prior") 
gjg_long<-gjg_long %>%
  separate(stim, into= c("social_issue", "signal", "company", "DV"), sep="_")

## shift dataset back to wide format ##
gjg_long <- spread(gjg_long, DV, resp)

names(gjg_long)

##  [1] "StartDate"             "EndDate"               "Status"               
##  [4] "IPAddress"             "Progress"              "Duration..in.seconds."
##  [7] "Finished"              "RecordedDate"          "PID"                  
## [10] "ResponseId"            "RecipientLastName"     "RecipientFirstName"   
## [13] "RecipientEmail"        "ExternalReference"     "LocationLatitude"     
## [16] "LocationLongitude"     "DistributionChannel"   "UserLanguage"         
## [19] "prolificID"            "consent"               "attentionCheck"       
## [22] "inc"                   "edu"                   "age"                  
## [25] "gen"                   "gen_3_TEXT"            "pol"                  
## [28] "pid"                   "area"                  "race"                 
## [31] "post_Real_Check"       "Ent_explore"           "openFeedback"         
## [34] "confusion"             "PROLIFIC_PID"          "STUDY_ID"             
## [37] "SESSION_ID"            "cluster"               "social_issue"         
## [40] "signal"                "company"               "compCur"              
## [43] "dyNorm"                "Ent"                   "issueCur"             
## [46] "noisyMin"              "OComp"                 "OFriend"              
## [49] "OGen"                  "OImport"               "OLegit"               
## [52] "OLike"                 "OMotivate"             "ORelImport"           
## [55] "OTrust"                "peerPress"             "Prior"                
## [58] "Profit"                "pubPress"              "Rep"                  
## [61] "riskAct"               "riskInact"             "SComp"                
## [64] "SFriend"               "SGen"                  "SImport"              
## [67] "SLegit"                "SLike"                 "SRelImport"           
## [70] "staticNorm"            "STrust"                "sufficiency"          
## [73] "supGen"                "surprise"              "vote"

colnames(gjg_long)

##  [1] "StartDate"             "EndDate"               "Status"               
##  [4] "IPAddress"             "Progress"              "Duration..in.seconds."
##  [7] "Finished"              "RecordedDate"          "PID"                  
## [10] "ResponseId"            "RecipientLastName"     "RecipientFirstName"   
## [13] "RecipientEmail"        "ExternalReference"     "LocationLatitude"     
## [16] "LocationLongitude"     "DistributionChannel"   "UserLanguage"         
## [19] "prolificID"            "consent"               "attentionCheck"       
## [22] "inc"                   "edu"                   "age"                  
## [25] "gen"                   "gen_3_TEXT"            "pol"                  
## [28] "pid"                   "area"                  "race"                 
## [31] "post_Real_Check"       "Ent_explore"           "openFeedback"         
## [34] "confusion"             "PROLIFIC_PID"          "STUDY_ID"             
## [37] "SESSION_ID"            "cluster"               "social_issue"         
## [40] "signal"                "company"               "compCur"              
## [43] "dyNorm"                "Ent"                   "issueCur"             
## [46] "noisyMin"              "OComp"                 "OFriend"              
## [49] "OGen"                  "OImport"               "OLegit"               
## [52] "OLike"                 "OMotivate"             "ORelImport"           
## [55] "OTrust"                "peerPress"             "Prior"                
## [58] "Profit"                "pubPress"              "Rep"                  
## [61] "riskAct"               "riskInact"             "SComp"                
## [64] "SFriend"               "SGen"                  "SImport"              
## [67] "SLegit"                "SLike"                 "SRelImport"           
## [70] "staticNorm"            "STrust"                "sufficiency"          
## [73] "supGen"                "surprise"              "vote"

gjg_long <- gjg_long %>% mutate_at(c("compCur", "dyNorm", "Ent", "issueCur", "noisyMin", "OComp", "OFriend", "OGen", "OImport", "OLegit", "OLike", "OMotivate", "ORelImport", "OTrust", "peerPress", "Prior", "Profit", "pubPress", "Rep", "riskAct", "riskInact", "SComp", "SFriend", "SGen", "SImport", "SLegit", "SLike", "SRelImport", "staticNorm", "STrust", "sufficiency", "supGen", "surprise", "vote","post_Real_Check", "pol", "edu", "inc"), as.numeric)

## Rename Vignettes
gjg_long <- gjg_long %>%
  mutate(companyLeaning = case_when(
    company == "Bud" ~ "R",
    company == "CAR" ~ "R",
    company == "Chic" ~ "R",
    company == "Dis" ~ "L",
    company == "Pop" ~ "L",
    company == "Star" ~ "L",
    company == "Tar" ~ "L",
    company == "Wal" ~ "R",
    company == "WNBA" ~ "L",
    company == "Fox" ~ "R",
    company == "NBC" ~ "L",
    company == "NBA" ~ "L",
    company == "NFL" ~ "R",
    TRUE ~ company  # if none of the above conditions are met, keep the original value
  ))

gjg_long$condition <- paste0(gjg_long$companyLeaning, gjg_long$signal)

gjg_long <- subset(gjg_long, !(condition == "CS" & company == "Chic" & !is.na(Rep)))

gjg_long <- gjg_long %>%
  mutate(condition = case_when(
    condition == "LS" ~ "Liberal \n\ Signal",
    condition == "LN" ~ "Liberal \n\ Control",
    condition == "RS" ~ "Conservative \n\ Signal",
    condition == "RN" ~ "Conservative \n\ Control",
    TRUE ~ condition  # if none of the above conditions are met, keep the original value
  ))

gjg_long <- gjg_long %>%
  mutate(companyLeaning = case_when(
    companyLeaning == "R" ~ "Right",
    companyLeaning == "L" ~ "Left",
    TRUE ~ companyLeaning  # if none of the above conditions are met, keep the original value
  ))


gjg_long <- gjg_long %>%
  mutate(company = case_when(
    company == "Bud" ~ "Budweiser",
    company == "CAR" ~ "NASCAR",
    company == "Chic" ~ "Chick-fil-A",
    company == "Dis" ~ "Disney",
    company == "Pop" ~ "Popeyes",
    company == "Star" ~ "Starbucks",
    company == "Tar" ~ "Target",
    company == "Wal" ~ "Walmart",
    TRUE ~ company  # if none of the above conditions are met, keep the original value
  ))

## Rename Conditions
gjg_long$signal <- as.factor(gjg_long$signal)
gjg_long <- gjg_long %>%
  mutate(signal = case_when(
    signal == "S" ~ "Signal",
    signal == "N" ~ "No Signal",
    TRUE ~ signal  # if none of the above conditions are met, keep the original value
  ))

gjg_long$social_issue <- as.factor(gjg_long$social_issue)
gjg_long <- gjg_long %>%
  mutate(social_issue = case_when(
    social_issue == "T" ~ "Transgendered Representation",
    social_issue == "C" ~ "Climate Change",
    social_issue == "L" ~ "LGBTQ Representation",
    social_issue == "B" ~ "BLM",
    TRUE ~ social_issue  # if none of the above conditions are met, keep the original value
  ))

signal_gjg_long <- gjg_long %>%
  filter(signal == "Signal")

  
  
# ## filter out rows for condition-vignette pairs people DON'T see
# gjg_long <- gjg_long %>% filter(is.na(p_approve) == FALSE)
# 
# view(gjg_long)

2 Attention Check

gjg %>%
  group_by(attentionCheck) %>%
  dplyr::summarise(n = n()) %>%
  mutate(freq = n / sum(n))

98 people passed the attention check.

3 Demographics

# Subset your data frame to include only the demographic columns
demo_gjg <- gjg[, c("gen", "race", "inc", "edu", "pol", "pid", "area")]

# Age
mean(gjg$age, na.rm=TRUE)

## [1] 42.22

sd(gjg$age, na.rm=TRUE)

## [1] 16.69185

# Loop through each demographic column and calculate frequency counts
freq_tables <- list()

for (col in names(demo_gjg)) {
  {
    freq_table <- as.data.frame(table(demo_gjg[[col]]))
    freq_table$Percent <- round(freq_table$Freq / sum(freq_table$Freq) * 100, 2)
    freq_tables[[col]] <- freq_table
  }
}

# Print the frequency tables
for (i in seq_along(freq_tables)) {
  if (!is.null(freq_tables[[i]])) {
    cat("\nTable of frequencies for", names(freq_tables)[i], ":\n")
    print(freq_tables[[i]])
  }
}

## 
## Table of frequencies for gen :
##             Var1 Freq Percent
## 1                   4    3.85
## 2 I identify as:    3    2.88
## 3            Man   50   48.08
## 4          Woman   47   45.19
## 
## Table of frequencies for race :
##                                         Var1 Freq Percent
## 1                                               4    3.85
## 2                                      Black   10    9.62
## 3                                Black,White    1    0.96
## 4                                 East Asian    3    2.88
## 5                     Hispanic or Latino/a/x    7    6.73
## 6           Middle Eastern and North African    1    0.96
## 7     Middle Eastern and North African,White    1    0.96
## 8  Pacific Islander or Native Hawaiian,White    1    0.96
## 9                                South Asian    3    2.88
## 10                         South Asian,White    1    0.96
## 11                           Southeast Asian    2    1.92
## 12                                     White   70   67.31
## 
## Table of frequencies for inc :
##   Var1 Freq Percent
## 1    1   15      15
## 2    2   26      26
## 3    3   20      20
## 4    4   21      21
## 5    5   13      13
## 6    6    3       3
## 7    7    2       2
## 
## Table of frequencies for edu :
##   Var1 Freq Percent
## 1    1    1       1
## 2    2   12      12
## 3    3   29      29
## 4    4   45      45
## 5    5   13      13
## 
## Table of frequencies for pol :
##   Var1 Freq Percent
## 1    1    3       3
## 2    2   12      12
## 3    3    9       9
## 4    4   23      23
## 5    5    7       7
## 6    6   24      24
## 7    7   22      22
## 
## Table of frequencies for pid :
##                  Var1 Freq Percent
## 1                        4    3.85
## 2            Democrat   53   50.96
## 3 Independent / Other   29   27.88
## 4          Republican   18   17.31
## 
## Table of frequencies for area :
##       Var1 Freq Percent
## 1             4    3.85
## 2    Rural   15   14.42
## 3 Suburban   51   49.04
## 4    Urban   34   32.69

4 Correlation

DVs <- gjg_long[c("SGen", "SLike", "STrust", "SFriend", "SComp", "OGen", "OLike", "OTrust", "OFriend", "OComp", "Ent", "surprise", "noisyMin", "Rep", "Profit", "peerPress",  "pubPress", "riskAct", "riskInact", "compCur", "staticNorm", "dyNorm", "SImport", "SLegit", "SRelImport",  "OImport", "OLegit", "ORelImport", "supGen", "OMotivate", "issueCur", "sufficiency", "vote", "Prior", "post_Real_Check", "pol", "edu", "inc")]

# Compute pairwise correlations
corr_DVs <- cor(DVs, use = "complete.obs")

colnames(corr_DVs) <- c("I think company is genuine", "I like company", "I trust company", "I think company is friendly", "I think company is competent", "Others think company is genuine", "Others like company", "Others trust company", "Others think company is friendly", "Others think company is competent", "Entativity", "Surprise", "Noisy Minority Motive", "Reputation Motive", "Profit Motive", "Pressure from Peer organizations motive", "Pressure from public motive", "risky to signal", "risky not to signal", "curious about company", "static norm", "dynamic norm", "I think it's important", "I think it's legit", "My weight of its relative importance", "Others think its important", "Others think it's legit", "Others weight of its relative importance", "supporters are genuine", "Post motivates others", "curious about issue", "sufficiency", "vote", "seen prior", "belief in manipulation", "political leaning", "education", "income")

rownames(corr_DVs) <- c("I think company is genuine", "I like company", "I trust company", "I think company is friendly", "I think company is competent", "Others think company is genuine", "Others like company", "Others trust company", "Others think company is friendly", "Others think company is competent", "Entativity", "Surprise", "Noisy Minority Motive", "Reputation Motive", "Profit Motive", "Pressure from Peer organizations motive", "Pressure from public motive", "risky to signal", "risky not to signal", "curious about company", "static norm", "dynamic norm", "I think it's important", "I think it's legit", "My weight of its relative importance", "Others think its important", "Others think it's legit", "Others weight of its relative importance", "supporters are genuine", "Post motivates others", "curious about issue", "sufficiency", "vote", "seen prior", "belief in manipulation", "political leaning", "education", "income")

# Plot the correlation matrix
corrplot(corr_DVs, is.corr = TRUE, type = "lower", lower = "circle", tl.cex = 0.7, insig = "label_sig", diag = FALSE)

5 factor analysis

factorDVs <- gjg_long[c("SGen", "SLike", "STrust", "SFriend", "SComp", "OGen", "OLike", "OTrust", "OFriend", "OComp", "Ent", "surprise", "noisyMin", "Rep", "Profit", "peerPress",  "pubPress", "riskAct", "riskInact", "compCur", "staticNorm", "dyNorm", "SImport", "SLegit", "SRelImport",  "OImport", "OLegit", "ORelImport", "supGen", "OMotivate", "issueCur", "sufficiency", "vote")]

corr_DVs = cor(factorDVs, use="complete.obs")
KMO(corr_DVs) #tests how suited data is for factor analysis

## Kaiser-Meyer-Olkin factor adequacy
## Call: KMO(r = corr_DVs)
## Overall MSA =  0.84
## MSA for each item = 
##        SGen       SLike      STrust     SFriend       SComp        OGen 
##        0.85        0.86        0.87        0.90        0.93        0.87 
##       OLike      OTrust     OFriend       OComp         Ent    surprise 
##        0.85        0.89        0.89        0.89        0.69        0.62 
##    noisyMin         Rep      Profit   peerPress    pubPress     riskAct 
##        0.85        0.79        0.79        0.79        0.75        0.66 
##   riskInact     compCur  staticNorm      dyNorm     SImport      SLegit 
##        0.67        0.84        0.82        0.76        0.84        0.84 
##  SRelImport     OImport      OLegit  ORelImport      supGen   OMotivate 
##        0.87        0.74        0.72        0.76        0.90        0.88 
##    issueCur sufficiency        vote 
##        0.83        0.91        0.81

cortest.bartlett(corr_DVs) #tests correlations between variables

## $chisq
## [1] 1750.954
## 
## $p.value
## [1] 8.170913e-131
## 
## $df
## [1] 528

ev<-eigen(cor(corr_DVs)) #gets eigenvalues (variance explained by each component)
ev$values

##  [1]  1.278851e+01  8.318158e+00  3.614394e+00  2.155692e+00  1.542332e+00
##  [6]  9.619915e-01  6.309853e-01  5.023408e-01  3.984108e-01  3.747354e-01
## [11]  3.411127e-01  2.240884e-01  1.779685e-01  1.555402e-01  1.526812e-01
## [16]  1.353896e-01  1.004078e-01  6.637891e-02  5.780821e-02  4.885560e-02
## [21]  4.625344e-02  3.636604e-02  2.956263e-02  2.855853e-02  2.332470e-02
## [26]  2.071588e-02  1.957961e-02  1.696961e-02  1.188525e-02  9.121680e-03
## [31]  6.485283e-03  3.399733e-03 -3.769339e-16

scree(corr_DVs) #number of factors until plot levels off

fa.parallel(corr_DVs, n.obs=50, fa="fa") #checks eigenvalues of factors against eigenvalues of identity (no correlation) matrix

## Parallel analysis suggests that the number of factors =  4  and the number of components =  NA

dat_fa <- na.omit(factorDVs)

5.1 4 factors

fit.4 <- factanal(na.omit(dat_fa),factors=4, rotation="promax", scores = "regression")
print(fit.4, digits = 2, cutoff = .4, sort = TRUE)

## 
## Call:
## factanal(x = na.omit(dat_fa), factors = 4, scores = "regression",     rotation = "promax")
## 
## Uniquenesses:
##        SGen       SLike      STrust     SFriend       SComp        OGen 
##        0.34        0.35        0.21        0.33        0.38        0.55 
##       OLike      OTrust     OFriend       OComp         Ent    surprise 
##        0.25        0.39        0.25        0.35        0.92        0.98 
##    noisyMin         Rep      Profit   peerPress    pubPress     riskAct 
##        0.64        0.57        0.55        0.47        0.35        0.87 
##   riskInact     compCur  staticNorm      dyNorm     SImport      SLegit 
##        0.67        0.50        0.79        0.89        0.16        0.19 
##  SRelImport     OImport      OLegit  ORelImport      supGen   OMotivate 
##        0.40        0.83        0.78        0.75        0.54        0.76 
##    issueCur sufficiency        vote 
##        0.55        0.33        0.71 
## 
## Loadings:
##             Factor1 Factor2 Factor3 Factor4
## noisyMin    -0.56                          
## SImport      0.94                          
## SLegit       0.91                          
## SRelImport   0.76                          
## supGen       0.60                          
## issueCur     0.67                          
## sufficiency  0.85                          
## vote         0.55                          
## SGen                 0.77                  
## SLike                0.79                  
## STrust               0.91                  
## SFriend              0.74                  
## SComp                0.60                  
## compCur              0.72                  
## OLike                        0.89          
## OTrust                       0.77          
## OFriend                      0.87          
## OComp                        0.79          
## peerPress                            0.67  
## pubPress                             0.80  
## riskInact                            0.58  
## OGen                 0.41    0.41          
## Ent                                        
## surprise                                   
## Rep                 -0.47            0.40  
## Profit              -0.43            0.47  
## riskAct                                    
## staticNorm   0.41                          
## dyNorm                                     
## OImport                                    
## OLegit                                     
## ORelImport                           0.42  
## OMotivate            0.44                  
## 
##                Factor1 Factor2 Factor3 Factor4
## SS loadings       4.93    4.52    3.36    2.56
## Proportion Var    0.15    0.14    0.10    0.08
## Cumulative Var    0.15    0.29    0.39    0.47
## 
## Factor Correlations:
##         Factor1 Factor2 Factor3 Factor4
## Factor1    1.00    0.18    0.47    0.16
## Factor2    0.18    1.00    0.30   -0.17
## Factor3    0.47    0.30    1.00   -0.16
## Factor4    0.16   -0.17   -0.16    1.00
## 
## Test of the hypothesis that 4 factors are sufficient.
## The chi square statistic is 899.04 on 402 degrees of freedom.
## The p-value is 4.82e-40

6 Plots + Inferential Stats

6.1 Business Perception

6.1.1 Main effect plots

business_percep_plot_list <- list(plot_cooker(gjg_long, condition, SLike, "I like company"),
                         plot_cooker(gjg_long, condition, STrust, "I trust company"),
                         plot_cooker(gjg_long, condition, SGen, "I think company is genuine"),
                         plot_cooker(gjg_long, condition, SComp, "I think company is competent"),
                         plot_cooker(gjg_long, condition, SFriend, "I think company is friendly"),
                         plot_cooker(gjg_long, condition, OLike, "Others like company"),
                         plot_cooker(gjg_long, condition, OTrust, "Others trust company"),
                         plot_cooker(gjg_long, condition, OGen, "Others think company is genuine"),
                         plot_cooker(gjg_long, condition, OComp, "Others think company is competent"),
                         plot_cooker(gjg_long, condition, OFriend, "Others think company is friendly"),
                         plot_cooker(gjg_long, condition, Ent, "Entativity"),
                         plot_cooker(gjg_long, condition, surprise, "Surprise"))

business_percep_plot_arranged <- ggarrange(plotlist = business_percep_plot_list, ncol = 4, nrow = 3)

business_percep_plot_arranged

### Mixed Effects Models

mod_SLike <- lmer(SLike ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_SLike)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: SLike ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1509.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.3151 -0.6595  0.1242  0.6659  2.6237 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.63163  0.7947  
##  social_issue (Intercept) 0.03135  0.1771  
##  Residual                 2.24943  1.4998  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                     3.936e+00  1.930e-01 2.185e+01  20.396 1.03e-15
## conditionConservative \n Signal 2.044e-03  2.143e-01 2.881e+02   0.010 0.992395
## conditionLiberal \n Control     8.229e-01  2.143e-01 2.881e+02   3.840 0.000151
## conditionLiberal \n Signal      6.559e-01  2.143e-01 2.881e+02   3.061 0.002416
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     ***
## conditionLiberal \n Signal      ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.555              
## cndtnLbrlCn -0.555  0.500       
## cndtnLbrlSg -0.555  0.500  0.500

mod_STrust <- lmer(STrust ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_STrust)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: STrust ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1438.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.4565 -0.4933  0.1181  0.5509  2.9990 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.16533  1.0795  
##  social_issue (Intercept) 0.04616  0.2148  
##  Residual                 1.60000  1.2649  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       3.7634     0.1994  18.7678  18.873 1.16e-13
## conditionConservative \n Signal  -0.2065     0.1807 288.0624  -1.142 0.254265
## conditionLiberal \n Control       0.6079     0.1807 288.0624   3.364 0.000873
## conditionLiberal \n Signal        0.3817     0.1807 288.0624   2.112 0.035562
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     ***
## conditionLiberal \n Signal      *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.453              
## cndtnLbrlCn -0.453  0.500       
## cndtnLbrlSg -0.453  0.500  0.500

mod_SGen <- lmer(SGen ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_SGen)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: SGen ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1407.2
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.89795 -0.58311  0.03022  0.58382  2.22289 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.2247   1.1067  
##  social_issue (Intercept) 0.0826   0.2874  
##  Residual                 1.4411   1.2005  
## Number of obs: 391, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       3.2553     0.2188  11.4222  14.881 7.85e-09
## conditionConservative \n Signal   0.2390     0.1721 287.4335   1.389    0.166
## conditionLiberal \n Control       0.9446     0.1715 287.1673   5.507 8.11e-08
## conditionLiberal \n Signal        1.2455     0.1715 287.1668   7.261 3.61e-12
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     ***
## conditionLiberal \n Signal      ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.391              
## cndtnLbrlCn -0.392  0.498       
## cndtnLbrlSg -0.392  0.498  0.500

mod_SComp <- lmer(SComp ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_SComp)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: SComp ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1404.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.8444 -0.4457  0.1281  0.5639  2.1917 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.8875   0.9421  
##  social_issue (Intercept) 0.0559   0.2364  
##  Residual                 1.5272   1.2358  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       4.3825     0.1965  13.7549  22.301 3.40e-12
## conditionConservative \n Signal   0.1873     0.1766 288.0521   1.061 0.289646
## conditionLiberal \n Control       0.8153     0.1766 288.0521   4.618 5.86e-06
## conditionLiberal \n Signal        0.6715     0.1766 288.0521   3.803 0.000175
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     ***
## conditionLiberal \n Signal      ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.449              
## cndtnLbrlCn -0.449  0.500       
## cndtnLbrlSg -0.449  0.500  0.500

mod_SFriend <- lmer(SFriend ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_SFriend)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: SFriend ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1429.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.2105 -0.4273  0.0976  0.6207  2.2974 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.68256  0.8262  
##  social_issue (Intercept) 0.02245  0.1498  
##  Residual                 1.74445  1.3208  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       4.3749     0.1743  26.0712  25.099   <2e-16
## conditionConservative \n Signal  -0.3046     0.1887 288.1052  -1.614   0.1076
## conditionLiberal \n Control       0.4536     0.1887 288.1052   2.404   0.0169
## conditionLiberal \n Signal        0.4433     0.1887 288.1052   2.349   0.0195
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     *  
## conditionLiberal \n Signal      *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.541              
## cndtnLbrlCn -0.541  0.500       
## cndtnLbrlSg -0.541  0.500  0.500

mod_OLike <- lmer(OLike ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OLike)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OLike ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1199.6
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.4504 -0.4101  0.1346  0.5700  1.9824 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.29855  0.54639 
##  social_issue (Intercept) 0.00805  0.08972 
##  Residual                 1.00655  1.00327 
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       5.26381    0.12382  31.68515  42.512   <2e-16
## conditionConservative \n Signal  -0.01803    0.14334 288.13308  -0.126    0.900
## conditionLiberal \n Control       0.15410    0.14334 288.13308   1.075    0.283
## conditionLiberal \n Signal        0.14545    0.14334 288.13308   1.015    0.311
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.579              
## cndtnLbrlCn -0.579  0.500       
## cndtnLbrlSg -0.579  0.500  0.500

mod_OTrust <- lmer(OTrust ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OTrust)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OTrust ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1219.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.5183 -0.5177  0.1079  0.5855  2.0254 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.5177   0.7195  
##  social_issue (Intercept) 0.0000   0.0000  
##  Residual                 0.9742   0.9870  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       4.8673     0.1234 285.0412  39.450   <2e-16
## conditionConservative \n Signal  -0.1531     0.1410 291.0000  -1.086   0.2786
## conditionLiberal \n Control       0.2245     0.1410 291.0000   1.592   0.1124
## conditionLiberal \n Signal        0.2959     0.1410 291.0000   2.099   0.0367
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal      *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.571              
## cndtnLbrlCn -0.571  0.500       
## cndtnLbrlSg -0.571  0.500  0.500
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

mod_OGen <- lmer(OGen ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OGen)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OGen ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1379.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.4677 -0.5773  0.0490  0.5850  2.4646 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.48881  0.6992  
##  social_issue (Intercept) 0.06916  0.2630  
##  Residual                 1.57842  1.2564  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       3.6947     0.1959   9.6595  18.858 6.07e-09
## conditionConservative \n Signal   0.2200     0.1795 288.0452   1.226    0.221
## conditionLiberal \n Control       0.7608     0.1795 288.0452   4.238 3.04e-05
## conditionLiberal \n Signal        1.0159     0.1795 288.0452   5.659 3.67e-08
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     ***
## conditionLiberal \n Signal      ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.458              
## cndtnLbrlCn -0.458  0.500       
## cndtnLbrlSg -0.458  0.500  0.500

mod_OComp <- lmer(OComp ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OComp)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OComp ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1238.8
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.0168 -0.4091  0.1198  0.5921  2.1091 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.537174 0.73292 
##  social_issue (Intercept) 0.001051 0.03242 
##  Residual                 1.024492 1.01217 
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       5.01001    0.12727  74.22598  39.364   <2e-16
## conditionConservative \n Signal  -0.09136    0.14460 288.21533  -0.632   0.5280
## conditionLiberal \n Control       0.24502    0.14460 288.21533   1.694   0.0913
## conditionLiberal \n Signal        0.12264    0.14460 288.21533   0.848   0.3970
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     .  
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.568              
## cndtnLbrlCn -0.568  0.500       
## cndtnLbrlSg -0.568  0.500  0.500

mod_OFriend <- lmer(OFriend ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OFriend)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OFriend ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1179.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.8927 -0.5155  0.1760  0.5806  2.7747 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.4583   0.6770  
##  social_issue (Intercept) 0.0000   0.0000  
##  Residual                 0.8805   0.9383  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       4.81633    0.11688 287.07060  41.208   <2e-16
## conditionConservative \n Signal   0.08163    0.13405 291.00000   0.609   0.5430
## conditionLiberal \n Control       0.32653    0.13405 291.00000   2.436   0.0155
## conditionLiberal \n Signal        0.29592    0.13405 291.00000   2.208   0.0281
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     *  
## conditionLiberal \n Signal      *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.573              
## cndtnLbrlCn -0.573  0.500       
## cndtnLbrlSg -0.573  0.500  0.500
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

mod_Ent <- lmer(Ent ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_Ent)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Ent ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1166.4
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.32328 -0.61847 -0.01501  0.58522  2.74723 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.268052 0.51774 
##  social_issue (Intercept) 0.003939 0.06276 
##  Residual                 0.929073 0.96388 
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       3.2457     0.1149  42.4499  28.249   <2e-16
## conditionConservative \n Signal  -0.1025     0.1377 288.1670  -0.745   0.4571
## conditionLiberal \n Control      -0.2565     0.1377 288.1670  -1.863   0.0635
## conditionLiberal \n Signal       -0.1239     0.1377 288.1670  -0.900   0.3691
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     .  
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.599              
## cndtnLbrlCn -0.599  0.500       
## cndtnLbrlSg -0.599  0.500  0.500

mod_surprise <- lmer(surprise ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_surprise)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: surprise ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1495.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.14437 -0.66930 -0.02707  0.62969  2.23094 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.73459  0.8571  
##  social_issue (Intercept) 0.05225  0.2286  
##  Residual                 2.12328  1.4571  
## Number of obs: 391, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                 Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       3.4948     0.2055  15.3377  17.006 2.28e-11
## conditionConservative \n Signal   1.1011     0.2088 287.7398   5.273 2.64e-07
## conditionLiberal \n Control      -0.8151     0.2082 287.3030  -3.915 0.000113
## conditionLiberal \n Signal       -0.3408     0.2082 287.3023  -1.637 0.102730
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal ***
## conditionLiberal \n Control     ***
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.505              
## cndtnLbrlCn -0.507  0.498       
## cndtnLbrlSg -0.507  0.498  0.500

6.1.2 Social Issue Facet

wrapped_plot_cooker(gjg_long, condition, SLike, "I like company", social_issue)

wrapped_plot_cooker(gjg_long, condition, STrust, "I trust company", social_issue)

wrapped_plot_cooker(gjg_long, condition, SGen, "I think company is genuine", social_issue)

wrapped_plot_cooker(gjg_long, condition, SComp, "I think company is competent", social_issue)

wrapped_plot_cooker(gjg_long, condition, SFriend, "I think company is friendly", social_issue)

wrapped_plot_cooker(gjg_long, condition, OLike, "Others like company", social_issue)

wrapped_plot_cooker(gjg_long, condition, OTrust, "Others trust company", social_issue)

wrapped_plot_cooker(gjg_long, condition, OGen, "Others think company is genuine", social_issue)

wrapped_plot_cooker(gjg_long, condition, OComp, "Others think company is competent", social_issue)

wrapped_plot_cooker(gjg_long, condition, OFriend, "Others think company is friendly", social_issue)

wrapped_plot_cooker(gjg_long, condition, Ent, "Entativity", social_issue)

wrapped_plot_cooker(gjg_long, condition, surprise, "Surprise", social_issue)

6.1.3 Political Affiliation Facet

wrapped_plot_cooker(gjg_long, condition, SLike, "I like company", pid)

wrapped_plot_cooker(gjg_long, condition, STrust, "I trust company", pid)

wrapped_plot_cooker(gjg_long, condition, SGen, "I think company is genuine", pid)

wrapped_plot_cooker(gjg_long, condition, SComp, "I think company is competent", pid)

wrapped_plot_cooker(gjg_long, condition, SFriend, "I think company is friendly", pid)

wrapped_plot_cooker(gjg_long, condition, OLike, "Others like company", pid)

wrapped_plot_cooker(gjg_long, condition, OTrust, "Others trust company", pid)

wrapped_plot_cooker(gjg_long, condition, OGen, "Others think company is genuine", pid)

wrapped_plot_cooker(gjg_long, condition, OComp, "Others think company is competent", pid)

wrapped_plot_cooker(gjg_long, condition, OFriend, "Others think company is friendly", pid)

wrapped_plot_cooker(gjg_long, condition, Ent, "Entativity", pid)

wrapped_plot_cooker(gjg_long, condition, surprise, "Surprise", pid)

6.2 Motive

6.2.1 Main Effect Plots

motive_plot_list <- list(plot_cooker(signal_gjg_long, condition, noisyMin, "Motive to cater to noisy minority"),
                         plot_cooker(signal_gjg_long, condition, Rep, "Reputation motive"),
                         plot_cooker(signal_gjg_long, condition, Profit, "Profit motive"),
                         plot_cooker(signal_gjg_long, condition, peerPress, "Pressured by peer companies"),
                         plot_cooker(signal_gjg_long, condition, pubPress, "Pressured by public"),
                         plot_cooker(signal_gjg_long, condition, riskAct, "Risky to send signal"),
                         plot_cooker(signal_gjg_long, condition, riskInact, "Risky NOT to send signal"),
                         plot_cooker(signal_gjg_long, condition, compCur, "Curious about company"))
                        
motive_plot_arranged <- ggarrange(plotlist = motive_plot_list, ncol = 4, nrow = 3)

motive_plot_arranged

### Mixed Effects Models

mod_noisyMin <- lmer(noisyMin ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_noisyMin)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: noisyMin ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 687.6
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.65550 -0.44656  0.09186  0.51593  2.06247 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.61934  1.2725  
##  social_issue (Intercept) 0.08838  0.2973  
##  Residual                 0.89462  0.9458  
## Number of obs: 195, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error       df t value Pr(>|t|)    
## (Intercept)                 4.47203    0.21889  8.92344  20.430 8.44e-09 ***
## conditionLiberal \n Signal -0.01863    0.13575 93.89400  -0.137    0.891    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.313

mod_Rep <- lmer(Rep ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_Rep)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Rep ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 651.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.4624 -0.4146  0.0872  0.5585  2.0508 
## 
## Random effects:
##  Groups       Name        Variance  Std.Dev. 
##  PID          (Intercept) 4.976e-01 7.054e-01
##  social_issue (Intercept) 3.277e-09 5.724e-05
##  Residual                 1.180e+00 1.086e+00
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error       df t value Pr(>|t|)    
## (Intercept)                  5.4184     0.1308 178.3135  41.409   <2e-16 ***
## conditionLiberal \n Signal  -0.0102     0.1552  97.0000  -0.066    0.948    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.593
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

mod_Profit <- lmer(Profit ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_Profit)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: Profit ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 698.8
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.94356 -0.41189  0.04733  0.59849  2.18120 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.993365 0.99668 
##  social_issue (Intercept) 0.004518 0.06722 
##  Residual                 1.279085 1.13097 
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)                  5.1945     0.1559 23.4412  33.310   <2e-16 ***
## conditionLiberal \n Signal  -0.1950     0.1616 94.5297  -1.207     0.23    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.518

mod_peerPress <- lmer(peerPress ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_peerPress)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: peerPress ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 681.1
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.79333 -0.39082  0.07806  0.56495  2.21734 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.23362  1.1107  
##  social_issue (Intercept) 0.01342  0.1158  
##  Residual                 0.99641  0.9982  
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)                  4.5819     0.1616 21.6773  28.354   <2e-16 ***
## conditionLiberal \n Signal  -0.1538     0.1426 94.5317  -1.078    0.284    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.441

mod_pubPress <- lmer(pubPress ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_pubPress)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: pubPress ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 706.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.3174 -0.5426  0.1042  0.6591  1.9928 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.9179   0.9581  
##  social_issue (Intercept) 0.0000   0.0000  
##  Residual                 1.4000   1.1832  
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error       df t value Pr(>|t|)    
## (Intercept)                  4.5612     0.1538 167.6999  29.658   <2e-16 ***
## conditionLiberal \n Signal  -0.2347     0.1690  97.0000  -1.388    0.168    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.550
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

mod_pubPress <- lmer(pubPress ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_pubPress)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: pubPress ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 706.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.3174 -0.5426  0.1042  0.6591  1.9928 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.9179   0.9581  
##  social_issue (Intercept) 0.0000   0.0000  
##  Residual                 1.4000   1.1832  
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error       df t value Pr(>|t|)    
## (Intercept)                  4.5612     0.1538 167.6999  29.658   <2e-16 ***
## conditionLiberal \n Signal  -0.2347     0.1690  97.0000  -1.388    0.168    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.550
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')

mod_riskAct <- lmer(riskAct ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_riskAct)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: riskAct ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 748.5
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.17595 -0.70722  0.00032  0.63821  2.04689 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.7797   0.8830  
##  social_issue (Intercept) 0.3668   0.6056  
##  Residual                 1.8957   1.3768  
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)                  3.6493     0.3450  4.0198  10.579  0.00044 ***
## conditionLiberal \n Signal  -0.2152     0.1968 94.9401  -1.094  0.27675    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.285

mod_riskInact <- lmer(riskInact ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_riskInact)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: riskInact ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 733.3
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.13162 -0.53229  0.02662  0.52198  1.97714 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.8209   0.9060  
##  social_issue (Intercept) 0.1976   0.4445  
##  Residual                 1.6996   1.3037  
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)                  4.0204     0.2741  4.9572  14.668 2.84e-05 ***
## conditionLiberal \n Signal   0.2372     0.1863 94.7815   1.273    0.206    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.340

mod_compCur <- lmer(compCur ~ condition + (1 | PID) + (1 | social_issue), data = signal_gjg_long)
summary(mod_compCur)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: compCur ~ condition + (1 | PID) + (1 | social_issue)
##    Data: signal_gjg_long
## 
## REML criterion at convergence: 742.5
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.23974 -0.55078  0.01905  0.56863  2.58970 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.39119  1.1795  
##  social_issue (Intercept) 0.04028  0.2007  
##  Residual                 1.50387  1.2263  
## Number of obs: 196, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                            Estimate Std. Error      df t value Pr(>|t|)    
## (Intercept)                  3.2152     0.1990 11.7375  16.153 2.23e-09 ***
## conditionLiberal \n Signal   0.2552     0.1752 94.8809   1.456    0.149    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr)
## cndtnLbrlSg -0.440

6.2.2 Social Issue Facet

wrapped_plot_cooker(signal_gjg_long, condition, noisyMin, "Motive to cater to noisy minority", social_issue)

wrapped_plot_cooker(signal_gjg_long, condition, Rep, "Reputation motive", social_issue)

wrapped_plot_cooker(signal_gjg_long, condition, Profit, "Profit motive", social_issue)

wrapped_plot_cooker(signal_gjg_long, condition, peerPress, "Pressured by peer companies", social_issue)

wrapped_plot_cooker(signal_gjg_long, condition, pubPress, "Pressured by public", social_issue)

wrapped_plot_cooker(signal_gjg_long, condition, riskAct, "Risky to send signal", social_issue)

wrapped_plot_cooker(signal_gjg_long, condition, riskInact, "Risky NOT to send signal", social_issue)

wrapped_plot_cooker(signal_gjg_long, condition, compCur, "Curious about company", social_issue)

6.2.3 Political Affiliation Facet

wrapped_plot_cooker(signal_gjg_long, condition, noisyMin, "Motive to cater to noisy minority", pid)

wrapped_plot_cooker(signal_gjg_long, condition, Rep, "Reputation motive", pid)

wrapped_plot_cooker(signal_gjg_long, condition, Profit, "Profit motive", pid)

wrapped_plot_cooker(signal_gjg_long, condition, peerPress, "Pressured by peer companies", pid)

wrapped_plot_cooker(signal_gjg_long, condition, pubPress, "Pressured by public", pid)

wrapped_plot_cooker(signal_gjg_long, condition, riskAct, "Risky to send signal", pid)

wrapped_plot_cooker(signal_gjg_long, condition, riskInact, "Risky NOT to send signal", pid)

wrapped_plot_cooker(signal_gjg_long, condition, compCur, "Curious about company", pid)

6.3 Message Perception

6.3.1 Main Effect Plots

message_percep_plot_list <- list(plot_cooker(gjg_long, condition, staticNorm, "Percent of support for issue"),
                         plot_cooker(gjg_long, condition, dyNorm, "Change in support for issue"),
                         plot_cooker(gjg_long, condition, SImport, "I think cause is important"),
                         plot_cooker(gjg_long, condition, SLegit, "I think cause is legit"),
                         plot_cooker(gjg_long, condition, SRelImport, "I think cause is important relative to other causes"),
                         plot_cooker(gjg_long, condition, OImport, "Others think cause is important"),
                         plot_cooker(gjg_long, condition, OLegit, "Others think cause is legit"),
                         plot_cooker(gjg_long, condition, ORelImport, "Others think cause is important relative to other causes"),
                         plot_cooker(gjg_long, condition, OMotivate, "Message motivates others to act"),
                         plot_cooker(gjg_long, condition, supGen, "Supporters genuinely care"),
                         plot_cooker(gjg_long, condition, vote, "I would want to know candidate's stance on this issue"),
                         plot_cooker(gjg_long, condition, sufficiency, "How much progress is needed"),
                         plot_cooker(gjg_long, condition, issueCur, "Curious about cause"))

message_percep_plot_arranged <- ggarrange(plotlist = message_percep_plot_list, ncol = 4, nrow = 4)

message_percep_plot_arranged

plot_cooker(gjg_long, condition, OMotivate, "Message motivates others to act")

6.3.2 Mixed Effects Models

mod_staticNorm <- lmer(staticNorm ~ signal*companyLeaning + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_staticNorm)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: staticNorm ~ signal * companyLeaning + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 3286.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.6587 -0.5489  0.0373  0.5406  3.5516 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 161.2    12.70   
##  social_issue (Intercept)  80.1     8.95   
##  Residual                 175.9    13.26   
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error       df t value Pr(>|t|)
## (Intercept)                       56.1702     4.8442   3.9393  11.595 0.000344
## signalSignal                       0.8600     1.8949 288.0053   0.454 0.650297
## companyLeaningRight               -0.7251     1.8949 288.0053  -0.383 0.702271
## signalSignal:companyLeaningRight  -2.4508     2.6798 288.0053  -0.915 0.361213
##                                     
## (Intercept)                      ***
## signalSignal                        
## companyLeaningRight                 
## signalSignal:companyLeaningRight    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) sgnlSg cmpnLR
## signalSignl -0.196              
## cmpnyLnngRg -0.196  0.500       
## sgnlSgnl:LR  0.138 -0.707 -0.707

summary(allEffects(mod_staticNorm))

##  model: staticNorm ~ signal * companyLeaning
## 
##  signal*companyLeaning effect
##            companyLeaning
## signal          Left    Right
##   No Signal 56.17024 55.44517
##   Signal    57.03021 53.85438
## 
##  Lower 95 Percent Confidence Limits
##            companyLeaning
## signal          Left    Right
##   No Signal 46.64615 45.92108
##   Signal    47.50612 44.33030
## 
##  Upper 95 Percent Confidence Limits
##            companyLeaning
## signal          Left    Right
##   No Signal 65.69433 64.96925
##   Signal    66.55429 63.37847

## shows estimates + CI, useful to see CI from mixed effects model, ggplot does not note confidence interval



## Significant ##
mod_dyNorm <- lmer(dyNorm ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_dyNorm)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: dyNorm ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1315.7
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.3724 -0.4004  0.0882  0.5604  2.6079 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.63849  0.7991  
##  social_issue (Intercept) 0.02818  0.1679  
##  Residual                 1.24312  1.1150  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       4.79380    0.16201  19.18172  29.589   <2e-16
## conditionConservative \n Signal   0.07648    0.15930 288.07408   0.480   0.6315
## conditionLiberal \n Control       0.04485    0.15930 288.07408   0.282   0.7785
## conditionLiberal \n Signal        0.31570    0.15930 288.07408   1.982   0.0485
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal      *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.492              
## cndtnLbrlCn -0.492  0.500       
## cndtnLbrlSg -0.492  0.500  0.500

mod_SImport <- lmer(SImport ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_SImport)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: SImport ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1339.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.4798 -0.3047  0.0160  0.5397  2.5378 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.6382   1.2799  
##  social_issue (Intercept) 0.1858   0.4311  
##  Residual                 1.0504   1.0249  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       5.75058    0.27183   6.75143  21.155 2.01e-07
## conditionConservative \n Signal  -0.06945    0.14644 288.01308  -0.474    0.636
## conditionLiberal \n Control      -0.07044    0.14644 288.01308  -0.481    0.631
## conditionLiberal \n Signal        0.11718    0.14644 288.01308   0.800    0.424
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.269              
## cndtnLbrlCn -0.269  0.500       
## cndtnLbrlSg -0.269  0.500  0.500

mod_SLegit <- lmer(SLegit ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_SLegit)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: SLegit ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1382.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.5063 -0.3451  0.1597  0.5359  2.1931 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.5122   1.2297  
##  social_issue (Intercept) 0.1377   0.3711  
##  Residual                 1.2546   1.1201  
## Number of obs: 391, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       5.64189    0.25033   8.24894  22.538 1.06e-08
## conditionConservative \n Signal   0.01575    0.16055 287.30768   0.098    0.922
## conditionLiberal \n Control      -0.12813    0.16004 287.12055  -0.801    0.424
## conditionLiberal \n Signal        0.15456    0.16004 287.12055   0.966    0.335
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.319              
## cndtnLbrlCn -0.320  0.498       
## cndtnLbrlSg -0.320  0.498  0.500

mod_SRelImport <- lmer(SRelImport ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_SRelImport)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: SRelImport ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1388.9
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.8190 -0.5051  0.1097  0.5166  3.5464 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.2591   1.1221  
##  social_issue (Intercept) 0.4524   0.6726  
##  Residual                 1.3141   1.1463  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       4.52859    0.37332   4.29347  12.130 0.000175
## conditionConservative \n Signal  -0.07760    0.16379 288.00689  -0.474 0.636028
## conditionLiberal \n Control      -0.09903    0.16379 288.00689  -0.605 0.545930
## conditionLiberal \n Signal        0.08268    0.16379 288.00689   0.505 0.614111
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.219              
## cndtnLbrlCn -0.219  0.500       
## cndtnLbrlSg -0.219  0.500  0.500

mod_OImport <- lmer(OImport ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OImport)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OImport ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1236.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.5210 -0.4068  0.0733  0.5099  3.7821 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.64203  0.8013  
##  social_issue (Intercept) 0.08901  0.2983  
##  Residual                 0.95800  0.9788  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       5.23074    0.19642   7.28819  26.630 1.57e-08
## conditionConservative \n Signal  -0.20329    0.13985 288.02366  -1.454    0.147
## conditionLiberal \n Control      -0.21819    0.13985 288.02366  -1.560    0.120
## conditionLiberal \n Signal        0.05975    0.13985 288.02366   0.427    0.670
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.356              
## cndtnLbrlCn -0.356  0.500       
## cndtnLbrlSg -0.356  0.500  0.500

mod_OLegit <- lmer(OLegit ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OLegit)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OLegit ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1261.4
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.8582 -0.4805  0.1182  0.5519  3.3244 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.4832   0.6951  
##  social_issue (Intercept) 0.1947   0.4412  
##  Residual                 1.0928   1.0454  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       5.15117    0.25446   4.74661  20.244 8.62e-06
## conditionConservative \n Signal  -0.06846    0.14937 288.01291  -0.458   0.6470
## conditionLiberal \n Control      -0.25264    0.14937 288.01291  -1.691   0.0918
## conditionLiberal \n Signal        0.05317    0.14937 288.01291   0.356   0.7221
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control     .  
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.294              
## cndtnLbrlCn -0.294  0.500       
## cndtnLbrlSg -0.294  0.500  0.500

mod_ORelImport <- lmer(ORelImport ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_ORelImport)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: ORelImport ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1336.3
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.96987 -0.60778 -0.03385  0.60595  2.48756 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.8045   0.8969  
##  social_issue (Intercept) 0.2568   0.5067  
##  Residual                 1.2534   1.1196  
## Number of obs: 391, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                   Estimate Std. Error         df t value
## (Intercept)                       4.138491   0.291875   4.789687  14.179
## conditionConservative \n Signal  -0.289941   0.160497 287.154180  -1.807
## conditionLiberal \n Control      -0.104759   0.159970 286.820441  -0.655
## conditionLiberal \n Signal       -0.006758   0.159970 286.820441  -0.042
##                                 Pr(>|t|)    
## (Intercept)                     4.28e-05 ***
## conditionConservative \n Signal   0.0719 .  
## conditionLiberal \n Control       0.5131    
## conditionLiberal \n Signal        0.9663    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.273              
## cndtnLbrlCn -0.274  0.498       
## cndtnLbrlSg -0.274  0.498  0.500

## Significant ##
mod_OMotivate <- lmer(OMotivate ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_OMotivate)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: OMotivate ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1404.6
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.22729 -0.69139  0.05749  0.65084  2.22130 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.742296 0.86157 
##  social_issue (Intercept) 0.004834 0.06953 
##  Residual                 1.601296 1.26542 
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                   Estimate Std. Error         df t value
## (Intercept)                      3.215e+00  1.585e-01  5.742e+01  20.283
## conditionConservative \n Signal  1.305e+00  1.808e-01  2.882e+02   7.216
## conditionLiberal \n Control     -3.804e-04  1.808e-01  2.882e+02  -0.002
## conditionLiberal \n Signal       1.918e+00  1.808e-01  2.882e+02  10.608
##                                 Pr(>|t|)    
## (Intercept)                      < 2e-16 ***
## conditionConservative \n Signal 4.76e-12 ***
## conditionLiberal \n Control        0.998    
## conditionLiberal \n Signal       < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.570              
## cndtnLbrlCn -0.570  0.500       
## cndtnLbrlSg -0.570  0.500  0.500

summary(allEffects(mod_OMotivate))

##  model: OMotivate ~ condition
## 
##  condition effect
## condition
## Conservative \n Control  Conservative \n Signal      Liberal \n Control 
##                3.214922                4.519457                3.214542 
##       Liberal \n Signal 
##                5.132711 
## 
##  Lower 95 Percent Confidence Limits
## condition
## Conservative \n Control  Conservative \n Signal      Liberal \n Control 
##                2.903285                4.207820                2.902905 
##       Liberal \n Signal 
##                4.821074 
## 
##  Upper 95 Percent Confidence Limits
## condition
## Conservative \n Control  Conservative \n Signal      Liberal \n Control 
##                3.526559                4.831094                3.526179 
##       Liberal \n Signal 
##                5.444348

# same plot as b4 but with error bars from allEffects CI

mod_supGen <- lmer(supGen ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_supGen)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: supGen ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1188.3
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -4.6667 -0.3848  0.0885  0.4963  2.9478 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.784105 0.88550 
##  social_issue (Intercept) 0.008071 0.08984 
##  Residual                 0.796911 0.89270 
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       5.59087    0.13473  50.14852  41.497   <2e-16
## conditionConservative \n Signal  -0.02834    0.12754 288.11925  -0.222    0.824
## conditionLiberal \n Control      -0.01854    0.12754 288.11925  -0.145    0.884
## conditionLiberal \n Signal        0.12216    0.12754 288.11925   0.958    0.339
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.473              
## cndtnLbrlCn -0.473  0.500       
## cndtnLbrlSg -0.473  0.500  0.500

mod_vote <- lmer(vote ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_vote)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: vote ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1317
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.5165 -0.4511  0.0835  0.5080  2.7257 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.09866  1.0482  
##  social_issue (Intercept) 0.07752  0.2784  
##  Residual                 1.09420  1.0460  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                   Estimate Std. Error         df t value
## (Intercept)                       5.618632   0.204349  10.504948  27.495
## conditionConservative \n Signal  -0.022975   0.149462 288.030094  -0.154
## conditionLiberal \n Control      -0.052142   0.149462 288.030094  -0.349
## conditionLiberal \n Signal        0.008751   0.149462 288.030094   0.059
##                                 Pr(>|t|)    
## (Intercept)                     3.96e-11 ***
## conditionConservative \n Signal    0.878    
## conditionLiberal \n Control        0.727    
## conditionLiberal \n Signal         0.953    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.366              
## cndtnLbrlCn -0.366  0.500       
## cndtnLbrlSg -0.366  0.500  0.500

mod_sufficiency <- lmer(sufficiency ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_sufficiency)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: sufficiency ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1134.5
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -2.97275 -0.38479  0.04583  0.50868  3.12244 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 0.80169  0.8954  
##  social_issue (Intercept) 0.05001  0.2236  
##  Residual                 0.65662  0.8103  
## Number of obs: 392, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                     3.966e+00  1.655e-01 1.102e+01  23.964 7.39e-11
## conditionConservative \n Signal 8.809e-02  1.158e-01 2.880e+02   0.761    0.447
## conditionLiberal \n Control     5.949e-02  1.158e-01 2.880e+02   0.514    0.608
## conditionLiberal \n Signal      3.674e-04  1.158e-01 2.880e+02   0.003    0.997
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.350              
## cndtnLbrlCn -0.350  0.500       
## cndtnLbrlSg -0.350  0.500  0.500

mod_issueCur <- lmer(issueCur ~ condition + (1 | PID) + (1 | social_issue), data = gjg_long)
summary(mod_issueCur)

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: issueCur ~ condition + (1 | PID) + (1 | social_issue)
##    Data: gjg_long
## 
## REML criterion at convergence: 1123.6
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -3.06129 -0.49188  0.03395  0.51309  2.83801 
## 
## Random effects:
##  Groups       Name        Variance Std.Dev.
##  PID          (Intercept) 1.5362   1.2394  
##  social_issue (Intercept) 0.1095   0.3310  
##  Residual                 0.5266   0.7257  
## Number of obs: 391, groups:  PID, 98; social_issue, 4
## 
## Fixed effects:
##                                  Estimate Std. Error        df t value Pr(>|t|)
## (Intercept)                       3.06417    0.22009   8.43175  13.923 4.13e-07
## conditionConservative \n Signal  -0.04874    0.10369 287.05897  -0.470    0.639
## conditionLiberal \n Control      -0.12966    0.10369 287.05915  -1.250    0.212
## conditionLiberal \n Signal       -0.02464    0.10404 287.16076  -0.237    0.813
##                                    
## (Intercept)                     ***
## conditionConservative \n Signal    
## conditionLiberal \n Control        
## conditionLiberal \n Signal         
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cndtCS cndtLC
## cndtnCnsrvS -0.236              
## cndtnLbrlCn -0.236  0.500       
## cndtnLbrlSg -0.235  0.498  0.498

6.3.3 Social Issue Facet

wrapped_plot_cooker(gjg_long, condition, staticNorm, "Percent of support for issue", social_issue)

wrapped_plot_cooker(gjg_long, condition, dyNorm, "Change in support for issue", social_issue)

wrapped_plot_cooker(gjg_long, condition, SImport, "I think cause is important", social_issue)

wrapped_plot_cooker(gjg_long, condition, SLegit, "I think cause is legit", social_issue)

wrapped_plot_cooker(gjg_long, condition, SRelImport, "I think cause is important relative to other causes", social_issue)

wrapped_plot_cooker(gjg_long, condition, OImport, "Others think cause is important", social_issue)

wrapped_plot_cooker(gjg_long, condition, OLegit, "Others think cause is legit", social_issue)

wrapped_plot_cooker(gjg_long, condition, ORelImport, "Others think cause is important relative to other causes", social_issue)

wrapped_plot_cooker(gjg_long, condition, OMotivate, "Message motivates others to act", social_issue)

wrapped_plot_cooker(gjg_long, condition, supGen, "Supporters genuinely care", social_issue)

wrapped_plot_cooker(gjg_long, condition, vote, "I would want to know candidate's stance on this issue", social_issue)

wrapped_plot_cooker(gjg_long, condition, sufficiency, "How much progress is needed", social_issue)

wrapped_plot_cooker(gjg_long, condition, issueCur, "Curious about cause", social_issue)

6.3.4 Political Affiliation Facet

wrapped_plot_cooker(gjg_long, condition, staticNorm, "Percent of support for issue", pid)

wrapped_plot_cooker(gjg_long, condition, dyNorm, "Change in support for issue", pid)

wrapped_plot_cooker(gjg_long, condition, SImport, "I think cause is important", pid)

wrapped_plot_cooker(gjg_long, condition, SLegit, "I think cause is legit", pid)

wrapped_plot_cooker(gjg_long, condition, SRelImport, "I think cause is important relative to other causes", pid)

wrapped_plot_cooker(gjg_long, condition, OImport, "Others think cause is important", pid)

wrapped_plot_cooker(gjg_long, condition, OLegit, "Others think cause is legit", pid)

wrapped_plot_cooker(gjg_long, condition, ORelImport, "Others think cause is important relative to other causes", pid)

wrapped_plot_cooker(gjg_long, condition, OMotivate, "Message motivates others to act", pid)

wrapped_plot_cooker(gjg_long, condition, supGen, "Supporters genuinely care", pid)

wrapped_plot_cooker(gjg_long, condition, vote, "I would want to know candidate's stance on this issue", pid)

wrapped_plot_cooker(gjg_long, condition, sufficiency, "How much progress is needed", pid)

wrapped_plot_cooker(gjg_long, condition, issueCur, "Curious about cause", pid)

7 Factor 1 (Participant perceptions of issue)

7.1 Issue Importance

Loading = 0.94 “I think anti-racism, racial justice, and racial equality are important.” (1 = Strongly Disagree, 7 = Strongly Agree)

• Consider getting rid of ceiling by saying “Very important” or “Extremely important”

plot_cooker(gjg_long, condition, SImport, "I think cause is important")

7.2 Issue Relative Importance

Loading = 0.76 “There are many important issues in society. To what extent do you consider anti-racism, racial justice, and racial equality an important issue, relative to other issues in society?” (1 = Significantly less than other issues, 7 = Significantly more than other issues)

• Gregg noted that responses to this may be influenced by presentation order/seeing the other issues

plot_cooker(gjg_long, condition, SRelImport, "I think cause is important relative to other causes")

7.3 Noisy Minority motive

Loading = -0.56 “I think Chick-fil-A made this social media post to cater to a small group of people who are very vocal about this issue.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(signal_gjg_long, condition, noisyMin, "Motive to cater to noisy minority")

8 Factor 2 (Participant perceptions of company)

Tried to parallel this to factor 3

8.1 Like

Loading = 0.79 “I like Chick-fil-A.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(gjg_long, condition, STrust, "I trust company")

8.2 Trust

Loading = 0.91 “I trust Chick-fil-A.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(gjg_long, condition, STrust, "I trust company")

8.3 Competence

Loading = 0.60 “I think Chick-fil-A is competent.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(gjg_long, condition, SComp, "I think company is competent")

9 Factor 3 (Participant perceptions of how others perceive the company)

Tried to parallel this to factor 2

9.1 Like

Loading = 0.89 “Others like Chick-fil-A.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(gjg_long, condition, OTrust, "Others trust company")

9.2 Trust

Loading = 0.77 “Others trust Chick-fil-A.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(gjg_long, condition, OTrust, "Others trust company")

9.3 Competence

Loading = 0.79 “Others think Chick-fil-A is competent.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(gjg_long, condition, OComp, "Others think company is competent")

10 Factor 4 (Company motive)

10.1 Pressure from general public

Loading = 0.67 “I think Chick-fil-A made this social media post because of pressure from the general public.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(signal_gjg_long, condition, pubPress, "Pressured by public")

10.2 Pressure from peer organizations

Loading = 0.80 “I think Chick-fil-A made this social media post because of pressure from peer organizations.” (1 = Strongly disagree, 7 = Strongly agree)

 plot_cooker(signal_gjg_long, condition, peerPress, "Pressured by peer companies")

10.3 Was it risky not to send this signal

Loading = 0.58 “I think from a public relations perspective, it was risky for Chick-fil-A NOT to make this social media post.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(signal_gjg_long, condition, riskInact, "Risky NOT to send signal")

11 Other interesting measures

11.1 Static Norm

loaded onto factor 1 = 0.41 “What percent of Americans do you think support anti-racism, racial justice, and racial equality?” (Slider scale from 0 to 100%)

plot_cooker(gjg_long, condition, staticNorm, "Percent of support for issue")

11.2 Dynamic Norm

Didn’t load onto any factors “How has support for anti-racism, racial justice, and racial equality changed recently?” (1 = Decreasing quickly to 7 = Increasingly quickly)

plot_cooker(gjg_long, condition, dyNorm, "Change in support for issue")

11.3 Motivate Others

loaded onto factor 2 = 0.44 “Chick-fil-A’s post motivates others to support anti-racism, racial justice, and racial equality.” (1 = Strongly Disagree, 7 = Strongly Agree)

plot_cooker(gjg_long, condition, OMotivate, "Message motivates others to act")

11.4 Reputation Motive

Loading = -0.47 onto factor 2, 0.40 onto factor 4 “I think Chick-fil-A made this social media post because they think it will help their reputation.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(signal_gjg_long, condition, Rep, "Reputation motive")

11.5 Profit Motive

Loading = -0.43, 0.40 onto factor 4 “I think Chick-fil-A made this social media post because they think it will help increase their profits.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(signal_gjg_long, condition, Profit, "Profit motive")

11.6 Surprise

Didn’t load onto any factor “I was surprised that Chick-fil-A made this social media post.” (1 = Strongly disagree, 7 = Strongly agree)

plot_cooker(gjg_long, condition, staticNorm, "Percent of support for issue")