Cleanup
experiment_two_DS_time <- read.csv("Experiment_2_Dataset_clean.csv")
experiment_two_DS <- read.csv("Experiment_2_Dataset_clean_noTime.csv")
corr_table <- readRDS("corr_table.rds")
m4 <- readRDS("m4.rds")
m5 <- readRDS("m5.rds")
m5_gen <- readRDS("m5_gen.rds")
m7 <- readRDS("m7.rds")
m8 <- readRDS("m8.rds")
m8_gen <- readRDS("m8_gen.rds")
m8_Age <- readRDS("m8_Age.rds")
dm0 <- readRDS("dm0.rds")
dm1 <- readRDS("dm1.rds")
dm2 <- readRDS("dm2.rds")
dm3 <- readRDS("dm3.rds")
dm4 <- readRDS("dm4.rds")
dm5 <- readRDS("dm5.rds")
demographicQuestions <- c("Age", "Gender", "Ethnicity", "Eth.ori", "Edu")
sjsQuestions <- c("SJS.1", "SJS.2", "SJS.3")
srpsQuestions <- c("SRPS.1"
, "SRPS.2"
, "SRPS.3"
, "SRPS.4"
, "SRPS.5"
, "SRPS.6"
, "SRPS.7"
, "SRPS.8"
, "SRPS.9"
, "SRPS.10"
, "SRPS.11"
, "SRPS.12"
, "SRPS.13"
, "SRPS.14"
, "SRPS.15"
, "SRPS.17"
, "SRPS.18"
, "SRPS.19"
, "SRPS.20"
, "SRPS.21"
, "SRPS.22"
, "SRPS.23"
, "SRPS.24")
srps.rcQuestions <- c("SRPS.1"
, "SRPS.2"
, "SRPS.3"
, "SRPS.4"
, "SRPS.5"
, "SRPS.6"
, "SRPS.7"
, "SRPS.8"
, "SRPS.9"
, "SRPS.10"
, "SRPS.11"
, "SRPS.12"
, "SRPS.13"
, "SRPS.14"
, "SRPS.15")
srps.dmQuestions <- c("SRPS.17"
, "SRPS.18"
, "SRPS.19"
, "SRPS.20"
, "SRPS.21"
, "SRPS.22"
, "SRPS.23"
, "SRPS.24")
ssesQuestions <- c("SSES.1"
, "SSES.2"
, "SSES.3"
, "SSES.4"
, "SSES.5"
, "SSES.6"
, "SSES.7"
, "SSES.8"
, "SSES.9"
, "SSES.10")
spiteQuestions <- c("Spite.1"
, "Spite.2"
, "Spite.3"
, "Spite.4"
, "Spite.5"
, "Spite.6"
, "Spite.7"
, "Spite.8"
, "Spite.9"
, "Spite.10"
, "Spite.11"
, "Spite.12"
, "Spite.13"
, "Spite.14"
, "Spite.15"
, "Spite.16")
Vign.Sex <- c("Vign.1", "Vign.2", "Vign.4", "Vign.7", "Vign.10")
Vign.noSex <- c("Vign.3", "Vign.5", "Vign.6", "Vign.8", "Vign.9")
Vign.Overall <- c("Vign.1", "Vign.2", "Vign.4", "Vign.7", "Vign.10", "Vign.3", "Vign.5", "Vign.6", "Vign.8", "Vign.9")
Real.Sex <- c("Real.1", "Real.2", "Real.4", "Real.7", "Real.10")
Real.noSex <- c("Real.3", "Real.5", "Real.6", "Real.8", "Real.9")
Real.Overall <- c("Real.1", "Real.2", "Real.4", "Real.7", "Real.10", "Real.3", "Real.5", "Real.6", "Real.8", "Real.9")
doplQuestions <- c('DoPL_1', 'DoPL_6', 'DoPL_11', 'DoPL_13', 'DoPL_14', 'DoPL_16', 'DoPL_5', 'DoPL_7', 'DoPL_8', 'DoPL_12', 'DoPL_17', 'DoPL_18', 'DoPL_2', 'DoPL_3', 'DoPL_4', 'DoPL_9', 'DoPL_10', 'DoPL_15')
dominanceQuestions <- c('DoPL_2','DoPL_3','DoPL_4','DoPL_9','DoPL_10','DoPL_15')
prestigeQuestions <- c('DoPL_5','DoPL_7','DoPL_8','DoPL_12','DoPL_17','DoPL_18')
leadershipQuestions <- c('DoPL_1','DoPL_6','DoPL_11','DoPL_13','DoPL_14','DoPL_16')
UMSQuestions <- c('UMS_1', 'UMS_2', 'UMS_3', 'UMS_4','UMS_5','UMS_6','UMS_7','UMS_8','UMS_9','UMS_11', 'UMS_12')
UMSIntimacyQuestions <- c('UMS_11', 'UMS_12')
UMSAffiliationQuestions <- c('UMS_1', 'UMS_2', 'UMS_3', 'UMS_4','UMS_5','UMS_6','UMS_7','UMS_8','UMS_9')
savedQuestionsBefore <- c('subjectID', "Age", "Duration..in.seconds.", "Gender", "Ethnicity", "Eth.ori", "Edu",
"DoPLSum"
,"dominanceSum"
,"prestigeSum"
,"leadershipSum"
,"UMSSum"
,"UMSIntimacySum"
,"UMSAffiliationSum"
,"sjsSum"
,"ssesSum"
, "Spite"
,"Vign.Sex"
,"Vign.noSex"
,"Vign.Ovr.Score"
,"Real.Sex"
,"Real.noSex"
,"Real.Ovr.Score"
,"srps.rc.rescale"
,"srps.dm.rescale"
,"srps.sum.rescale",
"Power",
"Vign.1",
"Vign.2",
"Vign.3",
"Vign.4",
"Vign.5",
"Vign.6",
"Vign.7",
"Vign.8",
"Vign.9",
"Vign.10",
"Real.1",
"Real.2",
"Real.3",
"Real.4",
"Real.5",
"Real.6",
"Real.7",
"Real.8",
"Real.9",
"Real.10")
savedQuestionsAfter <- c('subjectID', "Age", "Duration..in.seconds.", "Gender", "Ethnicity", "Eth.ori", "Edu",
"DoPLSum"
,"dominanceSum"
,"prestigeSum"
,"leadershipSum"
,"UMSSum"
,"UMSIntimacySum"
,"UMSAffiliationSum"
,"sjsSum"
,"ssesSum"
, "Spite"
,"Vign.Sex"
,"Vign.noSex"
,"Vign.Ovr.Score"
,"Real.Sex"
,"Real.noSex"
,"Real.Ovr.Score"
,"srps.rc.rescale"
,"srps.dm.rescale"
,"srps.sum.rescale"
,"Content",
"Vignette",
"Justification",
"Realism",
"Power")
Recoding variables
experiment_two_DS <- experiment_two_DS %>%
mutate_at(vars(locfunc(experiment_two_DS, "Gender")), ~as.numeric(recode(.,"Male" = 0, "Female" = 1))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "Edu")), ~as.numeric(recode(., "Prefer not to answer" = 0, "Primary School" = 1, "GCSes or Equivalent" = 2, "A-Levels or Equivalent" = 3, "University Undergraduate Program" = 4, "University Post-Graduate Program" = 5, "Doctoral Degree" = 6))) %>%
mutate_at(vars(locfunc(experiment_two_DS, 'Ethnicity')), ~as.numeric(recode(., "Prefer not to answer" = 0, "White" = 1, "Mixed or Multiple ethnic origins" = 2, "Asian or Asian Scottish or Asian British" = 3, "African" = 4, "Caribbean or Black" = 5,"Arab" = 6, "Other ethnic group" = 7))) %>%
mutate_at(vars(locfunc(experiment_two_DS, 'Eth.ori')), ~as.numeric(recode(., "Prefer not to answer" = 0, "Scottish" = 1, "English" = 2, "European" = 3, "Latin American" = 4, "Asian" = 5, "Arab" = 6,
"African" = 7, "Other" = 8 ))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "DoPL_1"):locfunc(experiment_two_DS, "DoPL_5")), ~as.numeric(recode(., "Strongly disagree" = 0, "Disagree" = 1, "Somewhat disagree" = 2, "Somewhat agree" = 3, "Agree" = 4, "Strongly agree" = 5))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "DoPL_6"), locfunc(experiment_two_DS, "DoPL_14")), ~as.numeric(recode(., "Strongly disagree" = 5, "Disagree" = 4, "Somewhat disagree" = 3, "Somewhat agree" = 2, "Agree" = 1, "Strongly agree" = 0))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "DoPL_7"):locfunc(experiment_two_DS, "DoPL_13")), ~as.numeric(recode(., "Strongly disagree" = 0, "Disagree" = 1, "Somewhat disagree" = 2, "Somewhat agree" = 3, "Agree" = 4, "Strongly agree" = 5))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "DoPL_15"):locfunc(experiment_two_DS, "DoPL_16")), ~as.numeric(recode(., "Strongly disagree" = 0, "Disagree" = 1, "Somewhat disagree" = 2, "Somewhat agree" = 3, "Agree" = 4, "Strongly agree" = 5))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "UMS_1"):locfunc(experiment_two_DS, "UMS_9")), ~as.numeric(recode(., "Strongly disagree" = 0, "Disagree" = 1, "Somewhat disagree" = 2, "Somewhat agree" = 3, "Agree" = 4, "Strongly agree" = 5))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "DoPL_17"):locfunc(experiment_two_DS, "DoPL_18")), ~as.numeric(recode(., "Not Important To Me" = 0, "Of Little Importance To me" = 1, "Of Some Importance To Me" = 2, "Important To Me" = 3, "Very Important To me" = 4, "Extremely Important To Me" = 5))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "UMS_10"):locfunc(experiment_two_DS, "UMS_12")), ~as.numeric(recode(., "Not Important To Me" = 1, "Of Little Importance To me" = 2, "Of Some Importance To Me" = 3, "Important To Me" = 4, "Very Important To me" = 5, "Extremely Important To Me" = 6))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "Spite.1"):locfunc(experiment_two_DS, "Spite.13")), ~as.numeric(recode(., "Strongly disagree" = 1, "Disagree" = 2, "Neither agree nor disagree" = 3, "Agree" = 4, "Strongly agree" = 5, "Strongly Disagree" = 1, "Somewhat disagree" = 2))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "Spite.14"):locfunc(experiment_two_DS, "Spite.16")), ~as.numeric(recode(., "Strongly disagree" = 5, "Disagree" = 4, "Neither agree nor disagree" = 3, "Agree" = 2, "Strongly agree" = 1, "Strongly Disagree" = 5))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SSES.1"):locfunc(experiment_two_DS, "SSES.3")), ~as.numeric(recode(., "Agree" = 1, "Slightly agree" = 2, "Neither agree nor disagree" = 3, "Slightly Disagree" = 4, "Disagree" = 5, "Slight Disagree" = 4, "Slight disagree" = 4, "Slightly disagree" = 4))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SSES.4"):locfunc(experiment_two_DS, "SSES.5")), ~as.numeric(recode(., "Agree" = 5, "Slightly agree" = 4, "Neither agree nor disagree" = 3, "Slightly Disagree" = 2, "Disagree" = 1, "Slight Disagree" = 2, "Slight disagree" = 2, "Slightly disagree" = 2))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SSES.6")), ~as.numeric(recode(., "Agree" = 1, "Slightly agree" = 2, "Neither agree nor disagree" = 3, "Slightly Disagree" = 4, "Disagree" = 5, "Slight Disagree" = 4, "Slight disagree" = 4, "Slightly disagree" = 4))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SSES.7")), ~as.numeric(recode(., "Agree" = 5, "Slightly agree" = 4, "Neither agree nor disagree" = 3, "Slightly Disagree" = 2, "Disagree" = 1, "Slight Disagree" = 2, "Slight disagree" = 2, "Slightly disagree" = 2))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SSES.8")), ~as.numeric(recode(., "Agree" = 1, "Slightly agree" = 2, "Neither agree nor disagree" = 3, "Slightly Disagree" = 4, "Disagree" = 5, "Slight Disagree" = 4, "Slight disagree" = 4, "Slightly disagree" = 4))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SSES.9"):locfunc(experiment_two_DS, "SSES.10")), ~as.numeric(recode(., "Agree" = 5, "Slightly agree" = 4, "Neither agree nor disagree" = 3, "Slightly Disagree" = 2, "Disagree" = 1, "Slight Disagree" = 2, "Slight disagree" = 2, "Slightly disagree" = 2))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SRPS.1"):locfunc(experiment_two_DS, "SRPS.15")), ~as.numeric(recode(., "Strongly agree" = 1, "Agree" = 2, "Disagree" = 3, "Strongly disagree" = 4))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SRPS.17"):locfunc(experiment_two_DS, "SRPS.24")), ~as.numeric(recode(., "Your Partner" = 1, "Both of You Equally" = 2, "You" = 3))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "SJS.1"):locfunc(experiment_two_DS, "SJS.3")), ~as.numeric(recode(., "Strongly disagree" = 1, "Disagree" = 2, "Somewhat disagree" = 3, "Somewhat agree" = 4, "Agree" = 5, "Strongly agree" = 6))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "Real.1"):locfunc(experiment_two_DS, "Real.10")), ~as.numeric(recode(., "Strongly agree" = 1, "Agree" = 2, "Neither agree nor disagree" = 3, "Disagree" = 4, "Strongly disagree" = 5))) %>%
mutate_at(vars(locfunc(experiment_two_DS, "Vign.1"):locfunc(experiment_two_DS, "Vign.10")), ~as.numeric(recode(., "Not justified at all" = 1, "Somewhat unjustified" = 2, "Neither justified or unjustified" = 3, "Slightly justified" = 4, "Very justified" = 5)))
Summing data
# DOPL----
experiment_two_DS$DoPLSum <- rowSums(experiment_two_DS[, doplQuestions])
experiment_two_DS$dominanceSum <-
rowSums(experiment_two_DS[, dominanceQuestions])
experiment_two_DS$prestigeSum <-
rowSums(experiment_two_DS[, prestigeQuestions])
experiment_two_DS$leadershipSum <-
rowSums(experiment_two_DS[, leadershipQuestions])
# UMS----
experiment_two_DS$UMSSum <- rowSums(experiment_two_DS[, UMSQuestions])
experiment_two_DS$UMSIntimacySum <- rowSums(experiment_two_DS[, UMSIntimacyQuestions])
experiment_two_DS$UMSAffiliationSum <- rowSums(experiment_two_DS[, UMSAffiliationQuestions])
# Spite ----
experiment_two_DS$Spite <- rowSums(experiment_two_DS[, spiteQuestions])
# SJS----
experiment_two_DS$sjsSum <- rowSums(experiment_two_DS[, sjsQuestions])
# SSES----
experiment_two_DS$ssesSum <- rowSums(experiment_two_DS[, ssesQuestions])
# Vign----
experiment_two_DS$Vign.Sex <- rowSums(experiment_two_DS[, Vign.Sex])
experiment_two_DS$Vign.noSex <- rowSums(experiment_two_DS[, Vign.noSex])
experiment_two_DS$Vign.Ovr.Score <- rowSums(experiment_two_DS[, Vign.Overall])
# Realism----
experiment_two_DS$Real.Sex <- rowSums(experiment_two_DS[, Real.Sex])
experiment_two_DS$Real.noSex <- rowSums(experiment_two_DS[, Real.noSex])
experiment_two_DS$Real.Ovr.Score <- rowSums(experiment_two_DS[, Real.Overall])
# SRPS----
experiment_two_DS$srps.rc.sum <- rowSums(experiment_two_DS[, srps.rcQuestions])
experiment_two_DS$srps.dm.sum <- rowSums(experiment_two_DS[, srps.dmQuestions])
# SRPS Rescaled----
experiment_two_DS$srps.rc.rescale <- ((experiment_two_DS$srps.rc.sum - 15)/(60 - 15)) * 3 + 1
experiment_two_DS$srps.dm.rescale <- ((experiment_two_DS$srps.dm.sum - 8)/(24 - 8)) * 3 + 1
experiment_two_DS$srps.sum <- ((experiment_two_DS$srps.dm.sum + experiment_two_DS$srps.rc.sum)/2)
experiment_two_DS$srps.sum.rescale <- ((experiment_two_DS$srps.sum - 11.5) / (42 - 11.5)) * 3 + 1
experiment_two_DS <- experiment_two_DS %>% mutate(Power = case_when(
experiment_two_DS$srps.sum.rescale < 2.420 ~ "Low",
experiment_two_DS$srps.sum.rescale > 2.420 & experiment_two_DS$srps.sum.rescale < 2.820 ~ "Medium",
experiment_two_DS$srps.sum.rescale > 2.820 ~ "High"
))
Renaming
experiment_two_DS <- rowid_to_column(experiment_two_DS, "subjectID")
newDF <- experiment_two_DS[savedQuestionsBefore]
analysisDF <- newDF %>%
pivot_longer("Vign.1":"Vign.10", names_to = "Vignette", values_to = "Justification")
tester1 <- newDF %>%
pivot_longer("Real.1":"Real.10", names_to = "Real", values_to = "Realism")
analysisDF['Realism'] = tester1["Realism"]
analysisDF <- analysisDF %>% mutate(Content = case_when(
analysisDF$Vignette == "Vign.1" ~ "Sexual",
analysisDF$Vignette == "Vign.2" ~ "Sexual",
analysisDF$Vignette == "Vign.3" ~ "Nonsexual",
analysisDF$Vignette == "Vign.4" ~ "Sexual",
analysisDF$Vignette == "Vign.5" ~ "Nonsexual",
analysisDF$Vignette == "Vign.6" ~ "Nonsexual",
analysisDF$Vignette == "Vign.7" ~ "Sexual",
analysisDF$Vignette == "Vign.8" ~ "Nonsexual",
analysisDF$Vignette == "Vign.9" ~ "Nonsexual",
analysisDF$Vignette == "Vign.10" ~ "Sexual"
))
analysisDF <- analysisDF %>% mutate(Vignette = case_when(
analysisDF$Vignette == "Vign.1" ~ 1,
analysisDF$Vignette == "Vign.2" ~ 2,
analysisDF$Vignette == "Vign.3" ~ 3,
analysisDF$Vignette == "Vign.4" ~ 4,
analysisDF$Vignette == "Vign.5" ~ 5,
analysisDF$Vignette == "Vign.6" ~ 6,
analysisDF$Vignette == "Vign.7" ~ 7,
analysisDF$Vignette == "Vign.8" ~ 8,
analysisDF$Vignette == "Vign.9" ~ 9,
analysisDF$Vignette == "Vign.10" ~ 10
))
analysisDF <- analysisDF[, savedQuestionsAfter]
analysisDF <- analysisDF %>%
rename("subjectID" = "subjectID") %>%
rename("Duration" = "Duration..in.seconds.") %>%
rename("Age" = "Age") %>%
rename("Gender" = "Gender") %>%
rename("Ethnicity" = "Ethnicity") %>%
rename("Ethnic_Origin" = "Eth.ori") %>%
rename("Education" = "Edu") %>%
rename("DoPL" = "DoPLSum") %>%
rename("Dominance" = "dominanceSum") %>%
rename("Prestige" = "prestigeSum") %>%
rename("Leadership" = "leadershipSum") %>%
rename("UMS" = "UMSSum") %>%
rename("UMS_Intimacy" = "UMSIntimacySum") %>%
rename("UMS_Affiliation" = "UMSAffiliationSum") %>%
rename("SJS" = "sjsSum") %>%
rename("SSES" = "ssesSum") %>%
rename("Vign.Sex" = "Vign.Sex") %>%
rename("Vign.noSex" = "Vign.noSex") %>%
rename("Vign.Ovr.Score" = "Vign.Ovr.Score") %>%
rename("Realism_Sex" = "Real.Sex") %>%
rename("Realism_NoSex" = "Real.noSex") %>%
rename("Realism_Overall" = "Real.Ovr.Score") %>%
rename("SRPS_RC" = "srps.rc.rescale") %>%
rename("SRPS_DM" = "srps.dm.rescale") %>%
rename("SRPS" = "srps.sum.rescale") %>%
rename("Vignette" = "Vignette") %>%
rename("Justification" = "Justification") %>%
rename("Realism" = "Realism") %>%
rename("Content" = "Content")
Fixing class
analysisDF$Vignette <- factor(analysisDF$Vignette)
analysisDF$subjectID <- factor(analysisDF$subjectID)
analysisDF$Spite_z <- scale(analysisDF$Spite)
analysisDF$SSES_z <- scale(analysisDF$SSES)
analysisDF$SRPS_z <- scale(analysisDF$SRPS)
analysisDF$SJS_z <- scale(analysisDF$SJS)
analysisDF$Justification_z <- scale(analysisDF$Justification)
analysisDF$Realism_z <- scale(analysisDF$Realism)
analysisDF$DoPL_z <- scale(analysisDF$DoPL)
analysisDF$Dominance_z <- scale(analysisDF$Dominance)
analysisDF$Leadership_z <- scale(analysisDF$Leadership)
analysisDF$Prestige_z <- scale(analysisDF$Prestige)
analysisDF$UMS_z <- scale(analysisDF$UMS)
analysisDF$UMS_Intimacy_z <- scale(analysisDF$UMS_Intimacy)
analysisDF$UMS_Affiliation_z <- scale(analysisDF$UMS_Affiliation)
Basic gglots
ggplot(analysisDF, aes(x = DoPL, y = Justification, color = Content)) + geom_point() + geom_smooth(method = "lm")
ggplot(analysisDF, aes(x = Dominance, y = Justification, color = Content)) + geom_point() + geom_smooth(method = "lm")
ggplot(analysisDF, aes(x = Leadership, y = Justification, color = Content)) + geom_point() + geom_smooth(method = "lm")
ggplot(analysisDF, aes(x = Prestige, y = Justification, color = Content)) + geom_point() + geom_smooth(method = "lm")
Initial Bayesian Analysis
m0 <- brm(Justification_z ~ 1 + (1|Vignette) + (1|subjectID), data = analysisDF,
warmup = 500, control = list(adapt_delta = 0.95)), iter = 8000, chains = 4, cores = 4, save_all_pars = T,
prior = c(prior(normal(0,1), class = 'Intercept')))
# models
m1 <- brm(Justification_z ~ Spite_z * Content + (1|Vignette) + (1|subjectID), data = analysisDF,
warmup = 500, control = list(adapt_delta = 0.95)), iter = 8000, chains = 4, cores = 6, save_all_pars = T,
prior = c(prior(normal(0,1), class = 'Intercept'),
prior(normal(0,1), class = 'b'),
prior(normal(0.151, 0.0723), class = b, coef = "Spite_z:ContentSexual")))
summary(m1)
m2 <- brm(Justification ~ Spite_z * Content + Realism_z + (1|Vignette) + (1|subjectID), data = analysisDF,
warmup = 500, control = list(adapt_delta = 0.95)), iter = 8000, chains = 4, cores = 4, save_all_pars = T,
prior = c(prior(normal(0,1), class = 'Intercept'),
prior(normal(0,1), class = 'b'),
prior(normal(0.151, 0.0723), class = b, coef = "Spite_z:ContentSexual")))
summary(m2)
m3 <- brm(Justification ~ Spite_z * Content + Realism_z + SSES + SRPS + SJS + (1|Vignette) +
(1|subjectID), data = analysisDF,warmup = 500, control = list(adapt_delta = 0.95)), iter = 8000, chains = 4, cores = 6,
save_all_pars = T,
prior = c(prior(normal(0,1), class = 'Intercept'),
prior(normal(0,1), class = 'b'),
prior(normal(0.151, 0.0723), class = b, coef = "Spite_z:ContentSexual")))
summary(m3)
m4 <- brm(Justification ~ Spite_z * Content * Realism_z + SSES + SRPS + SJS + (1|Vignette) +
(1|subjectID), data = analysisDF,warmup = 500, control = list(adapt_delta = 0.95)), iter = 8000, chains = 4, cores = 6,
save_all_pars = T,
prior = c(prior(normal(0,1), class = 'Intercept'),
prior(normal(0,1), class = 'b'),
prior(normal(0.151, 0.0723), class = b, coef = "Spite_z:ContentSexual")))
Bayesian Correlation
scale_corr <- brm(mvbind(SSES, SRPS, Spite, SJS, Dominance, Prestige, Leadership) ~ 1, data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6,
prior = c(prior(normal(0, 1), class = "Intercept")), save_all_pars = T)
cor_1 <- summary(scale_corr)
corr_table <- cor_1[["rescor_pars"]]
kable(corr_table, format = "html", booktabs = T, escape = F, longtabe = F, digits = 2, col.names = c('Estimate', 'Est.Error', 'l-95% CI', 'u-95% CI', 'Rhat', 'Bulk_ESS', 'Tail_ESS')) %>%
kable_styling(full_width = F) %>%
remove_column(6:8)
|
|
Estimate
|
Est.Error
|
l-95% CI
|
u-95% CI
|
|
rescor(SSES,SRPS)
|
-0.41
|
0.03
|
-0.47
|
-0.36
|
|
rescor(SSES,Spite)
|
0.19
|
0.04
|
0.12
|
0.26
|
|
rescor(SRPS,Spite)
|
-0.20
|
0.04
|
-0.27
|
-0.13
|
|
rescor(SSES,SJS)
|
0.26
|
0.03
|
0.20
|
0.33
|
|
rescor(SRPS,SJS)
|
-0.28
|
0.03
|
-0.34
|
-0.22
|
|
rescor(Spite,SJS)
|
0.24
|
0.04
|
0.17
|
0.31
|
|
rescor(SSES,Dominance)
|
-0.11
|
0.04
|
-0.18
|
-0.04
|
|
rescor(SRPS,Dominance)
|
0.03
|
0.03
|
-0.04
|
0.10
|
|
rescor(Spite,Dominance)
|
0.55
|
0.03
|
0.50
|
0.60
|
|
rescor(SJS,Dominance)
|
0.29
|
0.03
|
0.22
|
0.35
|
|
rescor(SSES,Prestige)
|
-0.01
|
0.04
|
-0.08
|
0.06
|
|
rescor(SRPS,Prestige)
|
0.24
|
0.03
|
0.18
|
0.31
|
|
rescor(Spite,Prestige)
|
0.13
|
0.04
|
0.06
|
0.20
|
|
rescor(SJS,Prestige)
|
0.02
|
0.03
|
-0.05
|
0.09
|
|
rescor(Dominance,Prestige)
|
0.22
|
0.03
|
0.16
|
0.29
|
|
rescor(SSES,Leadership)
|
-0.25
|
0.03
|
-0.32
|
-0.19
|
|
rescor(SRPS,Leadership)
|
0.29
|
0.03
|
0.22
|
0.35
|
|
rescor(Spite,Leadership)
|
0.01
|
0.04
|
-0.06
|
0.08
|
|
rescor(SJS,Leadership)
|
-0.07
|
0.03
|
-0.13
|
0.00
|
|
rescor(Dominance,Leadership)
|
0.31
|
0.03
|
0.25
|
0.37
|
|
rescor(Prestige,Leadership)
|
0.38
|
0.03
|
0.32
|
0.43
|
DoPL motives predicting risky behavior (by content) in addition to spitefulness
dm0 <- brm(Justification ~ 1 + (1|Vignette) + (1|subjectID), data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6
,prior = c(prior(normal(0, 1), class = "Intercept")), save_all_pars = T)
summary(dm0)
dm1 <- brm(Justification ~ Dominance_z * Gender + Prestige_z * Gender + Leadership_z * Gender + (1|Vignette) + (1|subjectID), data = analysisDF, iter = 8000,control = list(adapt_delta = 0.99), warmup = 500, chains = 4, cores = 6
,prior = c(prior(normal(0, 1), class = "Intercept"),
prior(normal(3, 1), class = "b", coef = "Dominance_z:Gender"),
prior(normal(0, 1), class = "b", coef = "Gender:Leadership_z"),
prior(normal(0, 1), class = "b", coef = "Gender:Prestige_z"),
prior(normal(0, 1), class = "b", coef = "Dominance_z"),
prior(normal(0, 1), class = "b", coef = "Leadership_z"),
prior(normal(0, 1), class = "b", coef = "Prestige_z")), save_all_pars = T)
summary(dm1)
dm3 <- brm(Justification ~ Dominance_z * Gender + Prestige_z * Gender + Leadership_z * Gender + Age + (1|Vignette) + (1|subjectID), data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6
,prior = c(prior(normal(0, 1), class = "Intercept")), save_all_pars = T)
summary(dm3)
dm2 <- brm(Justification ~ Dominance_z * Gender + Prestige_z * Gender + Leadership_z * Gender + SJS_z + SRPS_z + SSES_z + Spite_z + (1|Vignette) + (1|subjectID), control = list(adapt_delta = 0.99), data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6
,prior = c(prior(normal(0, 1), class = "Intercept"),
prior(normal(3, 1), class = "b", coef = "Dominance_z:Gender"),
prior(normal(0, 1), class = "b", coef = "Gender:Leadership_z"),
prior(normal(0, 1), class = "b", coef = "Gender:Prestige_z"),
prior(normal(0, 1), class = "b", coef = "Dominance_z"),
prior(normal(0, 1), class = "b", coef = "Leadership_z"),
prior(normal(0, 1), class = "b", coef = "Prestige_z")), save_all_pars = T)
summary(dm2)
dm4 <- brm(Justification ~ Dominance_z * Gender * Content + Prestige_z * Gender * Content + Leadership_z * Gender * Content + SJS_z + SRPS_z + SSES_z + Spite_z + (1|Vignette) + (1|subjectID), control = list(adapt_delta = 0.99), data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6
,prior = c(prior(normal(0, 1), class = "Intercept"),
prior(normal(3, 1), class = "b", coef = "Dominance_z:Gender"),
prior(normal(0, 1), class = "b", coef = "Gender:Leadership_z"),
prior(normal(0, 1), class = "b", coef = "Gender:Prestige_z"),
prior(normal(0, 1), class = "b", coef = "Dominance_z"),
prior(normal(0, 1), class = "b", coef = "Leadership_z"),
prior(normal(0, 1), class = "b", coef = "Prestige_z")), save_all_pars = T)
summary(dm4)
dm4_hdi <- bayestestR::hdi(dm4, effects = "fixed", component = "conditional", ci = .95)
dm4_hdi[sign(dm4_hdi$CI_low) == sign(dm4_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')]
dm5 <- brm(Justification ~ Dominance_z * Realism_z * Content + Prestige_z * Realism_z * Content + Leadership_z * Realism_z * Content + SJS_z + SRPS_z + SSES_z + Spite_z + (1|Vignette) + (1|subjectID), control = list(adapt_delta = 0.99), data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6
,prior = c(prior(normal(0, 1), class = "Intercept"),
prior(normal(0, 1), class = "b", coef = "Dominance_z"),
prior(normal(0, 1), class = "b", coef = "Leadership_z"),
prior(normal(0, 1), class = "b", coef = "Prestige_z")), save_all_pars = T)
summary(dm5)
loo(dm1, dm2, dm4, dm5)
DoPL motives predicting risky behavior (by content) in addition to spitefulness HDI
dm5_hdi <- bayestestR::hdi(dm5, effects = "fixed", component = "conditional", ci = .95)
kable(dm5_hdi[sign(dm5_hdi$CI_low) == sign(dm5_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')], format = "html", booktabs = T, escape = F, longtable = F, digits = 2) %>% kable_styling(full_width = T)
|
|
Parameter
|
CI
|
CI_low
|
CI_high
|
|
8
|
b_Intercept
|
0.95
|
0.74
|
3.27
|
|
18
|
b_Spite_z
|
0.95
|
0.06
|
0.24
|
|
5
|
b_Dominance_z.ContentSexual
|
0.95
|
0.01
|
0.28
|
Bayesian analysis m4
summary(m4)
## Warning: There were 2 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup
## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: Justification ~ Spite_z * Content * Realism_z + SSES + SRPS + SJS + (1 | Vignette) + (1 | subjectID)
## Data: analysisDF (Number of observations: 920)
## Samples: 4 chains, each with iter = 8000; warmup = 1000; thin = 1;
## total post-warmup samples = 28000
##
## Group-Level Effects:
## ~subjectID (Number of levels: 92)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.18 0.06 0.04 0.30 1.00 5343 3628
##
## ~Vignette (Number of levels: 10)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 1.33 0.40 0.79 2.33 1.00 7556 10705
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat
## Intercept 1.71 0.71 0.25 3.05 1.00
## Spite_z 0.12 0.05 0.03 0.21 1.00
## ContentSexual -0.10 0.63 -1.36 1.15 1.00
## Realism_z -0.10 0.06 -0.21 0.01 1.00
## SSES -0.00 0.00 -0.01 0.01 1.00
## SRPS 0.08 0.12 -0.16 0.32 1.00
## SJS 0.02 0.02 -0.02 0.05 1.00
## Spite_z:ContentSexual 0.07 0.05 -0.02 0.17 1.00
## Spite_z:Realism_z -0.11 0.05 -0.21 -0.02 1.00
## ContentSexual:Realism_z -0.02 0.07 -0.16 0.13 1.00
## Spite_z:ContentSexual:Realism_z 0.06 0.06 -0.06 0.19 1.00
## Bulk_ESS Tail_ESS
## Intercept 10635 14416
## Spite_z 27293 23384
## ContentSexual 9986 14758
## Realism_z 24794 22641
## SSES 34317 23121
## SRPS 30251 23206
## SJS 32126 22822
## Spite_z:ContentSexual 35115 23079
## Spite_z:Realism_z 24685 22889
## ContentSexual:Realism_z 25942 23012
## Spite_z:ContentSexual:Realism_z 26408 22494
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.99 0.02 0.95 1.04 1.00 25851 19058
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).
m4_hdi <- bayestestR::hdi(m4, effects = "fixed", component = "conditional", ci = .95)
kable(m4_hdi[sign(m4_hdi$CI_low) == sign(m4_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')] , format = "html", booktabs = T, escape = F, longtable = F, digits = 2) %>% kable_styling(full_width = T)
|
|
Parameter
|
CI
|
CI_low
|
CI_high
|
|
3
|
b_Intercept
|
0.95
|
0.26
|
3.07
|
|
6
|
b_Spite_z
|
0.95
|
0.03
|
0.21
|
|
9
|
b_Spite_z.Realism_z
|
0.95
|
-0.21
|
-0.02
|
m5 <- brm(Justification ~ Dominance_z + Leadership_z + Prestige_z + (1|Vignette) + (1|subjectID), data = analysisDF,warmup = 500, control = list(adapt_delta = 0.95)), iter = 8000, chains = 4, cores = 6, prior = c(prior(normal(0,1), class = 'Intercept'),
prior(normal(0,1), class = 'b')))
summary(m5)
m5_gen <- brm(Justification_z ~ Dominance_z*Gender + Leadership_z*Gender + Prestige_z*Gender + (1|Vignette) + (1|subjectID), data = analysisDF, warmup = 500, control = list(adapt_delta = 0.95), iter = 8000, chains = 4, cores = 6,
prior = c(prior(normal(0,1), class = 'Intercept'),
prior(normal(3, 1), class = 'b', coef = "Dominance_z:Gender"),
prior(normal(0, 1), class = 'b', coef = "Gender:Leadership_z"),
prior(normal(0, 1), class = 'b', coef = "Gender:Prestige_z")),
save_all_pars = T)
summary(m5_gen)
DoPL and Justification
summary(m5_gen)
## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: Justification_z ~ Dominance_z * Gender + Leadership_z * Gender + Prestige_z * Gender + (1 | Vignette) + (1 | Vignette)
## Data: analysisDF (Number of observations: 920)
## Samples: 4 chains, each with iter = 8000; warmup = 1000; thin = 1;
## total post-warmup samples = 28000
##
## Group-Level Effects:
## ~Vignette (Number of levels: 10)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.86 0.24 0.53 1.44 1.00 6759 11597
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.01 0.27 -0.53 0.55 1.00 5588 8686
## Dominance_z 0.08 0.03 0.03 0.14 1.00 27825 21036
## Gender -0.03 0.05 -0.13 0.07 1.00 35643 18845
## Leadership_z -0.00 0.03 -0.07 0.06 1.00 22254 20736
## Prestige_z 0.04 0.03 -0.02 0.10 1.00 22906 22421
## Dominance_z:Gender 0.01 0.06 -0.10 0.13 1.00 27002 20763
## Gender:Leadership_z -0.08 0.05 -0.19 0.02 1.00 22612 20831
## Gender:Prestige_z -0.01 0.05 -0.12 0.10 1.00 24251 20699
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.70 0.02 0.67 0.74 1.00 33649 19043
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).
m5_gen_hdi <- bayestestR::hdi(m5_gen, effects = "fixed", component = "conditional", ci = .95)
kable(m5_gen_hdi[sign(m5_gen_hdi$CI_low) == sign(m5_gen_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')] , format = "html", booktabs = T, escape = F, longtable = F, digits = 2) %>% kable_styling(full_width = T)
|
Parameter
|
CI
|
CI_low
|
CI_high
|
|
b_Dominance_z
|
0.95
|
0.03
|
0.14
|
Spite and DoPL
m7 <- brm(Spite_z ~ Dominance_z*Gender + Leadership_z*Gender + Prestige_z*Gender, data = analysisDF, warmup = 500, iter = 8000, chains = 4, cores = 6,
prior = c(prior(normal(0, 1), class = "Intercept"),
prior(normal(3, 1), class = "b", coef = "Dominance_z:Gender"),
prior(normal(0, 1), class = 'b', coef = "Gender:Leadership_z"),
prior(normal(0, 1), class = 'b', coef = "Gender:Prestige_z")),
save_all_pars = T)
m7_hdi <- bayestestR::hdi(m7, effects = "fixed", component = "conditional", ci = .95)
## Warning: Identical densities found along different segments of the distribution,
## choosing rightmost.
kable(m7_hdi[sign(m7_hdi$CI_low) == sign(m7_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')] , format = "html", booktabs = T, escape = F, longtable = F, digits = 2) %>% kable_styling(full_width = T)
|
|
Parameter
|
CI
|
CI_low
|
CI_high
|
|
6
|
b_Intercept
|
0.95
|
0.03
|
0.16
|
|
1
|
b_Dominance_z
|
0.95
|
0.45
|
0.58
|
|
3
|
b_Gender
|
0.95
|
-0.40
|
-0.17
|
|
7
|
b_Leadership_z
|
0.95
|
-0.20
|
-0.04
|
|
2
|
b_Dominance_z.Gender
|
0.95
|
0.10
|
0.37
|
|
4
|
b_Gender.Leadership_z
|
0.95
|
-0.34
|
-0.09
|
summary(m7)
## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: Spite_z ~ Dominance_z * Gender + Leadership_z * Gender + Prestige_z * Gender
## Data: analysisDF (Number of observations: 920)
## Samples: 4 chains, each with iter = 8000; warmup = 500; thin = 1;
## total post-warmup samples = 30000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.10 0.03 0.03 0.16 1.00 40628 23120
## Dominance_z 0.51 0.03 0.45 0.58 1.00 28732 23403
## Gender -0.28 0.06 -0.40 -0.17 1.00 35017 23530
## Leadership_z -0.12 0.04 -0.20 -0.04 1.00 21084 23402
## Prestige_z 0.03 0.04 -0.04 0.10 1.00 21678 24050
## Dominance_z:Gender 0.23 0.07 0.10 0.37 1.00 28824 24379
## Gender:Leadership_z -0.22 0.06 -0.35 -0.09 1.00 20611 21689
## Gender:Prestige_z 0.01 0.06 -0.12 0.14 1.00 24132 24383
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.83 0.02 0.79 0.87 1.00 42739 22720
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).
DoPL and Measures
m8 <- brm(mvbind(Dominance, Prestige, Leadership) ~ Spite_z + SSES + SRPS + SJS, data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6,
prior = c(prior(normal(0, 1), class = "Intercept")), save_all_pars = T)
summary(m8)
## Family: MV(gaussian, gaussian, gaussian)
## Links: mu = identity; sigma = identity
## mu = identity; sigma = identity
## mu = identity; sigma = identity
## Formula: Dominance ~ Spite_z + SSES + SRPS + SJS
## Prestige ~ Spite_z + SSES + SRPS + SJS
## Leadership ~ Spite_z + SSES + SRPS + SJS
## Data: analysisDF (Number of observations: 920)
## Samples: 4 chains, each with iter = 8000; warmup = 500; thin = 1;
## total post-warmup samples = 30000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Dominance_Intercept 6.50 1.80 2.96 10.02 1.00 25865
## Prestige_Intercept 3.99 1.73 0.61 7.39 1.00 28795
## Leadership_Intercept 5.60 2.07 1.57 9.67 1.00 25311
## Dominance_Spite_z 2.82 0.16 2.51 3.13 1.00 34724
## Dominance_SSES -0.15 0.02 -0.19 -0.12 1.00 31340
## Dominance_SRPS 1.89 0.48 0.94 2.84 1.00 27582
## Dominance_SJS 0.59 0.07 0.46 0.72 1.00 34688
## Prestige_Spite_z 0.48 0.15 0.19 0.78 1.00 33062
## Prestige_SSES 0.02 0.02 -0.01 0.05 1.00 34530
## Prestige_SRPS 4.17 0.47 3.27 5.09 1.00 31612
## Prestige_SJS 0.08 0.06 -0.05 0.20 1.00 29195
## Leadership_Spite_z 0.13 0.18 -0.22 0.48 1.00 30382
## Leadership_SSES -0.12 0.02 -0.16 -0.08 1.00 29064
## Leadership_SRPS 3.67 0.56 2.57 4.76 1.00 28015
## Leadership_SJS 0.05 0.08 -0.10 0.20 1.00 28129
## Tail_ESS
## Dominance_Intercept 22337
## Prestige_Intercept 23855
## Leadership_Intercept 23208
## Dominance_Spite_z 23523
## Dominance_SSES 24959
## Dominance_SRPS 25109
## Dominance_SJS 24874
## Prestige_Spite_z 24281
## Prestige_SSES 24176
## Prestige_SRPS 24442
## Prestige_SJS 23696
## Leadership_Spite_z 24159
## Leadership_SSES 24482
## Leadership_SRPS 23089
## Leadership_SJS 24565
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma_Dominance 4.58 0.11 4.38 4.80 1.00 36782 24459
## sigma_Prestige 4.39 0.10 4.19 4.60 1.00 33582 22374
## sigma_Leadership 5.30 0.13 5.06 5.55 1.00 34020 22710
##
## Residual Correlations:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## rescor(Dominance,Prestige) 0.16 0.03 0.09 0.22 1.00 31621
## rescor(Dominance,Leadership) 0.32 0.03 0.26 0.38 1.00 31943
## rescor(Prestige,Leadership) 0.34 0.03 0.28 0.40 1.00 36103
## Tail_ESS
## rescor(Dominance,Prestige) 24028
## rescor(Dominance,Leadership) 23826
## rescor(Prestige,Leadership) 24334
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).
m8_hdi <- bayestestR::hdi(m8, effects = "fixed", component = "conditional", ci = .95)
kable(m8_hdi[sign(m8_hdi$CI_low) == sign(m8_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')] , format = "html", booktabs = T, escape = F, longtable = F, digits = 2) %>% kable_styling(full_width = T)
|
|
Parameter
|
CI
|
CI_low
|
CI_high
|
|
1
|
b_Dominance_Intercept
|
0.95
|
3.09
|
10.13
|
|
3
|
b_Dominance_Spite_z
|
0.95
|
2.51
|
3.12
|
|
5
|
b_Dominance_SSES
|
0.95
|
-0.19
|
-0.12
|
|
4
|
b_Dominance_SRPS
|
0.95
|
0.96
|
2.86
|
|
2
|
b_Dominance_SJS
|
0.95
|
0.46
|
0.72
|
|
11
|
b_Prestige_Intercept
|
0.95
|
0.66
|
7.43
|
|
13
|
b_Prestige_Spite_z
|
0.95
|
0.19
|
0.78
|
|
14
|
b_Prestige_SRPS
|
0.95
|
3.30
|
5.12
|
|
6
|
b_Leadership_Intercept
|
0.95
|
1.59
|
9.69
|
|
10
|
b_Leadership_SSES
|
0.95
|
-0.16
|
-0.08
|
|
9
|
b_Leadership_SRPS
|
0.95
|
2.56
|
4.75
|
DoPL and Measures * Gender
m8_gen <- brm(mvbind(Dominance, Prestige, Leadership) ~ Spite_z*Gender + SSES*Gender + SRPS*Gender + SJS*Gender, data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6,
prior = c(prior(normal(0, 1), class = "Intercept")), save_all_pars = T)
m8_gen_hdi <- bayestestR::hdi(m8_gen, effects = "fixed", component = "conditional", ci = .95)
kable(m8_gen_hdi[sign(m8_gen_hdi$CI_low) == sign(m8_gen_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')] , format = "html", booktabs = T, escape = F, longtable = F, digits = 2) %>% kable_styling(full_width = T)
|
|
Parameter
|
CI
|
CI_low
|
CI_high
|
|
5
|
b_Dominance_Intercept
|
0.95
|
5.04
|
14.19
|
|
7
|
b_Dominance_Spite_z
|
0.95
|
2.62
|
3.31
|
|
10
|
b_Dominance_SSES
|
0.95
|
-0.23
|
-0.14
|
|
6
|
b_Dominance_SJS
|
0.95
|
0.59
|
0.89
|
|
4
|
b_Dominance_Gender.SSES
|
0.95
|
0.02
|
0.17
|
|
3
|
b_Dominance_Gender.SRPS
|
0.95
|
0.20
|
4.31
|
|
2
|
b_Dominance_Gender.SJS
|
0.95
|
-0.95
|
-0.33
|
|
25
|
b_Prestige_Intercept
|
0.95
|
2.45
|
11.13
|
|
27
|
b_Prestige_Spite_z
|
0.95
|
0.23
|
0.90
|
|
29
|
b_Prestige_SRPS
|
0.95
|
2.37
|
4.69
|
|
26
|
b_Prestige_SJS
|
0.95
|
0.08
|
0.37
|
|
24
|
b_Prestige_Gender.SSES
|
0.95
|
0.09
|
0.23
|
|
22
|
b_Prestige_Gender.SJS
|
0.95
|
-0.83
|
-0.24
|
|
15
|
b_Leadership_Intercept
|
0.95
|
6.38
|
16.92
|
|
11
|
b_Leadership_Gender
|
0.95
|
-22.33
|
-4.78
|
|
20
|
b_Leadership_SSES
|
0.95
|
-0.23
|
-0.13
|
|
19
|
b_Leadership_SRPS
|
0.95
|
0.76
|
3.60
|
|
14
|
b_Leadership_Gender.SSES
|
0.95
|
0.06
|
0.22
|
|
13
|
b_Leadership_Gender.SRPS
|
0.95
|
0.79
|
5.59
|
summary(m8_gen)
## Family: MV(gaussian, gaussian, gaussian)
## Links: mu = identity; sigma = identity
## mu = identity; sigma = identity
## mu = identity; sigma = identity
## Formula: Dominance ~ Spite_z * Gender + SSES * Gender + SRPS * Gender + SJS * Gender
## Prestige ~ Spite_z * Gender + SSES * Gender + SRPS * Gender + SJS * Gender
## Leadership ~ Spite_z * Gender + SSES * Gender + SRPS * Gender + SJS * Gender
## Data: analysisDF (Number of observations: 920)
## Samples: 4 chains, each with iter = 8000; warmup = 500; thin = 1;
## total post-warmup samples = 30000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Dominance_Intercept 9.46 2.33 4.88 14.06 1.00 17194
## Prestige_Intercept 6.87 2.22 2.48 11.17 1.00 16489
## Leadership_Intercept 11.63 2.70 6.33 16.89 1.00 16875
## Dominance_Spite_z 2.96 0.18 2.61 3.31 1.00 28409
## Dominance_Gender -5.12 3.88 -12.62 2.58 1.00 14702
## Dominance_SSES -0.19 0.02 -0.23 -0.14 1.00 20652
## Dominance_SRPS 0.84 0.62 -0.38 2.06 1.00 18850
## Dominance_SJS 0.74 0.08 0.59 0.90 1.00 25681
## Dominance_Spite_z:Gender 0.19 0.40 -0.59 0.96 1.00 28130
## Dominance_Gender:SSES 0.10 0.04 0.03 0.17 1.00 22063
## Dominance_Gender:SRPS 2.30 1.05 0.22 4.34 1.00 16176
## Dominance_Gender:SJS -0.64 0.16 -0.95 -0.33 1.00 24226
## Prestige_Spite_z 0.57 0.17 0.23 0.91 1.00 29089
## Prestige_Gender -5.94 3.72 -13.30 1.41 1.00 14229
## Prestige_SSES -0.04 0.02 -0.08 0.00 1.00 21998
## Prestige_SRPS 3.48 0.59 2.32 4.64 1.00 17920
## Prestige_SJS 0.22 0.07 0.08 0.37 1.00 24582
## Prestige_Spite_z:Gender -0.00 0.38 -0.75 0.75 1.00 25982
## Prestige_Gender:SSES 0.16 0.04 0.09 0.23 1.00 23564
## Prestige_Gender:SRPS 1.57 1.01 -0.43 3.57 1.00 15464
## Prestige_Gender:SJS -0.53 0.15 -0.82 -0.22 1.00 22610
## Leadership_Spite_z 0.24 0.21 -0.16 0.65 1.00 26199
## Leadership_Gender -13.53 4.50 -22.24 -4.67 1.00 14312
## Leadership_SSES -0.18 0.03 -0.23 -0.13 1.00 21252
## Leadership_SRPS 2.15 0.72 0.73 3.57 1.00 18556
## Leadership_SJS 0.02 0.09 -0.15 0.20 1.00 23764
## Leadership_Spite_z:Gender -0.67 0.46 -1.56 0.23 1.00 25240
## Leadership_Gender:SSES 0.14 0.04 0.06 0.22 1.00 21223
## Leadership_Gender:SRPS 3.16 1.23 0.75 5.55 1.00 15845
## Leadership_Gender:SJS 0.11 0.18 -0.25 0.47 1.00 21910
## Tail_ESS
## Dominance_Intercept 21671
## Prestige_Intercept 20383
## Leadership_Intercept 20028
## Dominance_Spite_z 23402
## Dominance_Gender 19110
## Dominance_SSES 22003
## Dominance_SRPS 22862
## Dominance_SJS 22732
## Dominance_Spite_z:Gender 23099
## Dominance_Gender:SSES 22201
## Dominance_Gender:SRPS 19830
## Dominance_Gender:SJS 23001
## Prestige_Spite_z 23554
## Prestige_Gender 17275
## Prestige_SSES 21963
## Prestige_SRPS 21164
## Prestige_SJS 23474
## Prestige_Spite_z:Gender 22343
## Prestige_Gender:SSES 23334
## Prestige_Gender:SRPS 18747
## Prestige_Gender:SJS 21649
## Leadership_Spite_z 23323
## Leadership_Gender 17234
## Leadership_SSES 22391
## Leadership_SRPS 21811
## Leadership_SJS 23074
## Leadership_Spite_z:Gender 23991
## Leadership_Gender:SSES 22449
## Leadership_Gender:SRPS 18504
## Leadership_Gender:SJS 21897
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma_Dominance 4.51 0.11 4.31 4.72 1.00 34438 23546
## sigma_Prestige 4.33 0.10 4.13 4.53 1.00 33077 23032
## sigma_Leadership 5.25 0.13 5.01 5.51 1.00 36326 24456
##
## Residual Correlations:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## rescor(Dominance,Prestige) 0.13 0.03 0.07 0.19 1.00 34418
## rescor(Dominance,Leadership) 0.32 0.03 0.26 0.37 1.00 33054
## rescor(Prestige,Leadership) 0.33 0.03 0.27 0.39 1.00 33469
## Tail_ESS
## rescor(Dominance,Prestige) 23737
## rescor(Dominance,Leadership) 23268
## rescor(Prestige,Leadership) 23120
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).
DoPL and Measures * Age
m8_Age <- brm(mvbind(Dominance, Prestige, Leadership) ~ Spite_z*Age + SSES*Age + SRPS*Age + SJS*Age, data = analysisDF, iter = 8000, warmup = 500, chains = 4, cores = 6,
prior = c(prior(normal(0, 1), class = "Intercept")), save_all_pars = T)
m8_Age_hdi <- bayestestR::hdi(m8_Age, effects = "fixed", component = "conditional", ci = .95)
kable(m8_Age_hdi[sign(m8_Age_hdi$CI_low) == sign(m8_Age_hdi$CI_high),
c('Parameter', 'CI','CI_low', 'CI_high')] , format = "html", booktabs = T, escape = F, longtable = F, digits = 2) %>% kable_styling(full_width = T)
|
|
Parameter
|
CI
|
CI_low
|
CI_high
|
|
5
|
b_Dominance_Intercept
|
0.95
|
20.07
|
44.23
|
|
1
|
b_Dominance_Age
|
0.95
|
-1.25
|
-0.39
|
|
10
|
b_Dominance_SSES
|
0.95
|
-0.38
|
-0.18
|
|
9
|
b_Dominance_SRPS
|
0.95
|
-7.54
|
-0.94
|
|
8
|
b_Dominance_Spite_z.Age
|
0.95
|
0.02
|
0.15
|
|
4
|
b_Dominance_Age.SSES
|
0.95
|
0.00
|
0.01
|
|
3
|
b_Dominance_Age.SRPS
|
0.95
|
0.07
|
0.30
|
|
25
|
b_Prestige_Intercept
|
0.95
|
-41.76
|
-19.34
|
|
21
|
b_Prestige_Age
|
0.95
|
1.01
|
1.81
|
|
30
|
b_Prestige_SSES
|
0.95
|
0.29
|
0.47
|
|
29
|
b_Prestige_SRPS
|
0.95
|
9.00
|
15.10
|
|
26
|
b_Prestige_SJS
|
0.95
|
0.29
|
1.21
|
|
24
|
b_Prestige_Age.SSES
|
0.95
|
-0.02
|
-0.01
|
|
23
|
b_Prestige_Age.SRPS
|
0.95
|
-0.44
|
-0.23
|
|
22
|
b_Prestige_Age.SJS
|
0.95
|
-0.05
|
-0.01
|
|
19
|
b_Leadership_SRPS
|
0.95
|
3.45
|
11.20
|
|
14
|
b_Leadership_Age.SSES
|
0.95
|
-0.01
|
0.00
|
|
13
|
b_Leadership_Age.SRPS
|
0.95
|
-0.27
|
0.00
|
summary(m8_Age)
## Family: MV(gaussian, gaussian, gaussian)
## Links: mu = identity; sigma = identity
## mu = identity; sigma = identity
## mu = identity; sigma = identity
## Formula: Dominance ~ Spite_z * Age + SSES * Age + SRPS * Age + SJS * Age
## Prestige ~ Spite_z * Age + SSES * Age + SRPS * Age + SJS * Age
## Leadership ~ Spite_z * Age + SSES * Age + SRPS * Age + SJS * Age
## Data: analysisDF (Number of observations: 920)
## Samples: 4 chains, each with iter = 8000; warmup = 500; thin = 1;
## total post-warmup samples = 30000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Dominance_Intercept 31.97 6.20 19.84 44.06 1.00 15301
## Prestige_Intercept -30.63 5.71 -41.80 -19.37 1.00 15760
## Leadership_Intercept -4.57 7.23 -18.71 9.61 1.00 14490
## Dominance_Spite_z 0.56 0.74 -0.91 2.00 1.00 28513
## Dominance_Age -0.82 0.22 -1.25 -0.39 1.00 14399
## Dominance_SSES -0.28 0.05 -0.38 -0.18 1.00 36329
## Dominance_SRPS -4.11 1.69 -7.44 -0.82 1.00 16074
## Dominance_SJS 0.34 0.26 -0.16 0.84 1.00 17489
## Dominance_Spite_z:Age 0.09 0.03 0.03 0.15 1.00 27632
## Dominance_Age:SSES 0.00 0.00 0.00 0.01 1.00 34730
## Dominance_Age:SRPS 0.18 0.06 0.07 0.29 1.00 14929
## Dominance_Age:SJS 0.01 0.01 -0.01 0.03 1.00 16613
## Prestige_Spite_z -0.98 0.68 -2.31 0.34 1.00 25959
## Prestige_Age 1.42 0.20 1.02 1.82 1.00 15079
## Prestige_SSES 0.38 0.05 0.28 0.46 1.00 34674
## Prestige_SRPS 12.15 1.56 9.09 15.20 1.00 16519
## Prestige_SJS 0.75 0.24 0.29 1.21 1.00 19303
## Prestige_Spite_z:Age 0.06 0.03 0.00 0.11 1.00 25342
## Prestige_Age:SSES -0.01 0.00 -0.02 -0.01 1.00 33898
## Prestige_Age:SRPS -0.33 0.05 -0.43 -0.23 1.00 15491
## Prestige_Age:SJS -0.03 0.01 -0.05 -0.01 1.00 18476
## Leadership_Spite_z 1.43 0.86 -0.25 3.10 1.00 27642
## Leadership_Age 0.41 0.26 -0.10 0.91 1.00 13685
## Leadership_SSES 0.01 0.06 -0.10 0.13 1.00 34463
## Leadership_SRPS 7.17 1.97 3.31 11.07 1.00 15257
## Leadership_SJS -0.10 0.30 -0.70 0.48 1.00 17756
## Leadership_Spite_z:Age -0.06 0.04 -0.13 0.01 1.00 27077
## Leadership_Age:SSES -0.01 0.00 -0.01 -0.00 1.00 33618
## Leadership_Age:SRPS -0.13 0.07 -0.27 -0.00 1.00 14116
## Leadership_Age:SJS 0.00 0.01 -0.02 0.03 1.00 16977
## Tail_ESS
## Dominance_Intercept 19449
## Prestige_Intercept 19327
## Leadership_Intercept 19917
## Dominance_Spite_z 22951
## Dominance_Age 18379
## Dominance_SSES 24743
## Dominance_SRPS 20687
## Dominance_SJS 20929
## Dominance_Spite_z:Age 22658
## Dominance_Age:SSES 24299
## Dominance_Age:SRPS 19666
## Dominance_Age:SJS 20914
## Prestige_Spite_z 21626
## Prestige_Age 17977
## Prestige_SSES 24070
## Prestige_SRPS 20276
## Prestige_SJS 22312
## Prestige_Spite_z:Age 22950
## Prestige_Age:SSES 24837
## Prestige_Age:SRPS 19495
## Prestige_Age:SJS 21552
## Leadership_Spite_z 22504
## Leadership_Age 18924
## Leadership_SSES 24934
## Leadership_SRPS 20887
## Leadership_SJS 20721
## Leadership_Spite_z:Age 21978
## Leadership_Age:SSES 25409
## Leadership_Age:SRPS 18886
## Leadership_Age:SJS 20252
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma_Dominance 4.46 0.10 4.26 4.67 1.00 49917 22042
## sigma_Prestige 4.09 0.10 3.91 4.29 1.00 50277 22139
## sigma_Leadership 5.19 0.12 4.96 5.44 1.00 46032 23326
##
## Residual Correlations:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## rescor(Dominance,Prestige) 0.17 0.03 0.11 0.24 1.00 48127
## rescor(Dominance,Leadership) 0.32 0.03 0.27 0.38 1.00 43176
## rescor(Prestige,Leadership) 0.32 0.03 0.26 0.37 1.00 47274
## Tail_ESS
## rescor(Dominance,Prestige) 23043
## rescor(Dominance,Leadership) 23593
## rescor(Prestige,Leadership) 23019
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).
Justification and Content type
ggplot(analysisDF, aes(y = Justification, x = Vignette, fill = Content, group = Vignette)) +
geom_violin() +
scale_shape_prism() +
scale_y_continuous(guide = "prism_offset") +
scale_x_discrete(guide = "prism_offset")
