descriptive_analyses.Rmd

---
title: "Descriptives"
author: "Pooya Razavi"
date: "last knitted: `r Sys.time()`"
output: 
  html_document:
    theme: cosmo
    highlight: textmate
    toc: TRUE
    toc_float: TRUE
editor_options: 
  chunk_output_type: console
---

This script provides a summary of the descriptive characteristics of the participants (i.e., demographics) and the event (e.g., cause of anger, ease of remembering).

```{r setup, include=FALSE, warning=FALSE}
#load libraries
package_list <- c("dplyr", "tidyr", "ggplot2")
lapply(package_list, require, character.only = TRUE)

#read in the data
df <- readxl::read_xlsx("C:/Users/pooya/Dropbox (University of Oregon)/Anger Dissertation/Prototype study analysis/ProcessedData_F21_W22_S22_F22.xlsx")

#the percentage function
  percentage <- function(var, includeNA = TRUE, arran = TRUE) {
      tabb <- table(var) %>% as.data.frame()
    
          if (includeNA == TRUE) {
               tabb$percentage <- (tabb$Freq * 100 / length(var))
          } else {
                tabb$percentage <- (tabb$Freq * 100 / sum(tabb$Freq))
          }
      
      colnames(tabb)[1] <- c("category")
      
          if (arran == FALSE) {
                knitr::kable(tabb)
            } else {
                tabb %>% arrange(desc(percentage)) %>% knitr::kable()
            }
}
knitr::opts_chunk$set(echo = TRUE)
```


```{r, factor-setup}
# gender
  df$gender <- factor(df$gender)

    levels(df$gender) <- c("female", "male" ,"nonbinary", "self-describe", "prefer not to say", NA)
# race/ethnicity
  df$ethnicity <- factor(df$ethnicity)
    
    levels(df$ethnicity) <- c(
        "American Indian or Alaska Native",
        "Asian",
        "Black or African American",
        "Hispanic, Latinx or Spanish Origin",
        "Middle Eastern or North African",
        "Native Hawaiian or Other Pacific Islander",
        "White",
        "Some other ethnicity or origin, please specify:",
        "I prefer not to answer.",
        "Multiracial",
        NA
    )
# person/entity who caused anger
    df$who_caused <- factor(df$who_caused)
    
    levels(df$who_caused) <- c(
        "romantic partner",
        "parent",
        "sibling",
        "child",
        "close friend",
        "coworker",
        "classmate",
        "acquintance",
        "teacher",
        "boss",
        "stranger",
        "public figure",
        "corporation/organization",
        "other",
        NA
    )
  #we can search for the word "roommate" for people who reported "other" as cause
  df <- df %>% 
            mutate(who_caused2 = if_else(who_caused == "other" & grepl("roommate", right_narrative), "roommate",
                                         if_else(who_caused == "other" & grepl("roommate", nonright_narrative), "roommate",
                                                 as.character(who_caused)))
                   )
    
  
```


```{r, data-exclusion}
#assigning values to factor levels
  df$NarrativeWritten <- as.factor(df$NarrativeWritten)
  df$NarrativeRelevant <- as.factor(df$NarrativeRelevant)
  df$Condition <- as.factor(df$Condition)
  
  levels(df$NarrativeWritten) <- c("No", "Yes")
  levels(df$NarrativeRelevant) <- c("No", "Yes", NA, NA) 
  levels(df$Condition) <- c("justified", "nonjustified", NA)

#Number of participants
  table(df$NarrativeRelevant)
  
#drop cases following preregistration
  df1 <- df %>% 
    filter(NarrativeWritten != "No") %>% 
    filter(NarrativeRelevant != "No") %>% 
    filter(!is.na(Condition))

```


# Demographics

```{r}
# age
  psych::describe(as.numeric(df$age))
  psych::describe(as.numeric(df1$age))

# gender
    table(df1$gender, useNA = "ifany")
    percentage(df1$gender, includeNA = FALSE)
    #participants who self-described:
    table(df1$gender_other)

# race-ethnicity
    table(df1$ethnicity, useNA = "ifany")
    percentage(df1$ethnicity, includeNA = FALSE)
    #participants who self-described:
    table(df1$ethnicity_other)
```

# Person/Entity who Caused Anger

```{r}
#overal
    table(df1$who_caused2, useNA = "ifany")
    percentage(df1$who_caused2, includeNA = FALSE) 

#for each condition
  cause_summary <- df1 %>% 
                      group_by(Condition, who_caused2) %>% 
                      summarise(n = n()) %>% 
                      pivot_wider(names_from = Condition,
                                  values_from = n) %>% 
                      mutate(justified_perc = round(justified * 100 / sum(justified), 2),
                             unjustified_perc = round(nonjustified * 100 / sum(nonjustified, na.rm = TRUE), 2),
                             overl_freq = justified + nonjustified,
                             overall_perc = round(overl_freq * 100 / sum(overl_freq, na.rm = TRUE), 2)) %>% 
                      arrange(desc(overl_freq))

  cause_summary %>% knitr::kable()    

#write.csv(cause_summary, "who_caused_anger_table.csv")

```

# Ease of remembering

```{r}
#descriptives
psych::describe(df1$ease_remember)
psych::describeBy(df1$ease_remember, group = df1$Condition)

#difference between the two conditions
  t.test(ease_remember ~ Condition, data = df1)
  effectsize::cohens_d(ease_remember ~ Condition, data = df1)

  
  df1 %>% 
    ggplot(aes(x = ease_remember, colour = Condition, fill = Condition)) +
    geom_histogram(alpha=0.4, position="dodge", binwidth = .4) +
    labs(x = "How easy was it for you to remember this incident?") +
    theme_minimal()
    

```