stem_analysis.R

library(readxl)
library(dplyr)
library(tidyr)

data <- read_excel("STEM Employment data & Salary.xlsx", sheet = "Final")

years <- unique(data$Year)

for (yr in years) {
  cat("\n====================\n")
  cat("Year:", yr, "\n")
  
  # Filter data for the current year
  year_data <- filter(data, Year == yr)
  
  # Chi-Square Test: STEM Background vs Gender (Total Employed Graduates)
  chi_table <- year_data %>%
    filter(!is.na(`Employed Graduates`)) %>%
    group_by(Sex, STEM) %>%
    summarise(total_grads = sum(`Employed Graduates`, na.rm = TRUE)) %>%
    pivot_wider(names_from = STEM, values_from = total_grads, values_fill = 0)
  
  if (nrow(chi_table) > 1) {
    chi_matrix <- as.matrix(chi_table[, -1])
    rownames(chi_matrix) <- chi_table$Sex
    
    cat("\nChi-Square Test (Gender vs STEM Graduates):\n")
    print(chisq.test(chi_matrix))
  } else {
    cat("Not enough data for chi-square test in year", yr, "\n")
  }
  
  # t-Test: Median Monthly Salaries by Gender
  stem_data <- filter(year_data, STEM == 1, !is.na(`Median monthly salaries & wages of employees (RM)`))
  
  if (n_distinct(stem_data$Sex) == 2) {
    cat("\nt-Test (Median Salary by Gender in STEM):\n")
    print(t.test(`Median monthly salaries & wages of employees (RM)` ~ Sex, data = stem_data))
  } else {
    cat("Not enough gender diversity in STEM data for t-test in year", yr, "\n")
  }
}