data-raw/play_with_data.R

In vanichols/CyChecks: Obtains and Analyzes Iowa State University Salary Data

data("all_sals")
data("sals18")
names(sals18)

library(tidyverse)
library(broom)

dat <- as_tibble(sals18)

dat %>%
  filter(department == "AGRONOMY") %>%
  filter(grepl('PROF', position)) %>%
  ggplot(aes(gender, total_salary_paid, color = position, group = position)) +
  geom_jitter(size = 2, width = 0.2, alpha = 0.5) +
  stat_summary(fun.y = mean, geom = "line") +
  stat_summary(fun.y = mean, geom = "point", size = 3) +
  facet_grid(~department)

# Calculate p-value of gender
dat %>%
  #filter(!is.na(department)) %>%
  filter(department == "AGRONOMY") %>%
  filter(grepl('PROF', position)) %>%
  filter(position %in% c("PROF", "ASSOC PROF", "ADJ ASST PROF")) %>%
  group_by(department) %>%
  nest() %>%
  mutate(mod = data %>% map(~lm(total_salary_paid ~ gender*position, data = .)),
         res = mod %>% map(~anova(.)),
         tidy = res %>% map(~tidy(.))) %>%
  unnest(tidy)




dat %>%
  #filter(department == "AGRONOMY") %>%
  filter(grepl('PROF', position),
         gender %in% c('M', 'F')) %>%
  group_by(organization) %>%
  mutate(n = n()) %>%
  filter(n > 2) %>%
  ggplot(aes(gender, total_salary_paid)) +
  geom_jitter(aes(color = position), size = 3, width = 0.2) +
  stat_summary(fun.y = mean, geom = "point") +
  guides(color = F) +
  facet_wrap(~organization)


dat %>%
  #filter(department == "AGRONOMY") %>%
  filter(grepl('ASST PROF', position),
         gender %in% c('M', 'F')) %>%
  group_by(organization) %>%
  mutate(n = n()) %>%
  filter(n > 2) %>%
  ggplot(aes(gender, total_salary_paid)) +
  geom_jitter(aes(color = position), size = 3, width = 0.2) +
  stat_summary(fun.y = mean, geom = "point") +
#  guides(color = F) +
  facet_wrap(~organization)

### models

# numbers of year working
yrs <- all_sals %>%
    group_by(name)%>%
    summarize(years_worked = n())
unos <- all_sals %>%
  filter(gender == "M" | gender == "F") %>%
  group_by(gender, position) %>%
  summarize(n = n())%>%
  arrange(n)%>%
  filter(n == 1)
all_pos <- all_sals %>%
  filter(!(position %in% unos$position)) %>%
  filter(gender == "M" | gender == "F")%>%
  filter(!(is.na(position)))%>%
  filter(position != "**") %>%
  mutate_if(is.factor, fct_explicit_na, na_level = "missing")%>%
  group_by(department, organization) %>%
  nest()

nos <- c("AG/BIOSYS ENG", "FOOD SC/HN-AGLS", "MANAGEMENT", "AG/BIOSYS ENG-E", "VET PATHOLOGY",
         "IT SERVICES CIO", "VP RESEARCH", "VP RESEARCH")

profs <- all_sals %>%
  filter(position == "PROF") %>%
  filter(gender != "*")%>%
  mutate(gender = fct_drop(gender, only = "*"))%>%
  filter(!(department %in% nos)) %>%
  mutate_if(is.factor, fct_explicit_na, na_level = "missing")%>%
  group_by(department, organization) %>%
  nest()


pos_model <- function(df){
  t.test(total_salary_paid ~ gender, data = df)
}
models <- map(profs$data, pos_model)