Salary by Highest attainment and Field of Study

Code

salary_plot <- function(tbbl, Highest){
  plt <- tbbl%>%
    ggplot(aes(x = weighted_mean, 
               y = fct_reorder(CIP, weighted_mean),
               text= paste0("Field of Study: ",
                            CIP,
                            "\n Weighted average median salary = ",
                            scales::dollar(weighted_mean),
                            "\n Weighted standard deviation median salary = ",
                            scales::dollar(weighted_sd)
               )
    )) +
    geom_point() +
    geom_errorbar(aes(xmin = lower, xmax = upper), colour="grey")+
    theme_minimal()+
    theme(axis.text.y = element_blank())+
    scale_x_continuous(labels=scales::dollar)+
    labs(x="Weighted average median salary",
         y="Field of Study",
         title=paste("Weighted average median salary, +/- one standard deviation for", Highest))
  
  plotly::ggplotly(plt, tooltip="text")
}

#get data------------------------------
edu_noc <- vroom(here("out","edu_noc.csv"))|>
  mutate(NOC_Description=str_sub(NOC, start=7),
         NOC=str_sub(NOC, 1,5))

median_salary <- read_excel(here("data","median_salary.xlsx"))|>
  select(NOC, median_salary=contains("calculated"))

joined <- left_join(edu_noc, median_salary)|>
  group_by(Education)|>
  summarise(weighted_mean=weighted.mean(median_salary, w=value, na.rm=TRUE),
            weighted_sd=sqrt(Hmisc::wtd.var(median_salary, weights=value)),
            lower=weighted_mean-weighted_sd,
            upper=weighted_mean+weighted_sd
            )|>
  separate(Education, into = c("CIP", "Highest"), sep=": ")|>
  group_by(Highest)|>
  nest()|>
  mutate(salary_plot=map2(data, Highest, salary_plot))

Here we join the LMO median salary data to the CIP-NOC table (98-10-0403-01), and calculate the weighted average and standard deviation of the median salary: i.e. for a given education, use the NOC counts as weights when calculating the mean and standard deviation.

Then for each level of highest attainment, we plot the weighted average median salary, and error bars for +/- one standard deviation (to indicate dispersion).

Code

joined$salary_plot[[1]]

Code

joined$salary_plot[[2]]

Code

joined$salary_plot[[3]]

Code

joined$salary_plot[[4]]

Code

joined$salary_plot[[5]]

Code

joined$salary_plot[[6]]

Code

joined$salary_plot[[7]]

Code

joined$salary_plot[[8]]