---
title: "ay23-24"
output: html_document
date: "2024-03-12"
editor_options: 
  chunk_output_type: console
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

```{r}
library(tidyverse)
primary <- c("#D1495B", "#EDAE49", "#00798C","#003D5B", "#30638E","#56B4E9", "#7F67B6", "#A4A8D1", "#380606")
```


#cleaning
```{r}
survey <- read.csv("Survey about job applicants, AY24-25.csv")

survey%>%
  select(-1:-6, -8, -10:-17, -X)%>%
  mutate(Q1 = str_remove_all(Q1, "~"),
         Q1 = str_remove_all(Q1, "\\+"),
         Q1 = as.numeric(Q1),
         Q2 = str_replace_all(Q2, "Science, Logic, and Mathematics", "SLM"),
         Q3 = str_replace_all(Q3, "^(?=\\s*$)", "not given"),
         Q2 = str_replace_all(Q2, "^(?=\\s*$)", "not given"),
         Q3 = factor(Q3, levels = c("Bachelors", "Masters", "Doctorate")))%>%
    filter(!is.na(Q1))%>%
  separate_wider_delim(cols = Q2, delim = ",", names_sep = "_sep", too_few = "align_start")%>%
  select(-6, -7)->survey_clean
```

#intro numbers
```{r}
survey_clean|>
  summary()
  
survey_clean|>
  arrange(Q2_sep1, Q1)|>
  view()
```

#boxplots
```{r}
survey_clean%>%
  ggplot(aes(x = Finished, y = Q1))+
  geom_boxplot()

quantile(survey_clean$Q1)

survey_clean%>%
  ggplot(aes(x = Q3, y = Q1))+
  geom_boxplot()

survey_clean%>%
  ggplot(aes(x = Q2_sep1, y = Q1))+
  geom_boxplot()
```

#overviews
```{r}
#overview: counts
survey_clean|>
  count(Q3)|>
  na.omit()|>
  rename(`Degree type` = Q3,
         Count = n)|>
  arrange(`Degree type`)|>
  gt::gt(caption = "Responses by program terminal degree")

survey_clean|>
  count(Q2_sep1)|>
  view()

survey_clean%>%
  arrange(Q3, Q1)%>%
  view()
```

#details--plot titles indicate what's being tracked
```{r}
survey_clean|>
  select(Q1, Q3)|>
  filter(Q3 != "not given")|>
  na.omit()|>
  arrange(Q3, Q1)|>
  group_by(Q3)|>
  mutate(rn = row_number(),
         med = median(Q1),
         avg = round(mean(Q1), 1))|>
  ungroup()|>
  ggplot(aes(rn, Q1))+
  geom_bar(stat = "identity")+
  geom_hline(aes(yintercept = med, color = factor(med)), linewidth = 4)+
  #geom_line()+
  #geom_smooth(linewidth = 2, se = FALSE)+
  facet_wrap(~Q3, scales = "free_x")+
  theme_minimal(base_size = 60)+
  ggtitle("Volume of applications by program status")+
  theme(axis.text.x = element_blank())+
  labs(color = "Median")+
  xlab("")+
  ylab("counts")+
  scale_color_manual(values = primary)->ay24_vol_status

ggsave(plot = vol_status_24, device = "png", filename = "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\ay24_vol_status.png", width = 40, height = 30, units = "in", dpi = 300, limitsize = FALSE, bg = "white")


survey_clean|>
  select(Q1, Q3)|>
  filter(Q3 != "not given")|>
  na.omit()|>
  arrange(Q3, Q1)|>
  group_by(Q3)|>
  mutate(rn = row_number(),
         med = median(Q1),
         avg = round(mean(Q1), 1),
         diff = avg - Q1)|>
  ungroup()|>
  ggplot(aes(rn, diff)) +
  geom_bar(stat = "identity") +
  facet_wrap(~Q3, scales = "free_x")

#contract type
survey_clean|>
  select(Q1, Q3, Q4, Q2_sep1)|>
  mutate(Q4 = str_replace_all(Q4, "^(?=\\s*$)", "x"))%>%
  filter(Q4 != "x" & Q4 != "Associate professor/tenured" & Q4 != "Fixed-term with no option for renewal (e.g. sabbatical replacement)")%>%
  na.omit()|>
  arrange(Q4, Q1)|>
  group_by(Q4)|>
  mutate(rn = row_number(),
         med = median(Q1),
         avg = round(mean(Q1), 1),
         diff = med - Q1,
         std = sd(Q1),
         upper = round(std, 1),
         lower = round(-1*std, 1),
         dist = max(Q1) - min(Q1),
         Q4 = str_remove_all(Q4, "\\(e.g. rolling contract\\)"))%>%
  ungroup()%>%
  ggplot(aes(rn, Q1, fill = Q3)) +
  geom_bar(stat = "identity", position = "dodge") +
  facet_wrap(~Q4, scales = "free_x")+
  theme_minimal(base_size = 60)+
  ggtitle("Volume of applications by contract type and program status")+
  theme(axis.text.x = element_blank())+
  labs(fill = "Program\nstatus")+
  xlab("")+
  ylab("counts")+
  scale_fill_manual(values = primary)->ay24_vol_contract_program

  ggsave(plot = ay24_vol_contract_program, device = "png", filename = "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\vol_contract_program_24_a.png", width = 40, height = 30, units = "in", dpi = 300, limitsize = FALSE, bg = "white")
  

survey_clean|>
  select(Q1, Q2_sep1, Q3)|>
  filter(Q2_sep1 != "not given")%>%
  na.omit()|>
  arrange(-Q1)|>
  group_by(Q2_sep1, Q3)|>
  summarise(avg = median(Q1), # change to median?
            max = max(Q1),
            min = min(Q1))|>
  mutate(Q2_sep1 = as_factor(Q2_sep1),
         aos_avg = median(avg))|>
  ungroup()|>
  ggplot(aes(Q3, avg, color = Q2_sep1))+
  geom_point(alpha = .8, size = 8)+
    geom_errorbar(aes(ymin = min, ymax = max),alpha = .7, linewidth = 5)+
  geom_hline(aes(yintercept = aos_avg), linetype = "solid", alpha = .2, linewidth = 5)+
  coord_flip()+
  facet_wrap(~ Q2_sep1)+
  theme_minimal(base_size = 60)+
  xlab("AOS")+
  ylab("counts")+
  ggtitle("Volume of applications by AOS and program status")+
  theme(legend.position="none")+
  scale_color_manual(values = primary)->ay24_aos_max_min_avg

ggsave(plot = ay24_aos_max_min_avg, device = "png", filename = "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\ay_24_aos_program_max_min_avg_c.png", width = 40, height = 30, units = "in", dpi = 300, limitsize = FALSE, bg = "white")


survey_clean|>
  select(Q1, Q2_sep1, Q3, Q4)|>
  mutate(Q4 = str_replace_all(Q4, "^(?=\\s*$)", "x"),
         Q2_sep1 = str_replace_all(Q2_sep1, "Metaphysics and Epistemology", "Metaphysics, Epistemology"))%>%
  filter(Q4 != "x" & Q4 != "Associate professor/tenured" & Q4 != "Fixed-term with no option for renewal (e.g. sabbatical replacement)")%>%
  filter(Q2_sep1 != "not given")%>%
  na.omit()|>
  arrange(-Q1)|>
  group_by(Q2_sep1)|>
  mutate(aos_med = median(Q1))|>
  ungroup()%>%
  group_by(Q2_sep1, Q4)|>
  mutate(spec_med =  median(Q1), # change to median?
            max = max(Q1),
            min = min(Q1),
         Q4 = str_remove_all(Q4, "\\(e.g. rolling contract\\)"),
         Q4 = as_factor(Q4))|>
   ungroup()|>
  ggplot(aes(Q4, spec_med, color = Q4))+
  geom_point(alpha = .8, size = 8)+
    geom_errorbar(aes(ymin = min, ymax = max),alpha = .7, linewidth = 5)+
  geom_hline(aes(yintercept = aos_med), linetype = "solid", alpha = .2, linewidth = 5)+
  coord_flip()+
  facet_wrap(~ Q2_sep1)+
  theme_minimal(base_size = 60)+
  xlab("AOS")+
  ylab("counts")+
  ggtitle("Volume of applications by AOS and contract type")+
  theme(legend.position="none")+
  scale_color_manual(values = primary)->ay24_aos_contract_max_min_avg

ggsave(plot = ay24_aos_contract_max_min_avg, device = "png", filename = "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\ay_24_aos_contract_max_min_avg_d.png", width = 40, height = 30, units = "in", dpi = 300, limitsize = FALSE, bg = "white")


survey_clean|>
  filter(!is.na(Q7))|>
  arrange(Q1)|>
  mutate(rn = row_number())|>
  rename(counts = Q1)|>
  mutate(Q7 = str_replace_all(Q7, "There were about as many applicants as I was expecting", "Expected"),
         Q7 = str_replace_all(Q7, "There were fewer applicants than I was expecting", "Fewer"),
         Q7 = str_replace_all(Q7, "There were more applicants than I was expecting", "More"))|>
  ggplot(aes(rn, counts, fill = Q7))+
  geom_bar(stat = "identity")+
  scale_fill_manual(values = primary)+
  facet_wrap(~Q3)
  theme_minimal(base_size = 60)+
  ggtitle("Volume of applications for TT jobs")+
  theme(axis.text.x = element_blank())+
  xlab("")+
  ylab("counts")->num_imp_23

ggsave(plot = num_imp_23, device = "png", filename = "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\num_imp_23.png", width = 40, height = 30, units = "in", dpi = 300, limitsize = FALSE, bg = "white")
```

#tables: not used in AY24-25
```{r}
survey_clean|>
  select(Q1, Q2_sep1, Q3)|>
  na.omit()|>
  group_by(Q3)|>
  reframe(std = sd(Q1), average = mean(Q1))|>
  ungroup()|>
  rename(status = Q3,
         `standard deviation` = std)|>
  mutate(across(where(is.double), round))|>
  gt::gt(caption = "Averages and dispersion") -> overview_p23

gt::gtsave(overview_p23, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\overview_p23.png", expand = 10)

survey_clean|>
  select(Q1, Q3)|>
  na.omit()|>
  arrange(-Q1)|>view()
  group_by(Q3)|>
 summarize(least = min(Q1), avg = mean(Q1), mid = median(Q1),most = max(Q1) )|>
  ungroup()|>
  mutate(across(where(is.double), round))|>
  rename(status = Q3,
         mean = avg,
         median = mid)|>
  gt::gt(caption = "Overview of applicantions by program status")->all_over_23

gt::gtsave(all_over_23, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\all_over_23.png", expand = 10)

survey_clean|>
  filter(Q4 == "Assistant professor/tenure-track")|>
  group_by(Q2_sep1)|>
  summarize(least = min(Q1), avg = mean(Q1), mid = median(Q1), most = max(Q1))|>
  ungroup()|>
  mutate(across(where(is.double), round))|>
  rename(AOS = Q2_sep1,
         mean = avg,
         median = mid)|>
  gt::gt(caption  = "TT position applications")->tt_23

gt::gtsave(tt_23, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\tt_23.png", expand = 10)

survey_clean|>
  filter(Q4 == "Postdoc" | str_detect(Q4, "Fixed"))|>
  group_by(Q2_sep1)|>
  summarize(least = min(Q1), avg = mean(Q1), mid = median(Q1), most = max(Q1))|>
  ungroup()|>
  mutate(across(where(is.double), round))|>
  rename(AOS = Q2_sep1,
         mean = avg,
         median = mid)|>
  gt::gt(caption  = "Postdoc and fixed-term position applications")->pd_fixed_23

gt::gtsave(pd_fixed_23, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\pd_fixed_23.png", expand = 10)

survey_clean|>
  filter(Q4 != "Postdoc" & Q4 != "Assistant professor/tenure-track" & str_detect(Q4, "Fixed", negate = TRUE))|>
  group_by(Q2_sep1)|>
  summarize(least = min(Q1), avg = mean(Q1), mid = median(Q1), most = max(Q1))|>
  ungroup()|>
  mutate(across(where(is.double), round))|>
  rename(AOS = Q2_sep1,
         mean = avg,
         median = mid)|>
  gt::gt(caption  = "Open and tenured position applications")->open_23

gt::gtsave(open_23, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\open_23.png", expand = 10)

  
  survey_clean|>
  filter(!is.na(Q7))|>
  count(Q7)|>
  rename(response = Q7,
         count = n)|>
  arrange(-count)|>
  gt::gt(caption = "Impressions of applicant volume")->imp_app_23
  
  gt::gtsave(imp_app_23, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\imp_app_23.png", expand = 10)
  
  survey_clean|>
  filter(!is.na(Q6))|>
  count(Q6, Q3)|>view()
  rename(response = Q6,
         count = n)|>
  arrange(-count)|>view()
  gt::gt(caption = "Time allotment for job")->time_job_23
  
  gt::gtsave(time_job_23, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\time_job_23.png", expand = 10)


```
#perceptions  
```{r}
survey_clean%>%
  select(Q3, Q4, Q7)%>%
  arrange(Q3, Q4, Q7)%>%
  view()
```


#top and bottom
```{r}
survey_clean|>
 select(Q1, Q2_sep1, Q3)%>%
  na.omit()%>%
  arrange(Q1)%>%
  add_count(Q2_sep1, Q3)%>%
  mutate(ends = ifelse(Q1 > 65 & Q1 < 159, FALSE, TRUE),
         end = ifelse(ends == TRUE & Q1<= 65, "Bottom", "Top"))%>%
  rename(`overall frequency` = n)%>%
  filter(ends == TRUE)%>%
  add_count(Q2_sep1, Q3, end)%>%
  mutate(ratio = round(n / `overall frequency`, 2))%>%
  
  rename(frequency = n,
         AOS = Q2_sep1,
         `degree type` = Q3)%>%
  distinct(AOS, `degree type`, frequency, end, .keep_all = TRUE)%>%
  select(AOS, `degree type`, frequency, `overall frequency`, ratio, end)%>%
  arrange(end, ratio, AOS, `degree type`)%>%
  gt::gt(caption = "Chararteristics of top and bottom quartile for applicant volume")->top_bottom

 gt::gtsave(top_bottom, "C:\\Users\\lassiter\\OneDrive - Gonzaga University\\R data\\PhilJobs_annual_outputs\\ay24_top_bottom_a.png", expand = 10)
 
 survey_clean|>
 select(Q1, Q2_sep1, Q3)%>%
  na.omit()%>%
  arrange(Q1)%>%
  add_count(Q2_sep1, Q3)%>%
  mutate(ends = ifelse(Q1 > 65 & Q1 < 159, FALSE, TRUE),
         end = ifelse(ends == TRUE & Q1<= 65, "Bottom", "Top"))%>%
   filter(end == "Top")%>%
   select(2,3)%>%
   fastDummies::dummy_cols()%>%
   select(-1, -2)%>%
   cor()-> bottom_cor
 
 corrplot::cor.mtest(bottom_cor, conf.level = .95) -> bottom_cor_sig
 
 corrplot::corrplot(bottom_cor, method = "number")
   
 
 #method = 'circle', type = 'lower', insig='blank',
  #       addCoef.col ='black', number.cex = 0.8, order = 'AOE', diag=FALSE
```

```{r}
DN|>
  count(Q3, Q2_sep1)|>
  group_by(Q3)|>
  mutate(tots = n / sum(n))|>
  ungroup()|>
  
  view()


DN|>
  select(Q2_sep1, Q3, Q1)|>
  filter(Q2_sep1 == "Open")|>
  select(-Q1)|>
  table()->AOS_status

(prop.table(AOS_status))
addmargins(AOS_status)
fisher.test(AOS_status)
chisq.test(AOS_status)

table(survey_clean$Q2_sep1)

```