Code (from history):

read.csv(“USMA_Progression.csv”) df<- read.csv(“USMA_Progression.csv”) df<- read.csv(‘USMA_Progression.csv’) read.csv(“/home/downloads/r stuff/USMA_Progression.csv”) setwd(“/Users/charlieburgess/Downloads/r stuff”) setwd(“/home/downloads/r stuff”) data <- read.csv(“USMA_Progression.csv”) View(data) data <- read.csv(USMA_data) data <- read.csv(USMA_data.csv) View(data) install.packages(“dplyr”) library(dplyr) USMAdata %-% head(10) USMA_data.csv <- transform(USMA_data.csv, femalespeers_recreated = ave(female, year, class, company_n, FUN = sum)) USMA_data <- read.csv(“USMA_data.csv”) USMA_data <- transform(USMA_data, femalespeers_recreated = ave(female, year, class, company_n, FUN = sum)) USMA_data <- transform(USMA_data, malespeers_recreated = ave(male, year, class, company_n, FUN = sum)) USMA_data <- transform(USMA_data, totpeople_recreated = ave(female + !female, year, class, company_n, FUN = sum)) totpeople_recreated <- ave(female + !female + 1, year, class, company_n, FUN = sum) USMA_data <- transform(USMA_data, totpeople_recreated = ave(female + !female + 1, year, class, company_n, FUN = sum)) USMA_data <- transform(USMA_data, totpeople_recreated = ave(female + !female + 1, year, class, company_n, FUN = sum)) library(dplyr) USMA_data <- USMA_data %>% group_by(year, class, company_n) %>% mutate(totpeople_recreated = sum(female + !female) + 1) USMA_data <- select(data, -totpeople_recreated2) data <- transform(USMA_data, company = substr(company_n, 1, 1), division = substr(company_n, 3, nchar(company_n))) data <- transform(USMA_data, company = substr(company_n, 1, 1),division = substr(company_n, 3, nchar(company_n))) USMA_data\(company <- substr(USMA_data\)company_n, 1, 1) USMA_data\(division <- substr(USMA_data\)company_n, 3, nchar(USMA_data\(company_n)) library(dplyr) USMA_data_1981 <- filter(USMA_data, year == 81) library(dplyr) USMA_data_1981 <- USMA_data[USMA_data\)year == 81, ] USMA_data_1981 View(USMA_data_1981) library(dplyr) companies_by_continuation <- USMA_data_1981 %>% group_by(company_n) %>% companies_by_continuation <- USMA_data_1981 %>% group_by(company) %>% summarize(continue_or_grad_rate = mean(continue_or_grad)) %>% arrange(desc(continue_or_grad_rate)) library(dplyr) company_rates <- USMA_data %>% group_by(company) %>% summarize(continue_or_grad_rate = mean(continue_or_grad)) %>% arrange(desc(continue_or_grad_rate)) View(company_rates) head(company_rates), 4) head(company_rates, 4) top4company_continuation <- head(company_rates, 4) View(top4company_continuation) library(dplyr) continuation_by_class <- USMA_data_1981 %>% group_by(class) %>% summarize(continue_or_grad_rate = mean(continue_or_grad)) arrange(desc(continue_or_grad_rate)) library(dplyr) > continuation_by_class <- USMA_data_1981 %>% library(dplyr)> continuation_by_class <- USMA_data_1981 %>% + group_by(class) %>% + summarize(continue_or_grad_rate = mean(continue_or_grad)) library(dplyr) continuation_by_class <- USMA_data %>% group_by(class) %>% summarize(continue_or_grad_rate = mean(continue_or_grad)) %>% arrange(desc(continue_or_grad_rate)) View(continuation_by_class) head(10) continuation_by_class <- head(continue_or_grad_rate, 10) continuation_by_class <- head(continue_or_grad_rate, 10 continuation_by_class_women <- USMA_data_1981 %>% continuation_by_class_women <- USMA_data_1981 %>% filter(female == 1) %>% group_by(class) %>% summarize(continue_or_grad_rate = mean(continue_or_grad)) %>% arrange(desc(continue_or_grad_rate)) View(continuation_by_class_women) women_per_class <- USMA_data_1981 %>% filter(female == 1) %>% group_by(class) %>% summarize(num_women = sum(female)) View(women_per_class)

Tables:

For questions 2/4. ‘USMA_data’, the original data table with added recreations of malespeers, femalespeers and totpeople, as well as the added company and division variables derived from company_n:

year class female femalespeers malespeers continue_or_grad company_n totpeople femalespeers_recreated malespeers_recreated totpeople_recreated company division
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
78 1 0 0 34 1 A-1 35 0 34 35 A 1
  1. ‘USMA_data_1981’, the same table but filtered down to only subjets in 1981, the only year that contains all 4 classes in full quantity. I used this data to create the rest of the tables:
year class female femalespeers malespeers continue_or_grad company_n totpeople femalespeers_recreated malespeers_recreated totpeople_recreated company division
81 1 0 3 31 0 A-1 35 3 31 35 A 1
81 1 0 3 31 0 A-1 35 3 31 35 A 1
81 1 0 3 31 0 A-1 35 3 31 35 A 1
81 1 0 3 31 0 A-1 35 3 31 35 A 1
81 1 0 3 31 0 A-1 35 3 31 35 A 1
81 1 0 3 31 1 A-1 35 3 31 35 A 1
81 1 0 3 31 1 A-1 35 3 31 35 A 1
81 1 1 3 31 1 A-1 35 3 31 35 A 1
81 1 0 3 31 1 A-1 35 3 31 35 A 1
81 1 0 3 31 1 A-1 35 3 31 35 A 1

6a. ‘top4company_continuation’, this table shows the 4 companies with the highest continuation rates in order:

company continue_or_grad_rate
D 0.9264249
A 0.9183673
H 0.9153077
B 0.9139842

6b. ‘continuation_by_class’, this table shows the classes with the highest continuation rates on order:

class continue_or_grad_rate
4 0.9712418
3 0.9572864
2 0.8891413
1 0.8478868

6c.’continuation_by_class_women’, this table shows the classes with the highest rate of continuation for the women in their class:

class continue_or_grad_rate
4 0.9838710
3 0.9090909
1 0.8120301
2 0.6847826
  1. ‘women_per_class’, I made an extra table for question 7 to try and find out more about the distribution of women:
class num_women
1 133
2 92
3 66
4 62

Written answers:

  1. Investigate the rows for which your recreation doesn’t line up exactly with the original columns. Any ideas what the issue might be? Do you trust the original or your recreation more?

I could only find one row where my recreations do not line up with the originals, row 73. I tried to find a pattern as to why this row did not work, perhaps it was because this student didn’t continue, because they are in company-number A-2, or because they had a sizable number of female peers. However when I looked at other rows that shared these variables in common with row 73, none of those rows had discrepancies with the old and new variables, so I am baffled as to why this is happening. I trust the original data more here because my row 73 shows that this person had 2 people in their classes, which cannot be right. I am interested why this is happening, or what the problem with the original data was supposed to be.

  1. Notice anything strange about the “random assignment” of women?

I have two observations about the random assignment of women; one is that class 1 has a lot more women than the others, more than classes 3 and 4 combined in fact, class 2 also has considerably more than 3 and 4. This might not be ‘strange’ though, these discrepencies can happen if women were truly randomly assigned to a class. The other thing that I noticed is that class 2’s women had a much lower continuation rate the other classes, perhaps there were factors in that class which led to it being harder for women to succeed.