The attached file "admit.csv" is a comma-separated values file with data on admissions to graduate school at the University of California Berkeley for Fall 1973. It includes admission decisions by gender of applicant for six large graduate departments. The variables are:
• dept admitting department (A, B, C, ..., F)
• gender = gender of applicant (F, M)
• admitted = admission decision (N, Y)
• count = number of applicants for the given combination of dept, gender, and admit
(a) Import the data into an R. data frame and display the first 10 rows of the data frame. (Hint: use "read.csv" - see help("read.csv") - or else set sep="," in read.table).
(b) Create a 2 x 2 x 6 table from the data frame to display the conditional 2 x 2 (gender x admitted) tables for the six departments.
(c) Compute the 6 estimated conditional odds ratios, conditioning on dept, and interpret how the odds of female applicants being rejected versus the odds for male applications varies across departments. Do they suggest a systematic bias in one direction?
(d) Create and display the marginal gender x admit table, and compute the estimated marginal odds ratio. According to this calculation, which gender has higher odds of rejection, and by how much, when we ignore differences in department admission rates?
(e) Display the 6 x 2 table showing the admission counts (N, Y) for the 6 departments. Also calculate the proportion admitted for each of the 6 departments.