Introduction

This report is on the UCB_Admissions data set, this is data pulled from 1973 UC Berkeley’s admissions system suspected gender bias where the school roughly admitted 44% of their male applicants and 35% of their female applicants. This data set in particular is interesting no only based on the topic of gender bias, but also due to the data presenting a phenomenon called Simpson’s Paradox. This paradox is a visual trend or result that is present when data is put into groups that make some of the data reverse or seemingly disappear when the data is combined

Graph 1: This graph displays the general number of applicants and there Status wether it be “Admitted” or “Rejected”

Numbers like these are typical for colleges, of course, varying in scale. This grpah would be consider the “granulated” or the most basic form in which the data can be graphed and displayed for viewing.

Graph 2: This set of graphs goes futher into the data providing whether an applicant is “Admitted” or “Rejected” and male or female.

In the graphs shown above we can see that compared to men, women seemed to be admitted less with about a 70% rejection rate. This also shows just how many men applied to UC Berkely compared to women who make up about 40% of applicants.

Graph 3: To take a closer look at the data lets include the Departments to which applicants appied to specifically

For this graph lets break it down by “Departments” of interest

Shown in Department A the acceptance rate is good with about 22% of women being rejected

Shown in Department B we can see from the graph that women were accepted at a higher rate than men

Shown in Department F We can see that the acceptance rate is very low but it has accepted women at a higher rate than men

Here is a similar grpah made using the same data. Does it look any different?

This is a prime example of the Simpson’s Paradox

Shown in Department A the acceptance rate is decent, but seemingly about all women that applied were accepted which is hard to see due to its small size.

Shown in Department B we can see from the graph that seemingly no women applied to this department perpetuating the false “gender bias”, but again the data is hard to see or isn’t shown due to the scale of the graph.

Shown in Department F We can see that the acceptance rate is very low but it has seemingly accepted only women, this is also incredibly false but according to the graph these are the fact.

For those into numbers here is a table depicting the information given in the previous graphs