UC Berkley Admissions

Looking at the UC Berkley’s admission rate it was said that they had a gender bias in 1973. In the graph you can see that there were more females that were rejected than they were admitted. While the males that applied were about half were admitted into the school and half were rejected. From the graph you can possibly conclude there was a gender bias happening in the admissions to the school because there were more females that were rejected than the males that were rejected.

Let’s really take a deeper look into the admissions rate at UC Berkley and see if there are any lurking variables. The graph bellow is the admissions rate broken down into gender as well as the sections that they applied for. Looking at section A you can see that more than 75% of the females that applied got accepted into the section. While still in section A the males that were accepted was less than the females. Section B is just a little bit less of the females that applied were admitted. Now, looking at section C, D and E you can see that more than half of the females that applied were rejected. Last is section D the males and females had a very similar rejection rate as well as an admitted rate.

From this graph I would conclude that there may not be a gender bias on the admissions. Females had a very good acceptances rate into two of the sections. The three sections had a lower admission rate, but it wasn’t as low as the very last section. In that last section it was the same for male and female, so it didn’t have anything to do with their gender. So, while yes you can say that less females were admitted in some sections than others there isn’t a complete gender bias in every section.

In conclusion, the Simpsons paradox is when we look at something but there are more things to consider, and it is not always the right thing to simplify the data. With that UC Berkley data they didn’t take a deeper dive into the data when it came to their admissions rate. That is why the conclusion came out that there was a gender bias. When you can see that by not including that lurking variable which means to not include an analysis, but it can affect the outcome (Jim). Adding that lurking variable showed that there isn’t a complete gender bias.

#Work Cited

Bremner, Kira. “Simpson’s Paradox and Interpreting Data.” Towards Data Science, 14 Aug. 2019, https://towardsdatascience.com/simpsons-paradox-and-interpreting-data-6a0443516765

Jim, Jim. “Lurking Variables: Definition, Examples, and Impact.” Statistics by Jim, Statistics by Jim, 2024, https://statisticsbyjim.com/basics/lurking-variable/.