Report 1: UCB Admissions
‘r Sys.Date()’
UC Berkeley and Simpson’s Paradox
Simpson’s paradox is a phenomenon in probability and statistics in which a trend appears in several groups of data but disappears or reverses when the groups are combined. A clear way to demonstrate this phenomenon is the case of UC Berkeley’s 1973 graduate school acceptance rates for men and women. upon preliminary searches, it appeared that there was a gender bias in the selection process.
The data from the graph above, hints towards the fact that females were not accepted as frequently as males, by quite some marginn. This data however was base of a rather large set of data. Once the schools statistician delved further into the data and the results, he decided to redo it but this time separating the departments up. In doing so, Bickel broke up the acceptance rates by department.
Factoring in the different departments
These new results given from the graph above indicates that there is a different side to the story than originally thoughtt. based of of the new graph and data, there is indeed a gender bias in favor of females, quite the juxtoposition the the original findings, where it was thought that there was a biased result in favor of men. Males are in fact more likely to be rejected in four of the six departments accounted for. The findings from the statistician was that women had an overall higher chance to apply throughout all the departments. This resulted in a lower percentage of women being accepted as a whole.
Simpson’s Paradox
This data and the results of which, found at UC Berkeley’s graduate scheme regarding their acceptance rates greatly portrays Simpson’s paradox. on first appearance, it looked as though males had a gender bias in their favor, however, after UC’s statistician, Peter Bickel, broke down the data into different departments, quite the opposite turned out. Females were actually the ones that had the gender bias in their favor. seeing the new and corrected data was aided with the help of adding the variable which allows us to see which department people physically applied to, also know as as a lurking variable. A lurking variable is defined as a “variable that is unknown and not controlled for; It has an important, significant effect on the variables of interest.”
Works Cited
https://www.statisticshowto.com/lurking-variable/
https://plato.stanford.edu/entries/paradox-simpson/#:~:text=Simpson’s%20Paradox%20is%20a%20statistical,population%20is%20divided%20into%20subpopulations.