The study [1] didn't "show" what you say. It
claimed there was a bias towards female
applicants.
To support this claim the study pointed out that the percentage of male
applicants _out of all male_ applicants that were granted admission was lower
than the percentage for female applicants, for all (examined) departments.
However, according to the study, there were 8442 male applicants and 4321
female applicants. So while a larger proportion of female applicants were
granted admission, it still meant that many fewer women were addmitted.
To clarify, a higher proportion of a smaller number can still be a smaller
number. 5% of 1000 is 50, 10% of 100 is 10.
Imagine we split a pie in 10 pieces, you keep 8 and I keep 2. I eat my 2, you
eat 4 of your 8 and then you accuse me of hogging the pie because I ate 100%
of my share while you ate only half of yours.
That is what the study actually showed happenned with admissions in Berkeley.
_________________________
[1] Sex Bias in Graduate Admissions: Data from Berkeley
https://homepage.stat.uiowa.edu/~mbognar/1030/Bickel-Berkele...