Re:Harvard discrimination lawsuit: data show penalization of...
by gizmo686 on Sunday June 17, 2018 @11:31PM
(#17335849)
Yes. In reality, we are talking about populations. It is still possible that the type of confounded you describe applies at the population level as well, which is why I qualified my statement with "might expect". This is an empirical question that requires empirical evidence to answer. Further, universities have this empirical evidence (although I am not aware of what it actually says).
I do recall hearing that (at some point in history) the army looked into this and found that black soldiers did better then their white counterparts with an equal score entering. Further, this is presisly the result I would expect to see in the face of generic anti-black bias. For instance, if there was some force that depressed black SAT scores by 10 points and had no other effect, then a black student would be expected to perform as well as a white student who scored 10 points better on the SAT.
I can picture many different things matching this overall impression. Perhaps Alice is the child of two university professors and has vaguely drifted to near-perfect scores without really having to work. Alternately, she has slaved over her books desperate to please her parents and is at the end of her rope.
Similarly, Bob might be a self-educated polymath who has transcended his background, or a crammer who has carefully done everything exactly right to produce the highest score possible, but is almost entirely uneducated outside of the areas that produced good-looking admissions.