Ray Fisman has a very interesting Slate article about a study of some students at the U.S. Air Force Academy, in which women performed better in their courses under female professors. The gender gap almost disappeared under female professors, but was fairly pronounced under males. Men performed better under male professors as well, but the effect was much smaller — that is, the gap disappeared mainly because the women got higher scores, not because the men got lower ones.

What troubles me about the piece is that it seems to assume that all grade differences are earned — that is, student performances exclusively, with no role for the biases of professors, lead to the grades. Therefore, female profs deserve credit for eliminating the gap.

Isn’t it possible that in this school, female professors go easy on women, or male professors go hard on them? One thing I’d like to know is how subjective the grading is. Fisman mentions a “rigid curriculum” and the fact that everyone takes the “same exams,” but doesn’t mention whether these exams have strict right and wrong answers.

Also, there’s a pretty serious “rats and sophomores” problem. Even if we’re limiting the discussion to higher education (and leaving out gender gaps elsewhere), the U.S. Air Force Academy does not have a representative sample of male college students, much less females.