My wife and I are both professional mathematicians. We had independently the same reaction to this report: the “researchers” are most likely ideologues, and they are using data based on mid-level tests, which do not test for high mathematical ability. Mathematicians tend to assess eachother in terms of “strength”, and these judgments are surprisingly repeatable from observer to observer. These assessments are NOT made on the basis of accuracy on routine problems, which is almost certainly the kind of problem found in the test used by these “researchers”. On the contrary, we’re interested in performance on the very hardest questions. In our experience, males heavily outnumber females on these questions. The study claims
“Looking at math ability alone, about one third of engineering Ph.D. candidates should be female, the researchers wrote. The fact that it’s only 15 percent suggests that old stereotypes have held women back, they believe.”
But assessment of math ability on the basis of accuracy performing routine problems is a very poor indicator of the math ability required in the professions in question.
Here’s a personal example, but one which I think is typical. Our son (now a star in a math PhD program) was an indifferent student in early math classes ( JHS math, HS algebra & geometry etc) and didn’t start to do well until the problems got really hard. At that point they got his interest, and his talent began to outweigh his carelessness and (mild) hyperactivity. By the end of HS he was typically the only kid in his class to actually ace the hardest problems on tests, but still got middling grades on these tests overall because of (characteristic) careless mistakes on problems he understood.
Both of us have seen this pattern numerous times, both with students and with colleagues. consider this sentence from the article:
“Males hold the lead in only a few disciplines: engineering, physics, chemistry, and computer science.”
What these disciplines have in common is they are heavily math loaded. (also some parts of economics). To a fair minded observer, this pattern is more persuasive then the crummy data used in this study. There’s no reason why math or physics departments would be more “sexist” than biology or psychology
