What I find mystifying in all of this research is that women who are able to pursue higher education in India, China, Russia, and Egypt do not seem to have the same difficulty with higher math that consistently appears in European, American, and Australian research. What gives?

Great article. Always wonderful to see a more careful analysis of data - and especially on a topic it's not fashionable to question the conventional wisdom on.

Although the NYT was certainly facile with the facts, I'm surprised you yourself gave glancing notice to the numbers for Asian students. Those seemed to me to be the most suggestive that, in white America at least, there is a cultural element that dissuades girls from pursuing their math potential. The only other explanations for the difference in Asian and White girls' performance is that a) Asian Americans have a cultural bias toward girls in math, or b) there's a racial variation between Whites and Asians that make Asian girls' brains more mathy. Both seem really unlikely.

Just so you know, I think there are cultural and biological reasons for why more guys go into math - that have nothing to do with sexism (or do so, but in a positive sense). I (a female) was a math whiz in high school. I was planning on studying math at Harvard but, although I got A's in my two math classes there, I was simply drawn to other topics - in the human sciences. I've also heard of a study that says similar things - that woman are turned away from the math and sciences, not because they are discouraged by lack of success, but because they don't find them as interesting or fulfilling as studying and doing other things.

That makes sense with my understanding of women - both biologically and culturally. We are naturally more social than men and fewer of us are willing to forgo social fulfillment (at work or at home) to spend massive hours working on highly theoretical abstract problems.

Aside from having broader interests, mathy women, I'm guessing, generally have an advantage over mathy men by being more employable in other areas. I'm someone who loves math geeks, but really - have you ever met any guy math professors? Can you really imagine them getting a job doing something else? Forget social work, but even running a business, playing with the boys on Wall Street or lecturing on philosophy would probably be a stretch.

Again, great to see you publishing smart, insightful work - as always.

Heather Mac Donald seems to have missed the point of the math study just as much as the New York Times did. One of the purposes of education is to idenify students who have the potential to excel in a field, and to help them to do so. The fact that statistically, more boys may excel at math than girls is no reason we should assume that "math is harder for girls" -- as Ms. Mac Donald titles her article. Rather, the top female students may excel as much as the top male students; they just may be fewer in number. However, they have an equal right to have their skills encouraged and developed -- and to be encouraged to seek out potentially lucrative math careers. Ms. Mac Donald's article seems just as intent on misreading statistics as the NYT article it criticizes.

You are spot on, as usual. Apart from the willful negligence of the NY Times, it is horrifying that Science would publish the source article with such a false title and deceptive conclusions. "Gender Differences Account For At Least Half of the Underrepresentation of Women in Science" would be a far more accurate title for the article's evidence.

One other aspect of the Science article caught my attention: How can the data show 1.25 percent of 219 Asians (in the chart labeled "The upper tail")? Two people out of 219 is 0.91 percent and three is 1.37 percent (the precise value given for females). I wasn't able to locate the raw data from Minnesota, but fractional persons suggest a data error. This leaves aside the question of whether a difference of less than one person can responsibly be presented as statistically significant, of course.