This is such an interesting - and politically dangerous - issue. Larry Summers, the former president of Harvard, made his job much more difficult than it needed to be by provocatively and stupidly discussing it in a ham-handed way. But the issue also shows how poor journalists are at reporting scientific data. The NYT article reports that:
The researchers looked at the average of the test scores of all students, the performance of the most gifted children and the ability to solve complex math problems. They found, in every category, that girls did as well as boys.
But that's just not what the actual article says. The article does say that there is no statistically significant difference between the
average performance of boys and girls. But it says that the variance is greater for boys than for girls. That means, pace what the NYT reports, that there are more boys in the upper tail (and presumably in the lower tail, too) of the distribution. When it comes to the most gifted mathematics students, in other words, and presumably the most incapable as well, the boys are overrepresented. Interestingly, the variance ratio (the ratio of male variance to female variance) differs by ethnicity. The article does not give us the full detail of the data, but it reports that for whites in grade 11 in Minnesota the variance ratio for those in the 99th percentile is 2.06. That means that there are roughly twice as many boys in that category as girls. By contrast, for Asian Americans in grade 11 in Minnesota the variance ratio at the 99th percentile is 0.91. In other words, there are slightly more girls than boys in that category. Still, when all the ethnicities are grouped together,
All VRs, by state and grade, are >1.0
The variance ratio is less than many have assumed, and not nearly significant enough to account for the fact that, for example, only 15% of the current students in engineering Ph.D. programs are women. But the article does report a difference in the tails. It makes no claim whatsoever to explaining this difference: it could be due to social typing, innate difference, or some combination of these and other factors. Obviously the significance of the results differs dramatically depending upon what causes them. But the main thing is that it's a subtle issue that the journalists are running roughshod over. Any discussion of the topic needs to go beyond simplistic cheerleading.
BB
