From the news release [1]:
"We tested some recently proposed hypotheses that try to explain a supposed gender gap in math performance and found they were not supported by the data," says Janet Mertz, senior author of the study and a professor of oncology at the University of Wisconsin-Madison.
Instead, the Wisconsin researchers linked differences in math performance to social and cultural factors.
The new study, by Mertz and Jonathan Kane, a professor of mathematical and computer sciences at the University of Wisconsin-Whitewater, was published today (Dec. 12, 2011) in Notices of the American Mathematical Society. The study looked at data from 86 countries, which the authors used to test the "greater male variability hypothesis" famously expounded in 2005 by Lawrence Summers, then president of Harvard, as the primary reason for the scarcity of outstanding women mathematicians.
...