It looks like he said, specifically:

> "...[there] is relatively clear evidence that whatever the difference in means—which can be debated—there is a difference in the standard deviation and variability of a male and female population..."

Sheesh, of all the things to be cancelled for...