> Like make up a characteristic where women differ statistically from men, but have the difference be about 1/2 the standard deviation for either group.

> there's no useful generalization to make from the clear evidence.

Unless the selection is on the tail ends. https://darrendahly.shinyapps.io/app1/