The article doesn't attempt to control for any variance that would lead to unequal outcomes even if there were no bias. For example, men make up the majority o

The article doesn't attempt to control for any variance that would lead to unequal outcomes even if there were no bias. For example, men make up the majority of people with very high (and very low) IQ's. If you randomly sampled the general population, binned to the IQ distribution of tenured professors, what gender ratio would you get?