Women are bad at higher level math in the same way men are violent. Not all are, but its a trend.
However you aren't allowed to say that for one group, but you are allowed for the other. So this has nothing to do about statistical accuracy, its just political pressure from one side.
It's nothing to do with political pressure and more to do with ill-formed comparisons.
You can't just substitute in random groups for another. In order to make a statement you need to know what you're talking about. And it seems like the HN crowd that so desperately wants to say blacks are more violent than whites, these people have no clue what they're talking about. If you go and look at history you'll quickly see how that statement is just ill-formed.
A woman walking alone at night who encounters a stranger does not care what generative process led to a group disparity, she cares whether she is likely to be in danger. It is politically palatable in polite society for her to be afraid of an unknown man on the basis of his sex. But it is not acceptable for her to even consider that a statistical disparity may exist on the basis of race, or take precautions on that basis, unless it is in the context of condemning society as solely responsible for creating that disparity.
A statement of empirical observation cannot be "ill-formed" unless you have appointed yourself ultimate arbiter over why a person might care.
However you aren't allowed to say that for one group, but you are allowed for the other. So this has nothing to do about statistical accuracy, its just political pressure from one side.