You can see how different cultures influence statistical comparisons relative to one another, but you still can't say what the statistics would be like in a neutral culture. I've seen many people (including study authors) jump to the conclusion that if culture affects a statistic relating to subgroup performance, then culture must be wholly responsible for the differences. Unfortunately, that's not a logical conclusion.
Well, if it's not genetic, and it's not cultural, then I'd be interested in your alternative hypotheses. Because we live in an "advanced" Western civilization which only, after millennia of feminine marginalization, have reluctantly allowed equality of opportunity for women for only the last 3-4 decades, and yet we can still see a math gap in only 35/65 participating nations... so it's most definitely NOT genetics.
