And you seem to be conflating two separate issues, romantic attraction, and ability to handle criticism.

Such is the apparent resolution of Tom Hunt's map of the territory.

The rational response would be to look into the matter to see which is in fact the case.

As I have already said, it is not a question of which is the case any more than the case of the unobserved falling tree is a question of sound or non-sound.

Well, you're link is more like, a woman after a lot of effort improves her criticism taking ability. And even then she would rather the culture change so that she doesn't have to exercise it. The question of whether criticism is necessary for the institution achieve whatever it's actual goals are is not addressed.

To represent my position or that comment's position as a claim that criticism is unnecessary in research or business is a straw man. You are not entertaining the possibility that women in general may adapt to the particular quality of criticism in predominantly male professions as that woman did, nor that a cultural change is possible, or furthermore, optimal.

Yes, if only there were a mathematical theory of how to update on that kind of evidence.

I use the word 'how' qualitatively, not quantitatively. I don't mean, "How well do women take criticism on a scale of positiveness?," I mean, "By what internal mechanisms and social norms do women tend to interpret and exchange criticism, and how is it different from the way men do?"

How about not going out of our way to actively promote more women in research, e.g., with all the "women in science" programs.

The problem with affirmative action is not encouraging minorities to participate in activities in which they have not historically participated; the problem with affirmative action is using a minority's membership test as the selection criterion as opposed to using the selection criteria that would maximize the intended effect of said activity. We can broadly encourage minorities to attempt to become researchers without letting people who are bad at research be researchers. Where affirmative action is suboptimal, 'negative action' is not the solution. It would be suboptimal for an NBA talent scout to exclude a seven-foot-tall white basketball player from consideration because he had precommitted to excluding white basketball players because being of African descent positively correlates more strongly with height than does being of European descent. It would be suboptimal for a senior researcher to exclude a woman with an IQ in the top 2% from consideration because he had precommitted to excluding women because having an IQ in the top 2% correlates more strongly with being male than with being female. Respecting the precommitment in those scenarios while keeping the ultimate size of your selected populations constant would not maximize the average height or IQ of your selected population, and the original purpose of the precommitments was to maximize the average height or IQ of your selected population; the precommitment in each case is a lost purpose. If you use race or gender to inductively infer someone's basketball-playing or research ability without using height or IQ when those data are available, then you are failing to update.

Except women appear to make up less than half of the intellectual elite, largely because they're IQ's have a smaller standard deviation.

Deary et al. (2006) estimate that ~1/3 of the top 2% of the population in IQ is female. 46.2 million people is non-negligible, to say the least.