[camccann]

The author argues that males have a wider intelligence distribution, and so though they are equally as intelligent as females on average, there are more of them at the far right end of the scale.

At both ends, actually, but that's largely irrelevant to the argument. Actually, I think the idea is that males have higher variance on a variety of attributes, due to higher expected utility, in an evolutionary sense, from trying high-risk, high-reward strategies (standard disclaimer: do not anthropomorphize evolution, do not equate evolutionary goals with conscious goals).

But there's a link here he never established, unless I missed it. Is it really true that the brightest of the bright go into STEM?

That's not necessary--only that success in STEM fields is more strongly correlated with intelligence than with unspecified other attributes. It doesn't matter what very smart people in general do; only that higher intelligence confers an advantage. Intra-field competition will take it from there.

It is not, I think, controversial that intelligence correlates with success in STEM fields, seeing as it correlates with success (to some degree) in almost every area of life, but as always, the details are complicated.

What if we established that the intelligence of people who go into STEM is on the right side of the distribution, but not the far right? Then there would be as many females as males at that intelligence level.

Not necessarily--that depends on the shape of all three intelligence distributions (both sexes, and STEM fields).

That said, though, the whole article was basically fluff--lots of raising questions, not much in the way of relevant research, and virtually nothing in the way of concrete predictions or testable hypotheses. Now, as far as I know, there is some established research behind the "wider variance" idea... but to be honest the article raises all kinds of red flags; it feels like something written by picking a conclusion, then looking for science-y-sounding things to support it.

[scotty79]

> Actually, I think the idea is that males have higher variance on a variety of attributes, due to higher expected utility, in an evolutionary sense, from trying high-risk, high-reward strategies (standard disclaimer: do not anthropomorphize evolution, do not equate evolutionary goals with conscious goals).

I think if anything it can be attributed to lower evolutionary obligation of males rather than to higher expected utility. Breeding for males is fairly easy so you can have a kid even if your IQ is 60. Child of woman of such low IQ would have much higher probability of dying.

Also males having single X chromosome have higher variability in traits influenced by genes on X chromosome as the defect of some gene of X chromosome won't be alleviated by correct gene at the other X (because there is no other X in males).

Personally I don't care if all Einsteins and Newtons will be male but since more and more scientific discovery rely on joint venture of tens or hundreds of scientists I really think that you should not pass up on opportunity to encourage half of the population to go into science just because there might be some relevant skewing in right tail of distribution.