[nostrademons]

There's a fairly obvious explanation that I didn't see mentioned anywhere in the abstract: assortative mating. People tend to have sex with people of roughly the same intelligence, socioeconomic status, and interests. As you go down the tails of the bell curve, there are fewer people of the same intelligence with similar interests.

If you have an IQ of 100, 68% of the population is within one standard deviation of you. Your chances of meeting someone you feel comfortable with in high school are pretty high. But if you have an IQ of 145 (3 stddev), only about 3% of the population is within one standard deviation of intelligence of you. Add to that the constraints of it being a member of your preferred sex, being reasonably attractive & socially compatible, and wanting to have sex, and it's no surprise that most such people don't find suitable mates until they enter college, which purposely brings together people of similar intelligence.

And if you have an IQ of 175 (5 stddev), less than 1 in 10,000 people is within one standard deviation of intelligence. (Though this is a bit meaningless, as IQ distribution has fat tails when you get past about 145.) Then add in that super-intelligent people often have very narrow and deep interests, and that males are overrepresented on both ends of the bell curve (there are both more male morons and male geniuses than female ones).

This also explains why the smartest men were most likely to hire a prostitute. If they had no chance of forming an emotional connection with a woman, why not pay her to leave afterwards?

Personally, I had about zero interest in 90% of girls in college. They were just inane. And this is at Amherst, which already selects for pretty smart people.

[tokenadult]

No, valid, currently normed IQ test yields a credible score of 175.

Anyway, IQ is orthogonal to rationality,

http://yalepress.yale.edu/yupbooks/book.asp?isbn=97803001238...

http://www.amazon.com/What-Intelligence-Tests-Miss-Psycholog...

and sufficient rationality might explore the implications of the submitted article without any reference to IQ tests.

[Rod]

That's an interesting analysis, but I disagree on some points. My ideas:

1) people tend to have sex with people of roughly the same attractiveness. Only a fool would fail to realize that there's a sexual market out there, and one's price in that market is mostly dictated by looks, not intelligence or interests. (disclaimer: I live in Southern California, and I realize that I may sound overly superficial).

2) Sex and stable romantic relationships are way different things. It's true that people tend to start relationships with people of roughly the same intelligence, socioeconomic status, and interests... but come on! I would not turn down a funny, hot girl just because she's an English major who's not interested at all in Quantum Field Theory! I probably would not want to marry her, but some NSA fun once in a while is refreshing...

3) Grad students at MIT should be rather smart, driven people. They should be surrounded by people of the same intelligence and interests, and they still don't get laid much. Your cute analysis does not explain this.

[nostrademons]

1.) I think it's both. People have sex with folks of the same attractiveness, but they don't ignore intelligence. Could you imagine an MIT grad student having sex with a trailer-trash bimbo of slightly-below-average attractiveness? (Yeah, that's a bad mental picture.)

3.) They've had less time to be surrounded by people of similar intelligence and interests. The virgin rate among MIT grad students is roughly the same as the virgin rate among average high schoolers. That's consistent with the hypothesis that MIT-caliber grad students can't find anyone until they get to MIT, and then they follow the normal rules of chance & dating & random hookups that most people do in high school.