A story that's recently been making the rounds has been this post mentioned in the Gene Expression blog in April about how "smart kids don't have sex". So far I've also seen it at slashdot, livejournal's "mathsex" community, and the Dilbert blog; it was also mentioned in Marginal Revolution back in April.
The Gene Expression post reproduces the plot which shows that "teens with IQs ranging from 75 to 90 had the lowest probability of virginity (the authors note this is also the same IQ range where propensity towards crime peaks)." "Teens" here means grades seven through twelve, so technically, at least in a lot of states, having sex is a crime for these people. And even if you ignore age of consent laws (as I think most teenagers are likely to do), there's still a strong message from parents and other authority figures that they shouldn't be having sex, and it's hard to find a place to do it. My suspicion is that both of these are tied in with some willingness to break rules. Why a willingness to break rules is correlated with a slightly below-average IQ, I don't know.
This post then goes on to consider college students, and to cite an article from the November 2001 issue of Counterpoint, a joint MIT-Wellesley publication, which has some charts purporting to show that likelihood of being a virgin is correlated with, um, scientificalness of one's major? (The choice of the word "scientificalness" is deliberate here, in that it's not a real word.) I'm inclined not to trust their results here because of sample size; they polled 287 Wellesley students and 236 MIT students, and break up the 287 students into 18 majors and the 236 MIT students into 15 majors. This means than, on average, they polled sixteen students in each major, and are then comparing those numbers.
The headline that grabbed me was the livejournal headline, "83% of undergraduate [...] students in math are virgins". 83%? Sounds like "five out of six" to me, or "ten out of twelve". Furthermore, 83% of Wellesley math majors were virgins; 29% of MIT math majors were virgins. Wellesley's a women's school; most MIT math majors are male. Similarly, they claim to be able to say which dormitories have more virgins than others; however they only mention eight of MIT's ten undergraduate dorms, and five of its several dozen Greek organizations. I can only assume that they got nobody from those places, which leads me to think that their samples from the dorms they did get a sample from weren't too big either.
And it's notoriously difficult to get information about people's sexual histories, even in an anonymous survey. One way I've heard of getting decent aggregate information is the following: instruct the subject to go into a closed room, and flip a coin. If it comes up heads, they are to answer the question (let's say it's "do you masturbate regularly?") truthfully. If it comes up tails, they are to flip the coin again, and answer the question "did the coin come up heads on the second toss"? From the results one can get a good estimate of the percentage of people who would have answered the first question "yes", but nobody feels incriminated.
I'm done criticizing the "study" in Counterpoint. (I remember criticizing it when I was an undergrad; I started at MIT in September of '01, and even after two months it was clear to me that something was fishy here.) But if "smart people" are having less sex, why? First, note that it's not "smart people" so much as "smart kids". There might be some correlation between "being smart" and "following rules", at least if you're using getting into a selective college as a proxy for intelligence. And as I said before, following the rules of society as a whole means not having sex outside a "committed relationship", whatever that is. But on the other hand, I think a lot of intelligence comes from knowing which rules to follow and which rules to break. MIT prides itself on creating the people who will lead the scientific and technological worlds in the future, and to get to that point people are going to have to break the rules, to ask the questions which haven't been asked, to do the things that the old farts have already said are impossible. What makes you think these kids aren't going to have sex?
Then again, "peer pressure" always seemed to me to be weak to nonexistent at MIT. So if college students are only having sex because they feel they "should" be having sex, then they're not going to.
As to whether mathematicians are more or less sexual than the general population, I can't really comment on this. I know I haven't had sex in a while, but other than that I can't really say; most of the people who I feel comfortable talking about sexual matters with are not mathematicians. There seems to be a tendency towards social awkwardness in the "smart kids" in high school, and less so in college, but I think after college it goes away; once people get out in the larger world they tend to find the people around whom they are comfortable. In the absence of the rule-breaking behavior I mentioned before, I would guess that the people having the most sex would be the ones right in the middle of the intelligence distribution, because people tend to have sex with people like them, and there are more people like the average people -- that's what being average means. (I am deliberately saying "average" instead of "mean", "median", or "mode", because intelligence is approximately normally distributed so these are all the same.) This is probably true for a lot of other things as well. The people having the most sex aren't the rich, beautiful people. The people having the most sex are probably of average financial status, average looks, average intelligence, ..., because finding a partner is easiest if you're Just Like Everyone Else.