Monday, May 18, 2009

terrible epidemiology of sex

The Myth, The Math, The Sex, from the NY Times

Someone sent this to me and asked for alternate explanations. How is it that men can report so many more sexual partners on average than women when men have sex with women?

There are a number of major reasons, with an important lesson to be learned that applies to many things outside this topic.

EDIT: So Annetta demonstrated my first example as wrong. I'll replace it here (oops...) by pointing out that the article refers to the median number of sexual partners, not the mean... which is IDIOTIC. You can explain the discrepancy by saying there are a lot more women having little or no sex, and a few women having lots and lots of sex, while men are more evenly distributed. In any case, mea culpa.

Another reason is that men die off younger, and older women may be likely to have had fewer sex partners than younger women and men because of different sexual mores. Thus, the increased number of old women in the sample relative to old men may skew women's numbers down. this may be especially true if women with fewer sex partners live longer than women with more sex partners, but men aren't as affected by that difference. I don't know if this is biologically accurate, just hypothesizing.

A third reason is the existence of gay men and women. Gay men and women may be a substantial portion of the population - as high as 10%, although no survey can get good numbers on this. Almost every study shows that there are more gay men than gay women, and that gay men, on average, are much more promiscuous than straight men. If homosexual intercourse is counted in the number, then the increased number of gay men in the sample may skew men's numbers up. If the definition of "sex" in these surveys requires penetration, then one instance of gay sex would count twice, and one instance of lesbian sex would not count at all. That's a big skew. Even if people are instructed only to count heterosexual sex, they may not only report heterosexual sex.

A fourth reason is that there are more women than men in the population, a difference which gets more drastic as you get older and into sexual maturity, so finding the "average" number of partners skews women's numbers down. If each of 5 men has sex with each of 6 women, then there will have been 30 partnerships, and each man averaged 6 partners and each woman averaged 5 partners.

I understand 7 lifetime partners (the male average) vs 4 lifetime partners (the female average) is a big difference, and much of it may be reporting issues, but a math professor such as the one in the article should understand a lot of this.

This kind of statistics and logic is a common problem in epidemiological papers, where doctors and public health people have learned outdated, inaccurate or incomplete statistics and then use it and publish it. Read sources carefully.

No comments:

Post a Comment