I had originally been planning to blog about the ownership of copyrights for the Unix operating system. But maybe no one else cares about that. Everyone’s minds are on more important topics, like sex. So imagine my pleasure in seeing that sex and math have had a rare convergence in the pages of The New York Times, in Gina Kolata’s article entitled
The Myth, the Math, the Sex.
Kolata quotes such leading mathematicians as Berkeley’s David Gale and UCSD’s Ron Graham (who memorably lectured at Gustavus in spring of 2006) to poke a big hole in survey data about sexual promiscuity. Apparently it is well established that the average man reports considerably more female partners than the number of male partners reported by the average woman. So are males more promiscuous? Well, there is this inconvenient mathematical fact that the number of male-female couplings needs to be constant, whether you count them by adding up all the females the males have coupled with or by adding up all the males that the females coupled with. So given roughly equal population sizes, there is no way that the averages can be very different.
A technical weakness in Kolata’s article is that one of the survey results it focuses on was a report of the median number of opposite-sex partners for each sex, and the article doesn’t address the distinction between mean and median. If the women have a much more skewed distribution, and hence have a mean much higher relative to their median than the men do, that could in principle account for the discrepancy. However, given the size of the differences reported, it does seem more likely that this is a case of math helping illuminate the unreliability of self-reporting.
[Update: Not surprisingly, I wasn’t the only one to bring up the issue of the median. Consequently, Kolata published a clarification regarding it.]
Leave a Reply