Sunday, March 21, 2004

Orgasmetrics

Steve Landsburg talks about the economics of orgasm (thanks to Alex for the pointer). I can’t improve on Landsburg’s explanation, so just read the article. However, I do think he makes one incorrect inference:
According to the 2000 Orgasm Survey (did you know there was a 2000 Orgasm Survey?), 72 percent of women have faked at least once in their current or most recent relationship, and 55 percent of men say they can tell when their partner's faking.

Apparently someone's deluded, though it's not clear whether it's the woman who overestimates her acting ability or the man who overestimates his perceptiveness.
Hold on, there. It wouldn’t be surprising to find that someone’s deluded, but the statistics cited don’t prove it. 55% is not the percentage of men who think they’ve caught a faker, or the percentage of recent sexual relationships that men perceive to have involved faking. 55% is the percentage of men who say they can detect faking when it occurs, and that is perfectly consistent with 72% of women faking.

For example, assume (for argument’s sake) that the pairing of women and men for sexual relationships is independent of women’s tendency to fake and men’s perceived ability to detect faking. Then the two figures together imply that approximately 40% (55% x 72%) of recent sexual relationship paired faking women with men who thought they could detect faking. If the men are correct about their detection abilities, we should expect 40% of men to have detected faking by their last partner. If a survey indicated that 40% of men had in fact detected faking in their most recent relationship, there would be no particular reason to think any men were deluded about their perceptiveness. If the survey indicated otherwise – if fewer than 40% of men said they had detected faking partners – then we would have evidence that some men are, in fact, deluded about their perceptiveness. (Either way, some men are being fooled, in the sense that some women are faking undetected. But being deluded, as Landsburg uses the term, means overestimating your perceptiveness or acting ability – i.e., thinking you couldn’t have been fooled when you were.)

However, even the evidence of fewer than 40% of men having detected faking by their last partner, if such evidence existed, would be tainted by the independence assumption. It might be that faking women are most likely to stay with men who are easily fooled, because nobody wants to get caught. A countervailing effect would result from women choosing to dump men whose inability to satisfy them leads them to fake more often. The balance of the effects could go either way, but suppose the first effect is larger. Then we would expect to get faking women paired mostly with oblivious men, and non-faking women paired with perceptive men. In this scenario, while there would certainly be many men getting fooled, few people would actually be deluded in Landsburg’s sense of the word, because most or all of the men being fooled would also recognize their lack of perceptiveness.

No comments: