Monday, August 13, 2007

Sex math

Candidates paid for votes in the Iowa straw poll. Does anyone realistically put any stock in the results?

Several polls about sex, where no one was paid, should engender the same skepticism.

A British study found that on average during their lifetimes, men reported 12.7 heterosexual partners and women 6.5, about half.

An ABC News poll conducted by TNS found that on average during their lifetimes men reported 20 sex partners, and women just 6. The same poll reported that only 54 percent of women condone sex before marriage, and for men it was 68 percent. So, let me get this straight: nearly half the women and one-third the men believe in no sex before marriage? That should be a tip off that there’s something terribly wrong with these polls.

Even the NY Times, who reported on inherent contradictions in sex poll results, got it wrong by equating mean and median.

For the record, the mean is the average. The median is the number both above which and below which half the results lie. The housing industry gets it right: they report both mean and median home sale prices. The mean is usually larger because the very high priced homes skew the average higher.

So what’s so wrong with those sex polls? Think about it: if we start with the same number of men and women, the average number of sex partners has to be the same for men as it is for women. There’s no way men have twice the number of sex partners.

But the polls do prove one thing: there’s no question they’d like to think so.

No comments: