>In an NYRB review Richard Lewontin pointed out that this discrepancy
>needs to be accounted for. There followed an exchange between him and
>the book's authors (some U. of Chicago sociologists I believe). I don't
>recall the exact arguments on either side. In general, I remember,
>Lewontin doubted the "scientific" status of such surveys.
This is especially unscientific - it's a web poll conducted by a condom company.
Doug