I'm kinda circling around Nicholas C. Edsall's Toward Stonewall: Homosexuality and Society in the Modern Western World (University of Virginia, 2003). I've been meaning to read it for some time, and I hadn't realized how long that was until I typed in the publication date just now. It's intended as an overview taking recent scholarship into account, and it might be that Edsall would have a different take on the essentialist vs. social constructionist controversy with which he opens the book if he were writing it now. As so often, Edsall seemed to be addressing that controversy dutifully rather than from a conviction that it's all that important, and as so often, he's not very clear on its import. I intend to return to that issue in another post, to update my own views, but for now I'll look at another topic that trips up scholars: Alfred Kinsey's research on human sexuality.
Edsall devotes several pages to Kinsey, and this passage jumped out at me:
Kinsey's statistics were based on questionnaires gathered from a large, random sample of ordinary people. (Unfortunately, his report on women, published five years later, was based on a far smaller sample and was not accorded the same authority or fame) [265].
That's a lot of errors to put into just two sentences. First, "questionnaires" suggests that Kinsey handed out paper forms for people to fill in. The problems with that method were well-known when he began his research, and Kinsey opted for in-person, face-to-face interviews based on a complex and flexible protocol that the interviewers memorized. You could call that a questionnaire, but it's really not the correct word.
Second, Kinsey's sample was large but it wasn't random. He used a method called stratified sampling instead. There was a lot of public debate about this when the Male volume was published in 1948, and it's hard to understand how Edsall could have missed it. (I recommend Peter Hegarty's chapter on the controversy over Kinsey's statistical methods in Gentleman's Disagreement [Chicago, 2013], but anything you read about Kinsey will deal with it.)
Third, the Female volume, published as Edsall says in 1953, wasn't based on a smaller sample than the Male. If anything, the sample was somewhat larger: six thousand women versus the Male volume's 5300. (In both cases, Kinsey didn't use all of the eighteen thousand histories he and his team had taken.)
Finally, the claim that the Female volume wasn't "accorded the same authority or fame" as the Male volume is pretty obviously false. The "authority" of both volumes was fiercely contested, but the Female volume outsold the Male, which had moved 200,000 copies in hardcover from a medical publisher; Sexual Behavior in the Human Female sold 270,000 hardcover copies in the first month, and unlike its predecessor was later issued by Pocket Books as a mass-market paperback. "Infamy" might be a better word than "fame," because Kinsey's claims about the number of American women who had pre- and extra-marital sexual experience outraged "decent" people; the furor led to the Institute for Sex Research losing its funding from the Rockefeller Foundation. (Ironically, women still were a lot less sexually active than men according to Kinsey, but still too active for the sensibilities of American moralists.)
I've often noticed that writers on human sexuality seem to feel no need to check their facts when they write about Kinsey. Some of this reflects confusion about "sexual orientation" and its relationship to sexual behavior, and can be put down to differences of interpretation, but much of it is a failure of factual accuracy, as with Edsall. I suspect it's partly because Kinsey, like Noam Chomsky, is a safe target, and nobody cares about facts; but even Kinsey's (and Chomsky's) fans get him wrong. And the errors I've found in Edsall's book should have been pretty easy to avoid. Academic publishers in particular are supposed to care about factual accuracy, but in at least certain instances they don't.
