[lbo-talk] Re: Diet Pills = Gay Babies

John Thornton jthorn65 at sbcglobal.net
Sat Dec 11 14:06:01 PST 2004


This is wild extrapolation from correlational data! Just as a correlation between ice cream consumption and violent crime on a given day is ludricious evidence that ice cream causes violent crime, a correlation between drug use during pregnancy and sexual orientation of offspring is ludricious evidence for the claim that "sexuality is determined in the womb".

The question the researchers here are dodging: what other social, psychological, and cultural factors distinguish the drug moms from the nondrug moms? Any of hundreds of confounds could be the true cause of the observed correlations here. Sloppy, sloppy--

Miles

At 09:42 AM 12/11/2004, Brian wrote:
> > And this is holiday homophbia as only Miles can serve.
>
>Like lots of others, you have the habit of leveling the "homophobia" charge
>when people disagree with you. But calling someone homophobic is not an
>argument--even in the cases where it's true. (And it's not--clearly not--in
>Miles' case). It's name calling.
>
>The point is that there are probably unobserved characteristics of moms who
>take these kinds of drugs (as opposed to those who don't) that might have as
>much to say about the sexuality of children as the drugs themselves. Let's
>say, for example, that women who take diet pills are also more likely to
>have a psychological or personality type that is itself strongly correlated
>with raising gay kids. If you don't try and control for this
>characterisitic, the drug correlations are completely misleading. What looks
>like the effect of the drugs is really the effect of the personality type,
>or some combination of the two. It's a basic principle of the use of
>statistics.
>
>Christian

The unfounded homophobia charge is getting old and tired and Brian seems to miss the question of why does Miles assume that this study did nothing to control for these things? When I first read the original post I did wonder what other possible influencing factors had been controlled for and how but I did not automatically assume that the researchers were unprofessional enough not to do so or were "dodging" this issue as Miles does. Perhaps Miles read the complete study before he posted? I have not.

John Thornton



More information about the lbo-talk mailing list