There's a new study making headlines over at AOL news. Walter Schumm, billed as a professor of family studies at Kansas State University, did a "meta-analysis" showing that children of GLBT parents are 1.7 to 12.1 times more likely to be GLBT themselves. AOL News assures us that this is "robust research" with "a detailed analysis that past almost aggressively ideological researchers never had."
Gosh, "meta-analysis" is such a big word! And his numbers have decimals in them! It must be really scientific.
People are more easily suckered by crap with numbers in it than by garden-variety crap.
Schumm's method here is similar to Paul Cameron's notorious obituary study, which has been debunked a million times and is still treated as truth on the homophobic right. Cameron also did a "meta-analysis" of children of GLBT parents, which was pretty much identical to Schumm's, and even using the same books. The AOL article hastens to assure us that Schumm's meta-analysis is much more scientific because it used ten books instead of three.
An actual meta-analysis means taking several studies with small samples and analyzing the data as a whole, in the hopes that any outliers will be mitigated by the use of multiple studies. Box Turtle Bulletin gives us a rundown of what Schumm actually did. He took ten books which were not studies, but popular-audience books about the children of GLBT parents. Schumm read them, tallied up the number of GLBT parents & children, and concluded that these were reliable numbers for the chances of GLBT parents having straight or GLBT children.
The books were not, of course, any sort of random sampling of children of GLBT parents. At least one of the authors, Abigail Gardner, stated:
In fact, I had made a point of having a roughly even number of straight kids and second generation (gay, bisexual or transgender) kids so that both views would be evenly represented in the book. In other words, because of the goals of my book, I deliberately aimed to have 50% of the kids interviewed to be queer. Not because it is statistically reflective of the population, but to give it balance of perspective.
Gardner said this in 2006, after Paul Cameron used (or rather misused) her book for his own study. Which means Schumm was well aware that he was committing scientific fraud.
I could do the same meta-analysis by reading seven Harry Potter books and tallying up how many children turn out to be wizards and witches.
Of course, there would be nothing intrinsically wrong with GLBT (or straight) parents having GLBT children, and it would lead to some interesting questions about genetics, among other things. But Schumm makes clear that he's pushing a narrative of choice and "recruitment:"
Schumm also finds evidence of gay mothers pushing their daughters, upset over a relationship with a man, to "try out women." But couldn't gay men also tell their sons this? Yes, but Schumm tells AOL News that most gay men have at some point been with a woman, so they understand why their sons might date them. Whereas the literature shows some lesbians "have a hatred of men that's intense," Schumm says.
Nope, no bias there, nothing but scientific rigor.
Who is Walter Schumm? He was one of two "expert" witnesses paid by the state of Florida to defend its ban on GLBT adoptions. (The other was George "whatever lifts your luggage" Rekers). Schumm came off badly in doing so, and was cornered into admitting that
"Gay parents can be good foster parents," and "The decision to permit homosexuals to adopt is best made by the judiciary on a case by case basis."
The AOL article quotes this, but neglects to mention that he also claimed that GLBT parents were almost 5 times as likely to have GLBT children (seems to be a theme with him, doesn't it?), and that this "statistic" wound up getting shredded by an actual expert.
Schumm tells AOL News that he agreed to testify as one of the state's witnesses only if his evidence was not "slanted" for or against gay rights.
And if that's what he told AOL news, it must be true - never mind that he was recruited and paid by the State of Florida for the explicit purpose of defending the adoption ban.
And in 2007, Schumm was chosen as one of the board members for a planned online publication, The Empirical Journal of Same-Sex Sexual Behavior, created by - small world, isn't it? - Paul Cameron and George Rekers.
Expect to see this meta-nonsense become meta-popular on the right. Forewarned is forearmed.