If you’ve had the “Regnerus study”, which supposedly demonstrated that same-sex couples were not as successful at raising children as their opposite sex counterparts, you will be grateful for this takedown of the study from the Southern Poverty Law Center.
First the background:
In July 2012, the journal Social Science Research published a study by University of Texas at Austin sociology professor Mark Regnerus that seemed to indicate that the children of LGBT parents are more likely to get involved with self-destructive behaviors like using drugs and to suffer from depression than those raised by heterosexual parents, despite many studies that have indicated otherwise. The Regnerus study was immediately trumpeted by anti-gay groups as proof that children are in danger in LGBT households. Just a day after its release, for example, it was cited in an amicus brief by the American College of Pediatricians, a tiny anti-gay breakaway from the main pediatricians’ professional association, that was filed in support of the federal Defense of Marriage Act (DOMA). At press time, the U.S. Supreme Court was weighing arguments on DOMA’s constitutionality.
But the study also engendered serious criticism. More than 200 other sociologists signed a letter to journal editor James Wright that decried the study’s allegedly faulty methodology, and major medical and psychological associations criticized it for similar reasons. Regnerus himself subsequently conceded that his study had problems (key reasons are detailed in the interview below), and he also said his analysis did not conclude that “gay or lesbian parents are inherently bad.” But at the same time, he has continued to defend the study’s results and to push them in anti-LGBT circles. This summer, he is to speak about it to a gathering sponsored by the anti-marriage equality National Organization for Marriage (NOM).
Then an interview with Darren Sherkat, professor of sociology at Southern Illinois University and a member of the editorial board of Social Science Research, was asked by the journal’s editor to “audit of the process of publishing the Regnerus study”:
Let’s get down to the details. What’s wrong with the Regnerus paper?
Regnerus and other right-wing activists have been fond of claiming that the study is “population-based” or a “national probability study.” As a scientist, I don’t even know what “population-based” means, and the data used in this study are by no means a probability sample. Regnerus’ data are from a large number of people recruited through convenience by a marketing firm — they are not a random, representative sample of the American population. Science requires random samples of the population, and that is not how this marketing firm collected their data.
Several scholars also have pointed to incongruities and outlandish values in the Regnerus study, such as people claiming hundreds of sex partners in the prior week. The online collection of data makes the veracity of responses even more problematic. The state of the art in family research would use a random sample of households and follow up with parents and children to see whether or not parental couplings impacted child outcomes — controlling for other potential influences like income, education, ethnicity, relationship stability, and the like.
Isn’t a key criticism also that the study doesn’t actually address children growing up in households of self-identified LGBT parents?
The key measure of gay and lesbian parenting is simply a farce. The study includes a retrospective question asking if people knew if their mother or father had a “romantic” relationship with someone of the same sex when the respondent was under age 18. This measure is problematic on many levels.
Regnerus admits that just two of his respondents were actually raised by a same-sex couple, though I doubt that he can even know that, given his limited data. Since only two respondents were actually raised in gay or lesbian households, this study has absolutely nothing to say about gay parenting outcomes. Indeed, because it is a non-random sample, this study has nothing to say about anything.