This is getting uglier.
Prof. Mark Regnerus has been giving interviews about his study on parents who’ve had same-sex relationships, saying things like this:
Well, in the generation that are adults now, kids raised in a same-sex household were more likely to experience instability and shifting household arrangements. For example, 14 percent of kids whose moms had a lesbian relationship reported spending more time in foster care, well above the average of 2 percent among all respondents.
That leapt out at me because the error is obvious: The second sentence in no way supports the first. Children whose “moms had a lesbian relationship” weren’t necessarily “raised in a same-sex household” — the children might have never even met their mother’s lesbian partner, much less have been raised by her. Jim Burroway has done some great work pointing this out, and I’d like to extend it. In fact, I’d like to go so far as to show that Regnerus himself admits that he has, well, nothing.
Regnerus’s team interviewed 15,058 people. Few of them had a gay parent; even fewer lived with their gay parent’s partner for a significant time; and fewer still came from what Regnerus calls a “‘planned’ gay family.”
|Respondents who:||“Lesbian” mother||“Gay” father|
|Had a parent in a same-sex relationship||175||73|
|Lived with parent’s same-sex partner more than 3 years||40||1|
|Came from “planned” gay families (estimated)||30 – 45||less than 1|
A couple points:
- Regnerus is fond of talking about “lesbian mothers” and “gay fathers,” but he defines them as adults who have ever had a same-sex romantic relationship, even if it only happened once, even if it only lasted a few days.
- Regnerus has no data on “planned gay families.” He derived those numbers from looking at “respondents who claimed that (1) their biological parents were never married or lived together, and that (2) they never lived with a parental opposite-sex partner or with their biological father.” The numbers are a guess.
Back to those numbers, though. Regnerus obviously can’t draw any conclusions male same-sex parenting based on a sample of less than 1. How about lesbian same-sex parenting? Is his sample of 30-45 respondents enough to significantly describe the broader population?
Here’s the kicker: Regnerus agrees with me. His article bemoans the low sample sizes of studies that offered up good results for same-sex parenting:
It is not surprising that statistically-significant differences would not emerge in studies employing as few as 18 or 33 or 44 cases of respondents with same-sex parents, respectively…Even analyzing matched samples, as a variety of studies have done, fails to mitigate the challenge of locating statistically-significant differences when the sample size is small.
Look at the numbers in that quote. Now look back at the numbers in the table. This is Regnerus telling us he’s got, as I said, nothing.
Now here’s why this is so ugly.
- In the study’s introduction, Regnerus frames it as an examination of same-parenting and a corrective to flaws in earlier, positive studies on same-sex parenting.
- But Regnerus’s data on same-sex parenting contains the same sample-size flaws for which he which criticized those other studies.
- So once he leaves his introduction and enters analysis, he abandons all pretense of studying same-sex parenting and focuses instead on parents who have ever had a same-sex romantic relationship, regardless of whether they raised a child with that same-sex partner.
- Nevertheless, he does not correct his introduction in order to frame the issue properly.
- And finally, he grants interviews to conservative outlets, claiming that his study shows the harm of same-sex parenting, even though his own words, in his own study, demonstrate that he knows his sample size is just too damn small to say anything with confidence.
Am I wrong to call this ugly? Prof. Regnerus could well be following this blog, given Jim’s excellent and well-publicized work. I hope the professor provides us an explanation and justification for what he’s telling the press.