Bad LGBT Parents? Bad UT study, audit finds

Below are excerpts from an interview with Sherkat this week
So you told the Chronicle of Higher Education Regnerus's study was “shit.”
I guess that's an unfortunate quotation, because following that I went into a very detailed statement on what are the theoretical, methodological and analytic expectations of a paper published in Social Science Research, which is one of the top journals in the world for social sciences. Most papers that are submitted are not accepted. And so how does a paper like this get into Social Science Research? The Regnerus paper was inadequate, grossly inadequate, in terms of its data, particularly given the small distribution of supposed gays and lesbians – there's only 1.7 percent of them in the overall sample taken. There are some issues and problems with the sample that, in my opinion, should have disqualified any study that used any data related to it. It doesn't matter if it's about gays and lesbians or about whether you own a gun, or whatever you might investigate with that data.

Next is the issue of his measures, measurement being another big thing in the social sciences. How did you measure, how did you conceptualize gays and lesbians? And there Mark was completely disingenuous. It was very disturbing and deceptive in how he labeled his tables, how he contracted some kind of measure of what he called “gay” or what he called “lesbian.” That measurement issue, even if he had perfect data, would have completely disqualified this paper. I mean, this is a top, top journal. You don't publish shit in a top journal. And this was shit.”

Your main question is an important one: how does something like this get past the peer-review process in a journal like SSR. You say the peer-review process here failed. Should we worry about peer-reviewed studies?
This is a problem with all of the sciences. And another caveat on that is that normatively in sociology we don't retract papers. I've had exchanges with people in biology and some of the other hard sciences where retraction is normative. But usually in those disciplines it's being retracted because someone deliberately made something up. Here, not that the study wasn't bad, but he didn't really make anything up. It's just not high enough quality to be published.

So a lot of people are like, “Well, why wouldn't this be retracted?” Well, that's not the way we do things in sociology. The process will weigh it out. People will respond – and they have.

This is not what usually happens. Peer review is not perfect, but peer review is good. And so, yeah. You can usually trust things. But let's not act like just because something's peer reviewed that it's somehow unassailable. And that's especially true in the social sciences.
On one hand, you don't want to say, “Oh, it's peer reviewed, so it's perfect.” On the other, you don't want to say peer-review is shit because Regnerus' paper got in. It's somewhere in between and it really has to be judged case by case. And in this case, peer-review didn't work well.

> Email Michael Barajas