Ethics Complaint Filed Against Regnerus


Scott Rose has filed an ethics complaint against Mark Regnerus, the author of the highly dishonest “study” on gay parents, and written a letter to the University of Texas detailing the ethical and scientific problems with that study. You can read the full letter here. A sample:

2)  UTA Professor Mark D. Regnerus alleges that he carried out a study comparing a)young adult children of “intact biological families,” with b) young adult children raised by homosexual parents up to the 1990s. However, to find study subjects, Regnerus worked through the company Knowledge Networks, which has a limited list of potential survey respondents from the general public. Regnerus had prior knowledge that Knowledge Networks would not be able to connect him with an adequate sampling of young adult children of gay parents raised by those gay parents up to the 1990s. Regnerus additionally knew that companies other than Knowledge Networks exist, which companies, with far greater likelihood would have been able to find him the class of study subjects he purported to want to compare to young adult children raised by “intact biological families.” Regnerus fraudulently classed as a present-day young adult raised by a “gay” parent up until the 1990s, anybody from Knowledge Networks list who said that their parent had ever had a “romantic relationship” with a same-sex partner. In other words, a study subject’s parent could have had a one-night stand with a same-sex partner, and Regnerus would count that person as having been raised by a homosexual parent. Regnerus told Deciutus that he “assumed” that if the person knew of their parent’s same-sex “romantic relationship,” that it would have been “substantial.”

Is that how scientists work, by “assuming” and not verifying?  Please note that one of Regnerus’s study conclusions is that family stability is good for children. That was hardly news. Yet, Regnerus set up his screening questionnaire, such that those people he surveyed and counted as having been raised by a gay parent, were virtually certain to have experienced family instability, as they knew of a parent having a same-sex affair…

In fact, in his written study, he wrote that the idea that children of gay parents do not have worse outcomes than those of heterosexual parents “must go.”  In sum, Regnerus dishonestly stacked the deck against gay parents, in a way favorable to the results his funders had paid him to produce. Nor was that the only way that Regnerus stacked the deck against his faux gay parents. Regnerus surveyed a disproportional number of children from broken African-American and Hispanic households, likely to be non-affluent families. He subsequently attributed to gay parents bad child outcomes that in reality were due to lack of financial resources. As an example, if a survey respondent said they are currently on public assistance, that counted as a “bad” outcome. Yet, unemployment at the time of the survey was such that there was, in the economy overall, one job opening for every five job seekers; we know that the ratio is worse among the poor.

There’s much more.

Comments

  1. Azkyroth, Former Growing Toaster Oven says

    Cue concern trolling about how we should care that any action taken in response to this will be seized on by the Reichwing as an excuse to feel persecuted in 3, 2, 1…

  2. jaxkayaker says

    That’s going to be an uphill battle.

    Was the study published in a peer-reviewed journal? If the study design is that flawed, that’s a big failure on the journal’s part. A rebuttal article or retraction might be a more achievable goal.

  3. abb3w says

    I’ll note that the problem does not appear intrinsic to the use of Knowledge Networks, per se; but rather, that kids raised by gay couples are an incredibly small fraction of the population, such that even the Knowledge Networks participant pool not a large enough population to be an oversample that allows filtering a statistically useful sample size of the target population. Knowledge Networks seems to be relatively adequate for studying modest sized sub-populations down to the ~1% level. But this is a group better measured in ppm than percent.

    I also don’t think other companies would have done much better. Oh, sure, if you threw a mountain of money at them, Gallup would cheerfully have set out ringing phones and even doorbells to acquire the monstrous oversample needed to have a sample of ppm-needles. But I think they would have had to systematically call almost every phone in the country in order to get it… which leads to quite a phone bill.

    I suppose some controls could have been done, to improve handling of the data scraped up off the floor. Comparing kids where a parent had a same-sex affair to those where a parent had an opposite-sex affair; comparing kids where the parents divorced and ended up in a stable same-sex partnership to kids where the parents divorced and ended up in a stable opposite-sex partnership. But even that probably would have left a statistically weak sample pool; and with comparisons dubiously helpful for figuring out what would happen with same-sex couples starting out together and trying to raise kids.

    And there’s also the question, are the difficulties due to the family structure, or the society they’re embedded in? Anyone able to turn up studies of mixed race marriages from the 1950s?

  4. Doug Little says

    I want to know how he could just gloss over the wealth issue. Surely if your study contained a disproportionate number of poor on one side it would be difficult to discriminate between the effects of wealth and what ever else you are studying.

    Fundies, getting it wrong since the enlightenment.

  5. eric says

    I’ll note that the problem does not appear intrinsic to the use of Knowledge Networks, per se; but rather, that kids raised by gay couples are an incredibly small fraction of the population, such that even the Knowledge Networks participant pool not a large enough population to be an oversample that allows filtering a statistically useful sample size of the target population.

    Yep. Just to add to that, this is probably one of those areas for which Steve Gould would’ve said “its the distribution, stupid.” A wide and significant overlap in the range of outcomes (between straight and gay parents) is much, much more important than any hypothetical small difference in average outcome.

  6. Aliasalpha says

    I once said no when offered a glass of milk, by his ‘standards’ that probably makes me lactose intolerant

  7. vmanis1 says

    On the subject of whether the paper was properly peer-reviewed, see some of the critiques that have been posted at Box Turtle Bulletin. One factoid that sticks in my mind is that the time between submission and publication was much shorter for Regnerus’s paper than for other papers published at similar times in the same journal.

    This does seem unusual.

  8. iknklast says

    “family stability is good for children.”

    Of course, the big problem is how they define family stability, which tends to be defined as two parents, one male and one female, who stay together through thick and thin. Oh, and it’s best if the mother is in the home, and the father works.

    Big fail. My two parents, one male and one female, stayed together, mom in the home, but could hardly be called stable. I’m sure there are others out there, though I have only anecdotal evidence. If there is a truly stable, loving household, it shouldn’t matter if the parents are the same sex, the same race, or the same religion. It shouldn’t even matter if there are two parents, if the household is truly stable and loving. Many traditional families are so unstable that a single gust of wind was enough to start a major conflagration, but for some reason, this particular thing seems to be overlooked in most of the studies.

    Not to mention, there are many problems that can occur as a result of societal pressures. My son, while living with his father who had a live-in boyfriend, might have done much better if it wasn’t for all the kids at school and their nonsense. My son didn’t see it as a problem; his “friends” did. Not to mention, the religious wingnuts on his mother’s side of the family (yes, my family is a freak show; I don’t have much to do with them anymore).

    Until we can separate the family from the society it exists in, any such study is probably going to be picking up a lot of problems and possibly interpreting them wrong.

  9. birgerjohansson says

    “Regnerus” sounds like he had Scandinavian ancestors. Once again, I apologise for the trouble our DNA is causing you Americans (why do so many right-wing nutters have Scandinavian names?).

Leave a Reply