Junk science in the courtroom

Nothing has brought home the dangers of having "junk science" presented as evidence in our courts of law than the current case in Michigan where the state is defending its ban on gay marriage.   The particular point here is that the state has based much of its case on the claim that children are harmed by having gay parents.

The star "expert witness" for the state was Sociologist Mark Regnerus, who has become the darling of the anti-gay forces -- religious conservatives, National Organization for Marriage, Family Research Council -- you know, that ilk.

The outrageous thing is that the Regnerus study has been thoroughly discredited by reputable scientists.  In response, the journal that published the Regnerus study had an outside group audit their paper review process.  They concluded that it should never have been published because of serious medthodological flaws, because its claims are not supported by the data, and because the author failed to honestly disclose the connections to funding from the Witherspoon Institute, a religious conservative organization.

Not only did Witherspoon give Regnerus a grant to do the study, but they advised him on what they were looking for and consulted with him during the study.  In addition, all three of the peer reviewers who recommended publication to the journal have connections with Witherspoon.

In short, it was a study designed to get a certain result, paid for and directed by a religious conservative organization, which also seems to have had a special inside track to get the report published in the jiournal Social Science Research.  The head of the audit committee told The Chronicle of Higher Education (the academic world's chief news publication) that Regnerus' study was "bullshit."

This all happened in 2012-13.  And yet, in 2014 the state of Michigan had Regnerus as its star witness in a most important case.    Why do they do this?    Because it is the only recent study that supposedly shows what the anti-gay marriage group wants to hear.  The over-whelming evidence is hundreds of studies that show same sex couples are just as good as parents as opposite-sex parents.

Here's the important thing about this Michigan use of Regnerus, however.   In all previous cases, his study has been cited in briefs filed with the case.   In Michigan, Regnerus himself took the stand and was cross-examined by an attorney who understood that it was junk.

Now, just to give a brief summary of the issue:  Regnerus claims that he evaluated the long-term emotional well-being of children of "same-sex parents" and found that they did not do as well as children reared by their biological parents.   But in fact, he did not evaluate children raised by same-sex parents.   Out of his study sample of 248, he found only two who had been raised by a stable same-sex couple.    All the rest of his sample were products of male-female couples in which one parent also had same-sex experiences that the kids knew about.

What this means is that Regnerus was comparing the effect of stable families with the effect of broken or troubled families.    It says nothing about the relative merits of stable same-sex couples as parents vs stable opposite-sex couples as parents, because he didn't have the same-sex couples to study.  In effect, he is comparing apples and oranges.

Under cross-examination, Regnerus acknowledged that his study is not conclusive and could not really make the claim that he does.   But he still believes what he believes, and he thinks that it would be shown by a study that did include same-sex parents.

Now that his study has been aired in court by a knowledgeable attorney, what really matters is what the judge believes.  Fortunately, it doesn't take a scientist to know that this is junk . . .  or, in the words of the journal audit committee's chairman,  "bullshit."

Ralph