No data? No problem

Yeah, comments about that flawed study of children of gay parents are still filling gay blogs. John Corvino, known as the Gay Moralist, summarizes the flaw this way:

Question: What do the following all have in common?

A heterosexually married female prostitute who on rare occasion services women

A long-term gay couple who adopt special-needs children

A never-married straight male prison inmate who sometimes seeks sexual release with other male inmates

A woman who comes out of the closet, divorces her husband, and has a same-sex relationship at age 55, after her children are grown

Ted Haggard, the disgraced evangelical pastor who was caught having drug fueled-trysts with a male prostitute over a period of several years

A lesbian who conceives via donor insemination and raises several children with her long-term female partner

Give up? The answer—assuming that they all have biological or adopted adult children between the ages of 18 and 39—is that they would all be counted as “Lesbian Mothers” or “Gay Fathers” in Mark Regnerus’s new study, “How different are the adult children of parents who have same-sex relationships? Findings from the New Family Structures Study” (NFSS).

Jim Burroway of Box Turtle Bulletin (the guy who wrote the analysis I discussed earlier) reports an email discussion between the paper's author and Straight Grandmother (also discussed earlier). The author tries to explain why he had to combine all those categories into one. Too few values in a category and the statistics don't play nice. Not enough money to make the category bigger. Straight Grandmother isn't buying. At least the author should have properly identified the category he had, not what he wanted it to be but no longer was.

Rob Tisinai (who contributes to BTB but I found this post on his personal site) looks at that issue of too few values in a category. So, how many are there? First of all, Tisinai admits it's a guess. Regnerus doesn't break it out. Let's work with that guess. Out of 15,058 people randomly selected for the study, the number raised by an intact lesbian couple for the duration of the child at home is perhaps 30-45. The number of kids raised by an intact gay couple is maybe 1. Yup, sample sizes too small for any type of analysis. Remember, these study subjects were born 2 to 3 decades ago.

Which leaves Tisinai with an ugly riddle: Regnerus makes a claim in his introduction about what he intends to prove. He then admits he doesn't have the numbers to back up that claim. Why then is he granting interviews with conservative news outlets claiming he has proven something while at the same time admitting he doesn't have the numbers to back up that claim?

Timothy Kincaid, also of BTB, thinks he knows what Regnerus is really up to. The control group in the study was Intact Biological Families, in which the parents are still married to each other when the child leaves home and the child is biologically related to both parents. Kincaid says this is not only the control group but the author's ideal family. His goal is to show IBFs are superior, so it doesn't matter that all other categories are sloppily lumped together. Put another way, showing gay families are inferior is only one step along the way to claim IBFs are superior.