How not to do research...

Given the list of complaints you find people making about biased studies on subjects, it doesn't seem that conservatives necessarily do better. Yesterday I came across an article entitled Study suggests risks from same-sex parenting.

The first question raised in my mind was whether the article's basis for comparison was reasonable, as their list of negative outcomes includes more children "currently cohabiting" and fewer "currently married" vs. those raised by married, biological parents. My immediate thought: how would that change if they compared the kids in those categories vs. those raised by cohabiting, divorced, or single parents. (If you were to look for the phrase "same-sex marriage" in the news in the past while you might find a reference or two... or 59,100 [Google News] - most of which has to do with its legality or lack thereof).

That was the first red flag, but if the LA Times is correct, the study is far more dubious:

his study does not actually compare children raised by same-sex couples with those raised by different-sex couples. The criterion it uses is whether a parent 'ever ha[d] a romantic relationship with someone of the same sex.' In fact, only a small proportion of its sample spent more than a few years living in a household headed by a same-sex couple. Indeed, the study acknowledges that what it's really comparing with heterosexual families is not families headed by a same-sex couple but households in which parents broke up. 'A failed heterosexual union,' Regnerus writes in the study, 'is clearly the modal method' — the most common characteristic for the group that he lumps in with same-sex-headed households.

The LA Times article then goes on to say the following:

The trouble is that no scholarly research, including the Regnerus paper, has ever compared children of stable same-sex couples to children of stable different-sex couples, in part because an adequate sample size is hard to come by.

Based on such a claim it would seem that there's also inadequate research to substantiate the claim ("the 20-year 'consensus'") made in the LA Times article that "two parents are better than one, not that parents have to be different genders." If the lack of scholarly research in support of a hypothesis applies to the Regnerus case, it would also seem to apply to such a claim.