Tales of Two Social Scientists

First tale: A tenured sociologist at a prominent research university, with a couple of books under his belt on related subjects, publishes the first-ever research, using a nationally representative sample, on the young-adult outcomes for kids raised by people who have same-sex romantic relationships. The results show that these young adults have more difficult life experiences across a host of variables involving their employment, education, dependence on public assistance, mental health, relationship success and sexuality, trouble with the law, and experience of abuse.

The study, which confirms that the “gold standard” for the rearing of children is the intact biological family, of married mom and dad staying together faithfully and raising their own offspring, naturally causes a furor. The peer-reviewed social science journal that published the study, along with commentary alongside it, commissions a member of its own editorial board (who has an openly hostile view of the study) to “audit” the peer-review process. He concludes that, as much as he dislikes the article, the journal did everything by the book in publishing it. The journal’s editor, who is likewise friendly to the cause of same-sex marriage, stands by its publication, and in a subsequent issue publishes a follow-up article by its author, who cogently defends and restates his findings.

Meanwhile, at the sociologist’s home institution, a “scientific misconduct” inquiry prompted by a hostile non-scholarly ideologue is undertaken, and the sociologist is cleared completely after a thorough review of his methods and even his correspondence. His double vindication, however—by the journal that published his work and the university where he works—makes little headway against the howling media denunciation of his “debunked” research. (This is a world where “we hate your results” means the same thing as “they’re invalid.”)

Two and a half years later, the same social science journal that published the original study publishes a “re-analysis” of the very same data set (that’s right, no new data collection, just reinterpretation of the admittedly valid stuff). The original study’s creator does one of the peer reviews of the re-analysis and generously green-lights its publication. The new article’s authors claim the original scholar committed “classification errors” because some of the same-sex relationships were very brief, even evanescent affairs, and so what he should have done is what they proceed to do: toss out data until they get a handful of same-sex households where a couple stayed together at least several years. Many of the bad outcomes are washed out in this data-laundering exercise. But, as the original author notes, this is controlling for the pathways, striving to depict as normal something that is almost nonexistent in any meaningful statistical sense: stable, faithful, long-term same-sex couples heading households with children in them and rearing them from early childhood to adulthood.

Still, the media are all over this “re-analysis” as, yes, finally, the “debunking” of the original study that they’ve told us twenty times has already occurred. Of course it is no such thing, and neither were all the previous alleged debunkings, which is why they have to keep seeking do-overs. When the dust clears, what we still know is that kids raised by their own married moms and dads have, on average, the best start in life. But your great-grandma knew that. Hurrah for social science.

Second tale: Defying all the best previous research on how readily people change their opinions, a young PhD student in political science at a top research university teams up with a senior scholar in his field at another top school to publish a brief report in America’s leading scholarly scientific journal that upends everything we thought we knew about the subject.

It seems that when teams of doorstep canvassers are sent out to chat with thousands of residents about same-sex marriage, swift and durable change occurs in the opinions on that issue held by subjects who chatted with a gay canvasser whose script told him to say that he’d really like to get married but the law won’t allow it. Straight canvassers attempting the same persuasion don’t get such great results. Using surveys ahead of time to find the people he wants to persuade, the lead researcher (the grad student) also commissions follow-up surveys after the canvassing visit. For weeks and months afterward, the positive effects of opinion change resulting from the chat with a gay canvasser endure, and even spread to other members of the subjects’ households.

When the study’s results are published, academic commentators and journalists alike are wowed by the finding. Just getting to people on a personal level and delivering a heartfelt persuasive appeal can change their minds. And they’re not just being nice to the person on the doorstep: the follow-up survey shows lasting results! The major newspapers and public radio rejoice at how easy it turns out to be to change public opinion on an issue that many had feared would be fraught with conflict for years to come. Hallelujah!

Other young political scientists, jazzed up by this breakthrough, attempt to “extend” the analysis, but find they can’t. The canvassing appears to have taken place, all right, but the all-important pre-canvassing and post-canvassing survey data? They appear not to exist at all. That is, the data used in the study suspiciously match another data set entirely; the prestigious gay-rights research institute at the grad student’s university says it did not fund any survey effort as he claimed it did; the student admits to having no such funding and to not having paid survey respondents as he claimed; the private firm allegedly employed to collect survey data says it has never heard of him or his study. The grad student tells his chagrined co-author that he “accidentally” deleted all of the original survey data.

The senior scholar who agreed to co-author the grad student’s work asks the journal to retract the article, because he has evidently become convinced that the young fellow faked all the survey data. The young PhD student says he will have answers soon. Meanwhile it looks doubtful that he will be awarded his doctoral degree (his advisor is on the case), or move across the country to the leading university that offered him an assistant professorship (where there are now sober second thoughts about the hire).

Of course I have been talking, in the first story, about the beleaguered but unbowed Mark Regnerus, the sociologist whose New Family Structures Study was published in Social Science Research in 2012. (Full disclosure: Regnerus’s NFSS was funded in principal part, but uninfluenced by, the Witherspoon Institute where I work.) And the second story, which broke this week, is about UCLA political science doctoral student Michael LaCour, whose co-author, Donald Green of Columbia, has asked for the journal Science to retract their much-ballyhooed December 2014 article.

Leave it to the New York Times to find someone willing to claim, ridiculously, that the Regnerus and LaCour cases are fundamentally similar as instances of “debunked” research. Except that in one case we have actual data, validly obtained, rich in their findings, about the interpretation of which the scholars are quarreling. In the other, we have strong reason to believe the data the young scholar claimed to be reporting didn’t exist at all.

The true similarity between the two cases is a rather different one. Regnerus offered the first example of sound social science questioning what the elite in the academy and media desperately want to believe—that same-sex marriage will have absolutely no negative fallout for the young and vulnerable. So of course his research had to be attacked, mischaracterized, or explained away by “re-analysis” at your local dry cleaner. Never mind that his results actually accorded with common sense and historical experience about parents and children.

LaCour, on the other hand, offered those same academic and media elites an astounding reversal of conventional wisdom on public opinion formation, but one that delighted them because it made their work look easier, their future brighter, and their pet cause more imminently triumphant. So of course this was the most exciting breakthrough in social science of the last year!

Until it wasn’t. I wish Michael LaCour well, though I cannot see how his vindication could be possible in this case, or his budding academic career salvaged. Meanwhile Mark Regnerus continues to advance the cause of sound social science.

Matthew J. Franck is Director of the William E. and Carol G. Simon Center on Religion and the Constitution at the Witherspoon Institute.