Analysis: New Study Did Not Prove That Gay Parents Are Better
Several liberal media organizations are reporting the results of a new same-sex parenting study which suggests that gay parents do a better job of raising children than the general population. There are four imporant points to understand about that study, however.
Here are a few of the headlines:
CBS News: "Children of same-sex couples healthy, well-adjusted, study finds"
NBC News: "Children of Same-Sex Parents Are Healthier: Study"
The Huffington Post: "Children Of Gay Parents Are Happier And Healthier Than Their Peers, New Study Finds"
Vox: "Largest-ever study of same-sex couples' kids finds they're better off than other children"
The study, though, does not warrant the conclusions suggested by those titles.
"Parent-reported measures of child health and wellbeing in same-sex parent families: a cross-sectional survey," by lead author Dr. Simon R. Crouch at The University of Melbourne in Australia, was published June 21 by the journal BMC Public Health. The co-authors were Elizabeth Waters, Ruth McNair, Jennifer Power and Elise Davis. Power is affiliated with La Trobe University. The rest of the authors are at The University of Melbourne.
The study found that children of same-sex parents scored higher on measures of general behavior, general health and family cohesion than the general population of Australia. The study also measured how often the parents felt stigmatized for being gay. A high number of stigmas was negatively correlated with measures of the children's physical activity, mental health and family cohesion.
Here are four important points to understand about the study:
1) The study did not use a random sample.
To make a generalizable conclusion about a population, scientific studies need a large, probability sample of the population, sometimes called "random sample" or "representative sample." A probability sample means that those surveyed are representative of the general population.
The Crouch study was based upon a convenience sample, or non-probability sample. Participants for the study were recruited through gay and lesbian community email lists and ads posted in gay and lesbian press. This means that the participants volunteered for the study and were not randomly chosen from the population.
The sample had 315 parents of 500 children. Most of the children, 80 percent, had a female parent complete the survey. Eighteen percent had a male parent, while the remaining parents described themselves as "other gendered."
As stated in the study: "Every effort was made to recruit a representative sample, and from the limited data available about same-sex parent families it appears that the [study's] sample does reflect the general context of these families in contemporary Australia."
Convenience samples can be an important research tool when probability samples are difficult to achieve. They can also help researchers design better studies and help them resolve issues with their research before conducting large scale studies. Social scientists understand, however, that conclusions about a general population should not be drawn based upon a convenience sample.
2) The study did not compare same-sex parents to biological parents.
Previous studies have shown that kids do best when they are raised by their biological parents and those parents are married. The Crouch study, however, compares its convenience sample of children raised by same-sex parents to the general population, which includes those raised by single parents, step parents, foster parents and other same-sex parents.
The study cannot conclude, therefore, that children raised by gay parents have better or worse outcomes than children raised in two-parent heterosexual households.
3) The study relies upon parent-reported outcomes.
The health and well-being of the children are based upon what the parents say they are. While these measures are being compared to other parent-reported measures, there are reasons that gay and lesbian parents might overstate their outcomes at a greater rate than the general population.
The survey was conducted while Australia is debating redefining marriage to include same-sex couples. Part of that debate deals with child-rearing. Government recognition of marriage should only be for a man and woman, proponents of traditional marriage argue, because this arrangement is best suited for the raising of children, which is a public good.
It is in the interests of gay marriage supporters, therefore, to show that gay couples can raise children just as well as straight couples. The gays and lesbians who volunteered to participate in the Crouch study likely understood the significance of the study. As a result, they may have inflated their results more than the average parent. Additionally, gays and lesbians who are raising children with poor outcomes may have been reluctant to participate in the study for similar reasons.
4) Studies using probability samples show poor outcomes for gay parents.
Two recent studies that did use probability samples showed some poor outcomes for children of gays and lesbians.
The New Family Structures Study at the University of Texas led by sociologist Mark Regnerus found, for instance, that those who reported that at least one parent had a same-sex relationship had poor outcomes along a range of variables. They were, for instance more likely to be depressed, unemployed, have more sex partners and report negative impressions of their childhood.
A study published last December by economist Douglas W. Allen looked at a 20 percent sample of the Canadian census and found that children from gay and lesbian families were less likely to graduate from high school than children raised by opposite sex couples and single parents.
The issue of gay parenting in highly politicized. In such an environment, liberal media tend to exaggerate the results of those studies that appear to confirm their biases and write hyper-critically about the studies showing different results. Conservative media have similarly focused more on reporting the research that confirms their biases.
There are some significant differences, though, between how Allen and Regnerus are presenting their findings compared to Crouch and other social scientists who say there are no differences between gay and straight parents. Unlike the "no differences" social scientists, Allen and Regnerus do not argue that their studies are conclusive.
Gay parenting is difficult to study because it is so new. In the history of human civilization, gay parenting has only recently become culturally accepted. To understand the effects on the children they raise, social scientists need more and larger samples and time — time for the kids raised by gays and lesbians to grow up and have outcomes that can be measured and compared to those raised by other family types. Allen and Regnerus point this out in their research and other reports.
For the time being, research has shown that biological, two-parent households provide, on average, the best outcomes for children compared to all other family types. Additional research has demonstrated the unique contributions of mothers and fathers to child development. (One study, for instance, found that fatherlessness harms the brain.) These studies should be sufficient to at least raise suspicion of the studies suggesting that kids raised by parents of the same gender have the same, or better outcomes as kids raised by both a mom and a dad.
The social scientists reporting "no differences," on the other hand, make sweeping generalizations based only upon their small, non-random samples that confirm their liberal biases. Liberal media uncritically follow them.
Some of Regnerus' liberal critics have also argued that his findings should be ignored because he is a conservative Catholic. Crouch, though, is a gay man raising two kids with his partner. Would these same critics suggest that Crouch's study should be ignored because Crouch is personally invested in the results?