Same-sex Survey Bias Addressed

, Richard Thompson, Leave a comment

In examining the effects that same-sex relationships have on children, social scientists have too often employed speculation and shoddy statistics, but a recently published study shows there is reason for uncertainty.

“The scholarly discourse concerning gay and lesbian parenting has increasingly posed a challenge to previous assumptions about the supposed benefits of being raised in biologically-intact, two-parent heterosexual households,” University of Texas associate professor of sociology Mark Regnerus wrote in his New Family Structures Study (NFSS).

Regnerus noted differences that included higher levels of criminality, lower levels of employment, poorer relationship quality with one’s current partner, and poorer mental and physical health.

This study, published by Social Science Research, shows a distinct difference between children raised by heterosexual parents and those raised by homosexual couples. From a sampling of approximately 3,000 Americans aged 18 to 39, the complied data compared how children raised in eight different family structures fared on 40 outcomes such as overall happiness, levels of household income, and involvement in criminal activity.

Regnerus found that, when compared with those who spent their entire childhood with both biological parents, the children of mothers who had same-sex relationships were different on 25 of the 40 outcomes tests while those whose fathers engaged in same-sex relationships were different on 19 outcomes.

“Why such dramatic differences? I can only speculate, since the data are not poised to pinpoint causes. One notable theme among the adult children of same-sex parents, however, is household instability, and plenty of it,” Regnerus wrote in his Slate-published article “Queer as Folk.”

In conducting his study, Regnerus set out to refute the “no differences” theory, the common misconception that children of gay parents are not significantly different from the children of heterosexual parents. Regnerus’ problem with this theory was the skewed sampling size used by many of the studies that support it.

“Many published studies on the children of same-sex parents collect data from ‘snowball’ or convenience samples,” Regnerus claims. “This sampling approach is a problem when the goal is to generalize to a population. The empirical claim that no notable differences exist must go.”

These “snowball” samples were used in social experiments such as the National Longitudinal Lesbian Family Study (NLLFS), which recruited its participants in women’s bookstores, in lesbian newspapers, and at lesbian events. Through the use of these inaccurate samples, such studies as the NLLFS became somehow qualified to make the claim that the children of lesbian mothers fared just as well, if not better, than children with heterosexual parents.

Regnerus was able to avoid the use of “snowball” samples by conducting his data collection through a research group known as Knowledge Networks, who conducted a panel that gathered random and representative data.

“The Knowledge Panel® is based on a sampling frame which includes both listed and unlisted numbers, those without a landline telephone and is not limited to current Internet users or computer owners, and does not accept self-selected volunteers. As a result, it is a random, nationally-representative sample of the American population,” said Regnerus.

While Regnerus’ study is in no way definitive, it offers a refreshing departure from non-representative studies and introduces some interesting statistics that must not go unchecked.

Richard Thompson is an intern at the American Journalism Center, a training program run by Accuracy in Media and Accuracy in Academia.

If you would like to comment on this article, e-mail