The Single-Sex Trick

The flaws in a new survey that praises girls-only and boys-only classes.

It sounds like great news! Children and teachers like their all-boys and all-girls classrooms, according to a recent survey of some 7,000 students by the South Carolina Department of Education, the nation’s leader in single-gender public schooling. In the survey—which has gotten plenty of media attention —76 percent of children between kindergarten and ninth grade reported increased self-confidence since they began participating in single-gender academic classes. They also rated more highly their chances of completing high school. The cheering was extra loud among girls and younger students: A whopping 93 percent of girls in kindergarten through second grade reported that their motivation had increased, compared with 58 percent of sixth-through-ninth-grade boys.

But let’s not mistake students’ opinions for evidence that separating boys and girls can close gender gaps in achievement—or even that it is in their best interest. These aren’t questions children can answer themselves. And this survey is especially misleading because of its serious methodological flaws.

The biggest problem with the survey is that the state Single-Gender Initiatives program that administered it—and which fervently advocates single-sex education—did not give students the option of answering “no change” when asking how their current attitudes compared with their feelings before they moved to single-sex classrooms. Forced to choose between “increase” and “decrease” for questions about their motivation and confidence since switching from coeducation, the students were more likely to pick the positive option. The younger students’ glowing answers to every question in particular reflect the instinctive desire to please adults in many kids of their age group. By contrast, about 42 percent of middle-school-aged boys responded that their motivation had decreased since switching to single-gender classrooms—a finding that is easily lost in the way the data are presented.

The survey is also stacked in favor of single-sex classes because it queried only children who are currently enrolled in them. By law, these are kids who deliberately chose single-sex classes over a coed option, as the new Title IX regulation permitting the expansion of gender-segregated instruction requires that enrollment be purely voluntary. Asking such children if they’re happy with their choice is a little like Pepsi asking cola drinkers who recently switched from Coke which soda they prefer. In other words, the survey lacks a control group. As far as we know, children in coeducational classes are equally or more confident and motivated as children in single-sex classes, but there is no way to know because they were not surveyed.

There’s another problem, too—the Hawthorne effect. As many educators know, this is a tendency to find improved results in schools simply by telling students and teachers that you are implementing and studying a change. In medical research, we have placebo-controlled drug trials, where it is common to find that a placebo produces nearly as much improvement as an active medication. The fact is, people want to believe in a new treatment or educational method, and their belief goes a long way toward improving things. While it’s great to hear that students feel good about themselves, we shouldn’t pretend that this survey proves that single-gender instruction is the reason.

With all due respect to self-confidence, what really matters is whether children are truly learning better in single-gender classrooms—and South Carolina’s survey provides no proof here. For a better answer we can turn to a detailed 2005 report by the U.S. Department of Education, which found little evidence of any difference in outcomes between gender-segregated and coed classrooms or schools. Some studies even report negative effects in single-sex settings, including more stereotyping by gender role, more behavioral problems, and lower academic achievement. Unlike the flawed South Carolina survey, the DoE report restricted its analysis to rigorous, scientifically valid data culled from more than 2,000 independent studies.

Communities are increasingly turning to single-sex education because they despair about the quality of their local public schools. But all-boys and all-girls instruction is a misguided fad, not a genuine solution for underachieving students, who would be better served with more educated, higher-performing teachers and smaller class sizes. School should be a place where children learn to respect and appreciate each other regardless of gender, just as they will need to do in their future families and workplaces.

Like  Slate on Facebook. Follow  us on Twitter.