MADISON -- As many American public school districts adopt single-sex classrooms and even entire schools, a new study finds scant evidence that they offer educational or social benefits. The study was the largest and most thorough effort to examine the issue to date, says Janet Hyde, a professor of psychology at the University of Wisconsin-Madison.
"We looked at 184 studies, representing the testing of 1.6 million students in grades K-12 from 21 nations, for outcomes related to science and mathematics performance, educational attitudes and aspirations, self-concept and gender stereotyping," says Hyde. "From these, we selected 57 studies that corrected for factors like parental education and economics, which are known to benefit children's school performance."
The study, published in the online Psychological Bulletin Feb. 3, used an analytical technique called meta-analysis, which draws conclusions from multiple studies of an issue. "We are trying to shed some light by putting together studies that applied different methods to different populations," says Hyde. "If you do this right, the whole becomes greater than the sum of its parts."
Hyde's co-authors were Erin Pahlke, who was a postdoctoral fellow at UW-Madison and is now teaching at Whitman College in Washington State, and Carlie Allison, who was a graduate student in psychology.
According to one estimate, thousands of U.S. public schools offered single-sex academic classes during the 2009-10 school year.
As school districts ponder decisions to adopt or suspend single-sex schooling, the scientific literature offers conflicting advice, Hyde says. "The problem is that there are lots of really poor quality studies, but also lots of really good ones, and they've never been separated out before. Especially in the United States, parents who choose single-sex schooling, on average, have more money and more education, which all predict performance. So if you find that the students are performing better, you don't know if it's due to the single sex education or the fact that they started out with these advantages."
Many of the studies focused on math and science, Hyde says. "One claim of single-sex schooling advocates is that, for girls, it will improve math-science performance because they are not mixed with boys who, it's claimed, dominate the classroom. But there is not any advantage, if you look at the controlled studies."
Other assertions made by advocates of single-sex schooling were also unsupported, Hyde says. "The claim that boys do better verbally in single-sex schooling, because they get squelched in a coed setting, did not hold up. And the claim has been made that girls will develop a better self-concept, but again there is no evidence for that."
Data was scarce regarding one disputed area: possible benefits for minority boys, Hyde says. "There has been some thinking that this would help ethnic minority boys, but we did not find enough studies covering that topic. We urgently need high-quality study of these programs that make careful comparisons with coed schooling, comparing students with equal resources, to see if the single-sex configuration really makes a difference."
If single-sex schooling does not have demonstrable benefits, it does have downsides, Hyde says. "There is a mountain of research in social psychology showing that segregation by race or gender feeds stereotypes, and that's not what we want. The adult world is an integrated world, in the workplace and in the family, and the best thing we can do is provide that environment for children in school as we prepare them for adulthood."
On a practical level, Hyde adds that single-sex schooling is "terrifically difficult and expensive. If you have a single-sex 8th grade math class for girls, you need another for boys, and a third that's coed. Public schools have better places to put their money."
The study has other policy implications, Hyde says. "Federal regulations permit single-sex schooling in public schools only if there is a compelling educational interest. The kids would have to perform better and the evidence does not show that they do."
--David Tenenbaum, 608-265-8549, email@example.com