A Reply to Sax on the Dangers of Single-Sex Schools
Cherry-picking findings does not make single-sex education effective.
Posted Sep 28, 2011
The article we published last week in Science has received much attention in the press (e.g., NY Times, Washington Post, etc.), and for good reason. Contrary to what Leonard Sax has led many parents, teachers, and principals to believe, large-scale scientific studies have all concluded that single-sex schooling is not more effective than coeducational schooling.
That is not to say that there aren't individual studies (among the thousands that have been conducted) that find advantages. But it is to say that the bulk of the evidence indicates that any positive effects of single-sex schooling found cannot be attributed to the single-sex school environment.
And, the claims about brain and learning differences that are touted in support of single-sex schooling are pseudoscience, in that findings are cherry-picked and misconstrued. Because single-sex schooling has not been found to be beneficial, it does not make sense to use precious public educational funds to create such programming.
In the article, in addition to concluding that there is no evidence in favor of single-sex schooling, we also provided empirical evidence of the potential dangers of single sex-schooling. One danger is that single-sex schooling makes gender a more salient dimension than it already is.
No one denies that gender is a functional dimension used commonly in everyday interactions. Many of the authors of the Science article have long histories of studying gender and its effects. To underscore the potential dangers of single-sex schooling, we discussed findings from a study that showed that when gender was made salient in coed preschool classes, children increased their tendencies to stereotype (Hilliard & Liben, 2010). It was included as only one example. There are others on this topic already in the literature (Bigler, 1995) and we are about to submit a new study for peer scientific review on older children that addresses this issue in single-sex versus coeducational contexts in particular.
In that study, adolescents who had single-sex classes showed greater increases in stereotypes over the course of the school year than children without exposure to single-sex schooling (Pahlke et al., 2011). There are many other studies that demonstrate that when children are assigned to different groups-even as simple as those defined by shirt color and groups that are created by lottery-they develop stereotypes and prejudices, liking their own group better and avoiding the other group (see Bigler and Liben, 2006, for review of some of this work).
The research we discussed about how children are negatively influenced by spending time with same-sex peers was presented by Dr. Sax as suggesting that this is a uniform finding for all children, which of course it is not. In fact, one study that we did not cite because of space limitations (Fabes, Martin, & Hanish, 2003) shows exactly that. Some boys were much more negatively influenced by time spent with same-sex peers and some boys actually showed increased academic and social competence (those boys who were high in self control). These findings suggest that we do not need to institutionalize gender-segregated classes; rather, we need to find ways to structure the coeducational school environment so that all children benefit.
Next, Dr. Sax criticized the age and sample-sizes of the children in the studies, but these studies were well-controlled, data-intensive studies with thousands of observations of children's behavior collected over a relatively long period of time (a school year), and were considered to be ground-breaking studies as shown by their being published in the top journals in the field of child development (Child Development, Developmental Psychology).
Dr. Sax also criticized us for not publishing effect sizes. If this is really the issue, we are happy to report that the effects sizes were in the .60 to .70 range for the Fabes et al., 2003 paper, suggesting moderate to large effects. But we suspect that this is not really the issue for Dr. Sax. In short, there is much research in support of the points made in our paper in Science but limitations in the length of the article prevented us from presenting all of the evidence.
Our point in highlighting these studies here is that, even in preschool, when boys and girls spend time in same-sex peer groups, they generally become more sex-typed in their behavior and attitudes. The fact that this is seen as early as preschool shows that this is the outgrowth of group processes-processes that occur whenever boys and girls spend large amounts of time in same-sex peer groups. Peers are very strong sources of social influence.
Because of these findings and many others like them, we need to be very careful and thoughtful about how we gather and arrange children in our classrooms.
In a new study conducted by one of the graduate students at ASU, we show that it is in same-sex peer settings that differences between boys and girls become pronounced (Galligan et al., 2011). When boys and girls play together, sex differences are minimized. So it is exactly the opposite of what Dr. Sax would have us believe: single-sex settings increase rather than decrease sex-typed responding.
Although Dr. Sax cites the Kessels and Hannover (2008) paper as evidence in support of the idea that single-sex schooling decreases sex-typing (for girls only), a careful reading of that paper reveals that most of their findings are not significant or rely on the use of difference scores which for decades has been criticized as having notoriously serious methodological flaws (Cronbach, 1958; Edwards & Cooper, 1990, and many more).
So, let's consider the bottom line. Dr. Sax has criticized our research and our conclusions, but he has failed to provide systematic body of scientific evidence that contradicts our conclusions. He picks and chooses his attacks without regard for the bigger picture or what the larger array of evidence tells us. He is, in essence, acting as prosecutor, judge, and jury, and that is not how science is conducted or moves forward.
Systematic review of research by qualified scientists (peer review) is one of the most fundamental tenets of science and provides the basis for establishing scientific merit. Claims that are based on false interpretations and personal opinions, however, do not have the power to diminish scientific knowledge.
And, although our arguments are supported by a large body of research (including many studies that were not cited in the Science article due to space limitations), we also have direct experience (contrary to Dr. Sax's claim) with schools that provide single-sex education. For example, Drs. Fabes, Martin, and Hanish are currently conducting a large-scale study of single-sex education that focuses on individual differences in response to single-sex and coed schooling. In another study underway, interviews were conducted with almost 70 principals across the US about their experiences with single-sex schooling. In addition, Dr. Bigler conducted a 3-year longitudinal study of a public single-sex school in Austin, Texas.
In response to Dr. Sax's recent invitation to attend his upcoming conference on gender-based educational practices, we offered to come and present our findings to the audience, but our offer was turned down. Almost 2 years ago, we invited Dr. Sax to give a talk at the biennial Gender Development Research Conference to allow discussion and debate on a panel that included three of the authors of the current paper. But he refused.
It is not surprising that Dr. Sax and his followers are upset with the conclusions of our article. However, our article was critiqued several times by anonymous peer reviewers who are experts in the field and who had diverse views on this topic. Dr. Sax has been criticized in the past for his misuse of science and the data simply do not support his position, including the US Department of Education's own large review of the evidence. Dr. Sax mostly draws upon obscure and misconstrued findings to advocate his position. This is not only unscientific, but it does a tremendous disservice to those who follow his advice.
Dr. Sax says he will publish a number of blogs in an attempt to undermine the credibility of our article. Given that we published our review in one of the most prestigious and respected journals in the field of science, we do not plan to respond to every one of the false claims that he makes. But this does lead us to the question: why is Dr. Sax so anxious to attack our conclusions, regardless of-or even in spite of-the evidence?
In the end, we stick firmly to the conclusions of our article in Science and our recommendations to the Obama administration. We hope the Obama administration sees beyond the rhetoric and pseudoscience in making future educational policy decisions.
Richard A. Fabes, Professor anad Chair, Department of Family and Human Development, Arizona State University
Carol L. Martin
Laura D. Hanish
Rebecca S. Bigler
Diane F. Halpern
Lynn S. Liben