Girl-Dominated Classrooms Can Improve Boys’ Early School Performance

The No Child Left Behind Act has set strict annual performance accountability goals for educators, while giving schools unprecedented freedom to experiment with new models to promote achievement. Limiting classroom size has been a frequent focus of efforts, given the sizable research on the benefits of smaller classes. Other models have worked to narrow the achievement gap between elementary school boys and girls. In “Resource and Peer Impacts on Girls’ Academic Achievement: Evidence from a Randomized Experiment,”(1) Diane Whitmore Schanzenbach exploits a large data set that randomly assigned youth to small or large classes to test whether boys in predominantly female classrooms performed better than boys in more mixed classrooms. Her findings suggest that there are early benefits to boys, which dissipate after third grade—perhaps because of changing classroom dynamics and academic shifts or changing peer pressures. The study also supports much past research on the benefits of small class sizes. Study Design Schanzenbach analyzed data from Tennessee’s Project STAR, a randomized experiment that assigned students to small classes (13–17 children) or regular-size classes (22–25 children) in kindergarten through third grade. The project was implemented at seventynine schools starting in 1985 and eventually included 11,600 students. Teachers were also randomly assigned to classrooms. Baseline analyses show that small classes improve school performance for students, especially from kindergarten through third grade, and the benefits are more pronounced and last longer for African-American girls. The improvements are also similar for boys and girls.

The Project STAR data allowed for the study of the effects of sex composition in the small classes since, by virtue of the small size, the difference of three or four more or fewer girls in a small class creates a large difference in the proportion of the classroom that is female. Schanzenbach’s study controlled for students’ gender, free-lunch (economic) status, and race, as well as whether the class size assigned is small or regular, along with school fixed effects.

Effects of Girls on Class Achievement

Prior to third grade, both girls and boys benefit from placement in a class that is small and predominantly female. Kindergarteners in small, predominantly female classes score an average of 2.3 percentile points higher than those in small classes that are predominantly male. Results of the effects of the sex makeup in first grade classes were not statistically significant. However, this is likely because at the time the Project STAR data were collected, Tennessee did not require children to attend kindergarten, and the influx of new students may have disrupted the classrooms enough to offset gains from the predominantly female classes. By second grade, students in the small, predominantly female classes posted scores that were 2 percentile points higher than the scores posted by students in the predominantly male classes. The positive impact of girls is even more pronounced in inner-city schools, where the increases in percentile points are about 50% higher than in all schools included in Project STAR.

Despite these positive early effects, by third grade students in small classes that are predominantly female score 2.4 percentile points lower than the other small classes—although the decline is entirely attributable to the scores of the boys in the classes. Schanzenbach identifies several possible reasons for the boys’ poor performance. One could be developmental changes in children as they near adolescence and peer pressure begins to influence behavior. The decline could also reflect the fact that students are no longer engaged in learning to read, but must now read and often work independently in order to learn. Because thirdgrade boys score on average about 0.2 standard deviations below girls in reading, boys in predominantly female classes may be left behind because the class is progressing too rapidly for them.

The Impact of High Scoring Peers and Girls on Individual Scores

Other studies generally find that exposure in the classroom to higher-scoring peers has a positive “spillover” effect on other students. Schanzenbach used the Project STAR data to attempt to determine whether the positive impact of girls in the classroom is likely derived from this spillover effect on boys of girls’ typically higher scores or from another factor. This analysis, which controlled for students who had been in small classes and thus had higher-than-average scores, indicates that being exposed to high-scoring peers improves a student’s own test score by 0.6 points for every one point increase in average peer scores. In addition, the impact of having a predominantly female class in second grade is a 1.3 point increase, which suggests that there is something about having girls in the class per se that improves outcomes for all students.

Policy Implications

Both boys and girls benefit from small classes (13–17 children). In addition, African-American girls appear to experience larger long-term benefits than other girls or African-American or white boys when they attend small classes from kindergarten through third grade. Regarding single-sex classrooms, the study can only extrapolate, given that the classrooms were only dominated by one gender, in this case girls, and not purely single sex. However, Schanzenbach finds that all students do better in female-dominated classrooms until third grade. At this point, boys fall behind and begin to perform worse than their peers in mixed classrooms. She also finds that the better performance is driven in part by the influence of higher-scoring girls in the class. However, something else—yet to be identified—is also contributing to the positive results through second grade.