A recent study examines how peer achievement and a classroom’s gender composition influence math achievement and student attendance rates.
The author, Ozkan Eren of Louisiana State University, uses data collected from a well-executed randomized experiment of middle and high school students in disadvantaged neighborhoods between 2010 and 2012. The sample included 5,320 students from eighty schools in twenty school districts.
There are three key findings: First, having a higher proportion of female peers in math classrooms improves the math test scores of female students, especially in less-advanced math courses, such as general high school math. A 10 percentage point increase in the proportion of female peers increases the average math test scores by 0.1 of a standard deviation. The gender effects on male students, however, are positive but insignificant.
Second, regarding peer effects, having higher achieving peers has no significant effect on female math test scores, but does improve the marks of males in the bottom two-thirds of achievement: A 1.0 standard deviation increase in peer achievement increases boys’ math achievement by 0.4 of a standard deviation.
Finally, having a higher proportion of female peers in math classrooms decreases the probability of chronic absenteeism among male students, but has no effect on female attendance. A 10-percentage point increase in the proportion of female students in math classrooms translates to 2.5 days decrease in total days absent from school among male students.
Schools should be careful what lessons they take from these findings. But the results do suggests that gender stereotypes of math ability impact girls negatively, especially those in low-level math courses. Therefore, improving female students’ confidence and their own perception of math ability may bolster their math achievement. And teachers may be wise to intentionally alter their perceptions of female students’ math ability.
SOURCE: Ozkan Eren, “Differential Peer Effects, Student Achievement, and Student Absenteeism: Evidence From a Large-Scale Randomized Experiment.” Demography (2017).
Christopher Yaluma is a doctoral student at the Ohio State University.