Boosting Number of Girls in Math Class Can Improve Test Results

• By Dian Schaffhauser
• 11/07/17

Does gender composition and peer achievement in high school make a difference to math test scores or school attendance? That's what a researcher from Louisiana State University set out to explore. Ozkan Eren, an associate professor of economics, earlier this year published evidence that having a higher proportion of female students in the classroom improves girls' math test scores in less advanced courses (such as general high school math). A 10-percentage point increase in the proportion of female peers lifted the average math test scores by 0.11 of a standard deviation.

An apparent side effect is that the presence of more girls in class also dampens student absenteeism among male students while having no effect on female absenteeism. Also, while having more girls in the class has a "positive" impact on math achievement by male students, Eren reported, it's not particularly measurable.

When the average math achievement score of the whole class rises, that doesn't boost overall math test scores by young women; however, male students in the bottom and middle quartiles of the class do perform better. A one standard deviation increase in peer achievement boosted boys' math achievement by 0.42 of a standard deviation, a "large improvement," according to the researcher.

In his conclusions, Eren suggested that having more female students in the class "is likely to make gender less salient," thereby possibly improving the self-concept girls have of their math abilities. It also explains why the effect dissipates as the math levels go up, he added, because the girls "gain more confidence" as they progress into more advanced courses. At the same time, he noted, girls appear to compete more effectively only when it's against their own gender. Pitted against the entire class boys tend to be more competitive (and do better on tests), while girls are less likely to respond (and don't do better).

Should math classes be single-gender then? Eren advised against it. As he wrote in a draft version of his paper, "Our results indicate that some degree of gender segregation at the classroom level may improve female achievement but that may be at the expense of male students. Moreover, gender segregation may also exacerbate student absenteeism for boys and may further trigger the risky behaviors."

The results are based on data collected from a previous randomized experiment involving 5,320 middle and high school students in 80 schools in 20 districts. The researcher used only the data for high school students and examined results from four types of math courses: general high school math, algebra I, geometry and algebra II.

Oren's research was published in Demography.