A review of test scores from 10,000 school district finds that gender gaps in math and English vary with community wealth and racial diversity.
When Stanford Professor Sean Reardon and his research team set out to take an unprecedented look at how elementary school girls and boys compare in academic achievement, they expected to find similar stereotype-driven patterns across all 10,000 U.S. school districts: boys consistently outperforming girls in math and girls steadily surpassing boys in reading and writing by a wide margin.
Instead, Reardon and his team of researchers at Stanford Graduate School of Education discovered wide variations in how girls and boys in grades three through eight perform from one district to the next. In some cases, girls did better in both math and reading. In others, boys had the advantage in math and almost matched girls on English-related subjects.
The swings in math scores were especially striking. Looking closely, the researchers uncovered a pattern: in affluent, highly-educated and predominantly white districts, boys outperformed girls in math. In poorer, more racially diverse districts, girls often outdid boys in math.
In reading and writing, however, the researchers found no correlation with local socioeconomic status or racial makeup. In almost every public-school system, girls came out ahead in reading scores, though to different degrees across communities.
The study, published online as a working paper, marks the first comprehensive analysis of gender achievement gaps at the district level.
"Our goal was to map the patterns of gender achievement gaps across the entire country in order to develop a better sense of what kinds of communities and school districts most commonly provide equal educational opportunities for girls and boys," says Reardon, the Professor of Poverty and Inequality in Education. We hope this information will help educators and policymakers eliminate educational gender disparities."
The findings were drawn from the Stanford Education Data Archive (SEDA), a massive online collection of roughly 300 million math and reading test scores from every public school in the United States from fall 2008 through spring 2015. Reardon, one of the creators of SEDA, has previously found that school systems with large numbers of low-income students have average academic performances significantly below the national average. He’s also shown that poverty alone does not determine the quality of a school district.
For the latest study, Reardon and his team built an analytical model that takes into account possible discrepancies in math and reading scores across states, including differences in test formats. Reardon and his collaborators have previously shown , for example, that multiple-choice questions favor boys while girls better solve open-ended problems.
The results were striking. In some of the country’s richest districts, boys on average outperform girls in math by two-fifths of a grade level but are half a grade level behind on reading and writing. In many of the poorest districts, however, the gap favors girls by one-fifth of a grade level in math and four-fifths of a grade level in English subjects.
The scholars also found that boys perform especially well in math in communities where adult males are more highly educated and earn substantially more than females.
In other key findings from the study:
- Girls surpass boys on reading and writing in almost every U.S. school district regardless of local wealth or racial makeup. In third grade, female students outperform boys by roughly half a grade level. By the end of 8th grade, girls are almost a full grade ahead.
- Gender achievement gaps in grades three through eight have narrowed significantly over the last decade. Math gaps, in particular, were notably smaller on average in 2015 than in 2009.
The new research doesn’t provide evidence as to why socioeconomic and racial conditions impact learning in elementary school. The authors also can’t say whether one race does better than another across subjects.
"The study shows that local conditions seem to partly shape gender achievement gaps," says Erin Fahle, a co-author of the study who earns her PhD in education policy from Stanford this month. "In particular, in some places female students appear to have an academic advantage; in others male students do."
New research opportunities
One hypothesis holds that, in wealthy families where the men earn much more than women, the stereotypical idea that boys are better at math and sciences and girls are better at reading and storytelling may be inadvertently bolstered when a son shows early interest in robots or girls stage a play. That’s because richer families have the resources to invest in, say, after-school science or theater programs.
"It may be easier for parents to reinforce stereotypical patterns in affluent places because they have more money to do so," says Reardon, who sits on the steering committee of Stanford’s Center for Education Policy Analysis (CEPA) and is also a senior fellow at the Stanford Institute for Economic Policy Research. "In less affluent places parents can’t spend the same kind of money and, therefore, may not reinforce those patterns as much."
Fahle adds that, while it’s important to know why girls do better than boys in math in some places and not others, it’s equally critical to figure out why boys consistently and almost universally lag in reading and writing.
"There’s a lot of evidence," says Fahle, who will soon join the faculty at St. John’s University in New York City, "that boys’ educational opportunities in English language arts are being constrained."
"We can’t say with this research why these patterns exist," says Reardon. "But the data provides an opportunity to further investigate what’s behind them and to think about steps that schools and communities can take to address the more troubling ones."
The paper’s additional co-authors were: Demetra Kalogrides, a researcher at CEPA; Rosalia Zarate, a GSE doctoral student; and Anne Podolsky, a researcher and policy analyst with the Learning Policy Institute.