Q&A: Dr. Joe Cimpian on Gender Achievement Gaps and Biases

young girl at school

The Brookings Institution recently published Dr. Joe Cimpian’s “How our education system undermines gender equity.” In his research-based piece, Dr. Cimpian discusses how as early as kindergarten, U.S. education systems undervalue women’s intelligence and generate discriminatory barriers that can follow women into graduate school their professional careers. Alex Clothier asked Dr. Cimpian about his research, how K-20 educators can examine their own biases, and what points policymakers should keep in mind while tackling the gender achievement gap.

How did you and your colleague, Dr. Sarah Lubienski, introduce your gender achievement gap research to teachers?

When Sarah reached out to these teachers initially, she said that we were looking at the male-female achievement gap in math and teachers’ expectations of boys and girls. Understandably, most people would interpret that as the gap in either grades or standardized test scores. And so, when we met with the teachers in person and were having a conversation, one of the teachers brought the state standardized math test scores from her students and said, “Now, I don’t even understand why you’re looking at girls’ math achievement. These are my students’ standardized test scores, and there are absolutely no gender differences. See, the girls can do just as well as the boys if they work hard enough.” Then she realized on her own that she was attributing girls’ success in math to hard work and assuming that boys are naturally good at math. And she said, “Oh my gosh, I just did exactly what you said teachers are doing. I see now why you’re studying this.” In my opinion, it’s great that she realized her own bias in this regard, but this story is important because it also illustrates the kind of evidence that teachers, policymakers, and many others look at when thinking about which group is advantaged or disadvantaged by school—and in this case, the evidence was test scores, and the results of those test scores revealed no gender differences to the teacher. So, looking at the tests alone, she likely never would have realized her bias because the bias does not necessarily materialize in the test scores, and that is what makes it difficult to raise awareness about this issue.

My Mother is an elementary school teacher and recently read your blog post (I confess that I forwarded it to her). She’s worried she may be reinforcing stereotypes in her classroom. What advice would you give her and other K-12 teachers who wanted to assess their own biases?

Some simple things can involve checking to see if she tends to call on boys more for difficult questions or if she encourages boys to take bigger risks in their mathematical thinking. Of course, sometimes boys will volunteer more for difficult items, and some of this volunteering may be due to a cumulative encouragement they received over the years that advantages boys, and some of it may be a result of their own interest and development. It’s really hard to know what is contributing to students’ engagement and confidence differentials. But it is a good first step to be mindful that stereotypical thinking (about which gender is good at math or is innately gifted) can create barriers for students, and then to see if you might be contributing to differential treatment.

Some other things that psychological research suggests can help improve diversity are fostering a growth-oriented mindset and providing a diverse set of role models in a field. In this context, fostering a growth-oriented mindset means encouraging students to recognize that mathematical thinking is something that can be improved over time with effort, rather than thinking that math is something someone is either born good at it or not. Regarding role models, teachers can make sure that they highlight the contributions of mathematicians and scientists from a range of gender (and racial, etc.) backgrounds, so that students don’t think that math is not for them because they are not of the gender (or race, etc.) of people in the field that they learn about in school.

Researchers and policymakers may have different approaches to viewing gender achievement gaps. What should they keep in mind while tackling the issue?

In almost any area of education policy research, the assumptions made in the research design can have a substantial influence on the results; and those research assumptions can be rooted in what we think is most important to understand. The findings of gender bias against girls that I report in my blog post may seem counterintuitive to some people who see boys lagging behind girls in school, achieving at lower rates, being disciplined at higher rates, and getting fewer advanced degrees. Those examples of boys faring worse are often based on research designs that don’t make any statistical adjustments. Those gender gaps are important to be aware of and are certainly worthy of concern; however, the source of the gender gaps resulting from those kinds of research designs isn’t clear, and so it makes it more difficult to pinpoint how policymakers and educators can do something about those gaps. In order to figure out some mechanism that policymakers and educators can act on, we often try to isolate one specific factor at a time, and hold the other factors constant across groups.

Thus, my blog post focused on the latter kind of studies, those which compare similar kinds of students to more plausibly isolate a single factor. More specifically, the thrust of my blog post was focused on evidence of bias and discrimination against individuals who are matched on background characteristics—meaning that the focal studies discussed were essentially looking at pairs of boys and girls who were achieving identical scores on math tests, of the same race/ethnicity, from families of the same socioeconomic status, having similar stated interests, within the same classrooms, and (when available) were rated to be equally engaged with school and equally well behaved. When we look among these matched sets of students, we can more easily isolate bias and discrimination because we have “controlled for” other differences between the groups. And it is then that we see this bias against young women’s intellectual abilities, underestimation of women, and women entering fields where they expect to encounter less discrimination.

Because my research is often concerned with isolating effects and comparing the most similar groups as possible to identify specific areas to improve equity, I have argued that policymakers and educators need to be aware of bias against young women in the education system. And there are rigorous, replicated, robust, nationally representative studies that find evidence of bias. (It may be worth noting that girls consistently outperform boys in reading/English language arts, which is indeed a worrisome trend, but our research finds no evidence of teacher bias in reading when comparing similar students; this is in sharp contrast to math.) But at the same time, it’s also important to keep in mind that young men are not succeeding in the current education system—that does not negate the bias against young women, but it does suggest that we need to more critically evaluate how the education system can be improved to benefit all genders, so that no one’s opportunities are limited because of their gender.


Dr. Joe Cimpian is an IHDSC faculty affiliate and an Associate Professor of Economics and Education Policy in NYU Steinhardt’s Department of the Applied Statistics, Social Science, and Humanities.