The possible existence of a gender bias in the classroom is not a new controversy. Research has shown that, consciously or not, some teachers treat students differently according to gender; they may give boys more (or different types of) attention, encourage boys more in certain subjects and girls in others, and otherwise interact with each gender differently.
Economists Victor Lavy and Edith Sand bring us an important continuation of that work with a National Bureau of Economic Research paper that explores whether students are exposed to gender bias during elementary school. It then examines whether that exposure has an impact on students’ later academic achievement.
Their study follows approximately three thousand elementary school students and eighty teachers from twenty-five different elementary schools in Israel. They first ask: Is there gender bias? In other words, do teachers believe one gender is academically stronger than the other when there’s actually no difference (or even if the preferred gender is actually doing worse)? The answer to this question is yes. By comparing a teacher’s assessment of a student’s performance in a variety of subjects to the student’s scores on external exams in the same subject, the researchers find that girls outscore boys on external math exams, but boys outscore girls on teachers’ own math tests. The researchers find similar results at the classroom level. (In English, girls are over-assessed relative to boys, but the difference is not statistically significant; in Hebrew, the researchers found no signs of bias at all.)
Second, the authors estimate the impact of that math gender bias. The findings are stark—the more positive bias a male student received in elementary school, the higher his eighth-grade test scores; the more negative bias a female student received, the lower her scores. And the effects persist throughout high school. Positive bias for male students increased the likelihood of graduation; negative bias for female students decreased it. The same goes for secondary test scores, the likelihood of a given student enrolling in advanced science and math courses, and even postsecondary enrollment, attainment, and wages. (Two important notes: There were a lot of insignificant results amidst the significant ones, and the effects of a biased classroom seem to be much more important than the effects of a biased student-teacher interaction.)
These are weighty (and troubling) results. A student’s exposure to bias in elementary school can have a lasting effect on not just that student, but ultimately the labor market and economy. Male students are not only on the positive receiving end of bias in math, but are also more likely to “shake off” negative bias, while female students tend to internalize negative feedback.
Interestingly, while the researchers found that older single teachers are more likely to favor boys, they did not report the impact of teachers’ respective genders on bias. This last point would be a particularly interesting line of analysis. As a female math teacher, I was very aware of gender bias against girls in math and science and went out of my way to consciously avoid it. (But would an analysis of my behavior actually show overcompensation and a negative bias toward my male students? Given that my male students had probably experienced positive bias in previous classes, would it have mattered?)
The researchers found classroom effects especially significant—perhaps teacher training programs should explicitly include gender in their courses on teaching in diverse classrooms. My own training for secondary math included instruction on teaching students of different ethnicities, language skills, and cognitive abilities…but never different genders. Clearly, it should have.
SOURCE: Victor Lavy and Edith Sand, “On The Origins of Gender Human Capital Gaps: Short and Long Term Consequences of Teachers' Stereotypical Biases,” National Bureau of Economic Research, Working Paper 21393 (January 2015).
