Implicit Stereotypes: Evidence From Teachers' Gender Bias

The gender gap in math substantially widened when students were assigned to math teachers who held stronger implicit stereotypes. 

• Girls, especially those with lower initial math skills, got lower test scores, chose less challenging academic tracks, and displayed lower self-confidence when assigned to teachers who associated men with sciences.
  • Math teachers who associated men with math widened the gender gap in standardized math test scores by one-third, compared to math teachers with the opposite associations. This effect was strongest for girls starting middle school with worse math skills (-0.095 standard deviation per 1 standard deviation in teacher bias), compared to boys with equivalent math skills (+0.040) or girls with better math skills (-0.068).
  • Math teachers who associated men with math made track recommendations that encouraged girls to self-select into less demanding high schools. This effect was driven by students with worse math skills, with an 11.4% increase in girls’ enrollment in the vocational track, as compared to almost no impact on girls’ enrollment in the scientific track. 
  • Math teachers who associated men with math widened the gender gap in self-confidence by 4.8 percentage points, leaving girls with lower self-confidence about math at the end of middle school, even controlling for their standardized test scores.
• Boys’ test scores, choice of academic track, and self-confidence were not affected by math teachers’ stereotypes.  
• The gender gap in reading was not affected by literature teachers’ stereotypes, with girls scoring 0.14 standard deviations higher in reading than boys at the baseline.

Already facing the stereotype that “women are bad at math,” girls were more susceptible to math teachers’ biases than boys – particularly girls with lower initial math skills. These findings are consistent with a model in which stigmatized groups underassess their own ability and underperform, fulfilling negative expectations about their achievements. In schools, implicit assumptions can form an unintended and invisible barrier to equal opportunity.

This study targeted 91 middle schools in Milan, Brescia, Padua, Genoa, and Turin. The study’s dataset included 1,390 middle school teachers, with main analysis focusing on 454 math teachers (81% female), 615 literature teachers (90% female), and 30,359 middle school students (48.3% female). To test gender bias about sciences and humanities, teachers were asked to participate in a survey (80% response rate), including the Gender-Science Implicit Association Test and a questionnaire. The study also utilized student survey data, and administrative information from the Italian Ministry of Education and INVALSI. Analysis focused on gender gaps within classes, as well as comparisons between students of the same gender, school, and cohort, who were assigned to teachers with different levels of stereotyping.