What Have We Learned About the Instructor’s Gender Effect on Student Evaluation of Teaching From Experimental and Quasi-Experimental Research?

Student evaluation of teaching (SET), one of the most widely used mechanisms for evaluating teaching worldwide, is arguably gender-biased, potentially obstructing women’s access to and advancement in academia. To date, a significant number of studies have focused on interpreting a difference in SET scores by gender as evidence of gender bias. The validity of inferences from many of these studies is in jeopardy if they lack a strong research design that effectively isolates various potential confounding factors. We conducted a systematic literature review of experimental and quasi-experimental studies testing the claim that SET scores are biased against women. Our search identified 21 relevant studies published between 2000 and 2021. We qualitatively analyzed the supporting theories that predict gender-biased SET scores, the study design features used to estimate the instructor’s gender effect, and the findings across studies. The two most dominant explanations across studies are gender stereotypes and expectancy violations theory. We found ingenious methods for simulating a lecture and manipulating the instructor’s gender. SET scales across studies feature up to five types of distinctive questions about teaching performance. Findings across studies suggest that women are more likely to receive lower SET scores than men. However, this review reveals that examining gender bias is far more complex than individual studies portray. We note that the lack of strong supporting theory and tensions in research designs affect the construct and external validity of findings across studies. We discuss implications for fair evaluation of teaching that protect the development of women in academia.