Average rating: | Rated 5 of 5. |
Level of importance: | Rated 5 of 5. |
Level of validity: | Rated 5 of 5. |
Level of completeness: | Rated 5 of 5. |
Level of comprehensibility: | Rated 5 of 5. |
Competing interests: | None |
The evidence for gender differences in student evaluations of teaching (SET) is mixed. However, if there is a difference, it is usually female instructors, who receive worse ratings. Although not a plausible hypothesis, but one that needs to be ruled out, is that the tendency to evaluate the teaching of female instructors somewhat more negatively, could be due to their lower teaching ability. A hihgly cited study by MacNell, Driscoll and Hunt (2015) appeared to rule out this interpretation. (As of April 2020, their paper had received 154 citations.) These authors conducted an experiment, in which they manipulated the perceived gender of instructors. Two assistant instructors, (one male and one female) in an online class each operated under two different gender identities. Regardless of actual gender, male identity teachers received higher evaluation on professionalism, promptness, fairness, respectfulness, giving praise and enthusiasm. Uttl and Violo (2019) questioned these findings on several accounts: They argued that one could hardly generalize to all male or female instructors based on findings with only two individuals. Furthermore, the sample of students in each condition was rather small, ranging from 8 to 12 individuals. But most critically, there were three outliers, who gave the lowest ratings on all SET items in the two female conditions. If one removed these outliers, the gender difference disappeared. Instead, students rated the actual female instructor higher than the male instructor, regardless of perceived gender. By pointing out that the conclusion of the MacNell et al. study have no empirical basis, Uttl and Violo do not only render a service to the discipline, they will also encourage other researchers to condcut empirically sound studies on this issue.