Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Employing a comprehensive, ten-year data set of class-level observations and institutional data from a large US university, we explore gender disparities in STEM fields by isolating differences in student evaluations of teaching effectiveness (SETEs). Faculty teaching face-to-face or hybrid classes receive significantly higher scores on average when compared to themselves online — a premium that is independent of gender. In STEM fields, the face-to-face premium for males is double that for females. The gender composition of the class matters. As the percentage of female students increases, SETE scores increase on average for females teaching in STEM fields but not for males. These results have implications for the ongoing problem of retention of women in STEM fields.