Contextualizing gender disparities in online teaching evaluations for professors.

Area of Science:

• Higher Education Research
• Gender Studies
• Educational Psychology

Background:

• Student Evaluation of Teaching (SET) is a critical tool for assessing teaching effectiveness in higher education.
• Existing research suggests potential gender biases in SET, but large-scale analyses are lacking.
• Understanding these biases is crucial for fair evaluation of professors' careers.

Purpose of the Study:

• To investigate gender disparities in Student Evaluation of Teaching (SET) using a large dataset.
• To examine how professor gender influences student comments and ratings.
• To apply role congruity theory and shifting standards theory to explain observed gender gaps.

Main Methods:

• Analysis of approximately 9 million Student Evaluation of Teaching (SET) reviews from RateMyProfessors.com.
• Multiple linear regression to assess the impact of professor gender on numerical ratings.
• Dunning log-likelihood test and BERTopic for analyzing textual comments and identifying key themes.
• Sentiment analysis to compare comment polarization across genders.

Main Results:

• Women professors received significantly lower SET numerical ratings compared to men across many fields.
• Qualitative analysis revealed distinct language patterns in student comments based on professor gender.
• Specific topics in SET comments were significantly associated with professor gender, aligning with societal role expectations.
• Sentiment analysis showed more polarized comments (both positive and negative) for women professors.

Conclusions:

• Student Evaluation of Teaching (SET) data demonstrates significant gender-based disparities.
• Gender role expectations appear to influence student evaluations and comments.
• Caution is advised when using SET for high-stakes decisions due to potential systematic biases against women professors.