Understanding university instructor responses to machine-learning analysis of mid-semester teaching evaluations

The study explores how instructors of large-enrollment university courses engage with the results of their mid-semester evaluations of teaching (MSET) analysed with the help of machine learning. Eleven participants from different departments at a Canadian university participated in the study and conducted MSET in their courses, which were analysed for major themes using topic modelling. During in-depth interviews, participants shared their perspectives on the results of machine analysis of their MSET, offering valuable suggestions on how to improve data visualisations and the accessibility of the generated report, as well as the benefits and drawbacks of using machine analysis. Overall feedback from instructors was positive, suggesting that machine learning can enhance instructor engagement with qualitative student feedback. This paper outlines the process of testing different topic modelling approaches to analyse student feedback and experimenting with various data visualisations—insights that may benefit others developing similar tools and seeking instructor input.