Can Machine Learning Help Instructors Make Sense of Student Evaluation Comments?

Updated Sep 10, 2026

Can automated summaries make student comments more useful to the people teaching a course? Valeriya Minakova and colleagues explored that question with eleven instructors from different departments at a Canadian university.

The participants collected mid-semester feedback in their courses. The researchers used topic modelling to identify major themes, then interviewed instructors about the resulting analysis. Participants’ feedback was broadly positive, and they suggested improvements to the visualisations and accessibility of the reports while discussing benefits and drawbacks. View the research record at McMaster University.

This is evidence about instructors’ responses to an analysis process. It does not establish that the tool improved student learning, that it was more accurate than all alternatives, or that the same results would follow in UK institutions.

Designing a local pilot

Our recommendation is to involve teaching staff in reviewing a small trial report. Ask whether its themes make sense, whether important exceptions are visible, and whether readers can move from a summary back to the underlying comments. Record both useful features and misunderstandings before extending the process.

Timing is another practical question: would the report reach staff while they can still discuss changes with the current cohort? See our education text-analysis guide for questions to ask when comparing approaches. These suggestions extend the paper’s usability question; they are not verified outcomes of a particular product.

Correction

Correction, 7 September 2026: Earlier response-rate figures, enrolment thresholds, algorithm details and specific interview reactions could not be verified and have been removed.

Reference

Minakova, V., Patterson, R., Potter, B., Wang, S., Wei, L., Xiao, Y., Junkins, C. and Greenberg, S. (2026). Understanding university instructor responses to machine-learning analysis of mid-semester teaching evaluations. Assessment & Evaluation in Higher Education, 51(4), 770–790. DOI: 10.1080/02602938.2026.2619428.