Professors will be able to use artificial intelligence to summarize student evaluations after faculty members voted in favor of it at a Feb. 2 meeting.
The program, which is administered by the higher education software company Watermark, will most likely go into effect for the next round of evaluations, according to professors who were at the meeting.
Professors can use the program to summarize Student Perceptions of Teaching (SPoT) surveys, which is how students currently fill out their course evaluations. Watermark will characterize student feedback in positives and negatives, as well as neutral and mixed, according to its website.
These sentiments are organized into pie charts, and the platform then uses AI to create a summary of results, according to Watermark’s website. The results are intended to be supplementary, and Watermark encourages its users to use their judgement to confirm evaluations.
The program serves to help an instructor save time, achieve a deeper understanding of their impact and implement more effective improvements to their teaching methods, according to Watermark’s website.
Professor Julie Woodzicka said she believes that Watermark will be most helpful to department heads.
“If they have 20-some faculty, that’s like 13,000 comments that they would have to read in one term,” Woodzicka said. “I imagine it will help them summarize more.”
Watermark is only being used to summarize by shortening and organizing student comments. It does not remember or use student comments to improve itself later, she said.
“It’s purely AI that helps with summarizing,” Woodzicka said. “It is a non-learning model. The AI can’t learn from itself.”
Watermark is not meant to be a replacement for professors reading student evaluations, Woodzicka said.
“That is an important thing for students to understand,” said Woodzicka. “For most faculty, I don’t think they’re going to use AI.”
But Professor Toni Locy said she encourages students to question if professors will read every evaluation.
“Students should ask, ‘Are you going to read what I write? Or are you going to let AI summarize it?’” Locy said.
Locy said she will not use AI to summarize student evaluations.
“You don’t have to. I’m not going to, because I don’t trust it,” said Locy, who said she spoke out against the proposal during the faculty meeting. “I don’t trust any of these platforms to preserve student confidence and the confidentiality of the material. This stuff seeps out.”
Locy said she won’t use AI because she wants her students to put effort into writing the course evaluations.
“I’m going to continue to read them because I want students to take the time to fill them out,” Locy said. “If I’m asking students to take the time to fill them out, then I’m going to take the time to read them. I think that’s only fair.”
Locy said Amanda Bower, a business administration professor, led the push to permit using AI.
But some students have already expressed reservations.
“Using this program would be an inaccurate representation of students’ perspectives,” Lela Ganske, ’29, said. “I write what I write in professor evaluations for a very specific reason. I want professors to see exactly what I have to say, not an AI version.”
Jackson Weiser, ’29, shared similar thoughts. “Evaluations are about as human as a process gets: you state your opinion on a subject’s goodness of fit.”
He said he believes summarizing can be beneficial but it should be done without AI.
The Holistic Teaching Assessment Committee (HTAC) will assess the program once it’s implemented. HTAC works to evaluate and revise how teaching effectiveness is measured, including SPoT assessments.
“Now that it has been approved, we will assess the utility of AI as a summary tool,” said Professor Nadia Ayoub, a co-chair of HTAC. “Our goal is to ensure best practices in teaching assessment.”
