Students describe potential and limits of AI in assessment feedback

Updated Sep 07, 2026

Updated, 7 September 2026: We checked the institutional abstract and accessible publisher passages. The study explores perceptions of potential use; it is not an experiment comparing feedback quality.

What the study examined

Sarah E. Rose, Louise Taylor, Gary Pheiffer, Zoe Fortune and Natalie Wilde recruited 25 volunteers from five UK higher education institutions, including a UAE campus, during March–May 2025. Seven focus groups and two interviews were analysed together using reflexive thematic analysis.

Participants saw possible value in individualised and apparently objective AI feedback while questioning accuracy and the loss of human expertise and dialogue. These are perceptions, not evidence that AI is objectively accurate, unbiased or effective.

The authors initially planned a UK/UAE comparison but combined the accounts after finding consistency. That decision does not establish cultural equivalence. The small, self-selected qualitative sample cannot estimate sector-wide preferences, and no feedback intervention or attainment effect was tested. One participant did not provide demographic details, so demographic counts refer to 24 people.

Questions for feedback design

Our suggestion is to explain what role a tool plays in producing comments and who can discuss or correct them. Invite students to describe what they find useful or uncertain without assuming that all objections have the same cause.

Evaluate accuracy against the task separately from perceived accuracy. If students describe a missing personal connection, ask what kind of interaction they need. Our governance checklist can help organise those questions. The study does not validate a commercial comment-analysis workflow or guarantee that visible oversight will increase trust.

Reference

Rose, S. E., Taylor, L., Pheiffer, G., Fortune, Z. and Wilde, N. (2026). Lacking the ‘personal touch’: students’ perceptions of generative artificial intelligence in assessment feedback. Assessment & Evaluation in Higher Education. Paper, online 6 May 2026.