AI feedback chatbot research highlights dialogue and human judgement

Updated Sep 07, 2026

Noorhan Abbas’s paper on an AI chatbot for interpreting assessment feedback was published online on 27 July 2026 in Assessment & Evaluation in Higher Education. Publisher listing.

Evidence available for this briefing: the paper’s identity and date were checked against publisher and DOI records. Its substantive text and abstract could not be retrieved during this review. The account below is limited to the author’s own public announcement.

What the author’s account supports

Abbas describes investigating how AI might help students understand, question and use academic feedback. The author reports hearing students’ perspectives on desirable and undesirable features, highlighting dialogue, contextual support, trust and the continuing importance of human academic judgement. Author’s publication announcement.

That announcement does not provide the sample, methods, numerical results or a measured improvement in feedback uptake. It therefore supports a discussion of design priorities, without establishing that a chatbot improves learning or that all students share these preferences.

Questions for a local pilot

Our practical suggestion is to define what support students can expect, and who can resolve a disputed or unclear academic judgement. Ask participants what they understood from the original feedback and what they still need to discuss.

If collecting comments about a proposed tool, distinguish anticipated usefulness from experience of using it. For a deployed pilot, agree how usefulness, accuracy and any learning outcomes will be assessed before drawing conclusions.

Document the assessment context and confidentiality requirements for any comment review. The student comment analysis governance checklist offers prompts for that planning. It does not validate a chatbot or replace academic responsibility for feedback.

Review note, 7 September 2026: replaced the earlier full-paper account with an explicitly limited author-summary briefing. Removed an unverified 69% quotation, NSS statistic, focus-group count and course-specific details. Source access remains insufficient to certify those earlier claims.

Reference

Noorhan Abbas (2026). Investigating student perceptions of an AI-powered chatbot to support feedback interpretation and uptake in higher education. Assessment & Evaluation in Higher Education. DOI: 10.1080/02602938.2026.2706016.