Halo Effects in the Student Voice: Unwanted Correlations

Updated Sep 07, 2026

A halo effect occurs when an overall impression influences ratings of separate qualities. Correlated survey answers alone do not prove it: good organisation and useful teaching materials, for example, might genuinely occur together.

Edmund Cannon and Giam Pietro Cipriani investigated this problem using student evaluations at an Italian university. They compared a question about lecture-room adequacy with independent room-capacity information. Detailed room data were available for one faculty. This was an observational analysis, not an experiment that changed question order.

The authors found evidence of halo effects, but the room-related responses still contained useful information about rooms. Their modelling suggested that independent room information explained more of those responses than the halo-related component. Giving the same answer to every question was also less common than in some earlier studies. Read the accepted manuscript, especially the room-size analysis and conclusion.

Their conclusion is specific: these findings did not establish a need to redesign evaluations because of halo effects. They do not prove that every survey is valid or that halo effects never matter for diagnostic feedback.

Applying the distinction

Our suggestion for evaluation teams is to define what each question is intended to tell you, then look for independent evidence relevant to that question. If students report unsuitable rooms, compare that feedback with timetabling and capacity information before attributing the result to a general impression of teaching.

Open-ended comments may give reviewers examples to investigate. This paper did not test automated text analysis as a way to detect or remove halo effects. Similarly, shortening a questionnaire is a design decision to evaluate locally, not a guaranteed solution.

Evidence and correction

Correction, 7 September 2026: Checked against the full author accepted manuscript. The earlier summary incorrectly implied that asking about rooms first influenced later answers. We corrected that account and narrowed claims about survey validity and text analysis to match what the study tested.

Reference

Cannon, E. and Cipriani, G. P. (2022; first published online 2021). Quantifying halo effects in students’ evaluation of teaching. Assessment & Evaluation in Higher Education, 47(1), 1–14. DOI: 10.1080/02602938.2021.1888868.