Stable teaching ratings can hide changing judgements

Updated Oct 27, 2026

An unchanged teaching evaluation average can conceal movement in individual students’ judgements. In Assessment & Evaluation in Higher Education, Ahmed Salim Abdullah Al-Abri examines that distinction in an Omani foundation programme.

What did the study test?

Teachers had argued that examination marks depressed their evaluations. Al-Abri tested ratings before examinations and after marks were released, alongside students’ satisfaction with those marks.

The study involved 411 distinct students, of whom 89 completed both administrations, between March and May 2024. Students evaluated their reading-and-writing teacher at an Omani public university. The sample was effectively drawn from a single campus and participation was voluntary.

The questionnaire covered perceived teacher personality, teaching evaluation, students’ sense of responsibility when evaluating, and self-study skills. It was translated into Arabic after piloting revealed difficulties with English wording.

What the findings show

The study did not detect a fall in ratings after marks were released. Among students measured twice, the average teaching rating barely changed. The comparison between students satisfied and dissatisfied with their marks also fell short of conventional statistical significance, although it was close to the threshold. The modest matched sample means these results do not prove that marks never influence evaluations.

A stable average did not mean stable individual judgements. Teaching ratings from the same students correlated at r = 0.50 across administrations, with a 95% confidence interval of 0.33 to 0.64. The median interval was 45 days. Changes in experience during the examination period may also have contributed to changing judgements.

Perceived personality was closely associated with teaching ratings. Across the distinct-student sample, the correlation was r = 0.67. Personality was the strongest predictor in the regression, but the author cautions that the measures overlap conceptually and come from the same respondents.

The instrument still needs further validation. Some measurement-model fit indices fell below the study’s stated thresholds. The repeated respondents were self-selected, and teacher or class identifiers were unavailable to account for students rating the same teacher.

For teams reviewing their own teaching feedback, our module evaluation comment analysis organises written responses into themes for module and programme reports.

References

[Paper Source]: Ahmed Salim Abdullah Al-Abri. "Do the numbers hold up? Grades, personality and the validity of student evaluations of teaching in an Omani foundation programme"

DOI: 10.1080/02602938.2026.2746332

Published online: 14 October 2026, Assessment & Evaluation in Higher Education.