Students’ evaluation of the quality of teaching using generalisability theory: A case of a selected university in Ghana

Students' evaluation of lecturers' quality of teaching has been a common practice in universities in Ghana and beyond. Data gathered from students are used to make vital decisions about lecturers such as promotion, training and development needs, among others. In recent times, the accuracy of students' ratings of teaching quality has been questioned by stakeholders due to several reasons. Previous studies have attempted to investigate this issue using classical test theory (CTT) which comes with its own flaws. Little attention has been paid to the applicability of Generalisability theory (GT) to students' evaluation of teaching in Africa. This study aims to assess the reliability of students' rating of teaching through the lens of GT. A three-facet partially nested random balanced design [(r x i x o): l] was adopted for this study. Student (rater), item, and occasion served as the facets and lecturer served as the object of measurement. Both G (generalisability) and D (decision) studies were conducted. The institution's evaluation questionnaire was adapted, validated and used for data collection. The sources of measurement error were accounted for by raters and lecturer-by-occasion interaction. Generally, the dependability index for the students' evaluation of teaching quality was low, signalling little trust for such data. It was recommended that a minimum of 25 students should be permitted to rate lecturers for each class using at least 20 evaluation items. Suggestions for further studies were made based on the findings.