Modeling unit non-response and validity of online teaching evaluation in higher education using generalizability theory approach.

Abstract

Unit non-response is a common phenomenon in online teaching evaluation in higher education institutions. However, little is known about the relationship between the rates of unit non-response and the quality of online teaching evaluation exercise. This study explored the incidence of unit non-response and how this phenomenon relates to the reliability of students' responses to online teaching evaluation. Adopting the generalizability theory approach, students' evaluation of teaching data from a university in Ghana was analyzed by conducting both generalizability study (G-study)- and decision study (D-study) analyses. The results revealed that unit non-response among students was predominant in online teaching evaluation exercise. The study demonstrated that higher rates of non-response among students were associated with high levels of measurement errors and low reliability of responses. The findings of this study have implications for the accuracy of online evaluation data obtained for decision-making in higher education contexts. The study calls on higher education administrators to embark on sensitization and awareness campaigns that target students on the need to actively participate in the appraisal of teaching at the university to address the issue of unit non-response.