Student evaluation of teaching is an established measure of teaching performance in nearly every university worldwide. It is also a subject of challenging social science research, with several methodological controversies. This diploma thesis addresses the specific administrative issue of when to conduct student evaluation surveys: throughout the semester, before the final exam, or after the final exam. This issue is an extremely important administrative and practical question that raises various methodological concerns. A literature review shows that surprisingly little attention has been paid to this issue. Only 32 studies have addressed the effects of different timings, and they generally found relatively few effects. In the empirical part, the data (n = 5.036) from an experiment at a selected university were analyzed. The experiment involved the replication of seven course-related items before the final exam and again after the exam grades were released to students. A statistically significant increase was found in the overall satisfaction in the post-exam survey, which can be partially explained by the exam experience. Test–retest correlations (median r = 0,49) showed considerable instability of responses and relatively low retest reliability. In addition, response quality was lower in the post-exam survey. These findings speak against asking the bulk of course-related questions after the exam, whereas exam-related aspects cannot be asked before the exam. From a methodological point of view, it thus seems that student evaluation surveys conducted before and again after the final exam provide the most comprehensive insight into student evaluation of teaching.
|
