Relevance & Research Question

Student evaluations of teaching (SET) gain more and more importance for university teaching, especially web-based SETs. To guarantee the validity of SET and to avoid sample-errors high response rates are necessary – which cannot always be reached. This study investigates different factors, which might give some explanations for nonresponse in SETs. We examined the social exchange-theory, salience, opportunity costs, survey fatigue and the survey mode (online vs. paper-pencil).

Methods and data

We contacted student representatives of 48 universities, additional student groups in social media, and sent approx. 900 invitations via the online-panel PsyWeb. The web-based survey was available for about four months. Participants provided reasons for non-response, information about response behavior, and attitudes. A total of 490 participants (69.59 % female; age: M = 24.10, SD = 4.07) were included in the final sample.

Results

First, our results show a significant influence of social exchange on responding in SET (Δ R²=.069, p < .001). Also the influence of salience (ΔR²=.201, p < .001) and survey fatigue (ΔR²= .078, p < .001) show significant influences on the participation in SETs. No significant effects were found for opportunity costs (ΔR² = .005, p = .166) as well as for the survey mode (ΔR²= .004, p = .209).

Added value

The results of our study can be helpful for online researchers and evaluation managers in reducing non-response. Notably, our findings stress the importance of communication between students: It influences a student’s evaluation behavior if a fellow student evaluates all of his or her lectures and courses. Thus, universities should indicate how many students take part in a current SET to motivate even more students. Furthermore, it is helpful to increase the students’ identification with their own university and SET, for example with special events or university-games. In addition, the consequences of the SET should be public for the students so they become aware of how they can influence the quality of teaching in their faculty. Finally, we could find no evidence for the often made assumption that an online administration of questionnaires leads to a non-response-problem in SETs.