Students’ evaluation of teaching (SET) is the most common way of assessing teaching quality at universities. Recently, online SET (instead of the traditional paper-and-pencil questionnaires) has become the established means for gathering students’ opinions. Obviously, online surveys have great advantages, but their greatest challenge remains the low response rates that might affect the representativeness of the sample. This paper contains a non-response analysis on online SET-data using multilevel logistic models. At the student level, course grade, program level, and the number of course evaluations a student was asked to complete are significant predictors for participation. Student's gender and study domain are not significant. The variance component of the course level is estimated to be 0. The implications for future research and SET-practice are discussed.
展开▼