Student evaluation of teaching (SET) is used widely in the higher education sector to elicit feedback about perceived teaching performance and effectiveness (Richardson, 2005). When using that feedback to reflect on their teaching approach and courses, educators may simply assume that the attributes evaluated are equally important. However, if this equi-importance assumption is not valid, teachers may be drawing incorrect inferences from the evaluation findings. Performance-Importance Analysis (PIA), derived from the services literature (Martilla and James, 1977), is a tool that can be used to compare a teacher’s perceived performance with the students’ views on the relative importance of each of the aspects evaluated. The main shortcoming of previous PIA applications to SET is the use of rating scales to derive measurements for both the performance and importance dimensions. Huybers (2009) discusses the drawbacks of the use of rating scales in SET and demonstrates the particular advantages of the Best-Worst Scaling (BWS) technique (Finn and Louviere, 1992; Marley and Louviere, 2005). The contribution of this paper is to show how BWS can be used as a tool in conjunction with PIA in higher education teaching and course evaluations.

Keywords: student evaluation of teaching, performance-importance analysis, best-worst scaling