Investigating sample bias in student evaluations of teaching

Student evaluations of teaching have increased in importance to universities in Australia over recent years due to changes in government policy. There has been significant debate in the literature as to the validity and usefulness of such evaluations and as to whether students who respond to the evaluations are indeed representative of the student population. A potential invalidating issue is self selection in the evaluation process. In this paper, we consider student evaluations of a large first year business statistics subject that had 1073 eligible students enrolled across four campuses at the time of the evaluation. The study is based on the 373 students (34.8%) who responded to the survey, and their final results. The evaluations were open for a period of six weeks leading up to and just after the final exam. The study looks in detail at the student population identifying such attributes as gender; home campus; course of study; domestic/international; Commonwealth Supported Place/full fee paying, etc. and then mapping these results to those of the students who responded to the survey.