Student Evaluation Of Teacher Performance: Random Pre-Destination

Abstract

The student evaluation, used to measure students’ perceptions of teacher performance, has been increasingly used as the predominant component in assessing teaching effectiveness (Waters et al. 1988), and the widespread movement of outcomes assessment across the country makes this trend likely to continue in the future (McCoy et al. 1994, AACSB 1994, SACS 1995).  Substantial research has been conducted with regard to the reliability and accuracy of student evaluation of teaching quality, and a considerable number of uncontrollable factors are found to bias the results of the evaluation rating.  This paper identifies one more factor.  Each student has an “evaluator profile”, which decreases the reliability of the student evaluation.  An “evaluator profile” is a persistent pattern of evaluating behavior that may or may not be consistent with the quality of the characteristic being evaluated.  Each class of students consists of a random sample of different evaluator profiles.  A student evaluation rating of a teacher’s performance is biased up or down depending on the concentration of high or low evaluator profiles present.  This paper further shows through simulation the degree to which student “evaluator profiles” impact the overall student evaluation rating of teacher performance. We find that there is evidence to support the “evaluator profile” conjecture, and that these “evaluator profiles” do in fact have the potential to change overall student evaluation ratings substantially.