Estimating Determinants of Student Evaluation Scores to Improve Teaching

Abstract

Abstract Student evaluations are used for both formative and summative assessment of teachers. This paper provides a method to make more effective use of these student evaluations by individual teachers. Data on three years of evaluations in two courses were used to develop regression models to explain overall effectiveness of teaching. The relative importance of explanatory variables changed with the course taught. Introduction Student ratings can be used for both formative and summative assessment of teachers in the same way that exams are used both to provide feedback to students so they can improve, and to evaluate their performance. In a meta analysis, Cohen (1981) found that the correlation between course ratings and mean student achievement was 0.47. He also found that global ratings, such as overall effectiveness, correlated more highly with student learning than more specific items. Therefore, while student evaluations do not provide a perfect measure of student learning, they are indicative. In any case, promotion and tenure committees generally rely on student evaluations rather than objective measures of learning. This paper provides a method to make more effective use of these student evaluations in order to: 1) improve one’s teaching and subsequent evaluations, 2) provide a framework for reflection which can be incorporated into a teaching portfolio. Student evaluations of teaching have been used for a number of purposes (Ory 2000). In the 1960s, they were promoted by students to improve public accountability and to help students made decisions regarding courses. In the 1970s, the primary focus was on development of faculty. In the 1980s and 1990s, teaching evaluation has been extensively used by administrators for promotion and tenure decisions. In a study of