Student Evaluations of Teaching as a Predictor of Teaching Effectiveness in a Selective Liberal-Arts College

Abstract

Previous research has shown mixed results regarding the relationship between Student Evaluations of Teaching (SETs) and teaching effectiveness. Although the literature on this topic is extensive, it appears that none of the scholarship was conducted at small, selective, undergraduate institutions in which teaching is prioritized. In the current research, we explored the association between student evaluations and teaching effectiveness at such an institution. Data were collected at a selective liberal-arts institution of approximately 2200 students in the Northeastern United States. Participants were 139 students who had enrolled in both Calculus I and Calculus II. Seven different professors taught Calculus I, while seven different professors taught Calculus II. Linear regression analyses indicated that, for one of the three subscales of the College’s SET, there was a positive association between SET scores of the faculty member in Calculus I and student grades in Calculus II. Machine-learning analyses confirmed this association and indicated that the relationship was strongest for students with overall grade-point averages near the cohort’s mean. Our research thus adds to the scholarship broadly but also suggests that assessment of SETs in smaller, selective teaching-oriented institutions is an important direction for future scholarship.