The impact of teaching on retention: Some quantitative evidence

Objective. The impact of student course ratings on retention has not been e xamined in previous research. This study begins to fill that gap. Methods. Using data on American University undergraduates and on the ratings of the courses in which they were enrolled, the study employs logistic regression to estimate the impact of the ratings, averaged over the classes, on retent ion, holding constant race, gender, GPA, receipt of aid, and other factors. Results. Compared to mid-rated courses, enrollment in consistently poorly rated classes significantly reduces the probability of retention. Unexpecte dly, enrollment in consistently top-rated courses also significantly reduce s the probability of retention. Conclusions. As expected, poorly rated cour ses signify customer dissatisfaction. However, students enrolled in the bes t-rated courses at AU appear equally dissatisfied. One explanation is that satisfaction is short-lived; students may get used to top-rated courses, an d soon look elsewhere for better courses.