A new study offers an unusual perspective on the value of a good teacher, in this instance the value of a good university professor. Two economists, Florian Hoffmann and Philip Oreopoulus, examine ten years of data at a large Canadian university in order to assess the impact of professors on students' decisions to 1) drop a course, and 2) subsequently take another course in the same subject. They offer a reasonable case that both factors correlate with college students' achievement and four-year completion rates.

Mirroring the research findings on grade school teachers, various attributes of professors do not appear to be good predictors of quality. A host of measures did not explain variations: teaching full or part time, research activity, tenure status, or high salaries of over $100,000 have little to no influence on students' achievement. This suggests that universities could save a bundle if student achievement was of primary concern--though many might wonder if achievement is even a concern in the current environment.

Evaluation ratings by students and instructors' sense of their own effectiveness were the biggest predictors of professor quality. Students assigned to higher-rated teachers are less likely to drop a course, are more likely to receive better grades, and are somewhat more likely to take similar courses in following years. And unlike with younger, more impressionable students, the impact of either a good or bad professor is imperceptible a year or two later.