Fred Thys ’80 reports for WBUR about the cost of Williams College. Fred is a knowledgeable and sympathetic alum but this story, at least, painted an incomplete picture. Perhaps later stories will flesh out the scene? In the meantime, let’s spend 10 (!) days dissecting this article! Today is Day 10.

But are the best-paid professors necessarily the best teachers?

Answer: No. Not even close.

Not Thys fault (?) but asking this question — and then failing to answer it! — pushed a variety of EphBlog buttons.

First, the College knows the answer! Why didn’t Thys ask Falk et al, and, if they declined to answer, report that fact? Williams collects a variety of information about teacher performance at the end of each semester, the famous Student Course Survey (SCS) forms (pdf):

Student opinion and peer review are both important in the evaluation of teaching. On the student side, the College mandates use of the Student Course Survey (SCS) for every course. It consists of a page of questions to which students give numerical ratings and a page inviting descriptive commentary (“blue sheets”). Some departments substitute their own list of questions for the generic blue sheets, and individual faculty members can choose to substitute for or supplement the blue sheets with their own more specific questions. Each faculty member receives the blue sheets as well as an analysis of his or her own quantitative results after their course grades have been submitted.

The College knows the exact SCS average score for every professor last year. It could easily calculate (and probably does calculate) those measures by level (lecturer, assistant, associate, full) or by pay-grade or by any other factor. What does the data tell us? (The College might not want to tell Thys that data by salary group, but it has no excuse not to release it by faculty rank, which is highly correlated with salary anyway)

Second, the College probably already knows the truth: after the first few years, there is no correlation between teaching quality (at least measured by SCS scores) and teacher experience/salary. The “best-paid professors” are not “necessarily the best teachers.” Highlights (based on conversations over the years — corrections/pointers welcome):

a) SCS scores have consistent course/major/division biases. It is much easier to score higher in, say, Division I than in Division III.

b) There is improvement over the first few years at Williams. It takes a while to get into the swing of a Williams classroom.

c) There is no further improvement (or decline) after year 4. Your student evaluations at 32 are the same as they are at 65.

Third, some might reasonably complain that “best teachers” does not equal “highest SCS scores.” Agreed! Then why doesn’t the College collect better data? Why doesn’t it hand out prizes for the best teachers? The College could collect better data, could try to identify the professors that do a particular good job at, say, supervising a senior thesis or supporting/mentoring students from disadvantaged backgrounds.

Why doesn’t the College do these things? The major trend in faculty recruitment/appointment/retention/promotion at Williams over the last 30 years has been to decrease the weight placed on excellent teaching and increase the weight placed on research.

I don’t like this trend, but it does help to explain why, if you want to talk about excellent teaching, the SCS is all the data you are going to have to use.

Print  •  Email