Evaluation inflation

Last week, Dean of the Faculty David Dobkin issued a report indicating that instructor evaluations — the bubble-sheets distributed by the registrar's office at the end of each term — have steadily risen since the University began keeping records in 1974.

"Clearly, something is wrong," read the report. "Ratings are skyrocketing, but performance is flat. This office still receives just as many complaints about dry lectures, boring precepts, tedious seminars, impenetrable readings, and uninspired preceptors. But the numbers keep going up."

Dobkin observed that the average course's "overall value" rating (from 1 to 5) increased from a 2.95 to a 4.76 over the past three decades."

Dobkin proposed that the figures be reduced by capping the percentage of courses in a given department that could be rated 5 (the highest possible rank) in each department. No word has yet been issued on the exact implementation details.

Since the meeting, some faculty have expressed concerns with the proposals.

"We all know that the numbers are high. Is the dean going to deny that Princeton has the best faculty in the country? We certainly don't select faculty from all points along the bell curve. Why should we expect them to fit that distribution when they arrive?" asked Garcia Gabriel, an assistant professor in the Spanish department.

"This is just going to put more stress on my junior colleagues," explained Michael Dawes, a professor of sociology. "You're an up-and-coming, high-flying, hotshot. Then you come here and find yourself slapped down to a 3.2 because your courses' ratings are capped? We'll be seeing a lot more faculty at the counseling center, believe me."

Students' explanations for the ever-soaring ratings varied. Alec Connage '05 explained that "even if the course was terrible, it's hard to rank the guy down if he's trying. English obviously isn't [assistant economics professor] Vladimirinov Koromitchovitchov's first language — why should I penalize him for that? If I give him too low a ranking, it might get noted in his file. Then, if the Committee doesn't like him, it might be used as an excuse to deny him tenure. He'll lose his job. He'll become homeless. Then, when I'm on Wall Street next year, I'll have one more drunken beggar to step over on my way to work."

William Young '07 concurred. "It's an issue of fairness — if I expect the prof to err on the side of better grades, wouldn't it be inconsistent for me to err on the side of lower ratings?"

"Capping rankings will only discourage collaboration." explained Patricia Eaton, an associate professor of physics. "If you know that only a certain percentage of courses in the department will get 5's, are you going to share your strategies with your colleagues and reduce your chances? The laid-back faculty-lounge atmosphere will go straight into the toilet."

David Webber, of molecular biology, explained that out-of-class research would most likely suffer if faculty were forced to compete for an "artificially small" pool of 5's. "If all of your time is spent turning overheads into downloadable PowerPoint slides because students can't be bothered to take notes, that totally kills your time in lab."

Other faculty worried about the implications of the proposals on their later careers. Helga Tomlin, of the English department, lamented the review process at certain schools. "At UVA, I've heard that they just plug your ratings and your publication count into a formula and reject everyone below a threshold. The ratings cap puts us at a disadvantage in these places."

Get the best of the ‘Prince’ delivered straight to your inbox. Subscribe now »

Dean Dobkin had explained in the memorandum that he had spoken with the faculty recruiters at other institutions and had received nothing but encouragement for his proposals. "In addition," the memorandum noted, "a faculty appointment at Princeton is not an instrumental good. It is not, and should not be regarded simply as a steppingstone to one's future career."

Other faculty were skeptical that a problem exists. Math professor Jordan Ellenberg, noting the nationwide concern for overgenerous evaluations, published an editorial in Slate calling the hand-wringing misplaced. Drawing upon a statistical model of course ratings, he concluded that the overvalued rankings still could be used to distinguish between individual faulty members "on a sufficiently large sample size."

Dobkin was unimpressed. "I'd advise him to stick to writing about Q-curves and Galois representations," he remarked in a press conference on Monday afternoon.

Dobkin concluded his remarks at the conference by observing that the students' overvalued ratings were in fact a disservice to the professors because they obscured the difference between merely good and truly great teaching, ruining any incentive to strive for the latter. As to whether reports of student-acclaimed teaching actually played a role in the tenure-evaluation process, Dobkin had no comment. Joseph Barillari is a computer science major from North Canton, Ohio. His column appears on alternate Tuesdays. He can be reached at jbarilla@princeton.edu.