Big Courses and the Problem of Parallel Universes

(Added: Those who are unaware of the cheating allegations at Harvard could read this quick summary from Bloomberg. It isn't the most complete--the Globe and the NYT have reporting that is fuller, but behind a paywall.)

Though I have weighed in on other blogs with some reactive comments about the Harvard cheating issue, I have been reluctant to go on the air or to launch an offensive since my knowledge of the course is limited. I have read some but not all of the news stories, and have talked to a student in the course. The professor's reputation is at stake, as well as the students' integrity, so I want to be careful about seeming to attack or defend either on the basis of inference or speculation. Still, a few structural and institutional things seem to be clear.

The course seems to have sent some mixed signals about the exam, with a TF holding office hours during the exam, in which students could hear each other's questions about their proposed answers, and the TF's responses.

Of course, the exam instructions were clear: No collaboration. But 60mph speed limit signs also could not be clearer. If you go 63 for an hour and people are passing you, and then you pass a speed trap and see that somebody who raced passed you at 75 has been pulled over, you get used to the idea that maybe you, everyone else, and the police all agree that 60mph signs don't literally mean what they say. It sounds to me like some students were going 75--they were cutting and pasting--and some of the other students who have been pulled in were going 63. Hopefully the Ad Board will respond in proportion, though I think those 63mph students have some justification for rolling their eyes about their culpability. Those who keep insisting that the students should be punished because the exam instructions were clear will, I trust, not complain if they are ticketed for driving 63, like everyone else,  in a 60mph zone.

Another thread (represented in a Globe editorial today, and an op-ed yesterday) suggests that if there is an institutional failure that contributed to the culture, it's that the course was too large. Again acknowledging that I know nothing in particular about this course or its professor, I think it's important to parse that a little bit--it has a germ of truth but it's not quite right. There are great courses that are large, and there are awful courses that are small.

First, why is a course large? If the course is large because the professor announces at the beginning that it is going to be a gut, there should really be no surprise that the students don't take it seriously, but the size itself is not the problem. It's that students put out in proportion to what is expected of them, and a professor who says that his course is a gut has announced that he doesn't expect much of his students. Of course they may misunderstand each other about what exactly that means, but the professor surely intends to signal SOMETHING if he says that.

A course can be large because it's a good course about an interesting subject. (Harvard's intro CS course is like that.) For certain courses, such as introductory economics, it doesn't matter that much whether it is a great course or not--the importance of the material will create demand, if it is the only intro to the field.

A course can also be big because the professor is famous. I'll bet Niall Ferguson draws crowds, even beyond what the high quality of his lecturing would produce.

Big courses, whatever the reason they are big, create management challenges. There are grading equity issues. There are problems with having adequate TF staffing, especially if enrollment is unexpectedly large.

And there is the problem that the guy in charge of the course may have no idea what is happening on the ground, among the students. That can't happen in a seminar, and is unlikely to happen in a lecture course of 20. It can happen very easily in a big lecture course with a lead instructor who is a celebrity or is very busy with other things, and has left the management of the course to his staff. It is hard to talk about this without naming names, and again I would stress that I don't know the first thing about Niall Ferguson and the way he runs his courses. But you have to wonder about a professor, affiliated with both Harvard and Oxford, who brings up a page beginning with the following Google results when you append "speakers bureau" to his name. (Only one of these lists a fee: $40,001 and up.)

I don't imagine that the professor involved in the current case is operating at that level. But isn't he taking his cues about personal and institutional values, as he works his way up the tenure ladder, from observing the behavior of his more celebrated peers?

One of the unhealthy divides in the faculty (not just at Harvard of course, but at any research university) is between those who lecture, and those who actually look forward to everything that teaching entails--not just lecturing, but to all the grungy work of educating students, in small numbers for sure, but in large numbers too. It is a lot of work to run a big course well--you in essence have to teach a second course, to the teaching staff. You need to tell them that you want them to take the temperature of the class and tell you the buzz. You yourself need, at least occasionally, to hang around the places where your students are talking over coffee or hacking on their laptops. You don't need to be a crack bureaucrat to run a big course, but you have to act like you care how the course is being experienced by your most ill-prepared and insecure students. (Of course, the top students too; they bring you the greatest joy. But they are not going to be your problem cases.) Because if those weaker students don't think you care, and your TFs pick up the same attitude, then your students will try to figure out how to navigate your course--what those 60mph signs mean--from their peers, from gossip, etc.

This is the problem of parallel universes in big courses. Not a new phenomenon--Lowell noted one of the problems Eliot had left him by hugely expanding the size of the College. "The personal contact of teacher and student becomes more difficult. . . . Great masses of unorganized young men . . . are prone to superficial currents of thought and interest, to the detriment of the personal intellectual progress that ought to dominate mature men seeking higher education."

The incentive and reward system is stacked against this kind of student-faculty engagement. There are papers and books to write to get yourself promoted and enhance your professional reputation. There are conferences to go to, or better yet for your standing in the academic community, to organize and run. You may be able to save something for your children's education if you take on outside work as an expert witness or a speaker or columnist, while your TFs are grading papers and holding office hours. And the only data about your teaching that you know will go into your tenure dossier, if you are junior faculty, are your Q guide ratings and enrollment numbers, so whatever else you do, you have an incentive to keep those high.

I enjoy those invitations to Faculty Dinners in the Houses, but have always found them rather odd--do the professors who go to them never talk informally to students otherwise? The trick in holding a big course together is to avoid the phenomenon of parallel universes, where the teaching staff talks to each other and the students talk to each other and the two groups have little to say to each other. The system of incentives in research universities is not set up to encourage mixing of those universes; there are plenty of great professors who do it, but they don't do it because it will show up on their annual Activity Reports. It is hard for me to imagine that half of a class could be charged with cheating in a course where the engagement and mutual understanding was at a healthier level.