A modest proposal (one having nothing to do with cannibalism)

Hmm.

In the past week or so I've had several really good conversations about this past semester's revisions policy with students in both Topology and Calc II. A few things are clear from these conversations:

1. A large number of students benefited from the policy of allowing unlimited revisions (on exams in Calc II and on homework in Topology). Moreover, they benefited in exactly the ways I would hope that they'd have benefited: they profess to being much more eager to look over their past mistakes, to understand what they did wrong, and to attempt (sometimes over and over again) to correct their errors.

2. A much smaller number of students played the system like a two-dollar fiddle, not taking adequate measures to prepare for the exams (or the homework), putting off studying various topics until after the exam's been given, graded, and returned (or until after time's been found to corner one of the class's brighter students and ask said student for a transcription of her or his homework).

3. Something needs to be done to counteract the inevitable effect of procrastination, so that I'm not socked with a hundred revised exams and homework sets to grade at the same time final exams come due. (I apologize if the mass of grading I had to do during the past week as more and more revisions came in made me a bit more snippy of late.)

One of my Topology students made the following suggestion (thanks, Karl!): include with each successive revision an "expiration date," which, once past, precludes further revision. For example, I might allow a student a week to complete each iteration of a revised assignment: say I hand back a graded assignment on Monday, May 10th; the next revision, should a student wish to undertake said revision, would be due on Monday, May 17th. If that date passes without revision, the student's out of luck. If the student does perform some revision and receives the paper back on Wednesday, May 19th, then he'll have until Wednesday, May 26th to effect further revision, and so on. It would be the student's responsibility to be present in class to pick up graded revisions.

This would definitely help address the third observation above. Moreover, I think it would encourage more students to take advantage of the revisions: instead of putting them off until the inevitably busy end of the semester, students would be pressed to perform revisions right away and receive the benefits thereof.

The second observation above remains unaddressed. What to do about those who play the system? I will say that to some extent the chickens came home to roost when the final exam was graded: it was clear to me from performance on the final exam which students cruised on through with little concern for actual understanding of the concepts treated in the course. However, perhaps it might be appropriate to provide some sort of system of diminishing returns, at least for the lower-level courses (like Precalc, Calc I or Calc II) in which I might allow unlimited revisions: the first round of revisions will earn 1/2 credit back, the next 1/3, and so forth, until after two or three rounds there's really no purpose, from a "credit" standpoint, to revise. I'd like to think the best students would continue to revise anyway, striving for a perfect theoretical mark and a more and more perfect understanding.

Anyway...that's where I stand right now.

FYI: I just kicked off my summer leisure-reading season by starting Jonathan Kozol's Ordinary resurrections: children in the years of hope, and although I'm only about thirty pages in (and enjoying it immensely), it's making me more and more aware of the need for "ethnographic" methods in the "hard" sciences. I suspect this coming summer will see me write a number of posts on this blog in which I discuss the humanistic side of mathematics, in which I experiment with what Laurel Richardson would call "creative analytic practices." Stay tuned!