Tuesday, March 22, 2011

Cheating: a side note

These are problems with the instructor or classroom environment, not necessarily the student. In this model of the world, students are just trying to do the best they can, given the environment they face. If the pressures are too great, the communication and enforcement of the rules too weak, the long-term objectives too muddled and the motivation for the subject is too little, who can blame the student for cheating?
Yeah, it feels to me like the social rhetoric surrounding cheating makes it easy to pass too much of the blame onto the students, who -- for the reasons we've been discussing -- aren't necessarily in the best position to prevent it. If we replaced (or supplemented) "How could you" with "how could I let you"...maybe a lot less cheating would occur. The best classroom environment is one in which people can't cheat, or in which they will surely be caught and therefore don't cheat, or in which the penalty of cheating is so high that nobody dare cheat. (There's a tension here though: You have to be willing to punish people who cheat, even if you privately realize that, in truth, cheating makes its own existence morally ambiguous).

So maybe we should be pushing harder on teachers. But if so, how? How do you incentivize a teacher to catch cheating? The trouble is, the teacher is also the one who monitors cheating. How can you punish them for failing to catch something when they're the only one who knows whether there was something to catch or not? Who watches the watchman? It is an interesting problem, and not without solutions. Let's plant students to cheat every so often, and punish the teacher for failing to catch them!


  1. What if teachers weren't allowed to proctor their own exams or grade homework for their own classes? Another teacher would do the proctoring and the grading. They would be financially rewarded for catching cheating.

    On a side note, what about when teachers (or the watchman in general) has an incentive to cheat? I believe this type of rampant cheating among teachers is mentioned in Freakonomics.

  2. Greg,

    At first I wasn't sure what swapping proctors accomplished, but let's think about rewards for a minute. We're in a bit of a funny environment: while a teacher's negative reports ("there was no cheating") are not verifiable, positive reports effectively are, assuming a teacher is not willing to make false accusations. So while we can't trivially punish her when there's cheating, we can trivially reward her for *catching* cheating.

    But of course, our goal isn't really to catch as many cheaters as possible. The difference between teachers and police is that the teachers are both the legislators and the enforcers: they create the classroom environment *and* enforce its rules. Perhaps if we switch proctors and have rigid external standards for what happens to cheaters (as we certainly wouldn't want a teacher to make the punishment lax, so that more cheating and more cheating-catching will occur), then there is nothing left of the exam environment in the proctor's control, besides detection of cheating. Then rewards for catching would seem to work as cleanly as punishments for failing to catch.

    Actually this will not take us the whole way though. We still need the approach of the source I linked to at the bottom.

    There is also the concern that students would resent teachers out for profit who bust them, in a way they might not if teachers were instead forced to bust cheaters to save their own skins.