April 24, 2014

avatar

On the Harvard “Cheating” Scandal

The news that Harvard is investigating more than 100 students on charges of unauthorized collaboration on a take-home exam has, predictably, led many commentators to chime in. No matter who you are, a story like this is likely to trigger one of your hot buttons, whether it’s the declining moral standards of kids these days, the moral core of elite educational institutions, the inherent injustice of top-down rulemaking, or whatever. Not to mention that the course was “Introduction to Congress.”

Much of the commentary has been sensible, and some has even managed to withhold judgment on the underlying “cheating” issue, about which few details are known to the public. Having served on Princeton’s Discipline Committee (which heard all cases of academic fraud except those involving in-class exams), I know how fact-specific these kinds of cases can be. Exactly what did the instructions say? Exactly what did the students do? If they claim to have misunderstood the instructions, was it a reasonable mistake, or did they willfully distort the instructions and avoid asking for clarification? And so on. Depending on the facts, there might not have been any “cheating” at all.

Perhaps the strangest commentary on the Harvard incident was Farhad Manjoo’s piece at Slate, whose argument can be summarized by its title and subhead: “There is no Harvard cheating scandal. The students should be celebrated for collaborating.” He argues that even if the students had clear instructions not to collaborate at all on the exam, and even if they collaborated completely and openly, they should not be punished. Why? Because people collaborate in the real world, and college should be like the real world.

Even leaving aside the fact–and useful lesson–that breaking a rule can get you punished even if the rule is stupid, Manjoo’s argument doesn’t hold water.

It’s true, of course, that collaboration is important, and that schools should be teaching people how to work in groups. I often ask students to work in groups, not only to teach collaboration skills but also because group assignments can be larger and more complex–and hence more realistic–than single-student assignments. But this doesn’t imply that everything in my course needs to be collaborative.

For one thing, people in the most collaborative professions, such as team sports, do in fact train alone sometimes. Basketball is a collaborative game–there is always a team on the court trying to cooperate toward a common goal–but basketball training involves a mix of group and individual activities. Players might practice their shot alone, or lift weights alone, or run alone, to make themselves better contributors to the group. It stands to reason that academic training would be the same, encompassing some mixture of group and individual work.

More to the point, the Harvard work was an exam, and exams are more about evaluation than about training. If the coach wants to evaluate basketball players to decide which ones to choose for the team, at least part of the evaluation will measure their individual skills. Can the player consistently make open three-point shots? How high can the player jump? How far can the player run without getting tired? It won’t do to let the group of players pick one person to do the shooting, another to do the jumping, and a third to do the running.

I don’t know if any Harvard students deserve punishment for breaking the rules. But I do know that if they were clearly instructed not to collaborate on the exam, and they collaborated anyway, it’s fair–and necessary–to punish them.

Comments

  1. dr2chase says:

    I’m surely biased, but I’ve long been a fan of Rice’s honor code pledge. Short, easy to remember, to the point, not very ambiguous. It would be interesting to know what Dan thinks of it, if he is comfortable expressing an opinion. (“On my honor, I have neither given received any aid on this _____” — and whatever is “pledged”, that’s the rule. Homework was often not pledged, and collaboration there was common.)

    On the other hand, at least one account I’ve read of the Harvard thing implied that it was a “mere” instance of insufficient citation.

  2. Barry Kelly says:

    Read TLP’s take on this: http://thelastpsychiatrist.com/2012/09/the_harvard_cheating_scandal_i.html

    The only other thing I’d say about moral standards / honor codes etc., is to refer to the worst argument in the world (and I believe it is, but before I read this post, I didn’t have a name or specific place to point to for it): http://lesswrong.com/lw/e95/the_worst_argument_in_the_world/ .

    There’s a lot of snap judgements made of every facile kind, including yours. Everyone optimizes their thought patterns to fit it into the most available story pattern they have, and finish it up with some trite truism like “if they disregarded instructions on an exam, they should be punished”.

    • Ed Felten says:

      I have to admit that I don’t understand TLP’s argument. He/she seems to think that everything about Harvard, the course, and the students is bogus, so it doesn’t matter if anybody cheats. In which case, why should he/she care one way or the other what Harvard does?

      • Nathan T says:

        Wow, to see Ed Felten who’s analytical thinking skills I have admired for many years now admit to not understanding something surprises me. I really liked your article, but was very curious about this other article when I read that you didn’t understand the argument.

        Let me have a hand at decoding TLP’s argument, at least as I understand it.

        Key phrases: “What does the professor want?” “everyone in that class cheated:” “with no internalization of the ‘knowledge’”

        The real question from which he bases his entire argument (or was it just a rant against college?): “I am forcing you to ask whether 125 people simultaneously cheating might be indicative not of a sudden resurgence of Satanism but an outbreak of encephalitis, with Professor Platt as Patient Zero?”

        His argument isn’t that they cheated or did not cheat–nor whether or not they should be punished, but whether or not the Professor is capable of “teaching” the students in a way that actually “imparts knowledge.”

        He first premises the article with the rule that there would be no collaboration with other students, professor, aids, etc. on the exam, in other words they did indeed cheat, but his argument is that they were forced to cheat because there was no other way to take the exam. The professor failed to impart knowledge, but instead used a faulty system of grading to determine whether or not people pass or fail or what grades they get. The only possible way then for them to get any grade was to copy either from each other or from the professor himself, they had to collaborate to know “what does the professor want?” They had to cheat.

        A comment on his article articulates the view in a simple paragraph: “Yes but don’t you think the fact that ‘this isn’t news’ and that everyone knows you’re supposed to give teachers want they want rather than actually gain knowledge and critical thinking ability is sort of a problem with college education?”

        Is college really to gain an education or is it merely to get the graduate paper. Is there knowledge to be gained? I suspect there is, but I myself did not find college a good method of imparting said knowledge and chose to gain my own education elsewhere. For instance from reading your blog posts which entail critical thinking and spark my own critical thinking skills.

        And, yet, you are a product of such academia system that TLP rants against. What does that say? I think it only indicates why you did not understand TLP’s argument that it wasn’t so much an argument about the scandal at all, as it was a rant (against the nature of college courses and those who end up becoming professors and the nature of research).

        • Ed Felten says:

          I think you’re agreeing with me that TLP wasn’t really making an argument at all, just a rant. For me, it fails the main test for a rant, which is that it should be funny. (I’m not offended by it, just disappointed to have spent time trying to figure out what the point was.)

  3. paul says:

    I think there’s an even more apposite example of the limits to collaboration in the real world. Workers, even from different companies, collaborating to complete a construction project: good. Workers from different companies collaborating to decide the supposedly-competing bids they will submit to a client: very, very bad.