Thursday, 21 July 2011

Catching cheaters

Panos Ipeirotis notes the extreme cost to lecturers of pursuing cheaters.
By the end of the semester, 22 students admitted cheating, out of the 108 enrolled in the class.

The process of discussing all the detected cases was not only painful, it was extremely time consuming as well.

Students would come to my office and deny everything. Then I would present them the evidence. They would soften but continue to deny it. Only when I was saying “enough, I will just give the case to the honorary council who will decide” most students were admitting wrongdoing. But every case was at least 2 hours of wasted time.

With 22 cases, that was a lot of time devoted to cheating: More than 45 hours in completely unproductive discussions, when the total lecture time for the course was just 32 hours. This is simply too much time.

When 1 out of 5 students in the class being involved in a cheating case, the lectures and class discussions became awkward. For the rest of the semester there was a palpable anxiousness in class. Instead of having friendly discussions, the discussions became contentious. Not a pleasant environment.

This, of course, had a direct effect to my teaching evaluations. Instead of the usual evaluations that were in the region of 6.0 to 6.5 out of seven, this time my ratings went down by almost a point: 5.3 out of 7.0. Instead of being a teacher in the upper percentiles, I was now below average.

The Dean’s office and my chair “expressed their appreciation” for me chasing such cases (in December), but six months later, when I received my annual evaluation, my yearly salary increase was the lowest ever, and significantly lower than inflation, as my “teaching evaluations took a hit this year”. (And I publish in journals not endorsed by Business Week, but I will leave this story for another time.)
Josh Gans explains that optimal structures take enforcement out of lecturers' hands so that enforcement is more credible.
So what should the administration do to change the equilibrium? First, take the enforcement of cheating out of the Professor’s hands. Employ a policing person who is the one who receives the Turnitin results and decides what to do about it. Make it their job and you solve the free-riding problem right away. You also allow cross-course, repeat offender detection. And you change the cheating equilibrium to one where enforcement is expected. Finally, the Professor has their hands clean and so suffers no conflict over teaching evaluations.
Canterbury isn't far from this. Everything is run through TurnItIn. Anything that looks dodgy, I hand off to the Head of Department. There's still a short meeting between the student, the HoD, and me, but it's short and the HoD is the bad guy. And there's reporting within the College (and hopefully extending within the University) on proven cases of plagiarism such that track records get established; we then can apply schedules with increasing marginal penalties.

This was all getting pretty reasonably firmed up in the pre-quake world. I'll be using TurnItIn again this semester; hopefully, I won't have to haul anybody into the HoD's office. But TurnItIn has embedded itself nicely here; after a few years, most students have cottoned onto that they can't get away with the dumber forms of cheating.


  1. I was an associate dean at Carleton for 6 years. The prof who suspected cheating would consult with the Chair of Department, who would then turn the case over to the associate dean. Two associate deans would then interview the student, and decide on the case, and the penalty. The ombudsman was also normally present at that interview, if the student wished.

    Most students would immediately admit what happened when shown the evidence. So meetings were short. Many would plead extenuating circumstances. We kept university-wide records, and repeat offenders faced much stiffer penalties. Because they knew this (we warned them), there were very few repeat offenders. And this meant we could be relatively lenient with first offenders, who were often clued out or sad cases. And we could be relatively consistent in the penalties we imposed, unlike the profs.

    It's a very bad idea all round to have the profs handle instructional offences.

  2. @Nick: The problem at Canterbury was that the Provosts who handled disciplinary matters were seen to be very soft absent extraordinary levels of evidence. So we had to bring a pile of enforcement to the department or college level. And from there it's been working back up.

  3. We associate deans had a bit of a reputation for being a bit soft too. But in part that's because we got to hear the student's side of the story too, and not just what the prof saw.

    There's a general principle here. Assume the probability of student i's cheating, Ci, is a positive function of the student's degree of innate evil Ei, and the student's extenuating circumstances Xi. Ci=Ei+Xi.

    Let E and X be normally and independently distributed across the student body. Those caught cheating will tend to be more evil than average, but will also have more extenuating circumstances than average. We got to observe (with error) those extenuating circumstances, and the prof didn't. So our estimate of the cheating students' average level of evil (and hence penalty) would be lower than a prof who did not observe the right tail of the distribution of extenuating circumstances, and assumed the average student caught cheating had the population average level of extenuating circumstances.

    (Wish my econometrics were better, so I could formalise this argument better.)

    For example, some students cheated where it would be certain they would be caught. Like plagiarising the class handout on "how to avoid plagiarism". Obviously still unclear on the concept. Remember the McNaughton/M'Naghten Rule? "Would you have done it if a policeman were standing at your shoulder?" In some cases, there *was* a policeman standing at the student's shoulder!

    We didn't allow profs to deal with it, even if they wanted to. (And we could always suggest that profs who didn't like our rulings apply for the job of associate dean!)

  4. I think our problem was that the Provosts would put next to zero weight on prior offences that hadn't been taken through the Provosts' office, and we dealt internally with anything that wasn't worth the rather high costs of taking things through the Provosts.

    A decent equilibrium is everything goes through the provosts right away, but that requires way lower cost of getting cases into that system. Another decent equilibrium is having everything handled through the Heads of Department with records going into a university-wide system for catching repeat-offenders across departments. Mixed systems cause problems.

  5. So in our case, the profs all also saw the extenuating circumstances before sending anything on; I'm not sure that asymmetric information between committee and prof was responsible for the provosts being seen as having a light touch.