What to Do With Cheating Students?

(Photo: Alberto G.)

I’m nearly certain that a pair of students cheated on my final exam—the probability they had so many identical answers on the multiple-choice exam is infinitesimal.  If I pursue them, it takes me time, and there’s no assurance they will be found guilty.  If I don’t, I’ll feel badly about giving them an undeserved grade.  Even for fairly risk-averse students, cheating seems like a good idea.  I doubt that most cheating is caught; and unless the penalty is very severe (expulsion) and/or the students’ costs of contesting the accusation are high, and both are very well-publicized, the incentive to cheat for students with weak consciences seems overpowering. To salve my own conscience I’ll report them, although it’s probably a waste of my time; but I doubt that reporting them will deter their future cheating or deter others very much.

Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.



View All Comments »
  1. fraac says:

    Hidden due to low comment rating. Click here to see.

    Disliked! Like or Dislike: Thumb up 4 Thumb down 15
    • Amanda says:

      While I appreciate your honesty here, you do not have “very high moral standards” according to the rest of that sentence.

      Well-loved. Like or Dislike: Thumb up 14 Thumb down 0
  2. Bill says:

    The similarity of answers is insufficient. Do you have other evidence that they were either cooperating (both cheating) or that one cheated off of the other? In the latter case, only the one should be reported.

    Thumb up 2 Thumb down 2
  3. kerry says:

    What would Joe Paterno do?

    Thumb up 3 Thumb down 4
  4. Eric M. Jones. says:

    Gee! I seem to have lost your exams. You two will have to take the test again or take an incomplete.

    Well-loved. Like or Dislike: Thumb up 10 Thumb down 3
  5. Hauke says:

    If you can’t catch them red-handed you can’t do anything. Yes, the probability is infinitesimal but it exists. I don’t know about your specific examination regulations but don’t you have any chance for an oral examination or do you have to let them pass when they reach a specific score?

    Reporting them is still a good idea as it will rise awareness to the issue even if these students will come out lucky this time.

    Thumb up 0 Thumb down 1
  6. right and wrong says:

    It’s your moral obligation to turn in both of the cheaters, no matter what it may cost you in terms of time or effort. Good luck! All of us who DIDN’T cheat in college are behind you, and we thank you!

    Thumb up 2 Thumb down 1
  7. Colin Fredericks says:

    As a fellow teacher, I urge you to report them.

    Most cheating is not caught.

    Most cheating that is caught is not reported.

    Because of this, someone who cheats in your class is almost certainly cheating in every other course they’re taking. The few times I’ve caught someone cheating, they’ve been doing it elsewhere too.

    This isn’t just about your conscience, this is about your integrity as an instructor. Students who cheat through a course don’t learn; I’ve seen substantial evidence for this in educational research. Beyond the ethical issue of cheating is the need to have what we say about our students mean something. Please, don’t be the professor who gives a pass to students who don’t know the material.

    Thumb up 3 Thumb down 0
  8. Cdub says:

    “the probability they had so many identical answers on the multiple-choice exam is infinitesimal”

    Infintesimal? Unless you were giving out individualized tests (this was common in my high school, but I don’t see it much at my college), I’d expect students who have sat in the same class and been taught the same material by the same person to have identical test answers more often than one would expect by chance. I don’t have my stats textbook in front of me, can someone describe the calculation you’d need to determine probability of this? Would this be a binomial distribution problem?

    Thumb up 2 Thumb down 0
    • thomas says:

      Student responses on each question are not randomly selected and should gravitate towards the correct answer (assuming the questions are not too hard). In addition, incorrect responses tend to be close to, but not as correct, as the correct response option, hence student choice may gravitate towards specific response patterns (frequency distributions for student responses would likely show a non-random distribution of responses from your students).

      Probability of identical scores would be increased if two people studied for the test together, but more importantly the probability increases as students get more questions correct (as there is less scope for distinct patterns of response).

      This means that overall, multi-choice questionairres lend themselves to identical patterns of response. And identical response sets are not a good measure of whether or not someone is cheating, especially when the students are performing well.

      Getting identical response sets is a numbers game, and you stack the odds if you only look at high performing students.

      If you do still suspect cheating If consider examining how well their “cheating” test scores correlate with previous measures of performance and whether they had previously obtained identical results.

      Thumb up 2 Thumb down 0
      • Sam Prince says:

        Presumably if you are going to go down the route of using multi-choice questions (I guess it’s seen as a necessary evil in US universities; it’s much more rare in the UK) software could come to the rescue to identify possible cheats.

        Given a large enough group of students taking an exam, you could find the distribution of answers for each question and from that work out the likelihood of each of the incorrect answers being given. From there it wouldn’t be too much of a stretch to find pairs of students with improbably well-correlated wrong answers. Assuming the cheating must be over-the-shoulder or side-by-side peeking, you could test its reliability quite thoroughly before going live with it by checking how often alarm bells rang with students sat in close proximity vs. students sat at a distance.

        Thumb up 0 Thumb down 0