2 ms·

I think there’s a difference between looking at an old exam and its answers, figuring out why each was right and wrong, and learning from it, and just finding a

• strken 3y ago
  The interesting part here is that at least some of the students might have thought it was an old exam from the same course, used it for studying, learned totally nonsensical answers, and then gotten confused in the actual final. Evidently not all of them straight-up copied the answers, or they would have gotten much higher scores. This is apparently still a violation of his institution's academic honesty policy, but there are interesting questions about what happens if he e.g. poisoned one member of a study group who went on to teach bullshit to the other members, giving them a higher chance of picking the wrong answers and putting them over the threshold without ever knowingly cheating. Of course this is irrelevant, as the p<0.01 test gives an up to 62%[0] chance that someone in his class of 96 students could be falsely accused of cheating, and is unethical in itself. Apparently one student managed to be right on the threshold, which the professor has chosen to interpret as "the student possibly cheating" rather than "me possibly ruining my student's academic life for no reason by choosing a test with lots of false positives and applying it to a class of 96 students". [0] 1-(0.99^96), and though the real chance is probably a lot lower if the questions are obviously wrong, it's still a stupid threshold to choose