3

I took a university course where the final exam was a multiple-choice type with the answers going into a "scantron" sheet that gets fed into a machine. This is reasonable, since there were about 600 students taking it.

My final grade for the exam was such that I was about 1% short of the next letter grade for the course overall. After all the marks were posted, the exams were made available for pickup to students. I picked mine up and found two questions that I was sure were right but were marked wrong, so I went to see my professor to review the exam and ask about them.

He gave me two answer keys to check - one was the master scantron that was fed into the machine to tell it the correct answers. The other was the master paper exam with the questions (same thing the students got on the exam), where the correct answer was circled. I checked the two questions I was interested in and sure enough the paper exam agreed with my answers, but they were marked incorrectly on the master scantron that was used to grade every student. In other words, whoever filled out the answer key scantron made an error on at least 2 questions and filled in the wrong bubble.

I showed the prof the error and he seemed completely unfazed and not surprised and simply said that yes, he agrees that I picked the right answer and since this actually bumps me up a letter grade he'll fill out the form for the university to adjust my already-posted course grade. He then simply put away the two answer keys and bid me good day.

While I have no concrete proof, I feel like he already knew about the errors and me finding them was not a surprise. However, it seems that he does not intend to regrade all 600 students and is only dealing with the ones that actually come in and spot the error. Is this an acceptable approach? So far he has sent no communication out to the class regarding this and to my knowledge most of the students are unaware of the error.

  • 4
    In his place, I re-did all the markings. Not 600, but 93, written answers (statistics). A hassle, to be sure, but the only way to be fair. And yes, I had to do that every time a student managed to change my opinion on the markings... – Fábio Dias May 4 '18 at 18:21
  • 1
    Would you have a way to know if the professor individually contacted students who were at borderline grades about reviewing those answers? Contacting the whole class would be unnecessary. – cactus_pardner May 4 '18 at 21:11
  • 1
    @cactus_pardner I wouldn't know, but at the very least he didn't contact me. Although it's possible that I saw him before he got a chance to do so, or that I was actually the first to spot the error. His behavior made me think that he already knew about it, but I'm just guessing. – Egor May 4 '18 at 21:21
  • 2
    you could spread the information and make 600 students come to the office to see their results ;-) - re-grading will be a matter of minutes ;-)))) – OBu May 5 '18 at 14:23
  • 1
    @OBu and isn’t social media wonderful... – Solar Mike May 5 '18 at 14:28
10

For me, all the papers should be re-graded with the corrected marking scheme - ethically correct and supported by any future checking.

Edit: And as per Fabio’s comment - have had to do similar corrections due to my error or poor question so it happens but not too often...

6

Ethically, of course he should regrade. I don't even see how this is in question; how could it possibly be fair to assign grades based on faulty information and fix it only for those who complain?

This is a huge hassle. I'm not going to defend the professor; he made a mistake and he should fix it. But my guess is that regrading would bump most people up by two points, so he would have to adjust the curve by two points, and only a small number of students' grade would change (some up and some down). Then he would have to submit grade-change forms which, in this quantity, could raise some eyebrows. It's easy to see why the professor would rather just fix it for the small number of students who complain.

Whether you can force him to grade it properly is largely down to your individual university administrators. The department chair, the dean, and the ombudsman should technically be concerned about this, but it's anyone's guess whether they'll choose to care given that it will only affect a few students with borderline grades anyway.

  • 2
    You're assuming the exam is graded on a curve, which might not be the case. – Massimo Ortolano May 4 '18 at 20:46
  • True. In fact, either way, I doubt the prof would want to lower students' grades weeks after the fact -- he would probably do grade increases only...which makes doing regrades even less appealing for him (though he should do it anyway). – cag51 May 4 '18 at 23:32
  • 6
    It’s not that big of a hassle; he only needs to regrade the exams of students whose course grades could be changed by a regrade. Even in a class of 600 this is likely in the small dozens. – JeffE May 5 '18 at 13:02
5

Since these exams were graded by a machine, it should be easy to correct the grading of those exams that the professor still has. If that's a large proportion of the total, then the exams should be re-graded. But if most of the exams have already been picked up by the students, then there's not much the professor can do to achieve fairness.

  • They actually keep the originals on file. If you come in to pick up your exam you can either review it right there and give it back or you can ask for a photocopy of your answer key and they'll print one out for you. – Egor May 4 '18 at 21:38
  • 1
    With all the originals available, I'd certainly prepare a corrected answer key and feed it and the exams through the machine again. The only issue I'd find difficult is whether I should lower any grades as a result of the corrected grading. (Probably very few students will be so close to a borderline between two grades, so there will be very few grade changes in either direction, but, as long as a machine can do the tedious work, I might as well take advantage of that to try to repair the error.) – Andreas Blass May 5 '18 at 0:01
3

The effort that should go into correcting things is proportional to the magnitude of the problem. If the mistake on the exam is a one-point error that amounts to, say, 0.1% of the final grade for the semester, I’m only going to look at the exams of anyone who would be directly impacted by such a small change. If it’s 5% of the final grade, I’d need to look at everybody’s exam just to make sure everybody was graded fairly.

  • The exam is 90 questions, so this was 1/45th of the exam. The exam is worth 50% of the course grade, so it's just over 1% of the overall course grade. – Egor May 7 '18 at 22:41
  • In that case, I'd look at the exams of everybody within that 1 point above or below the grade cutoffs. (We only use A/B/C style grades at my school, so this would cut down my work about 80% or so. – aeismail May 7 '18 at 22:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.