I am a graduate student who teaches their own class. I have a supervisor who I report to for everything, including final grades. We have a meeting every semester at the end to review our final grades and for my supervisor to see that I did everything correctly.

In my second semester, my supervisor looked through my grades before they were submitted or published. At the end, she said: “Let’s look at your borderline students.” She noticed that a student had a borderline C, say 73.5%, which is barely above the passing threshold of 73% for this course, and had us look at his final exam. The student got 66.5% on the exam, and she said to me: “Are you sure you want to pass him?” At the time I was thinking: “Well, the student did earn the passing score.” However, my supervisor kept suggesting: “Are you sure?” I looked up to my supervisor who was the one who taught me to be a teacher, and like anyone at their job listening to their superior, I felt pressured to say yes, because you do what your supervisor tells you to do, and I was told to take off points from his final exam to lower his final grade to 72.36%. The sole reason for this was not to let the student pass. My supervisor also mentioned specifically not to keep it higher than 72.5%, because students will ask for a bump. So, in the end, I ended up giving him that 72.36%, a C−, which is a failing grade: They cannot move on the the next-level class.

The discussion happened before the grades were submitted or published, so the student never knew he originally scored 66.5% on the exam. There were no other students with borderline scores.

I felt extremely guilty because I knew who this student was; they didn’t do much of the work, they didn’t always come to class, but they knew the topic and did the bare minimum to pass. I went back to my office right after our meeting and was tempted to change it back to a C. But knowing my supervisor, I knew there was a chance I could’ve gotten in trouble because my supervisor might’ve noticed.

I feel extremely bad for doing this, I felt like I was pressured to do this and couldn’t turn back.

Was the supervisor's decision ethical, academically, and legally sound?

  • 1
    Answers in comments and presumably answered requests for clarification have been moved to chat. Please read this FAQ before posting another comment. Also please check the chat whether you concern has already been voiced.
    – Wrzlprmft
    May 17, 2020 at 12:38
  • 3
    Responders seem to be assuming that your supervisor knew that student -- that she looked at the name on the final exam and personally didn't want him to pass. The other alternative is it was simply the low score. Maybe clarify that? May 18, 2020 at 3:14
  • 2
    In my undergraduate Mathematics degree, we were warned this could happen from the outset (albeit in the course's equivalent of small text) and, if I recall correctly, set out a formal framework for it happening. It is possible this behaviour is mentioned somewhere in the written material, or elsewhere, that is given to students?
    – Isaac
    May 18, 2020 at 8:42
  • 3
    Please note the above FAQ about Why do moderators move comments to chat and how should I behave afterwords? In particular, there is no way to move comments to chat more than once, so we have no choice but to delete additional discussions or answers-in-comments (whereas comments posted to the chat are essentially permanent).
    – cag51
    May 18, 2020 at 21:17
  • 2
    What subject was this in? How subjective is the exam mark? May 19, 2020 at 9:57

16 Answers 16


Since you ask, specifically, about legality, I'll say that it is unlikely that it would be a matter of criminal law anywhere, though, of course, I can't know the laws of every jurisdiction. But it could well be a matter of civil law and if this were known could easily generate a lawsuit against everyone involved.

I think that a person would be very hesitant to apply such a "rule" to the child of some powerful person. It depends, I think, on people either not knowing what really occurred or being so powerless as to be completely unable to prevent it.

But, as I've been saying in comments here, it is a profoundly unethical act and would, in my view be grounds for at least a reprimand, and even termination as an act of moral turpitude if repeated. It turns what should be a well determined process in to a form of chaos where we are just at the mercy of the powers.

An instructor needs to set a standard before the course begins and needs to make that public. You can make it as hard as you like and you can interpret it as strictly as you like, but you can't deviate from it to the detriment of an individual. I have no complaint about generous interpretations as most of us, I hope, consider generosity to be laudable and to believe that students can learn and improve.

I won't condemn you personally, however, since you may have had no agency here and may be as much a victim as the student involved. But I will condemn the professor, who should, at a minimum be reported to the dean for this behavior. Especially since the course is a gateway course and changing their grade to make it failing when it was not also denies them the opportunity to go on and, one hopes, improve.

It was especially pernicious, here, that the induced change in the average was set not only to unethically push the student below the passing grade, but further down so that it would be less likely they could appeal.

My advice to you personally, is to find a better mentor as soon as you can. This person is teaching you to do harm.

And, I don't know how you can make it up to the person you damaged. It may be too late, even to request a grade change. But the dean should be informed and if it is only you that knows that the professor is behaving this way, it is probably your responsibility to so inform them. I won't suggest that you do, of course, since it might have too adverse an effect on your own career, but there is an imbalance in justice that hasn't been corrected.

  • 43
    Turpitude, pernicious, agency, lol. I love the writing. I suppose I should expect this from academia stack exchange.
    – Joel
    May 17, 2020 at 12:33
  • 29
    @Joel that is why Buffy has earned the rep shown....
    – Solar Mike
    May 17, 2020 at 13:01
  • 19
    This may not take into account the cultural setting. There are definitely cultures where grades are expected to be invented by the instructor based on a number of factors of which the mathematical points - on a published scale or not - are only one factor. That isn't something I personally like or would choose to participate in if at all possible, but it exists. This question does ask about the US and I might say this answer fits the US -- perhaps it might make that clear.
    – Mike M
    May 17, 2020 at 19:38
  • 2
    Agreed. Capricious punishment, where someone gets punished even though they could reasonably surmise that they wouldn't be, is very wrong indeed. May 18, 2020 at 1:53
  • 3
    @CGCampbell, and others. Note that there is one data point here. One grade in one course has ended the potential career of someone. One can certainly advise a person that a field may not be right for them, but this was something else. I've actually had students with a very poor academic history turn into superstars in a difficult course.
    – Buffy
    May 18, 2020 at 19:04

Any you in my answer is general. I am discussing the general ethics or the ethics of your professor’s actions here, not yours.

You have a formal grading system (exams with points, a grading rubric mapping percentages to grades, some formula to derive the final percentage from the exam and other contributions, etc.) to make grading transparent and thus more objective and fair. The alternative is determining the grade by gut feeling only, which you would probably not be allowed to and which opens the door for numerous unfair biases, be they conscious or not. You can regard this system as an imposed control mechanism, but you can also regard it as a tool to control your own inevitable biases. Like the state’s legal system you want to uphold this system for its own intrinsic value.

The grading system is inevitably far from perfect and there are already plenty of ways to be unfair within the system. However, going outside the system is a good indicator that you are being unfair. This does not only work as a sanity check for yourself, but also for any internal or legal procedures: Breaking this system could be considered to evidence what is called consciousness of guilt; it is similar to destroying evidence. At the very least it shifts the burden of proof to you that you had good reasons to break the system. Now, I am not very familiar with the internal and legal requirements for courses in the US, but I would roughly expect them to require that every course has the requirements for passing set out beforehand or similar thus providing a formal and possibly legal foundation for the grading system and with it a point of attack for internal and legal action.

In your case, your professor explicitly and clearly ignored the system because she did not like the result. She did not even bother to revisit the student’s exam with extra scrutiny – which would still be problematic unless she did it for everybody with borderline grades in every exam and made this transparent. By lowering the grade even further than necessary for failure, she even took extra steps so the exam would not be revisited (another evidence for consciousness of guilt). This is clearly unethical by the Golden Rule or Categorical Imperative: You do not want to be a student in a system where professors can arbitrarily lower your grades because they think you do not deserve them.

If you conclude that your grading system was too lenient because it let a student pass who should not have, the right course of action is to change the difficulty next time. Having one student pass a single course they shouldn’t have is not the end of the world.

  • 7
    +1 for the last sentence. Usually, people on this site give instead the nonsense argument "if a student passes who don't deserve, they could cause the death of millions".
    – user111388
    May 18, 2020 at 16:41
  • 4
    @user111388: The argument that I often see is that systematically having students get away with incompetence, cheating, etc. is bad and I concur with that. However, this is not systematic.
    – Wrzlprmft
    May 18, 2020 at 16:44
  • @user111388 this seems to imply that "gaming the system" is a valid approach for students and should not be reprimanded - and the only course of action is applying increasingly more difficult patches to your grading system to make it airtight against cheaters and probably worse for it. I seriously cannot say if this kind of system leads to overall better education than a system which relies more on individual subjective treatment of students. Both bring their own problems.
    – Falco
    May 20, 2020 at 10:06
  • @Wrzlprmft calling this behavior intransparent hinges on the definition of grading which is presented to the students. If the formulation is "If your answer is not strictly 100% correct, I will try to gauge the correctness of your answer by gauging your understanding of the material and guessing 'how right' you are" - the behavior seems transparent and ethical. - And I know not many exams, where a partially correct written answer can be graded objectively.
    – Falco
    May 20, 2020 at 10:09
  • 1
    @Falco: The question is not about grading partial answers or individual exam tasks; the question is about grading the entire course performance. In the situation in question, the grade was determined by “the student does not deserve to pass” and setting the exam grade to an arbitrary value to achieve this (without caring about the contents of the exam).
    – Wrzlprmft
    May 20, 2020 at 10:48

TLDR: it’s not technically illegal, but as for whether it is ethical and/or academically appropriate, it’s complicated. (That’s why the answer below is longer than I normally like answers to be — sorry for that.)

Your story reflects a tension that exists between two different notions that one might define for what it means for a student to “deserve” to pass a class:

  1. The student satisfies an objective, formal set of criteria: that is, the student jumps through a predefined set of hoops (exams, homework etc) and produces a set of grades that according to some numerical averaging scheme gives a number that’s higher than some predefined threshold like “73”.

  2. The student performs in a way that actually satisfies the professor that the student understood the material at a minimal level that will enable them to keep progressing in their studies by taking and doing reasonably well in follow-up courses, applying the knowledge they learned , and finishing their degree without the degree being a worthless piece of paper.

Students invariably interpret “deserving” to pass in the first sense of satisfying a formal set of criteria. But anyone who’s been teaching for more than a couple of years has encountered cases where a student just barely satisfied the formal criteria for passing, but left the professor with grave doubts as to whether the student actually learned enough to deserve to pass in the second sense of making the professor feel it is “right” to pass them. In some cases the professor will seriously wonder if the student has learned anything at all.

The point is that this formal system of exams and grading and passing thresholds is imperfect and can be “hacked”, and some students (a small minority, to be sure) in fact get by by making supreme efforts to hack it and obtain passing grades by faking a semblance of knowledge rather than acquiring actual, genuine knowledge. The result is that at the end of the day the professor might look at a student’s grades, look at their exam, and see an inconsistency between what the grades tell them should happen and what the professor believes in their heart should happen.

(Note: this inconsistency can happen in both directions. In your story the grades said to pass and the supervisor thought the student should fail, but in other cases the professor might want to pass someone who has a just barely failing grade.)

Now, this poses a real dilemma for professors, and they develop different strategies for dealing with this problem. Some possible ways include:

  1. Grade on a curve. That is, decide what the passing threshold is only at the end of the semester. This gives you the ability to decide where in the ranked list of students to draw the line between “deserves to pass” and “doesn’t deserve to pass”, so it offers a partial solution to this dilemma.

    On the other hand, this method doesn’t let you tweak the pass/fail outcome of an individual student in a way that doesn’t affect any other students.

    There are other problems associated with grading on a curve. Both many students and many professors don’t like the concept and feel it is ethically and pedagogically flawed.

  2. Do what your supervisor did. That is, allow yourself to tweak the grade of the student in your question a little bit, “fixing” the inconsistency in a way that causes the formal pass/fail outcome to match with your gut feeling of what is the “right” outcome.

    The drawback here is that you are breaking an implicit promise you are making to your students that grading will be done impartially, with each question being graded independently of any other question. This is ethically fraught, and if a student finds out their grade has been tweaked in such a way, they are likely to be extremely upset and aggrieved.

    On the other hand, if I’m being completely honest, changing a point or two here or there is essentially a meaningless operation. If a student got 14 points out of 20 on a question, who is to say that that’s a more “scientific” or accurate measurement of their knowledge than giving them 13 or 15 points instead? (I’m imagining a scenario where the student’s mistake is slightly different than the mistake of any other student, so that by making this change you are not creating unequal treatment of different students.) So I cannot unequivocally declare that it is blatantly unethical to go back and review a student’s exam, ask yourself “are you sure that this student really deserves to pass?” and allow the answer to dictate a 1-point change to the grade in a question somewhere, when this is done in a good faith attempt to produce the most just outcome.

  3. Do nothing. That is, accept that whatever formal criteria you defined for passing in your syllabus represent your best approximation, flawed and hackable though it might be, for what it means to “deserve” to pass. If a student satisfies the criteria, you let them pass and ignore your inner gut feeling.

    The disadvantage is that in some cases you may indeed let an undeserving student “get away” with passing when they shouldn’t (which by the way may not be doing them any favors, and may be setting them up for larger failures down the road in their studies and in their career). But the big appeal is that this is an “honest”, objective approach that doesn’t require subterfuge or covert manipulation of grades as your supervisor pressured you into using.

Summary. As you can see, decisions about passing and failing students are ultimately very complicated. Every strategy one chooses for dealing with this situation ends up being problematic in some way and causes some measure of injustice. I don’t think there is a unique set of ethical principles by which one can decide whether your supervisor’s approach is right or wrong. At best, one can say that this approach is somewhat ill-advised and violates the principle that in a modern university grading should be done as impartially as possible. Personally I can’t bring myself to condemn the behavior more severely than that; perhaps others might.

If it’s any consolation regarding your sense of guilt about that incident, consider the very real possibility that passing the student in that scenario might not have been in their own best interest.

  • 24
    Like others, you are suggesting that your "gut feeling" is more important and even more valid than what the students actually do. This is a recipe for discrimination and other pernicious acts.
    – Buffy
    May 17, 2020 at 12:16
  • 12
    @Buffy you’re twisting my words. I never said gut feeling is “more important”. I also never said gut feeling is more valid, although since you bring it up, yes, sometimes it’s more valid (and other times it’s less valid). I also never denied that the professor’s behavior is problematic, although I don’t quite get what you mean about “discrimination and other pernicious acts”. Yes, obviously discrimination is unethical, but that’s not what’s being discussed here. My answer is premised on a professor changing the student’s grade because they sincerely believe the student did not deserve to pass,
    – Dan Romik
    May 17, 2020 at 14:11
  • 9
    @Dawn thanks, that’s a reasonable criticism and I mostly agree. And yet, if you take this logic to an extreme, grading cannot be truly objective unless it is assigned to a computer. Is there not a “potential” for human graders to be subconsciously influenced by a student’s handwriting, grammar, punctuation or any number of other things that in (say) a math exam ought to be irrelevant? At what point do we allow professors to exercise subjective judgment? Isn’t that what we pay them to do in the first place? Otherwise, why aren’t exams graded algorithmically?
    – Dan Romik
    May 17, 2020 at 15:09
  • 24
    Many disciplines have highly subjective marking: Arts, Humanities, Languages and Literature, Philosophy etc. Rigid, stratified criteria are not only unattainable but problematic in many contexts. Discretion will always be part of marking.
    – user117109
    May 17, 2020 at 15:22
  • 10
    (Cont'd) Having and following a narrowly defined grading rubric doesn't remove subjectivity from grading. At most it moves the subjectivity to the rubric design.
    – JeffE
    May 18, 2020 at 1:18

This is of course why many systems (but not commonly in the US I understand) insist on blind double marking: that is, all marking is done by two people, neither of whom know the identity of the student they are grading. All of our systems and samples of our marking are inspected by "External Examiners": that is respected academics in the same field, but from a different university. These examiners will also look carefully at the work and marking of students classed as borderline (you need at least 69.5 for a 1st class mark, but less than 68.5 for a 2nd class mark - 68.5-69.5 is officially "borderline") - and various factors might come into play - records of illness or personal problems, etc.

I say all this because it demostrates the lengths some systems go to avoid exactly the situation described, and should tell you something about the implied desirability of academics applying their judgement to something other than the piece of work in front of them.

How to prevent people passing when they don't genuinely understand? Setting assessments where you can't pass without understand is a start. But we generally include a grading criteria along the lines of "Student clearly demonstrates undrstanding", but works as long as you don't know how the student is, but wouldn't work well in an unblinded situation.

Grades can only be appealed on the basis of process - the academic judgement of the grader is absolute, but you'd better follow proceedure to the letter.

  • 5
    This doesn't answer the question at all.
    – usr1234567
    May 17, 2020 at 21:11
  • 3
    So are you saying that by adding a “student clearly demonstrates understanding” requirement to my syllabus I will have secured for myself a blanket permission fail any student I feel like failing (based on their anonymizes exam and other assignments) without further explanation and regardless of their actual grades? If that’s not what you meant, can you clarify how exactly this requirement is used in practice?
    – Dan Romik
    May 17, 2020 at 23:01
  • 1
    Each question is marked on a rubric with 6 criteria: 1. Coverage of areas relevant to the question, 2. Information is accurate 3. Sufficiently detailed information provided, 4. Answer structured in a way to bring out relevance and actaully answer the question 5. Demonstrates understanding 6. Provides evidence of reading beyond the lecture material. The grade in each criteria must be explained/evidenced with a written explaination at the end of each answer and examples highlighted in the students work. I teach biochemistry, genetics and molecular biology BTW. May 18, 2020 at 8:58
  • @usr1234567 Its an example of the lengths that we go to to avoid the possibility of what has happened here, thus providing evidence for how we regard such actions. May 18, 2020 at 9:00
  • @DanRomik "adding a “student clearly demonstrates understanding” requirement to my syllabus I will have secured for myself a blanket permission fail any student I feel like failing" - I did not read the answer like that. I read that it would be okay, as long as you did not know whose work you are marking. So you can't fail any student you (don't) like, but rather any work that doesn't show understanding. I agree that it should be okay iff the marking is done in a blind way, so it's impossible for you to be biased by student's character. That is unless the course is supposed to test character.
    – Sebi
    May 18, 2020 at 18:11

Echoing some others' remarks: no, this is unethical, immoral, and unfair. The course materials should have described the rules of the game, and if someone scored certain points according to those rules, then, ... well, they did.

Yes, it is true that "grading systems" do not reliably reflect "mastery of material" (whatever that supposedly means, anyway!), but that is insufficient reason to change the rules of the game after people have already signed on.

Yes, it is true, especially, that single numbers do not readily reflect what we'd like them to. Hm. Ok, so if we do want them to reflect something-or-other, such as "adequacy to follow the subsequent course", etc., then serious rethinking of the grading scheme is necessary. (This is apart from "gaming the system", but, yes, "gaming the system" is another consideration!)

But, in any case, from my viewpoint (in math, in the U.S.), changing the rules at the very end is crazily unfair, ... Don't do it. (Especially not if there is potentially some personal aspect...)


My apologies for the length and the style of what follows, but the questioner seems to be concerned about something I myself have struggled with for many years of teaching, namely, how to give just and fair grades.

It is certainly valid, because the grade will stand.

A grade is also a certificate, to the student, to the administration (who will give a degree or admit to examinations based on fulfilling requirements) and to the consumers, such as grant agencies who want to restrict funding to "good" students, or to future employers.

Even in the hard sciences, grading has a subjective component, especially when it comes to partial credit. Not too many people who grade are aware how an unacknowledged bias can influence their decisions and personally, I am worried about this.

Even with the best of efforts, grades are not an accurate representation of a students capabilities. Sometimes, we faculty even use "effort" in our grading decisions, whether we explicitly admit this on the syllabus or not. Anyone who is graded will be influenced on how the grades are calculated and so sometimes grades egregiously fail to reflect the student's capability in -- let's say -- ordinary differential equations and rather their capability to "game the system".

All this being said: In your situation, you were involved in the manipulation of a grade that took out-of-scheme information into account, namely your impression of the student, based on the final exam. If pressed, you and your supervisor can presumably justify your grading decision without admitting the manipulation.

You can argue that the grade is a means to a goal, namely to obtain information on how the student did and whether the student needs to repeat the class. Your impression is that the means failed to reach the goal. Therefore, you substituted for the failure by correcting the outcome. In fact, you could have done the student a favor. (I still remember vividly at the beginning of my teaching career of hearing another faculty describe the situation of a student who had passed Calc 1 and Calc 2 by learning all procedures by heart, but then was unable to perform in Calc 3, because their memory did not suffice. The advise was for the student to start over again. This could have been avoided by failing the student who was gaming the grading system to his ultimate disadvantage.) Now, if I teach the class that comes afterwards, I would be unhappy of having students who only formally require the prerequisites, because much more likely than not, I have to fail them, they will repeat the class, and they are still set up for failure. Or I might not care at all, because the prerequisite is really one of "academic maturity".

You can also argue, as you seem inclined to do, that manipulating grades breaks the implicit or explicit contract in the syllabus that promises students to get a passing grade for doing certain things. Allowing yourself to break the contract allows all sorts of mischief such as punishment for obnoxiousness up to discrimination based on irrelevant criteria such as gender or race.

In a situation like the one you were in (about to pass a student who in your best opinion should not be passed), you were in a dilemma. In an oral exam such as a Ph.D. qualifier, committees tend to use their knowledge of the student's capabilities in coming to a decision, whether they admit it or not, though some institutions go to great length to "objectify" the decision. In a large class, when personal knowledge of the student is sketchy, I personally would be very hesitant of using it, just because I am aware that I might have unrecognized biases. From my personal experience (assigning grades for almost forty years) I do not think your question can be answered definitely. However, it seems to me that the intentions of your supervisor were beneficial, asking you to consider the consequences of your action. Besides, if you are new to the game, failing someone should be very hard for you and it should never get much better. Thus, you might have a bias towards lenient grading that your supervisor recognized. Of course, this starts to be all speculation.

Manipulating a grade away from the borderline is common practice to prevent being exposed to the pressure from the student who sees the passing grade in reach. I consider it to be a bit cowardly, but certainly understandable.

A final piece of advice from someone who agonized over grades. You can only do your best, and after you have done your best, you just stop agonizing.

  • One way to describe the course of events presented by the OP: "They preliminary graded the exams, their superior asked them to reevaluate the exams of cases where a small difference in grading would decide pass/fail. The exam was re-evaluated and a previously wrong grade was rectified to the correct grade." - the old grade is not necessarily more "right" just because it was your first opinion.
    – Falco
    May 20, 2020 at 10:29

Looking at the final exams of barely passing students is a thing. It's done for just the reason your supervisor said: we don't want students moving onto the next required course if they're completely unprepared. My first supervisor did the same thing. Obviously this is when the course is part of a sequence and the first day of the next will absolutely require they know certain fundamentals. And also when the final is comprehensive -- looking it over gives a decent sense.

The easiest, fairest-seeming way is playing with the curve. You get a clump of 5 students near a 71% final percent and pull their final exams. 4 are merely bad, and 1 is from a student with suspiciously good homework and 35% on the final. Oh, well. 71% passes. Then you go down to the students with 69%. Their finals are all shockingly bad. Maybe the scores weren't so low, but the grading system gave far too much partial credit. You're not sure they understood many of the questions. So 69% is a fail. In fact, as your supervisor notes, you set 71.1% as the lowest passing score. That way the 69%'s don't feel as bad.

In theory, you're doing the same thing, but backwards. I assume this is the one student near the cutoff. For whatever reason, you can't raise the curve, so you cheat a little to lower the score. It's icky, but the end result is the same. It sounds like standard practice in your dept (but I've never heard of it being done). If another student with a lower final % didn't receive the same treatment -- that's more dubious.

In my case, I've never done heroic, slightly questionable efforts like that. I'll let them pass by the numbers, but often write to let them know they failed the final and why that matters. But since they got a terrible grade, they were often planning to retake or switch majors.

My other bit of advice, which is too late: don't look at who the student is until you've definitely decided either way.

  • 2
    If they have a system in place already, which they do since there are specific percent grades that signify passing, then deviating from that system cannot be ethically justified. If one wanted to have that kind of authority, then they should say explicitly in their syllabus something like "The final decision of passing or failing a student is solely at the discretion of the professor and may not be strictly reflected by the grade earned in the class through homework and exams". I don't think such a standard would fly, but at least it would be honest
    – Kevin
    May 18, 2020 at 22:32

Ethical it is not. It is only ethical to grade a student for their performance. Everything else is personal - and pretty much wrong. In your case, you will discriminate a student for their perceived attitude.

Imagine the distribution of the student grades. It will be some smooth distribution. Now, draw the "passing" line. Is there a significant inflection around it? Well, you can't analyse a single class this way, but still...

  • 2
    It is indeed extremely ethically murky to assign a grade based on how an individual student should have performed rather actual performance. The instructor should be wiser than support (much less suggest) this. May 17, 2020 at 21:38
  • 1
    I doesn't sound personal: the OP didn't seem to know the name until afterwards, and the supervisor picked "borderline" students (so based on final %'s, right?) Or do you mean wanting to help the student once they looked at who it was? May 17, 2020 at 23:16
  • 3
    @OwenReynolds “ Are you sure you want to pass him?” certainly sounds like they knew who the student was. Moreover, there is no reason to revise down unless it’s personal, else the pass grade is effectively inoperative. May 18, 2020 at 0:20
  • 2
    @OwenReynolds If you are right then passing the final exam should be a requirement for passing the course, and it should be stated as such. May 18, 2020 at 2:46
  • 1
    "Imagine the distribution of the student grades. It will be some smooth distribution." False. May 18, 2020 at 6:48

I would like to offer a dissenting perspective. Sure, it's bad optics and all that, but in the end of the day, making a correct, just decision cannot be all that unethical. And borderline cases are, by their nature, those where both letting the student pass and failing them are correct and just. Some elaboration:

  • grading is an extremely noisy procedure, especially if it involves human judgement. When grading final exam (in math), I quite often can justify giving anywhere between 25% and 75% of credit for a partial solution, at my discretion. BTW, I do grade inconsistently unless I make special effort to check myself.
  • therefore, when people are saying "the rules are set in advance", that's wishful thinking. The rules might say "73% is are enough", but the only thing this means for sure is that if a student submits perfect solutions to >73% problems, they've passed. For weaker students bordering on a fail, this is a moot point - they earn a lot on partial credit.
  • consequently, the scenario of a student who "studied just enough to pass the course, but to their astonishment failed and now feels robbed" has nothing to do with reality. If a student has studied just about enough, in practice they are accepting a gamble, in which luck with the questions, with who are grading, and whether they do it before or after lunch, can have major consequences. To be on the safe side, study a bit more!
  • there's no meaningful difference between 72% and 73% that's not erased by the noise. It's amazing that people here see no wrong in allowing 1% of the grade have major consequences, and are so eager to judge the ethics of a poor TA without questioning why the system that forced him or her into this situation is allowed to exist in the first place.
  • in view of the above, reviewing student's work in the borderline case, to try and decrease the noise effect, makes much sense. The measures like "the overall impression of the exam paper" and "the overall impression of the teacher on the student" surely are bad optics and have shortcomings, but practically they may be valuable inputs, much, much more significant than the difference of 1% in the score.
  • if in the end it's still borderline, then so it be, both failing and passing are just. It's not your fault you are forced to make decisions of major consequences in the situation where you can't decide meaningfully.
  • 1
    I disagree with the conclusion that the decision made was the ethical choice (largely because the intent was to make the student fail), but you raise some very important points here that should absolutely be considered when making the decision. +1
    – Ray
    May 18, 2020 at 16:37
  • 1
    It is true that marking is noisy but that’s why most instructors will have a “bump zone” that will allow students just under threshold to pass if they are in this zone. But you’re suggesting that, even if a student passes prima fasciae on merit, one should use the noise argument to fail this student. This simply invalidates the required passing grade. May 18, 2020 at 19:27
  • 2
    @ZeroTheHero, having a "bump zone" is no different from just lowering the threshold to the lower boundary of that zone, so that does literally nothing meaningful. Also, you seem to miss the point: there's no "required pass grade" to invalidate. A pass grade of 73% means nothing unless we specify in advance how difficult will the test be and how strict the grading, which is never done to the degree of detail sufficient to distinguish 72% from 73%.
    – Kostya_I
    May 19, 2020 at 6:53
  • there is a important distinction: students are who passed “on their own” are not made to fail, i.e. students always get the benefit of the doubt. But there is (presumably) a published passing grade: the minimum grade is not decided “on the spot”. May 19, 2020 at 12:14
  • 2
    @tbrookside, sorry, there is nothing in the post that indicates that noisyness has been accounted for (and it's intrinsically impossible to "account for" it - what would it even mean?), nor any indication of bad faith or conflict of interests.
    – Kostya_I
    May 19, 2020 at 19:06

I disagree with those who say this is a complicated issue, this is flatly unethical. I cannot speak to legality, but I would say that this is more than likely illegal if the student is a part of any protected group.

There was a standard set at the beginning of the course. That standard said, "if you score x% at the end of the course, you can move on." Unless there was a specific provision built in for scoring a low final exam score, then you broke the contract made at the beginning of the semester. The proper approach moving forward would be to express your discomfort with breaking the rules you set, and escalating to a higher level supervisor or HR department if you feel that strongly.

  • 2
    The point of my “it’s complicated” argument is that while the behavior is problematic (I called it “ethically fraught”), the alternatives are also problematic in different ways. Thus, whether it’s perceived as ethical or not depends on the framing of the question. Suppose I asked “is it ethical for a professor to give a student a grade the professor sincerely does not believe, after a careful evaluation of the student’s work, that the student has earned?” I bet a lot of people would have a problem with that. A lot of people have ethical problems with grading on a curve, which is another ...
    – Dan Romik
    May 18, 2020 at 13:32
  • 1
    ... of the standard solutions to the quandary in the question. My conclusion is therefore that it’s not a black or white issue. Professors agonize over failing and passing students because it’s such a consequential decision, where a mistake on either the side of wrongly failing or wrongly passing someone can be harmful to them. Each strategy for dealing with these decisions has problematic aspects to it and can be seen as at least mildly unethical in its own way. Perhaps the supervisor’s strategy is more unethical than the other options, I don’t know. But “flatly unethical”? I’m not seeing it.
    – Dan Romik
    May 18, 2020 at 13:38
  • 2
    Yes, and we can argue from here to the end of time whether or not grading on a curve and what not is unethical, but that ignores my reasoning for why this is flatly unethical. The rules set up were changed at the witching hour for this student. While I would argue that being this arbitrary is always more unethical than grading on a curve, that is completely beside my main issue with the actions taken.
    – Joel Hines
    May 18, 2020 at 13:42
  • 1
    Sorry, I’m unconvinced. You cannot weigh the ethics of an action without considering the alternatives. If I am forced to choose between actions X, Y and Z, you are not going to convince me that I must avoid X by arguing that X is “flatly unethical” but refusing to discuss the possibility that Y and Z may also be unethical and that therefore the correct question is really about weighing the trade offs between the different actions. (But you may end up convincing me if you are willing to discuss that more nuanced question.)
    – Dan Romik
    May 18, 2020 at 13:54
  • 2
    Unfortunately, I think we won't see eye to eye here. My point is completely unrelated to the other options, so I can't argue them. If he told the students at the beginning of the semester he wasn't going to grade on a curve, and then arbitrarily threw in a curve on the final exam, then my point would still hold. We create contracts with students to define our expectations. They can be as vague or as specific as you want, but you have to stick with the expectations you laid out for your students. At the very least, they need advance notice of changes to those expectations.
    – Joel Hines
    May 18, 2020 at 14:07

Another point I feel is very relevant to the issue was only addressed by one poster. It is what is going to happen after the student passes? I believe this was primarily on the mind of the supervisor when this decision was made.

It is the specifics of this that need to be answered in order to really pass judgement here. Was the student about to graduate and then go into the workforce with subpar skills? Was the student about to go into an extremely difficult course for which the current class is a prerequisite? I feel like either of these cases could give a professor pause to examine the performance of the student. This is also dependent on the academic standards of the university/school in question.

In summary, I side fully with people saying this is not a black and white issue. I come from college education in hard sciences/mathematics, and I feel like this opinion will likely be the majority in these departments. Personally, I feel like my primary ethical concern is to produce qualified individuals. If there is some ethical dilemma where this concern conflicts with another, then producing qualified individuals is likely to be the winner.


My syllabus says explicitly that my grades are assigned by my best judgement of what the student learned, and the numbers are a strong guide to that judgement, not a replacement for it.

  • Fine. Clear. I suspect you get arguments, though.
    – Buffy
    May 18, 2020 at 20:07
  • 1
    While this may be how you run things, that isn't how their syllabus was set up. They have a specific percent grade that represents passing or failing. They didn't just give the student a failing grade, they retroactively graded their exam lower than was deserved in order to drop them below that pass/fail line. Saying that other people run their classes in other ways doesn't address the morality of the act the OP is discussing
    – Kevin
    May 18, 2020 at 22:28
  • That’s very interesting. Can you say something about how often you use this discretion to deviate from what the dry numbers say? And, as @Buffy said, when you do that, how do the students for whom you did that react, if they are even aware that that’s what you did?
    – Dan Romik
    May 19, 2020 at 3:25
  • 1
    In practice, I deviate up regularly - probably something like 5% of students. I deviate down from the numbers very rarely, only once I can remember in 15 years of teaching. I deviate up inconsistently - in the sense that someone with a lower number gets a higher grade - around once every 3 or 4 years. I don't think it's been noticed. (This is not counting academic dishonesty cases, where the numbers are just thrown out the window.) May 19, 2020 at 5:32

I and a friend of mine had a strikingly similar experience. This happened in the first year as a graduate TA.

My friend, let's call him J, and I, along with other first year TAs were assigned as graders for Calculus II. We were given quizzes and exams together with some rough grading guidelines, and we were responsible for coming up with partial credit schemes for problems for which students had show work. The professor generally trusted our work and didn't really bother us... until the end of the semester. After reporting the final grades, the professor called for an emergency meeting. We had to come back to campus around dinner time to re-examine the final exams. The professor asked us to pick out the exams of all those students who barely made a C. She then reviewed these exam in detail and questioned our grading decision problem by problem. We were being way too generous, in her opinion. Eventually, she got tired, and unlike us, she actually need to eat dinner. So she left and require us to finish regrading before we can go home. Even though she did not say explicitly, the message was clear, we had to turn as many C's into D's. Note that for this course, D was basically a failing grade at least at the time because most courses that require Calculus II as prerequisites would require "C or higher" anyway. So basically, we were asked to fail as many students as possible. We got the message, and redesigned some partial credit scheme. Eventually, we turned a huge number of C's into D's (I don't remember the exact number, but it was certainly well over 100).

After that, we felt terrible about it. Not only our professional opinions were not taken seriously, we also felt bad for the students, and most importantly, we felt we became a part the academic mafia.

So far our story is almost identical to OP's, right?

I wasn't a person with strong moral foundation, so I felt bad for awhile and then just ignored that. My friend, J, however, just couldn't let it go. A week or so later, J confronted the professor on her shady behavior and eventually reported this unethical procedure to the department head, which is what most of the answers (so far) are suggesting the OP should do (at least morally speaking).

In my story, however, my friend J didn't do very well. My reporting, J didn't become a hero. Indeed, J turned out to be very wrong.

It turns out, the department head already knew about and approved the professor's re-examination/re-grading of the final exam. The professor felt our grading was too generous for a long time, but she didn't/couldn't correct in time. It was only after the final exam she realized that a few new TAs' generosity would allow hundreds of students pass Calculus II without actually being sufficiently competent. So the final exam was the only chance for her to correct that.

It also turns out, she did what she did only because the department head had been nagging her about this overly high passing rate.

It also turns out that the department head started to worry about passing rate being too high only because the physics and engineering department had been complaining --- the math department produced a huge number of students who cannot handle maths used in advanced physics and engineering. Indeed, the physics department threatened to open their own "advanced mathematics" course to replace Calculus II.

It also turns out that physics and engineering department complained about their students' poor math foundation only because the dean of the college decided to address the high major-change rate for physics and engineering --- many students study physics for a year or two and then get discouraged and switch to some "easier" major mainly because they stuck on mathematics.

I can keep going along this chain of events. At every step of the way, someone made a right decision: Surely the dean has to address unusually high turnover rate. Surely the head of physics department has to demand that their students have good mathematical training. Surely the head of math department has to worry about grade inflation. And surely the professor for Calculus II has to consider learning objectives and not just counting points.

Of course, as new graduate TAs, we couldn't see any of these higher level objectives... we were only worrying about partial credit assignment one problem at a time.

So I guess my point is...... well, actually, I don't have a point. But my and my friend's experience could serve as a reference. The OP should consider the possibility that he/she (especially if he/she is a new TA) may not have access to all the information. There is a small possibility that the professor is trying to correct a disconnect between grading scheme and learning objective.

I'm sharing this story because I thought about this experience a lot lately as I start to see the difficulty in translating abstract learning goals into concrete grading schemes, and the OP's situation is almost identical to what we experienced. In my story, my friend J wrecked his career: soon after that incident, he lost his TA position. With no one taking him as RA, he was forced to drop out.

  • The supervisor in yourstory should have told from the beginning "fail as much students as possible because the dean says so and the dean is happy to answer an inquiry about his decision" instead of just "fail as much students as possible".
    – user111388
    May 19, 2020 at 7:12
  • Sometimes the final part of these stories is the students find out, or hear a rumor ... and don't see any problem. There's no uproar. In fact, they're often glad to know their hard work for an honest C actually means something. May 19, 2020 at 15:53

The things you should consider:

  • How much has the student and/or his family invested in this training? Not just money, but effort and time.

  • What has your decision cost the student? Include the ongoing costs.

  • Did the published criteria for passing the exam include your feelings?

As a teenager, I had my maths exam decreased due to the teacher not liking my writing (minor disability, tremors in fingers). This in turn nearly made me miss out on an apprenticeship. Fortunately for me, the teacher wrote this on the exam, and I was able to show this to my prospective employer.

  • 6
    Passing a course is not based on (real or perceived) effort. It is based on proficiency (or at least approximate proficiency) in the topic. The cost to the student should not be a factor. May 18, 2020 at 3:53
  • @ZeroTheHero, actually if falsely failing them forces them to change the course of their lives then it is relevant. That should be their choice, not that of some person in authority. They have stolen an investment in time and money.
    – Buffy
    May 18, 2020 at 18:04
  • @ZeroTheHero I agree, and that's part of what makes OP's action here unethical. It's clear they made their decision based on their feeling that the student could have made a greater effort. It's clear that if they had the feeling that the student had made the maximum possible effort and worked at the limit of their abilities, the exact same course work would have merited a passing grade.
    – tbrookside
    May 19, 2020 at 16:55

It depends on the syllabus and protocol. In principle, it is widely accepted that borderline cases can be revisited. As long as the process for doing so is still based on academic merit, reasonably consistent between candidates in similar borderline situations (this is not to say that the outcomes must be the same), and carried-out without fear or favour, such practice seems ethical.

In fact, I happen to know of one examination protocol that requires the examiner to change the raw marks if the total of all components sums to just below a grade boundary (as a result, it is impossible for anyone to get a total mark one point below the amount needed for the next grade up). The rationale, as I understand it, is that numerical scores are not absolutely infalliable, so, where the total is borderline, the examiner is required to go back and reconsider carefully whether he/she may have been slightly too harsh or slightly too generous, and ask himself/herself "overall, is this candidate worth a pass?". That is to say, a holistic re-evaluation is almost encouraged.


If you think it is ethical, you should tell the student about it. If you think it is not ethical, you should definitely tell the student about it. What’s the worst that can happen?

(I know people who would take you to court or have their parents take you to court. I also know people how would meet you in a dark alley way. Don’t do things that you wouldn’t want the other person to know).

  • If you think it is ethical, you should tell the student about it. Why? There isn’t a universal principle that students deserve to know about any ethical decision a professor makes while running their class. Professors have all sorts of completely ethical tricks to manage students’ unrealistic expectations and sometimes outright delusional and/or manipulative tendencies. As one example, when grading on a curve, there is a certain technique to set grade cutoffs in a way that minimizes student complaints and feelings of disappointment. That’s ethical, but students don’t need or deserve to know.
    – Dan Romik
    May 20, 2020 at 18:17

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .