New APPS

Art, Politics, Philosophy, Science. Mostly Philosophy.

recent posts

about

    Why should we grade final exams?

    Lucky-jimAre students aware that most of their professors can accurately predict their final grades prior to the final exam or paper? Is this just hubris on my part, or do people in the biz long enough get very good at this?

    If it's not hubris, then it raises a genuine practical ethics problem. If I already know what the student is going to make, why do I feel morally obligated to grade the paper? I can't figure this out.

    I'm clearly obligated by prudence to grade the papers as the LSU hair and teeth men* would be very unhappy with me for not doing so. But would they be right to be unhappy?

    Maybe the problem is that if the students knew I wasn't grading their final papers, then they wouldn't do the work. So perhaps not doing the grading would involve deception? This seems pretty weak to me though. I could just not say anything one way or the other, or honestly say that I might** grade the person's exam or paper. Knowing their exams might be graded would motivate them enough.

    A few students want comments on their work and we owe them that, but the overwhelming majority of them don't. So why not just hold onto the papers and grade them later if a student wants comments?


    The only thing I can come up with is that I think maybe that people just have an obligation to bear witness to particular things. Most of us feel an obligation to watch Schindler's List, even if we are going to find it an unpleasant experience.*** A colleague of mine in Religious Studies just did an independent study on religion and healing. One student interviewed her sick grandmother and the other student interviewed an LSU student with a terminal illness. One of the weird things about the class is just how much the sick people loved having someone listen to them. Maybe reading student work is like that. Our job isn't just to ensure that they develop certain intellectual virtues while credentialing them for wage servitude. We're also supposed to just listen to them. If that's right, it does seem like a worthwhile part of one's vocation to me.

    I'm not sure that these kinds of concerns could convince me to actually enjoy grading though, which makes the thing pretty disanalogous to other forms of bearing witness. I mean if you find it uniformly unpleasant to talk to sick people, then you are not going to be the right kind of listener. So maybe there really is no good reason to grade final exams.

    [Notes:

    *What people who do the actual work in the tech industry call the suits. In addition to having wonderful hair and teeth, they are usually tall as well.**** The world doesn't make very much sense.

    **Not in the Eddie Haskell-like philosophical sense where there's just a possible world out there where me or a convincing copy do grade them, but rather as a statement about how the actual world could turn out. I don't think it's correct to label this kind of possibility "epistemic" possibility. To say that something might happen is not to say that for all we know it will happen.  I'm out of date on the literature on this point, so maybe none of this is controversial. I know there's some new work on modal realism without possible worlds that goes beyond the actualism/Lewisian divide chronicled in Divers' Possible Worlds, as well as new two-dimensional work on how to have an actuality operator in the language. That stuff is probably relevant. . .

    ***Several people (I think all non-philosophers) have told me it has a happy ending when I tell them I want to see it but always find the prospect too depressing. I don't get this. How could it possibly have a happy ending? Did Spielberg cast Robert Benigni? For an entertaining critique by someone who really hates the film for these and related reasons, go here.

    ****Some hair and teeth men, like Steve Ballmer, don't actually have hair.]

    Posted in

    26 responses to “Why should we grade final exams?”

    1. n Avatar
      n

      When I was an undergrad, I got into an argument with a professor over his grading practices that I felt unfairly disadvantaged me. To his credit, he implemented blind review of homework — we only wrote our names on the back of our homework and not with the content on the front — which did mitigate some, but not all, of the problems. I also changed what I was doing to accomodate his grading scheme.
      There are too many issues besides the actual learning of philosophy, like problems with grading schemes or personal disagreements with professors, to leave it up to the professor’s predictive skill (or bias). On tests there is less ambiguity, which can make up for other issues.

      Like

      Reply
    2. Tim O'Keefe Avatar
      Tim O’Keefe

      Are students aware that most of their professors can accurately predict their final grades prior to the final exam or paper?
      That’s not my experience, at least not with a great deal of confidence. And think about the following sort of case: a student is doing pretty good but not at all great work and has an 86 average going into the final exam, which is worth 25% of the grade. There is a good chance that the student will end up with a B, but if an 87 average gives him a B+, he needs to get only a 90 on the final to pull his grade up. That sort of improvement on the final exam doesn’t seem at all mind-boggling. And I’ve had plenty of students slack off at the end of the semester and end up with final exam or final paper grades significantly below their earlier average.
      On a somewhat-related note, I think it’s a good practice to grade papers and exams anonymously, e.g., by having students put their student ID#s on them instead of their names, so that your perception of their quality isn’t biased by your expectations of how well the student is likely to do.

      Like

      Reply
    3. Jon Cogburn Avatar
      Jon Cogburn

      We don’t have plus or minus grades at LSU.
      Probably from one to five students in a fifty student logic class will surprise me by getting their stuff together in time for the final. In upper level classes not so much, I’ve just read so much of their writing by the time they turn in a final paper.
      I probably should have made the OP in terms of just cursorily looking at the final paper. If you know the student you can usually read the first page and glance at the rest and know what the grade is going to be. Yet we read the whole things attentively, often writing comments that will never be read (cue Eleanor Rigby).
      When I was a graduate student at OSU we had the plus/minus system, and I found that it made grading a much more neurotic experience both for me and the students. There was this huge culture of grade grubbing where a certain percentage of students would try to nickel and dime the professors to get to the next highest grade. I also found that it made it a lot harder to decide grade distributions when I scaled the grades, because there weren’t as many big gaps where it made sense to make the divide. Without the pluses or minuses I can start with the standard ten point scale, and then make it more generous where there are natural divisions in the distribution.
      But an advantage might be that you have a lot less certainty with pluses and minuses (in addition to the other possible advantages).
      I do anonymity for logic exams, but it doesn’t make any difference for final papers because the students have already gotten the topics approved by me so I already know who is going to do what. I wish there were some way around that but it’s just too helpful for me to guide them in the process. A number of my students (undergraduates and MA) have gotten things published and I don’t think this would be the case if we didn’t talk about paper topics, relevant articles, and composition strategies in the classroom.

      Like

      Reply
    4. Joe Avatar
      Joe

      If the final exam/paper is about 1/4-1/3 of the grade, how can you ‘know’ what the student will get? Many students have a rather steady level of performance, many do not (performing at A on one assignment, B on another, and so on), so it seems that the final exam/paper could make a lot of difference! People are not robots: some students ‘get’ the class or material only at the end, some who most of their efforts into the final project, and some slack off at the end. In any case, the scenario were final exam/paper does not matter seems to me to be either one in which the professor decides on the grades before the exam based on his “knowing the student” or where the grading of final exams is taken lightly (to avoid what we all find somewhat unpleasant – reading repetitive and not particularly exciting stuff) and done so as to fit in with the pattern of previous work.

      Like

      Reply
    5. Jon Cogburn Avatar
      Jon Cogburn

      Joe,
      I think what you write certainly happens. However, it could also be the case that the professor is much more conscientious about getting students feedback throughout the semester.
      In upper level classes, my students have written assignments due three times a week. This semester I’ve read and graded every single one, this is forty or so two to five page papers for each student before the final paper is due. I’ve also talked extensively with many of them about their final papers. The professors I found most helpful at University of Texas and the Ohio State University graded an enormous amount through the semester. Many of my colleagues do the same. The ones who do (especially those of us who have mandatory rough drafts for the final papers; I do this sometimes) have a much better idea how the students are going to do at the end.

      Like

      Reply
    6. bzfgt Avatar
      bzfgt

      Not that I’d do this, but if you already have a good-looking grade spread that won’t be driven too high by an ‘A’ on most of the final papers, you could just glance through the papers and give ‘A’s to everything that looks reasonably decent without having to really read them…

      Like

      Reply
    7. Roberta L. Millstein Avatar
      Roberta L. Millstein

      First, let’s separate grading exams from putting comments on them — we might do the latter only at the request of interested students, given that not all are interested.
      Then, as others have noted, I have definitely seen students whose grades go up and down throughout the quarter. For those students especially, I’d be hardpressed to guess how the final will go.
      Third, and I might be atypical here, I tend to assign different sorts of things during the quarter than I do for the final. So, for example, I assign papers throughout the quarter, which require independent thinking, whereas the final exams place more emphasis on being able to explain course material. So, a student might well be able to do the latter even though struggling at the former.
      So, yeah, I need to grade final exams. Alas.

      Like

      Reply
    8. Jon Cogburn Avatar
      Jon Cogburn

      Yeah, it’s clear to me that how you teach is going to effect how sure you are about students’ grades.
      I’m a little bummed that the antecedent of the conditional (if you know a student’s final grade before they take the final exam/paper, you still have an obligation to carefully review the final exam/paper) is so implausible to people that the conditional itself is not seen as interesting.
      Certainly in some cases you know how the student is going to do. Say they did so poorly that it is impossible for them to pass. Do you still have an obligation to read and grade their final? I think so. Or say they’ve done so well that they’d have to fail the final to not get an A. You look at the paper/exam and it’s clear in two seconds that they are not going to fail and so will have an A. Should you still grade the thing? I think so. But it’s very hard for me to think of sound reasons for this other than something like a sort of general obligation to listen and attend to what people have to say in circumstances such as this.

      Like

      Reply
    9. Mike Avatar
      Mike

      Exams help us to understand a face about this world. It is that we have to sit in exam only without peers, Friends, or any one others. It is a lesson about life.

      Like

      Reply
    10. Sara L. Uckelman Avatar
      Sara L. Uckelman

      Are students aware that most of their professors can accurately predict their final grades prior to the final exam or paper?
      I think this depends very much on the type of exam/paper and how much it is weighted. When I was a TA, and hence grading assignments designed by other people and weighted differently from course to course, there were certainly some courses where it wouldn’t have been possible for me to predict the grades of the final assignments, or the final grades without the grades of the final assignment (in particular, I regularly had logic students who finally managed to bring up their grade to a passing grade at the final).
      On the other hand, there was one English course I took where we had two medium length term papers plus weekly short responses, such that a week or two before the final, the professor asked us, “Would you mind if we simply cancelled the final exam? I’ve read enough of your work by now that I’m satisfied I can judge it accurately, so unless any of you really want to take the exam, we won’t have it” — which I thought was a remarkably clear-headed thing to do. If you really can predict the grades without grading the final, why have the final?

      Like

      Reply
    11. Allan Olley Avatar
      Allan Olley

      There are many reasons marking exist, three that come to mind off hand are: to allow the instructor justification to vouch for the competence of the student, for the student to gain feedback to allow them to learn and grow, and thirdly to allow a paper trail to allow transparency in evaluating whether the first two goals have been achieved.
      If we only needed to do the first one then I don’t see the need for you to do the marking once you’ve established the level of competence or incompetence of the subject.
      For student feedback it may not matter or it may depending on attitude, motivation etc.; even for a certain to fail student the one minor success might still give some insight or reasons to go on (or insight where to go next in their life), and the one error of an otherwise perfect student might still be something they would benefit from confronting.
      Transparency only matters in one sense if the thing ever comes up for review, but then it may matter a great deal.
      If you are sure that you are not going to fall of the tightrope does it make sense to bother with the safety net? I’m inclined to put the net out anyway and that seems much like the obligation to mark everything.

      Like

      Reply
    12. r Avatar
      r

      Jon @7: For what it’s worth, in the case you describe–where the student has done so well or so poorly that they would need a miracle to change their grade, and you can immediately tell no miracle is forthcoming, my intuition is that there is no obligation do a thorough, best-effort-accurate, point-by-point grade. If they request comments, sure, but otherwise I don’t see any problem with saving time by just skimming and moving on.

      Like

      Reply
    13. Matt Avatar
      Matt

      Like a lot of others, I think there is very good reason to be more skeptical about the first part of the claim than is suggested here. But, even if you’re right on that, I think another part is missing in the discussion- the student has, per hypothesis, actually done the final exam or paper here. If he or she has done the work, it seems to me to be unacceptable in several different ways to not make a good-faith effort to grade it. If you think you can accurately evaluate the student w/o this piece of work, then it seems unfair to make him or her do it if they don’t want to, and certainly unfair (and dishonest) to represent that the work will matter if it won’t. So, if you want to follow this path, I think you need to stop one step sooner, and not give the “final” exam or paper or make it optional with guidance as to the likely impact. Doing otherwise seems unethical to me.

      Like

      Reply
    14. Peter Alward Avatar
      Peter Alward

      Insofar as comments on written work are designed to help students do better on future assignments — rather than justify grades –there is no reason to give comments on final exams/ assignments. But if your final assignment is of a different kind than those the students completed during the rest of the semester — an exam rather than a research paper, e.g. — then one cannot predict how students will do on the former on the basis of how they did on the latter. After all, success on the final assignment may require different skills or abilities than success on the latter. So, if like me you offer a final exam but only paper assignments during the regular semester, you should grade the final but you need not provide comments.

      Like

      Reply
    15. Derek Bowman Avatar
      Derek Bowman

      But presumably your class exists as part of a larger university education, indeed as part of a larger life experience. And presumably what you teach is designed to be of some relevance either to that education or that life. So comments designed to help student do better can be aimed at doing better at the tasks (or with the information) taught in your class – they need not simply be about doing better on “future assignments in that very class.”
      The better reason not to put comments (unless asked) is simply that students are unlikely to read them.

      Like

      Reply
    16. Mike Avatar
      Mike

      Tim,
      I’m sure no one is especially reliable in predicting course grades, prior to the final exam, given a distribution of weights like the one you mentioned. But I’m not sure Jon was saying that anyone has that sort of reliability. We (maybe I should speak for myself) are pretty reliable in predicting the final exam grade for most students. B students predictably remain B students, similarly for A students, and so on, right through the course. It is really rare in my experience that a C student gets an A on the final exam (almost never), and i don’t think I’m unique in this respect. It’s even rare in my experience that B students become A students on the final exam. It’s (obviously) not that it can’t happen, but that there seems to be a continuity to the effort students put into preparing for each exam which is evinced by the scores they receive.

      Like

      Reply
    17. Mike Avatar
      Mike

      So, if you want to follow this path, I think you need to stop one step sooner, and not give the “final” exam or paper or make it optional with guidance as to the likely impact. Doing otherwise seems unethical to me
      That’s hard to follow. Even if the instructor’s omniscient, students are required to do a certain amount of work to pass the course. The fact that this sort of instructor knows the final grade prior to grading doesn’t make it unethical to require (in order to pass the course) that a student actually complete the course work. Suppose you choose not to take the final exam and argue that it’s unnecessary, since the instructor already knows what your grade would be. The reply is that, true, in that counterfactual world in which you do take the exam, this instructor knows your grade is (say) B (without grading it). But your actual grade is F, and he knows that too, since you didn’t take the exam. That’s a really good reason to take the exam.

      Like

      Reply
    18. C. Avatar
      C.

      One thing might be the granularity of the grading scale. I’ve found myself moving further and further in the direction of simplifying my grading of final exams. Rather than trying to differentiate 5 shades of grades for each definition question, or 10 shades for each argument explanation question, or 30 shades for each essay, (driven by the silly 100 point scale), I’ve moved to simply grading say 3 categories for each definition etc.
      This has transformed my end of the semester grading-jail.
      I’ve started to move in the same direction with essays w–a good electronic rubric takes care of 80% of my evaluation and then I only comment on several key points. This way I still grade all of the work, but since the purpose is not primarily formative feedback, I simplify things.

      Like

      Reply
    19. Matt Avatar
      Matt

      But Mike, as I understood Jon, he thought he could accurately tell the final grade from earlier work. Typically, the instructor decides what work must be done. (That’s been so for every class I’ve taught, at least.) If Jon thinks he can tell what the grade would be w/o the final exam, and so grading it is unnecessary, then I don’t see why taking it would be necessary, especially as it would only be the instructor who decided the “amount of work [needed] to pass the course.”
      (If anyone thinks they can very accurately predict who will do well on a final exam just from talking with the students or in-class behavior, I would strongly suspect that he or she should quickly move to anonymous grading, as he or she is very likely exhibiting some biases. But I don’t think that Jon was claiming this ability, and a merely hypothetical ability isn’t really worth discussing here.)

      Like

      Reply
    20. Mike Avatar
      Mike

      If Jon thinks he can tell what the grade would be w/o the final exam, and so grading it is unnecessary, then I don’t see why taking it would be necessary…
      Here’s why. There are two possibilities (i) you take the exam or (ii) you don’t. If (ii), then Jon knows that your grade is an F, even if had (i) been true, your grade would have been a B and Jon would have known that. So, if you don’t take the exam, you will fail the course, though you would have passed it (had you taken it). That’s why you should take the exam.
      It’s a little like this. I can predict that you will win the Pulitzer if you finish your book. You say, if you know that then its wrong to make me write the book. That’s a strange thing to say. I can also predict that if you don’t finish the book, you will not have done anything worthy of a Pulitzer.

      Like

      Reply
    21. Sophie Avatar
      Sophie

      What!? You’re telling me you’ve never had a student ace the final after performing poorly in the course? That seems unlikely. Even if you haven’t, we as teachers have a duty to read and evaluate our students’ work, right?

      Like

      Reply
    22. Jon Cogburn Avatar
      Jon Cogburn

      It depends on how the course is structured.
      I tried to set up a thought experiment above because I agree we have the duty to do so, but am not certain about the ground of that duty.
      Again, the antecedent to the conditional needn’t be nearly as hyperbolic as I initially suggested for the thought experiment to bite. There are clearly some cases where we know either before grading the thing or with a very minimal perusal what the final grade is going to be. In those cases I still feel an obligation to read the student’s work as carefully as time and energy permit. But, if (big if!) in fact I do have such an obligation, why do I? I think that answering this might reveal something about what it is to be a teacher, which is in part being a good listener and reader. Though at this time of the year this can be a genuinely frustrating and sometimes maddeningly unrewarding part of the vocation.

      Like

      Reply
    23. Sophie Avatar
      Sophie

      I thought you were talking about logic exams, where there’s little room for bias. But I think it might be even worse not to look closely at final papers or essay answers, where there’s a lot more room for our opinions about a student’s ability to impair our judgment of their work.
      Anyway, it certainly is a frustrating time of year, but it’s our job to get the best out of students- not that it’s so easy or anything. I feel like I suck at it sometimes- I very often don’t write comments on finals for the very reason that most students don’t seem interested in them. But I wonder if that affects the fairness of my grading. I’m going to write comments- short ones on all the papers this semester just to make sure I’m not favoring anyone.

      Like

      Reply
    24. Matt Avatar
      Matt

      I don’t really see how that can work, Mike Your example would work if we were looking at why a student would want to take an exam that was already assigned, but that wasn’t the example I was interested in, nor, I think, the one Jon should be interested in. That question was whether, given that he thinks he knows what the outcome will be w/o the exam, Jon should assign it. Your example doesn’t address that at all, nor does the book prize example. Real book examples of course assume there is a book, but classes and exams are not like that. It would be more similar if this was “Jon’s prize for people he thought could write a book he thought was great”. That prize, of course, would not presume any book was actually written, so the example doesn’t really work, and we’re still left with the question of, if Jon thinks he knows what grade people should get w/o the final exam, he bothers, and is justified in, assigning it.
      (I’d say again that, if he purports that the exam will matter, he has an obligation to grade it, and that if he really doesn’t think it matters, he ought not assign it.)

      Like

      Reply
    25. Christian Marks Avatar
      Christian Marks

      Perhaps the next time I’m refereeing a paper, I will rely on the author’s prior publication record instead of doing the job I’m not paid to do.

      Like

      Reply
    26. Jon Cogburn Avatar
      Jon Cogburn

      Ha! That’s preposterous.
      Anyone who read the debate between the Journal of Philosophy editors and the hoi polloi (protesting lack of blind review) in the APA Proceedings a decade or so ago knows that reviewers and editors at journals that don’t triple blind review have a good will, know they have a good will, and also suffer from no heuristic biases whatsoever.
      What’s the point of philosophical training if it doesn’t blind you to your own imperfections?

      Like

      Reply

    Leave a comment