I recently taught an undergraduate seminar course, in which each student has to read a research paper, summarize it and present it to the class. In the first lesson I explained to the students the components of the grade: attendance, summary writing, and presentation quality.

Most of the students did great jobs: attendance was near full, summaries were good, and presentations were great. Many of them did more than I expected - they contacted the paper's authors to get more information, presented movies and demos, engaged the class in discussions, etc. So, when I wrote the grades to myself, most of them were between 90 and 100.

But then, when I told this to the department vice-chair, he told me "what have you done? You will cause a grade-inflation! The department policy is that the average grade on seminar courses should be at most 85!" I totally understand this policy - grades that are too high might show that the course was too easy, and they do not sufficiently distinguish between good and better students. Also, they might be unfair to students who took the course in previous years.

However, I am not sure what I should do now. I haven't published the grades to the students yet. Should I just re-scale the grades so that 90 becomes 70? I feel this is somewhat unfair to the students who worked hard for their presentations. Are there better options?

  • 1
    Comments are not for extended discussion; this conversation has been moved to chat. Do not post answers in comments. Please see this FAQ before posting another comment – Wrzlprmft Sep 6 at 15:32
  • 3
    Is this kind of assignment (and its relative weight in grading) typical for this kind of seminar in your department? – 1006a Sep 6 at 18:04
  • 6
    How detailed were the grading standards on the syllabus? Was there a rubric specified for the paper/presentation assignment? Was the grading schema (formula to convert numeric grades to letters, if any) publicized at the start? – Daniel R. Collins Sep 7 at 14:10
  • 11
    Has your department vice-chair (or other heads of the department) specified what they expect you to do for this semester's students? – Kevin Kruse Sep 7 at 18:57
  • 2
    Allow me to play the devil's advocate: If the students are used to courses in this department, then perhaps the reason they tried so hard is because they know other students will also be trying hard, and they assume they will be evaluated relative to their peer group. Did you tell them in the first lesson that they would not be compared to typical performance in the class? – Matt Sep 11 at 13:11

13 Answers 13

One really wants to say that the vice-chair is an idiot, but I will refrain. The policy is idiotic in any case. I've taught at places in which nearly every student excels on every measure I could devise. Why would I want to pit one student against another for the purpose of an artificial "average"? They weren't average.

If you have the twenty best people (students or employees for example) in the world and you measure them in any single way, then half of them will be below average. Just how fine can you make the graduations so that someone can be "called" worse than someone else.

On the other hand, if the average, as measured over several runnings of the class, is around 85 and if there are clear differences in behavior and outcome, then "rewarding" everyone equally is also idiotic.

You set a standard. People met it. The standard was not that you must do "better" than someone else. If you change it now you have an ethical failing.

I hope you have enough standing in the university and in the profession that you can stand up to such unfair and unethical suggestions. If your group of students was exceptional there is no reason not to mark them as exceptional.

If you want to lower the average in future, with a different group of students, make the course more demanding.

But if you change the grading structure after they have finished their work then you are doing evil, not teaching.

  • 115
    @Buffy It's actually much, MUCH worse than that. Stack ranking means the best way to keep your job is to sabotage someone else's work. You just changed everyone from working towards a measurable goal (which already has weaknesses, like gaming the metric), to everyone trying to work against each other (which is outright toxic). – Nelson Sep 6 at 0:23
  • 20
    @BenNorris you, and I think also Buffy, are talking about something that sounds quite different from what OP is asking about. There is no indication in the question that OP is considering doing anything like “assigning grades in a different way than what he told the students he was going to”. He is considering recalibrating the grades, i.e., changing the function mapping quality of presentation/writing/etc to a numerical score. If that function was not explicitly told to the students, he has every right to change it up to the very last minute. In fact, such last minute decisions ... – Dan Romik Sep 6 at 2:27
  • 15
    ... are extremely common, and I have never seen anyone argue that they are unethical, inappropriate, evil, or that a grade grievance based on such a decision has any chance whatsoever to succeed. What matters from the point of view of ethics is that: 1. the instructor follows to the letter their own grading policy that they explicitly communicated to the students, and 2. the instructor’s decisions are made in good faith and out of a sincere desire to have the grades most fairly and accurately represent their evaluation of the students’ abilities. – Dan Romik Sep 6 at 2:30
  • 71
    @DanRomik You call it recalibrating, but this certainly seems like grading on a curve: Because everyone did well, we're recalibrating the average to fit what we think it should be. "Many of them did more than I expected - they contacted the paper's authors to get more information, presented movies and demos, engaged the class in discussions, etc." Your grade: C, Although you did exceptional work and exceeded expectations, others did more/better than you did at exceeding expectations. Tough luck. Talk about moving the goalposts. – TemporalWolf Sep 6 at 20:27
  • 29
    @DanRomik: I will argue that (while extremely common), the practice you describe is indeed unethical, inappropriate, and evil. E.g., my letter to the editor of Thought & Action in 2006. Most best-practices I've seen recommend documenting the number-to-letter formula up front on the syllabus (e.g., Yale). The "recalibrating" argument is simply word-play. The standards for a particular letter grade should be, and in many cases are, made known beforehand. – Daniel R. Collins Sep 7 at 3:52

From a mathematical point of view, the problem that they are trying to solve with a policy that attempts to enforce a maximum/minimum average class value is the wrong problem. Statistically, you will have plenty of variation from class to class as to good classes and bad classes. If the general guideline is that you want an "average" student to receive an 85, that is fine, but you will get classes where your students excel abnormally, and you will get classes that are a bit behind the grade, and fluctuate the other way. This is even more common, in fact, because when an entire class is doing badly, they have more trouble getting help from their peers, whereas when a class is excelling as a whole, friendly tutors are easy to come by and the performance of the under-performing students rises.

Forcing any individual class to conform to an exact standard like that flies in the face of statistics. I would take my department head and go talk to the head of the statistics department, and rework the policy so that it takes into account things like standard deviation within a class, the tendency toward positive or negative feedback cycles, and outliers.

This is an academic setting. It can be easily recognized that the existing policy does not meet scientific rigor, and there are experts a few rooms or halls away who can help.

  • 18
    That’s a very nice way of framing the issue. Indeed a blind insistence on a fixed average grade that ignores statistical variations in performance levels between different groups of students would be deeply misguided. – Dan Romik Sep 6 at 15:55
  • 3
    @DanRomik Remind me to invite you to visit some UK universities if you have sabbatical leave coming up... ;-) – Yemon Choi Sep 6 at 22:47
  • 3
    Good points here, I will add (as one who was assigned the lower performing classes & students) that classes below the average also end up covering less material or having to pick or choose which material to cover in depth (because you need to scaffold learning to cover gaps in their understanding). While, those performing above the average allow you to cover more in depth as there is both less scaffolding (less gaps) but you can go more in depth with the fewer students who need that. @DanRomik I have definitely seen this in US universities. – JGreenwell Sep 7 at 11:14
  • Downvoted because what you are saying flies in the face of statistics. The bigger the 2 groups are more likely it is that they will have the same average on same problems. Not to mention that even if differences exist they are not of the magnitude that OP describes. – NoSenseEtAl Sep 13 at 11:22
  • @NoSenseEtAl "The bigger the 2 groups are the more likely it is that they will have the same average" Yes, exactly. The original poster did not post their class size, but many graduate seminars have limits of 15 or less students. That is no where near a high enough sample size to expect any kind of real conformity. In a class that small, each student accounts for 7% of the class average, meaning that just 1 or 2 extraordinary students can completely throw the average. If this were a class of 300, then large numbers apply, but in a small class, there is just too much deviation. – kashim Sep 13 at 22:53

Based on your description of what took place, it sounds like the reason most of the students will get a grade between 90 and 100 is because according to your best judgment they genuinely deserve it. If that is indeed your belief, you don’t need to do anything other than to publish the grades as they are. After all, the department charged you with teaching a class and grading it in the most fair and accurate way and to the best of your abilities, and that is what you have done. I would simply recommend, before publishing the grades, to make sure that you are not violating any formal policies the department may have that mandate a certain distribution of grades or range of average grades; and, since you are going against the vice chair’s opinion, it may also be advisable to either discuss the matter some more with the vice chair and try to reach a consensus as to the desired course of action, or to make sure you are comfortable with going against the suggestion of your senior and that you will not suffer any adverse career consequences as a result, before proceeding.

If on the other hand the premise that the students in fact deserved the good grades you were going to give them is not something that you strongly believe in or are sure you can defend if challenged, then I think it’s reasonable to follow the vice chair’s suggestion and curve down the students’ grades to reach an average grade that is in keeping with historical norms in the department. After all, if this year’s students are no more talented than, and worked no harder than, last year’s students, then having the average grade remain the same across successive years does sound like the most fair grading methodology.

To summarize, it all depends on your personal convictions regarding what the students deserve. Try to do what you thinks is the most fair for the students, while being attentive to the broader concerns of others in your department.

  • 64
    I think changing the grading structure after the course ends, to make it less generous, is completely inappropriate and can't be defended. It is breach of contract, in my view. The fact that the students don't yet know their grades is immaterial. If I work hard and believe I've done all that was required for a grade and you tell me, no, you get a lower grade, I don't think you can defend the decision. – Buffy Sep 5 at 22:46
  • 11
    @Buffy grades haven’t been published yet, so all that’s happening is that OP is making an internal decision between himself as to the calibration of the grades. Instructors do that all the time and it’s neither inappropriate nor requires defending, so I don’t understand what you’re talking about. It’s not “no, you get a lower grade”, it’s “here is the numerical score that I’ve decided, after considering all relevant factors, most faithfully represents my evaluation of your work”. – Dan Romik Sep 5 at 23:54
  • 40
    From the student's standpoint that sounds like a game with secret rules in which their work and effort is only one component of grading. The other being some hidden and mysterious process. I don't grade like that. My students knew exactly where they were at all times and had an opportunity to improve their grade if they didn't like what they saw. (Retired now, so past tense). I can play evaluation games at a lower level of granularity (this paper is worth 80...) but when the course ends, they get what they earned. No question. – Buffy Sep 6 at 0:00
  • 8
    Still (strongly) disagree. See the note about granularity above. You are dangerously close to making grading purely subjective. You earn what I say you earn. It is a bad practice. There should be no confusion in the student's mind about their standing and progress. – Buffy Sep 6 at 0:04
  • 18
    @Buffy I like and have upvoted both this answer and yours. I take issue with your comment here that Romik is "dangerously close to making grading purely subjective". In the course of my 60 years of teaching I came to believe that grading really is mostly subjective. I know mine was, after staring at spreadsheets with spuriously precise numbers. – Ethan Bolker Sep 6 at 1:23

You are in a less than ideal situation.

This happened to one of my undergraduate professors, who I had for a year long course (actually, two back-to-back semester long courses). The department told him his average grade for the course was too high after the first semester. The professor told the department he would change his grading in future years, but would keep the same grading for the second semester course because we were the same group of students. This was my professor's first year teaching at my university. I respect him and still keep in touch with him to this day.

If you are unwilling or unable to stand up to your department, I would turn the question on your department and ask them how to handle the situation and make them own the decision if they want you to lower grades. Also, hopefully you have a mentor at your university who can help you with local politics. I would add the warning you might accidentally burn bridges in your department if you do not handle this well. 

Grades are too high for the department - what should I do?

Make the class harder next time you teach it.
Set the bar higher in terms of learning objectives and challenge your students to meet that higher bar. Continue whatever teaching strategies you were using to promote that engagement, learning, and quality work among your students.

If there is a formal policy requiring a certain distribution of grades, obtain that policy and share it with the students. Share with the students what their grades would be without that policy and what they are with it applied. Tell them where the policy comes from (i.e. which authority/group within the university) and, to the extent you understand it (maybe ask the vice chair for some help here) the reasoning behind it. Then the students can organize to lobby for a change in the policy if they want (as other comments/answers have highlighted, there are good reasons to do so).

If there is not a formal policy requiring a lower distribution, issue the grades as they stand and just use this comment from the vice chair as a feedback point for you to raise the bar for the next class.

If you think the bar is already quite high, and students are meeting it with exceptional work, share some of that exceptional work with the vice chair and offer to ask students' permission to highlight their work in appropriate public promotions of the department's educational programs as examples of what students from your department are successfully able to do.

The most important outcome is students gaining knowledge and abilities they didn't have before. If you can show impressive success at that, and the department can show this to the dean, provost, prospective students, current students' potential employers, etc., the educational mission is a demonstrable success.

Regardless of the grading policy, be sure to give the students the positive feedback you feel they deserve. They worked hard and exceeded the standards you set and deserve praise for that.

  • 5
    +1 Sunlight is the best disinfectant. – JeffE Sep 6 at 19:29
  • 4
    +1 for "Share the exceptional work", Sometimes an inspiring instructor meets a hard-working, creative class, and a high number of A's is the result. – J.R. Sep 10 at 15:07
  • 2
    Back when I was in the military, I was in a Data Processing A school and has the (mis)fortune to be in a class full of people who knew what we were doing. Yes, we totally blew away the lifetime average of the school, and until the program was revamped due to rate changes, we were mentioned to having had set the highest they've seen, and dared other classes to do better. Don't change it to a curve. There is nothing in the curriculum or subject matter that has them competing with each other, so that should not be part of the grading. – whiskeyfur Sep 10 at 19:22

Giving high grades doesn't cause grade inflation. Grade inflation is where you give higher grades for the same quality of work than the year before. I assume he does understand that but what he wants is to avoid having to explain in the future why grades went down one year.

I think a good way for your department to handle this would be to have a number of other members grade a random sample of your class and move the average accordingly. Better still would be them at the same time grade some previous years samples so the grades are 'calibrated' based on the previous grades.

If there is more than one class, with more than one teacher/lecturer/tutor marking, then you can cross-mark a sample of each group to establish a baseline. This is known as moderation.

You could do a similar exercise if, in previous years and under the same syllabus/criteria, there is student work available as a comparison.

There is a school of thought that says you should never give 100% for a submission because there is always something that could be improved. I disagree with this because if the student addresses all the criteria and meets or even exceeds the standard set then they deserve the credit.

It's like going to work, doing your job to the best of your ability, delivering what was asked (or more) and your employer saying that you're only going to be paid 90% of your salary because there is always something that you could have done better. Nobody would stand for that.

It's the same flawed argument that if you grade a large enough sample of students on a normal distribution then half of them are below average. You don't fail half of the students on this basis because there are modifiers to the sample - not everyone has to take the class, the group of students in the class have progressed to a certain level and those who could not meet the standard are no longer in the group, etc. The students in the class are already at the upper end of the distribution.

The students should not be penalised for delivering what was asked. I would discuss how to address it in the future but submit these grades as they are.

Let's set aside your vice chair's use of the term "grade inflation". No, that term means something else, as PStag has already pointed out. The important thing to know that he's telling you your grade distribution is off from department norms. He's saying he thinks you're an "easy A".

It's up to you to decide what to do with that information. He may or may not be right and you may or not care, especially depending on whether it's a grade on an assignment where everyone gets an A (so what) versus a final grade in a course.

You're the instructor, you decide your students' grades. This is solely your responsibility and your call, nobody else's. If you're satisfied your grading is fair and correct, it doesn't matter what anyone else thinks. You have to call 'em as you see 'em.

But most instructors do like to keep their grading in line with their colleagues'. Most of us would like to avoid students thinking there's a grade advantage to one instructor over another. Also, there are a lot more students than instructors and we expect that with a class-sized sample, we can be more confident of what the distribution of A's and B's should be than of whether an assignment or exam we just wrote will turn out to be harder or easier than we expected.

So you might reach out to your colleagues or bring up the question at your next department meeting, asking others in the department how they set their curves. It might be helpful to know more about how your colleagues decide grades.

But let me get to the real bottom line: If everyone gets an A, it is likely, as WBT argues, that the material was too easy. It is your job to match the material to your students so as to be able to make that distinction. When I read your description of the assignment, it sounded easy to me and I wasn't surprised by what happened.

If it were me, I'd make the project more difficult somehow, add more structure, more hoops for the students to jump through, a more fine-grained rubric with pickier grading, whatever it takes to push the students enough that you can differentiate their performances.

  • +1 on all counts. I'll note that being "demanding" and being "generous" aren't inconsistent. But it usually takes a smidgen of "inspiring". – Buffy Sep 7 at 12:17
  • "we expect that with a class-sized sample, we can be more confident of what the distribution of A's and B's should be"... I was disabused of this notion a few years ago. I had two sections of the same course, meeting on the same days, receiving identical lectures and exams, etc. On the (uniform department) final exam, 23% passed in one section and 60% in the other. (This being consistent with other assignments/exams; N = 25 in each section.) I'm pretty sure that student performance is often dependent on the section they enter. – Daniel R. Collins Sep 10 at 5:37
  • @DanielR.Collins Twenty-five is a very small class. – Nicole Hamilton Sep 10 at 10:41

If the grades are deserved as you state, and within the marking framework for the course, then hold the line and keep the grades as they are. It is not the fault of the students that they all achieved well within the framework that was set before the semester started, therefore, they should not be 'retrospectively punished' in their grades for doing what they were asked to do, and ultimately doing it well.

This situation does pose a series of opportunities:

  1. Restructuring the coursework/assignments
  2. Reframing acheivement levels in the marking criteria
  3. Department wide review to ensure parity across the student assignments for different courses

In all my lecturing roles, these periodic reviews were undertaken on a fairly regular basis e.g. every year for small adjustments, every 5 years for a full course overhaul.

In your case, as I said, you must maintain the grades as they were deserved in this instance, but for next years cycle of teaching, make sure you have set clear expectations in what is needed to be done to acheive each grading level through a marking criteria matrix (and make sure the students can access this), and structure the coursework, so that additional work can earn extra credit. Students at different levels of abilities will then work at their appropriate level and that will be reflected in a broader distribution of grades.

If however, they all complete the work within the framework you set, complete the extra credit work, and all at a high standard - then they deserve the high grades.

The sentiment of the vice-chair is really quite alarming and very sad: He should have said "What have you done? The engagement with your course is fantastic! I think it would be really valuable if we could have a discussion about your overall approach, and your methods of teaching and assessment and see if we can implement some permanent changes for the seminar in future years, and perhaps on a wider scale in the department."

Of course that's in a perfect world where everyone has plenty of time for workshopping and professional development, and the vice-chair isn't stressed out and run off their feet and isn't able to see past the next fire they have to put out. :)

That, said, I think you should take this tack: Basically, talk about how happy you are with how it's gone, with how engaged the students are, and how this certainly gives you the perfect opportunity to increase the difficulty/scope next time and that you have some guidelines/recommendations for the course in the future to help make it so effective every time it's offered.

Basically, act slightly naive, but completely positively: relentlessly keeping focus on the benefits that have come solely from your hard work, and those that are still coming

The question of distribution of a good (in this case grades) has been extensively discussed in public finance. Warning: if you consider the way economists usually handle questions of distribution cynical, you may not want to continue reading and save yourself from throwing a fit.

Here is how this problem would be addressed from a public economics standpoint. Consider obtaining data on the performance of the students in other classes. If the students consistently scored highly, you may present this as evidence to the department that high grades are justified. If your course is new and has an exotic title you may have a sample of students that is not representative of the overall student pool. This violates the idea that grades should be independent across courses but throwing away (potentially) useful data would be irrational.

However, you may also find out that your students actually perform very poorly in other classes. In this case, you may ask yourself the question whether the criteria by which you score the course are likely to be uncorrelated (or even negatively correlated) with performance in other courses. For example, if you teach an acting class to econ majors...

A third reason why students performed well could be that your teaching style is very motivating or the topic of intrinsic interest to the students. In this case, you need to decide for yourself (or by department policy) what you consider the function of grading the students. (This is where the cynical bit starts.) If you consider grades only to be a job market signal of a student's ability, you may not want to give high grades if the achievement was not due to ability but due to your teaching style/course content. Notice that in most countries grades have exactly that function. Better grades mean (on average) easier admission to further degrees, more job interviews, better jobs. If instead you consider grades to be a measure of the degree to which a student understands the material, you may insist on giving high grades. Just remember that in this case you are perhaps slightly distorting the job market outcomes.

Now here is a strange bit from public finance. You may also care about interpersonal equity between students. If you reward students with high grades, then students who could not attend your class (class size restrictions, limited information) are disadvantaged. Here things become a bit complex. First of all, you would need to check if your student sample is currently worse off or better off than the remaining student pool. Next, you need to balance the efficiency loss due to a worse signal against a potential improvement in interpersonal equity. Luckily, if you do not have any information about the expected lifetime income of your student pool, you may simply ignore this aspect.

tl;dr: I would be highly skeptical of my own grades if they were to vastly differ from the department's and consider it important to find out the reasons for such differences. Once you have found out the reason, you will know whether you want to act on it.

  • 2
    downvoters care to elaborate? – HRSE Sep 10 at 1:24

It's a tough situation that you are in... You feel like your class deserved the grades you gave them, while the department wants to limit the grades.

I assume that you had a grading rubric for the various subtasks? If yes, great...If no, ok too. There's a difference either way. The good thing for a rubric is you can later point to it and justify grading -- both positive and negative.

My thoughts are to either go to your supervisor and say that this class deserves the higher grades because while people were required to do X they did Y, which was beyond the scope of what they had to do, so they deserve more.

Or, re-calibrate the grades so the people who minimally met the requirements would get an 85%..if they exceeded the minimum, they'll get more -- and rightly so.

This is your first problem:

I feel this is somewhat unfair to the students who worked hard for their presentations.

You think just about your students. What about students that got 80 for the work that you are willing to give 95? Is that fair to them?

Your second problem:

Course is too easy. Now it is too late to fix this, but for the future you should make sure students have hard time "maxing out" easily. Not to torture them for fun, but to make sure that the spectrum of outputs is recognized(instead of half of the results piling on around 95).

This brings us to your current situation. You should rescale the grades, but unfortunately this will hurt some students because since course is easy precision of the measurement was bad. But it is unfeasible for you to do anything about it now.

protected by Wrzlprmft Sep 8 at 9:10

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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