I teach an elective chemical engineering mixed-level class. About 30% are 3rd year and 4th year undergrads each and the remaining 40% is a mix of Masters and PhD students.

I want the grading to be fair to everyone. Clearly the advanced students have an advantage in terms of pre requisites etc.

Wondering if others have dealt with this sort of situation before. Our grading is relative and using a curve.

Would grading the undergraduate students on a different curve be an option?

Any thoughts?

  • 15
    Why is it "clear" that the grad students have seen more prerequisites?
    – Buffy
    Commented Jan 23 at 11:51
  • 9
    In similar courses I've taken, to get the graduate level credit, there was usually more/additional homework, papers, etc but the lectures/tests/base homework were the same as the undergraduate level work. Commented Jan 23 at 20:36
  • 4
    I don't know what it's like in the USA, but my understanding is that here in the UK PhD students taking taught courses don't tend to care much about the marks and as a result tend to get lower marks than they got in similar courses as undergrads. Commented Jan 24 at 13:40
  • 3
    @PeterGreen: here in Germany, PhD students are sent to undergrad courses because they presumably haven't mastered the subject. E.g. I know of biologists who went to do their PhD in chemistry. They had to do take undergrad exams/courses for particular subjects (related to the PhD project). The idea was to make sure they have the same level of knowledge as the chemistry students. Commented Jan 24 at 16:20
  • 1
    @PeterGreen: Yup same thing in the US. It's not that they couldn't get better grades (although sometimes that's true too, because it's not in their field), it's just that they have other stuff they're supposed to be putting more of their efforts into.
    – user541686
    Commented Jan 24 at 23:53

7 Answers 7


Rather than different "curves" or grading schemes for the two groups, you could teach two courses (see below) in one with different tasks and exams for the two groups. Some of the tasks could be common to both groups and some could be more advanced for the grad students. The lectures could be the same and the readings different, for example, and more challenging exams for one group.

But, curving, in the sense that for one person to get a better grade someone else needs to get a worse grade is, IMO, unethical. There should be a standard that if met results in a defined grade and the path to meeting that standard should be clear to everyone. Everyone should be able to get "full marks".

Not every group has a predefined skills distribution that the grading scheme forces on them. Few do, actually.

For this to work, however, the university would probably need to create such a split course in its records: CE401 and CE601, say with different people being given credit for different courses, even if they "meet" together. Absent that, you will have difficulties justifying such a split scheme. If a grad student gets a B while an undergrad gets an A while doing (or learning) less but for the "same" course you will hear about it, and not in a good way. Without such an explicitly split course, I suggest you get specific permission from the university, say at the "Dean" level.

  • 7
    I think that's a very good suggestion. I had that for one of the course during my master. We were mixed with 3rd/4th year undergraduate, with the same course, but different tasks, and extra question on coursework/exams. Specifically, our exams was more research oriented. I also TA for this course during my PhD and it seemed to work fine in term of pre-requisite and work done by the students of both groups
    – JackRed
    Commented Jan 23 at 12:47
  • 1
    This also gives the option to the undergrads to earn the grad level credit if they are good enough for it.
    – quarague
    Commented Jan 24 at 10:57
  • 1
    +1 just for the last paragraph. If this wasn't done from the start you may be losing out on extra pay for teaching graduate students. A possible alternative could be to have all the graduate students sign up for self-study courses where the classroom material is the base and additional work is assigned that would fit a graduate course. Make sure this self-study course can be used to meet their graduation requirements.
    – gns100
    Commented Jan 24 at 16:11
  • 1
    The most sensible approach. Either the material is the same, in which case their mastery of it must be evaluated uniformly, or you give them different material. If it's the same and it feels like it's either much too hard for the juniors or much too easy for the seniors (defined as grades being much too low or much too high when a uniform scale is used), the real question is why one group or the other is in the course, but that's not necessarily in OP's power to change. And if it's different, the credit should presumably be different. Commented Jan 24 at 23:36
  • 1
    I took a statistics class like this as an undergraduate; the graduate students had extra questions on the homework and exams, and more challenging projects. One suggestion I might add: allow undergraduates to try the graduate questions for bonus credit :D
    – elutionary
    Commented Jan 25 at 16:23

I do this all the time. Grad courses that are split with senior undergrads, and open across Chemistry, Physics, and Materials Science and Engineering.

If your course is split, i.e. grad and undergrad with separate course codes, you typically need to provide some differentiation in terms of undergrad vs grad to justify having two separate but concurrent classes. You should not provide differentiation based on their background (prior courses or experience). You need to provide a consistent and fair learning environment.

When I set my differentiation for course approval, and in practice, I provide the same content for everyone, but I build in opportunities for higher order understanding for the grads. For example, if I provide a writing assignment, the portion where the student is asked to synthesize a concept based on literature is weighted heavier, or intended to be longer for grads. Similarly, when I ask an open-ended question on an exam, I expect a deeper level of answer from the grads. As long as you are clear in your rubrics about this, then it is easy to run. Just make sure your expectations are consistent and expressed in the syllabus.

Otherwise, I have two pieces of advice:

(a) do not create too much differentiation, especially not in the day to day of how the class is run or the content taught. It's better for everyone if the course is cohesive and everyone is involved in all content. If you have concerns about lack of content/context for the undergrads, what I personally do is record a couple of videos and upload it to the course Canvas page for undergrads (and really grads as well if they want), to view and get caught up.

(b) if you are allowed to, drop the curve, it's just a nonsensical way to run a course....

  • 3
    For open ended questions where more in depth understanding is expected from grad students I think using different curves actually makes sense. It removes a potential grading error (evaluating a student on the wrong criteria), and allows students to more easily compare their success (ex an 75 is an 75 in either case but is a C for a grad student and only a B for an undergrad). For undergrads planning to continue their studies it also makes it easier to know if they learned the material well enough for the advanced degree they ultimately want or need to plan on retaking it later. Commented Jan 23 at 17:27
  • 2
    +1 for (b). Grading on a curve is toxic and encourages students to see their learning (and life in general) as a competition and a zero-sum game, rather then embrace collaboration and recognise win-win outcomes. Commented Jan 24 at 13:05

At my institution at least, there would be two course numbers assigned for the same lecture and the graduates are supposed to do more work and show more success for the same grade. If it is officially a single class, then grading differently and even offering different assignments would open me up to complaints of discrimination.

Besides, often senior undergraduates are better prepared than first year graduate students (who might come from a different academic background).


Having been an undergraduate in Computer Science who took graduate level courses, my suggestion would be to know your students and their technical capabilities.

In computer science specifically, a number of graduate students start their graduate program there without any support from a lesser degree, and in a course about analysis of algorithms, junior and senior undergraduates would have a major advantage on them. In this case what Buffy suggests, and teaching two separate courses would likely be best for both groups, and as you get to know your graduate students, you can promote them into the harder course as they prove themselves capable.

For example, in my case, on the final exam, the professor put a question on the exam asking us to do a topological sort of a graph with 7 nodes in a horizontal line that all had edges pointing right in some fashion, such that: "A, B, C, D, E, F, G" was one valid sort, and he admitted to the class that 80% of the graduate students and none of the undergraduates got that question wrong. The professor was quite upset at the level of curve he would have to give to allow for even some of them to pass. All through the course he been complaining that some students just weren't prepared for his course, and in reality all I got from his course was a nice textbook with some advanced algorithms, and a professor that I could go to when he had time to ask questions.

I disagree that it should be split between different course codes, as likely this has already been done at your institution, has it not? Is it not the case that the undergrads are taking the graduate level course as an elective? If this is not the case, then this should definitely be done if possible, though maybe your institution has problems with limited class size in this case.

If it is the case where your grad students in your field are more technical than your undergrads, while it is not ethical to let them teach as of yet, can you perhaps have them lead small research groups in a class project setting? In this way you can have different grading and goals for undergrads and grads, but still be overall assigning the same thing to both.

All in all, I honestly think it comes down to knowing your student body, and what they can individually accomplish. Build on that, and you should go farther than you would if you simply tried to force a one size fits all solution on them.


How is it fair, if two people achieve different results (like test and assignment scores) but both get 75%?

One way to handle it, institution permitting, is to "code share" the course. The Master's and Ph. D. level people register for EE542. The undergrads register for EE442. They go to the same room at the same times and listen to the same lectures, but the EE542 people get some more challenging assignments and maybe some different exams.

Two people of different levels of experience can each get 75% in the course: one has it in their transcript as having passed EE442, the other as EE542. No monkey business took place with how their work was graded, like one being more strictly graded than the other, or adjusted through a curve; they actually did different work in accordance with what they registered for.


Not only is it entirely normal to grade grad students and undergrads on a different curve, but in fact, sometimes professors do this in mixed- and grad-level courses to 'protect' the grad students from the undergrads, rather than the other way around. (!) There are a few reasons for this, such as the following:

  • Undergrads are supposed to focus their attention and efforts on getting good grades; grad students (at least for research programs) are supposed to focus on research, and are generally encouraged to not worry about grades as much. (Faculty often say this quite explicitly, though perhaps not in writing.) It would be hypocritical to say this and then expect both groups to obtain comparable grades.

  • Undergrads (sometimes) have lower requirements for such courses (e.g., no final research paper), and thus naturally have an advantage. Though in this case it's simply wrong to grade them on the same curve, IMO.

  • Undergrads that take courses intended for grad students can sometimes be especially strong in that area (sometimes even already doing research in the field and even published material on it), compared to grad students who might have only just started in that field, or even be completely unfamiliar with it and merely taking it for the sake of a requirement.

Of course, the reverse direction can be a concern depending on the course as well; there are certainly courses where grad students would have a massive advantage.

There is some judgment call to be made here, so I'm not saying this makes sense for every course by any means. However, in general: if you have reason to believe one group is likely to naturally have an advantage over the other, you're well within your rights (again, in general; I don't know about your particular institution) to grade on two separate curves. Just make sure all students understand this ahead of time to avoid needless stress/confusion/etc.


This question is proof yet again that grades are meaningless...

What is the purpose of your course work? Is it to teach the students a certain set of skills/ability/knowledge? Or is it a meaningless class that is meant to fill a checkbox?

If it is a meaningless class that fills a checkbox. You could just assign grades at random... If its not I would grade the students based off of how well they have learned that knowledge/developed that skill. Instead of grading them on a curve. If this means that every students gets an A because they are able to do xyz, then every student gets an A. If this means that every student got an F because they cannot do xyz, then every student gets an F. (*though every student getting an F means that you are a bad teacher).

  • 4
    I don't think this question provides any such proof. What if the purpose of grades is to motivate learning? Then it definitely makes sense to have different motivation standards for different types of students.
    – Bryan Krause
    Commented Jan 23 at 17:55
  • 1
    Grades motivate students? That is news to me... I thought that recent research proved the opposite journals.sagepub.com/doi/pdf/10.1177/1469787418819728
    – Questor
    Commented Jan 23 at 18:35
  • 3
    I think you need to reevaluate the word "prove" in describing social science research. Yes, grades motivate students. Every student who asks what will be in the exam is motivated by their grade. I'm not claiming it's the best motivation. I am confident that busy students balancing lots of demands on their time will neglect courses that they can pass without effort.
    – Bryan Krause
    Commented Jan 23 at 18:40
  • 1
    Out of curiosity did grades ever motivate you to <bold>learn</bold> or did they just motivate you to pass a test...
    – Questor
    Commented Jan 23 at 18:43
  • 6
    Hard to pass a well designed test without learning.
    – Bryan Krause
    Commented Jan 23 at 18:44

You must log in to answer this question.

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