In my undergraduate courses (discrete math, linear algebra), students' backgrounds vary significantly. One group has a strong math foundation from high school, while the other, less prepared, struggles with rigorous assessments. Making exams too challenging leads to higher dropout rates among the latter, but easier exams result in too many high grades, blurring distinctions between these groups.

I'm seeking advice on balancing assessment rigor and fairness without using grading curves. How can I fairly assess a diverse student body while maintaining academic integrity?

  • 3
    $\begingroup$ What are the expected outcomes for your course? What should the students be able to demonstrate at the end? Are there courses available for the 'less prepared' to take to become more proficient? Students not being prepared to tackle a given course is not a good reason to water down the course. They should be directed to a more introductory course to become prepared. $\endgroup$
    – Jon Custer
    Dec 14, 2023 at 15:48
  • 2
    $\begingroup$ There’s quite a bit flexibility of what to expect from students. The syllabus could say “learn to count with binomial coefficients”. But I can easily give a problem that one out of twenty can do. Or I can give one nearly everyone can solve. But it’s not easy to find a problem right in between. $\endgroup$
    – user19945
    Dec 14, 2023 at 15:57
  • 2
    $\begingroup$ Most people taking calculus don't need to ever learn "what is math", although that will not be a popular opinion here. $\endgroup$
    – Jon Custer
    Dec 14, 2023 at 16:03
  • 1
    $\begingroup$ @JonCuster Heck, most people in college don't need to ever learn, period. OTOH, some can take great advantage of what they learn. Even at the highest levels, "what is math" is mainly problem-solving. That's certainly the case in calculus. Most workers are just recipe-followers. We don't need that many problem-solvers. $\endgroup$
    – user12357
    Dec 14, 2023 at 18:13
  • 2
    $\begingroup$ This is a problem faced by every computer science teacher. For some students, computer science just "clicks" and they find it extreeeeemely easy at the undergraduate level. For other students this is as hard as learning a foreign language. Grades can easily range from 0 to 100%. A classic trick is to include lots of small problems in the exam, in ascending order of difficulty, hoping that every student will solve at least the first problem and even the best students will find the last problem interesting. $\endgroup$
    – Stef
    Dec 17, 2023 at 16:27

1 Answer 1


This is essentially not a problem that a single instructor can solve on their own.

  • The current top pedagogical theory in these cases is to provide "corequisite" or some kind of outside supplemental support to help the students who need to catch up. (IMO, the in-practice results of this have yet to be proven.)
  • I would look closely at requirements for the next courses to follow (if any), and where students need to be at the end of your course to succeed at the next step.
  • You should consult with colleagues, chair, program coordinators, etc., on the expected proficiency level at the end of your course, historical and expected pass rates, how the institution responds to high pass or fail rates, etc.
  • Once this requirements-gathering phase has been done, my core advice would be to set reasonable standards for success (in your class and to enter the next level), and hold fairly to those expected standards.
  • Give test questions which demonstrate the basic, essential, expected skills. Avoid questions which are overly clever or "interesting" from the instructor's POV.

Ultimately if the context and institution are letting low-skills students into your courses, who cannot possibly succeed at the expected level, and there aren't the needed outside supports to help, then the individual instructor cannot fix this. Sometimes the job is to give students the taste and opportunity to try a certain step, and give the difficult lesson that they're not ready to digest it at the current time. There's a passage from Ken Bain's What the Best College Teachers Do that sticks with me:

Susan Wiltshire, a classics professor at Vanderbilt, captured a sentiment we heard often. Her classes, she explained, were in her view like a great meal she had prepared, and she simply wished to invite her students to the dinner table.

Ultimately, and especially in a scaffolded STEM discipline like mathematics, the instructor needs to maintain certain standards for students to succeed at the next level. If the institution is funneling students into a situation where they can't succeed -- then, unfortunately, they won't succeed.

As a brief example, I teach at an open-admissions community college which is simultaneously part of an R1 university (every credit we award is a full credit for both). We've been told in official guidance, "The college must allow enrollment of students who are not skills proficient" in our lower-level, corequisite credit courses (emphasis in original). Failure rates in courses throughout our program are about half. In some cases of remedial courses I've had semesters where the median grade for all students in the semester was circa 5%. So there are certain contexts where high failure rates are simply an expected part of the institution's operation; you should find out your context and calibrate accordingly.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.