On Grading

This is a short essay that I constructed some years ago in response to a student's complaint about the difficulty of a particular exam, along with her feeling that her exam grade didn't reflect how well she actually knew the material.

Grading is, without a doubt, the most odious part of teaching. I would dearly love to teach a course in which grades didn't matter. Indeed, as a tenured professor - which means, of course, that the University can't do much to me that is bad - I've been sorely tempted, as an experiment, to announce at the beginning of class, that everyone can choose his or her own grade based on how much they feel they know. Teaching would then be quite sweet, as there would be no need for anyone to complain about much of anything, gradewise in general, and examwise in particular. Unfortunately, I'm pretty sure that I know what the bottom-line outcome would be, which is that people would learn less from the class. So I haven't done it.

Where does that leave me? I have to provide grades, which means that I have to concoct a necessarily arbitrary grading system. Part of the arbitrariness is deciding how long and how hard the exam should be. These decisions, in turn, depend on the equally arbitrary decision about what an optimal mean grade should be. Should it be 100? Obviously not, because that would be equivalent to simply assigning everyone a 4.0 in advance. Should it be 50? Again obviously not, because everyone would flunk which would unfairly cause everyone major hassles. So what should it be? Well, in general, I've settled on 70 is an optimal mean exam grade: Because of the forgivingness of the grading system, 70 isn't low enough that it will hurt lots of people, nor is it so high that it will remove incentive to study.

I generally don't meet this goal: For one reason or another, the exam means usually (although certainly not always) turn out to be higher than 70. I think the reason for this is that I don't like to be the object of student furor, so that influences me to design exams to be higher than what seems optimal from a pedagogical standpoint.

Of course this goal of 70, and the system that engenders this goal is imperfect in many ways, the most disturbing of which is that many students feel that their grade doesn't reflect what they actually know. Probably for other students who happen to be exceptionally good test-takers. their grade reflects more than they actually know. But I know of no way to remove these imperfections: If I did, I would; it is no more fun for me to deal with student complaints than it is for the students to get the grades that engender these complaints.

I realize that all of this grading philosophy probably won't satisfy you. But I also hope that it provides at least a little insight about why grading systems - this system and all systems - will never be universally acclaimed no matter how they're designed.