Over at Dot Physics, Rhett is pondering grading curves:

Should you grade on a curve or not? If you are student, the answer is clear: go by whatever the instructor does. Otherwise, you have a choice. I don’t like to tell other instructors or faculty what to do because I respect their freedom. For my classes, there is no curve. Why? Well, the question really is: “why grade on a curve?” I don’t know the exact reason for particular instructors, but I can come up with some possible reasons.

My first few years teaching, I worried about this quite a bit. I talked to different faculty in the department about what they did, and to a few people in other departments. In my third year or so, though, I realized that it just didn’t matter.

The catalyst for this realization was a student who wasn’t happy with his grade. He’d been a decent student in the class, and I thought the grade I gave him was perfectly respectable, but he was hoping to go to med school, and thought he should’ve been a step higher.

I was bothered by this, because he had been a good student, so I went back and recalculated his grade using every one of the methods other people had told me about, from explicit numerical “curves” to setting the mean for the course at a target letter grade, and going one letter up or down for each standard deviation away from that mean. And every one of them came back the same way.

Since then, I’ve been much more casual about the way I assign letter grades. The relative weights of the different assignments are clearly set out in the syllabus, and I try to make the grading standards as clear as I can, but when the time comes to convert the numbers in the spreadsheet into letters for the Registrar, I just go with numerical values: 95% is an A, 85% is a B, 75% is a C, with pluses and minuses spaced roughly equally between those. I may shift the line between B+ and A- (or any other adjacent pair of grades) if I think a particular student deserves a higher grade than strict numerical cut-offs would indicate, but that’s a tiny effect.

And as long as I’ve been doing my job well during the term, the grades fall out more or less where you would expect. There is always a student or two whose score is around 95%, to get an A, and the weakest students come in around the C/D boundary. I don’t think I’ve had anybody finish the course and earn an outright F– I’ve occasionally had to fail students because they failed to turn in labs (we have a policy that students must have a passing lab grade to pass the course, and that all lab reports must be handed in to get a passing lab grade), and I’ve had students who were headed for an F drop the class, but I don’t think anyone has been so dismal they didn’t deserve even a D.

I think that’s really the key. If I’m doing my job, and properly matching the content of the tests to the content of the class, everything works out fine without needing to curve anything.

When students ask, I do tend to say “The person with the highest total grade in the class will get an A,” which works remarkably well to reassure students who are hoping for a curve. I don’t tell them that that’s because there’s almost always somebody in the class who gets A grades on all the individual assignments, and not because I curve the grades to get there.