How Exactly Does the 'Class Curve' Work?

We really didn’t use it in my high school–except once. For whatever reason (perhaps the class had done especially bad on this one test), the teacher said he’d use it. Up until then I had only heard of it, and it was interesting what he did. He put the highest score on a list and said that was the “A” range. Then proceeded to divide the list into the other grade ranges. His method obviously wasn’t very technical, even if he apparently knew what he was doing. But it was my only exposure to this method.

So how exactly does the class curve work? Specifically, is there any mathematical formula that is usually used? I really would like to know.

:slight_smile:

It’s pretty much up to the teacher. Some teachers have a pet system that they’ll use, while others will decide how to choose the grades on an assignment-by-assignment basis (or a semester-by-semester basis, if it’s the final grades that are in question.)

While I can’t speak for all teachers, the general method at my high school and college worked like this:

The highest score on the test was considered a 95 or 100 and however points were added to it to get there were added to everyone’s test. For example, highest score is a 72. 28 points are added to that test to make it a 100, so everyone else gets 28 points as well.

YMMV.

I think the term “curve” is short for the bell curve, i.e. the Normal Distribution. If we give a test to a bunch of students, we’d expect the distribution of scores to resemble the Normal Distribution. The teacher, then, would calculate the average and make that a C, go up so many standard deviations and make that a B, and so on. People who score, not just unusually well, but very unusually well are “breaking” the curve by skewing the distribution and making the average move up, thus causing people who would get a C to get a C- or worse.

In practice, I’ve only seen one instructor actually grade that way, and they’re pretty much free to define the grading system however they wish.

My physics teacher in high school used a “curve” in which the highest grade was considered 100 and the other grades were multiplied by the same amount. (Say 80% was the high grade. 80% * 5/4 = 100%, so all grades are multipled by 5/4.) Grading was then done on the usual 92/84/etc. scale. Never happened in the AP class because on any given test at least a couple of us would get a 99. The intro class (slightly less complicated than Rocks for Jocks), on the other hand, made frequent use of it.

As far as a real bell curve goes, I think you take the standard deviation of the grades in the class.

1 sttdev on each side of the mean is a C
1-2 sttdev on the high side is a B
more than 2 stddev on the high side is an A
1-2 sttdev on the low side is a D
below 2 stddev on the low side is failing

At least, I think that’s how it works. As has been said, teachers/professors can do whatever they want, really.

There is no one true answer, it is not a rigid concept. There’s probably as many answers as teachers.