People on this thread keep talking about “forcing” grades into a curve. I have seen 3 different types of curving.
- In large college courses where the exams are designed to be difficult, the actual raw scores usually do fall in a bell shaped curve. In such classes when I took them, a line was drawn at the median. We all knew the curve was set to a B- or whatever, that was announced in advance. So, above the line, B-, below the line, C+. More lines were drawn further out. I am not sure exactly how their placement was determined, but I never looked at this data and thought, wow that doesn't seem right. It looked very reasonable to me. So yes, these students were being compared to each other, but the variation in difficulty of the exams was far greater than the year to year variation in quality of the students, so it would be less fair in some absolute sense to grade students simply on their percent grade on those exams.
- Most professors are smart enough to realize that in a small class you may not see such the abovementioned distribution of scores. Lines are drawn between obvious breaks with the goal of assigning the same grade to students whose performance was closest. Attention is given to how the students performed in some absolute sense, such that grades might be higher in a good class and lower in one whose performance disappointed the prof. I'm not sure this process is any different than what happens when an English professor reads papers and puts grades on them. "This paper was pretty good and I gave an A- to the last paper that was about the same. A-".
- The typical high school teacher "curves" by simply adding extra points to everyone's score to get the scores into the desired range to then apply typical high school grading standards to. I've most often seen this done by method 3a, adding enough points to get the top student to 100, which could mean rather large fluctuations in curving from year to year or teacher to teacher and be rather unfair to the other students depending how much of an outlier the top student might be. And talk about competition--this approach could result in some nasty social repercussions for said student, whose identity is often known. I've also seen teachers use method 3b, where the top student is allowed to go well over 100, because even more points are added to everyone's grades, to achieve the desired number of As or to save kids from getting low grades, not sure which. And then there's the method which I can only call 3 unacceptable, which involves forcing the top students to feed answers to lower scoring students in some kind of regrade process and then allowing the lower scoring students to replace their bad grades with the grades provided by the better students.
The AP teachers are in a bit of a bind when it comes to grading because on the one hand, students get a 5 for getting something like 60 percent of the questions correct, but on the high school scale, they need usually at least 90 percent. So, is it fair to students to administer questions designed for the much more lax AP grading scale, and then grade their performance on these questions according to a much more stringent high school scale? Some type of curving seems only fair to the students.