|
Proper "curving" of grades relies on the assumption that grades will be distributed among a group of students in a pattern that roughly matches a normal bell curve distribution.
It is not a grade given based on how many points you achieved on a test but rather a grade given based on how you performed in relation to the other students.
So, for example, the average score achieved might be given a C, with all scores out from that score (in either direction) given Cs until 50% of the class has a C. Then the next 20% of students on the high end get a B while the next 20% of students on the low end get a D. Then the top 5% get an A while the bottom 5% get an F.
Under the strictest form of this system is it is impossible for every student to get an A. Even if every student scored 100% they'd all get a C. But theoretically it also means that even if every student scores less than 50% some would still get As.
Now, abstolutely strict curves really never happen but this was the basis on which GPAs were originally given comparative value for ranking students. Since all students were always graded on a curve and all went through the same curriculum there was consistency allowing for more realistic class ranking (as opposed to how you go about comparing an A in auto shop for Student A to an A- in AP Calculus for Student B).
Anyway, what most of my teachers in high school did was not a curve. They adjusted the scale. They said "ok, whatever the highest score in the class is, we'll call that 100% and an A. Then anything within 10% of that score will get an A as well. Anything between 80% and 89% gets a B, and so one.
|