Exams
| Here are the results from one class's midterm exam: A+: 3, A: 8, A-: 3 B+: 2, B: 6, B-: 2 C+: 2, C: 3 D: 3 F: 5 That's with a curve. Without a curve, the average was a 75% and the median a 78%. Half the class would have been in the C-range or below, which would produce a whole lot of unhappy campers (and therefore bad evaluations). While this might be taken as grade inflation in action (after all, nine people scored in the A-range without a curve, three of them with perfect scores), in other disciplines curves are the norm. The question is where and how the curve should be drawn. I'm not entirely sure. This one breaks down to 38% A's, 27% B's, and 35% C's. For comparison's sake, essay grades typically come out to about 15% A's, 70% B's, and 15% C's. Any thoughts? Also, do y'all publicize the breakdown in scores, or just the cutoffs for different grades? Or nothing at all? |
by Simplicius @ 8:12 PM
At 2/21/2006 11:37:00 PM,
Hieronimo wrote…
I do not publicize the breakdown of scores, and I don't usually curve it, except occasionally I give everyone an extra 4 points or something like that (not a real curve). But usually there are enough people scoring 90% or above to make me feel like it should not be curved.
I never get complaints about the exam grades, I think because of the seemingly quantitative nature of the grade. They do get pissed that I actually ask them to know something for the exams.
At 2/22/2006 10:32:00 AM,
Hieronimo wrote…
Oh, I also don't explicitly tell them the cutoffs for different grades, I guess because I assume they all know the usual 95, 92, 88, 85, 82... system. I don't get many questions on it.
At 2/22/2006 11:54:00 PM,
Glenn A. Primm wrote…
how about you grade them on their performance?
if they suck, well, tough.
At 2/23/2006 02:46:00 PM,
Hieronimo wrote…
G. Randy,
There's certainly an appeal to the "whatever % you get is what you get" policy, and I use that one sometimes. On the other hand, it's not always easy to tell if you've made the exam too difficult to be "fair"--that's why in math and science classes, curves are regularly used. The assumption is that, while a single professor's decisions about the level of difficulty in making up exams can vary widely and often unconsciously, in a relatively large class the level of ability of students over time won't vary *that* much. Thus if you usually get an exam average of, say, 82%, but suddenly on one exam it's 65%, odds are this is because you inadvertantly made the exam unfairly difficult, not that the class happened to be 17% below the norm. So that's the rationale for curving when you get anomalous results.
At 2/23/2006 09:46:00 PM,
Hieronimo wrote…
PS How does someone manage to be both Randy and Prim simultaneously? An achievement worthy of Isabella's prone and speechless dialect such as move men.
Scribble some marginalia
<< Main