Wednesday, June 28, 2006

Grading thoughts.

Did I inflate the grades on the final exam? The short answer is "yes", but I think I have good reasons for it. When grading, there are really two issues in play. First is between rewarding improvement and holding tight to academic standards. Second is reasonableness of expectations.

With regard to the first, generally, improvement is to be encouraged. So, a student who started out with a D and ends up with a B on the last assignment should, overall, have a final grade that tends towards the B -- but, only insofar as there can be things found in the last assignment to justify a slight grade inflation.

With regard to the second, if students are taking an accelerated course (5 weeks vs. 10 weeks), then it's likely that there's only two or three assignments (instead of five or six). This means that early assignments will probably get disproportionately weighted. In a regular-length course, grades for early assignments will be weighted less than later ones, because it's expected that students take a little time to get into a course and really reach their actual level of ability. In an accelerated course, though, this is not possible because there are fewer assignments. So, to compensate, a little grade inflation on later assignments is necessary.

So, despite initial misgivings, overall I'm happy with how the grades turned out. I think those who deserved to be rewarded were rewarded deservedly, and the disproportionality was at least somewhat corrected for. And all this without sacrificing standards.


Anonymous said...

I work hard to get B and B+s in my philosophy classes. Got 1 B+ and 1 A- so far, the rest Bs. (Ancient Philosophy and American Philosophy, respectively.)

Philosophy should be hard. Don't sacrifice standards.

ADHR said...

I take the point, but there are, fortunately or not, other considerations in play. First, I'm grading first and second-years who are, on the whole, not really that bright. Some are very good, but most are quite mediocre. They won't stay in philosophy, though, and some won't even stay in university. So, I have to adjust what I expect downwards to account for the students I have to deal with. (Why can't I just fail everyone? Well, that gets into the second point....)

Second, the faculty has regs that I'm forced to comply with. Not just on things like averages, but also on distribution (no more than a particular percentage above A; no less than a particular percentage below C; etc.). I can break these regs, but would then have to justify the abnormal distribution to the instructor, who would then have to justify it to the dean's office. This is a big hassle, so it's usually better to just try to keep it close to what we're required to achieve.