On 02/12/2011 01:55 PM, Kirsten Manning wrote in part:
2) Remaining objective in grading questions that are qualitative.
That's an important objective.
Here are some possibly constructive ideas (certainly not an
exhaustive list)
a) Sometimes when faced with a big stack of papers, it helps
to do a pre-grading "dry run" as follows: For the first ten
or so papers, I get a blank piece of pink paper and set it
beside the student's paper. I grade the work as best I can,
writing on the pink paper and /not/ the student's paper. Then
I staple the two together and put them on the bottom of the
stack.
By the time I get to the 11th paper, I have (relatively speaking)
more of a clue about what I'm doing, so I just grade the paper in
the usual way. When I get to the bottom of the stack, I grade
those papers in the usual way, and then check to see how well
my end-of-run grading compares to my dry-run grading. I
detach and discard the pink sheets.
I do this whenever I feel unsure of my footing. It is also good
advice to give to a new crop of TAs who are serving as graders
for the first time. It increases the workload by some small
percentage, while increasing the quality of the result by a
much larger percentage.
b) As a related point: Hopefully within the first ten or so you
will find somebody who nailed the problem "just right". You can
photocopy that solution, underline the key points, and use that
as your grading-key. (I can't bring myself to use the word
"rubric" in this context. It's a pointless abuse of the word.)
If there are more than one method of solution, photocopy one of
each. Tape them to the wall over your desk for ease of reference,
if necessary.
This adds practically nothing to the workload. It also gives
you a big head start in preparing the "official solutions"
handout. This handout is IMHO an important part of the learning
process.
Keep the official solutions handouts from year to year, if there
is any chance you (or a colleague) will be reusing the questions.
c) If there are N unrelated questions, grade Q1 of all papers,
then grade Q2 of all papers, et cetera. Short term memory is
better than long-term memory. This adds nothing to the workload.
Note that suggestions (a) through (c) promote consistency, which
is not quite the same as objectivity, but is still a step in the
right direction.
d) Everybody gets a paper handed back. If somebody didn't turn
in the assignment, put his name on a blank piece of paper and
hand it back. This helps reduce the amount of stink about
"lost" assignments. The first time it happens, the culprit is
too surprised to come up with a timely objection, especially
in front of his peers. The second time it happens to the same
guy, he might be better prepared, but he doesn't want to call
attention to the fact that his assignments keep getting "lost"
while nobody else's do.
e) Be scrupulously alert for eccentric solutions. It is quite
possible for a student's solution to bear no resemblance to your
solution while being entirely correct. As an obvious example,
if you solved it in terms of third law action/reaction, somebody
else could solve it in terms of conservation of momentum. The
problem is, there are lots of less-obvious examples.
f) Mistakes will happen. Encourage students to push back on any
dubious grading. One good point is that you learn a tremendous
amount from the push-back. An even larger point is that it is
super-important for them to know you /want/ to be fair. They
will tolerate fallibility but not un-fair-mindedness.