John D wrote in part:
"Here's a possibly-constructive suggestion for an alternative:
-- Assign homework.
-- Collect the homework. Correct the homework and pass it back. But
don't grade the homework. There is a world of difference between
correcting a paper and grading a paper.
-- At frequent intervals (maybe even every class, especially early in
the year) have an in-class quiz consisting of one or two problems verrry
similar to homework problems. Grade the quizzes.
The point is that anyone who _understood_ the homework will do well on
the quiz. (Sad to say, you will have to _explain_ the relationship
between homework and quizzes; otherwise there will be some who guessed
it would be OK to blow off the homework, and won't easily figure out
where the quizzes are coming from.)"
I have a colleague that takes this approach, and it seems to work
reasonably well for him in intro classes. The major downsides, and all
fixes have downsides, is the class time it takes away to administer the
quizzes, we have 41 lectures in our 3credit lecture 1 credit lab intro
courses a semester, not a lot of time to cover the syllabus. The other
downside is that the quizzes have to be very very much like the HW, and
that can serve encourage rote memorization. Naturally, if these
down-sides aren't worse than the problem than the approach is worth
trying I suppose. As I said, my colleague currently likes the approach.
I think it would work less well in an advanced class, as it would be
hard to dream up problems for the quiz that take only 5-10 minutes of
class time.
McDermott wrote in part:
"Is there some rule that requires assigning and grading homework? Seems
to me that the tests would indicate how well a student has learned. As
to having the test solutions,"
My 1st year chem course in college was like this, HW was assigned but
not graded or even marked. Answers were posted to all assigned
problems. We knew that it was a good idea to try to work those
problems, *in order to learn the material*. I stress the last bit,
because it was *not* the case that the test problems were the same or
nearly the same as the HW.
I agree with McDermott statement about the usefullness of fuzziness in
student knowledge regarding their grade. Probably the physics courses
in college where I learned the least was one of these self-paced then
take a test over the unit style courses, the grade being determined by
how many units you finished. You always new precisely what you needed
to "memorize" and "regurgitate" on the tests and where you stood for
your grade. I wanted an "A", I'm also lazy, I earned my 'A' about 1/3
of the way through the quarter, and then naturally stopped. I have
always gotten more out of classes where I didn't precisely know what was
needed to achieve the highest mark. (as an aside, I suspect in a highly
grade inflated environment, this would have been much less motivating
for me.)
And lastly, without quoting, I mostly agree with R. Tarara's most recent
post.