Here are two more ideas for lab grading, both of which I like.
1) The department here at Valparaiso grades extensively based on an
abstract. Lab notebooks (or this year, carbon copies) are turned in
with the original data, sketches, etc. Also, a few pertainent graphs
are asked for. But 50% - 75% of the grade is for the abstract,
typically 100-200 words long, typed.
The abstract accomplishes several things. For the grader, all the
key information is collected (or at least should have been) in one
place. It must be typed, so it is legible.
For the student, they are forced to express what they were actually
trying to do, what details are critical, and how good their results
were.
Spot checks of the original data are made, but their summary is the
main criterion.
2) At the end have the students answers one extra question (incluing
an explanation):
What simple change in the experiment
would most improve the results?
The idea is to get the most "bang for the buck". Perhaps a critical
measurement could be done with calipers instead of a ruler - but using
an interferometer would most likely not offer any appreciable
improvement. Perhaps a measurement could just be repeated to get
better statistics. Perhaps....
The idea is to force them to think about designing experiments,
propagation of errors and available resources.
--- Tim
----------------------
Timothy J. Folkerts email: tim.folkerts@valpo.edu
Department of Physics phone: 219-465-2134
Valparaiso University
Valparaiso, IN 46383