Thomas O'Neill raises some good questions about whether in-class exams truly
measure a student's ability and/or the likelihood of success as a scientist.
Others frequently raise this type of question as well. Indeed, when we got
visited by the Chancellor of Education for Ohio a couple years ago he
stressed the need to evaluate students as team members rather than
individuals. After all, working on a team is what they are likely to do
once they get jobs.
However, I have a difficult time buying that for a couple reasons.
(1) Especially in high school and also the first couple years of college, we
are trying to achieve some sort of "core" training. It's difficult to be a
physics team member if you haven't achieved some degree of mastery of
calculus and basic freshman/sophomore physics. If we are going to do
team-work grading in college it ought to be in the senior year, or perhaps
junior/senior years.
(2) It is very difficult to judge an student's understanding when work is
not done in an environment in which we know the student worked alone and for
a definite amount of time. Just think of all the things students do when
they work on physics problem sets:
(a) They leaf back through the chapter trying to find an example close
enough to the problem that they can copy the solution format with new data
from the assigned problem. If necessary they can go to previous chapters to
look up stuff they should have mastered several weeks ago.
(b) They seek help from classmates.
(c) They seek help from the teacher or graduate assistant (when available).
(d) They might take a very long time to do a problem that ought to take 15
minutes.
(e) If they know the answer, they work backwards to try to find a way of
getting the correct answer from the given data.
Now... all of these techniques are good things for the student to learn...
and I assume they are learning these things as they work on problem sets and
lab reports. But I see enough similar problem solutions and read enough
similar lab reports to know I am frequently unsure how much of what I am
seeing actually comes from the student's mind versus how much went from his
eyes and ears directly onto the paper. Sometimes, when grading
problems/labs from two or three students I know worked together, I want to
give one grade then divide the credit by two or three.
I believe my best feeling for what a student can do comes from exams which
have a mixture of problems (numerical solution with work shown, or equation
derivation expected) and essay questions. I feel somewhat in the dark if I
give a multiple choice exam. Of course I have the privilege of small
classes so I can grade exams other than multiple choice, although it takes a
long time and some nights I wish I had given a multiple choice exam.
I guess I just don't know what else to do. Apparently I am not alone. We
still use ACT and SAT score for college admission and scholarships. Medical
schools still use the MCAT and dental and optometry schools have similar
exams. The GRE exam is less popular than it was, but many graduate schools
still require it. The accrediting agencies for colleges are now doing the
same thing the high schools have had to do for years... we must assess our
program and "prove" we are meeting some level of education consistent with
what we advertise, and the best accepted measure of this is standardized
exams in the field of the student's major.
Michael D. Edmiston, Ph.D. Phone/voice-mail: 419-358-3270
Professor of Chemistry & Physics FAX: 419-358-3323
Chairman, Science Department E-Mail edmiston@bluffton.edu
Bluffton College
280 West College Avenue
Bluffton, OH 45817
Thomas O'Neill said, in part:
<SNIP>
So I will ask the question that has always bothered me: Once
a
physicist/chemist/engineer/scientist leaves school, when will he/she/it
ever take another test? Because the issue of extra credit and testing
assumes that testing is a better method of evaluating student mastery of
the material. Tests revolve around solving known problems with limited
resources in a fixed time. Most professionals that I am aware of, do
not approach problem-solving with those skills. Indeed, the first step
to successful professional problem-solving (What has been done on this
problem by others?) is called cheating in a test context.
If a person was actually good at test-taking, how does that
translate into useful professional skills? I certainly believe that the
mastery of the material must be demonstrated by the student, but is a
two-hour exam the best way to do it? Put it another way, does doing
well on the Physics AP or Physics SAT II correlate with success in a
scientific or engineering career? Should colleges (particularly the
sciences) move to some other method of evaluation? Because I figure
that I am emphasizing testing because my students will encounter it at
college.
Note that I do not have answers to these questions and I do use
timed tests in my class.