Gus Gassmann <[EMAIL PROTECTED]> wrote in sci.stat.edu:
>exams are [...] not realistic
>decision making, and a certain amount of error is essentially forced on the
>students by creating this unrealistic environment. It is the professor's
>responsibility to recognize this, and to recognize the severity of each error
>in giving appropriate feedback and partial credit to the students.
Agreed. I think there is _something_ to be said for the validity of
testing under pressure, but while deadlines are a reality in the
world, deadlines shorter than overnight are extremely rare.
Only a minority of students is active in a particular class; in some
classes it's a tiny minority or even zero in a large lecture hall.
The only time the instructor can see whether students actually
understand the material is on quizzes and exams. This is an
opportunity, _the_ opportunity, to root out a student's
misunderstandings. I probably spend too much time(*) writing
diagnosis-type information on students' test papers ("You seem to
think A, but actually B is true. See page N in textbook."), but the
fact that they know they may have partial credit gives them
incentive at least to flick their eyes across the comments.
Keith and Cimperman have some good things to say in points 2 through
5 of their article "The Hidden Script". It's posted on my Web page
at
http://www.acad.sunytccc.edu/instruct/sbrown/math/faq.htm
(with their permission, of course).
(*) Every time I decide I'm wasting my time doing this and resolve
just to mark number of points earned on each problem, some student
comes to me with a question or comment about my written comments. I
know it's a minority that are reading them, but it's not always the
students I predict.
--
Stan Brown, Oak Road Systems, Cortland County, New York, USA
http://OakRoadSystems.com/
Support our troops: bring them home!
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
