Re: grades, assessments, etc.

["John S. Denker" <jsd@MONMOUTH.COM>]

"RAUBER, JOEL" wrote:
> The local supporters of "assesment" point out that you no longer can trust
> grades to provide such an assesment, hence the need for other assesment
> instruments.  I interpret this to mean that because of grade inflation we
> need other "grade-like" assesment instruments.  Thus adding a whole new
> layer of paper work that we as faculty and administrators have to produce.

Having served on many committees that look at transcripts,
I can tell you that grades are very hard to interpret.
Rampant and non-uniform grade inflation makes this worse,
but that's by no means the only problem:  There are so
many uncontrolled dependencies on school, teacher, class
syllabus, etc. that I despair of ever seeing a meaningful
GPA, with or without inflation.  Bottom line:  don't let
somebody use the "grades aren't meaningful anymore" argument
to bamboozle you into doing something that you wouldn't
otherwise be doing.  I doubt that grades could _ever_ have
been trusted for the purposes some people seem to want them
to serve.

Assessments e.g. standardized tests are not 100% good, but
they're not 100% bad, either.

Obvious remarks:

When there's a test, teaching to the test is more-or-less
inevitable.  If it's a bad test, this is a disaster.  If it's
a good test, this is not a big problem and actually has some
profound advantages.

 From the teacher's point of view, having an externally-imposed
test is good for motivation:  You can say to the students:
"We're working together to conquer this test."  Contrast
this with the case where the teacher designs the test and
assigns the grades:  This creates something of an adversarial
relationship in the classroom (which inteferes with learning)
and creates a conflict of interest (which leads to grade
inflation).

(Indeed my philosophy is that we are working together so
that the students will have a lifetime of happiness and
productivity, and the grades/test/whatever are just a short-
term proxy for that long-term objective ... but students
aren't necessarily known for their long-term thinking, so
having a good proxy is useful.)

Note that in British universities there is a heavy emphasis
on final exams (which are set by a university board, not by
the individual course professors).  So there's centuries of
data available for those who are interested.

Weaknesses of the tripos-based approach include
 -- promoting slacking during the year followed by cramming
    for the exams.
 -- not immune to grade inflation.
 -- risk of too much power concentrated in board of examiners;
    need for "quality feedback process".

No-win situations include:
 -- no way to grade elective special-topics courses.  I have
    no idea what a grade is _supposed_ to mean in a course
    like this.  Special topics in basket-weaving?  Special
    topics in quantum cosmology?