[net.cse] Grades

mckeeman@wivax.UUCP (01/03/84)

I recommend a separation of concerns.  Grades attempt to
linearize a multidimensional, nearly unmeasurable quantity.
UC Santa Cruz did one thing right: it allowed the
instructor all of the Power of English to give the grade,
which permitted such things as "clearly understood the
important concepts... "was unable to demonstrate knowledge
in an exam situation...  "an excellent lab technician, and
colleague...  and whatever else the instructor was able to
write.  

Effect #1.  Students do not come and argue about their grades.
Effect #2.  Students are willing to help each other.
Effect #3.  Reduced anxiety in giving grades (no invidious choices).
Effect #4.  (Somewhat) more work for the instructor.
Effect #5.  Some graduate schools will not admit gradeless students.

And so on.  The avoidance of a numerical or letter grade
elevates the discussion of grading to levels where
motivation, equity, ethics, and style show through.

There is, incidentally, an enormous body of
discussion/knowledge on this topic resident in the UCSC
campus, in memos, minutes and memories.

/s/ Bill McKeeman, Wang Institute

budd@arizona.UUCP (tim budd) (01/03/84)

        It is difficult to convince students that they are not in competition
with one another.  I always tell my classes that if they all deserve A's (ie,
they all learn what I think they should learn) they will all get A's.  They
are not in competition with each other.  Nevertheless, I have had troubles.
One semester I announced at the beginning of the class that grades would be
based on an average of three prelims and a final.  When time came to hand out
grades, I agonized over several borderline cases - including some individuals
who had done generally well on all exams but one.  So I recalculated grades
based on dropping the lowest prelim, and found a few (very few) grades shifted.
So then I decided the fairest thing would be to give each student the HIGHEST
of the two grades.

        Well, it then turned out that some students got B's with lower overall
averages than students who got C's.  One of the students stuck in the middle
(with a C) heard of this and came in to complain.  His argument was that I
was not being true to my original word.  My argument was that HE got the grade
he was going to get anyway, he was not in competition with this other student,
and overall (with the exception of one exam) I thought the other student was
better.

        Based on this experience, my policy now is to NOT announce how grades
will be calculated - I want to reserve my right to weight different exams
differently or take extenuating circumstances into account.  Nevertheless,
at any time any student can come to me and ask what their current overall
average is (ie, if I were handing out grades at that time what they would
get).  This gives students the information they want without tying my
hands too tightly.  I'm not sure I like this, but grades seem to be a large
evil anyway and this seems to be the lesser of many many evils.

mwolf@yale-com.UUCP (Anne G. Wolf) (01/03/84)

Grades should be a means, not an end.
-------------------------------------

1. The Purpose of a Course

    I think that a course should have only one goal, to  arrange  things
so that the students learn the material taught in the course.  For those
who believe that the particular material in certain courses will  become
out-of-date  in  the  near  future, I could change the purpose to giving
students practice in acquiring knowledge  of  a  certain  kind  (on  the
assumption  that  the experience of learning one thing now will assist a
student in learning something else in the  future).   However,  I  don't
think that this distinction is very important.

    It is also not important whether the effort comes  mostly  from  the
students  or  mostly  from  the  person teaching the course.  It is just
important that the information be transferred one way or another.

2. The Purpose of Grades

    Grades  help  learning  in  two  ways.   The grades given during the
course give the students a way to tell the extent  to  which  they  have
been  successful  in  learning  the  material.   The  final  grade is an
incentive for  students  to  do  whatever  is  necessary  to  learn  the
material, even when it is not fun.

    I think it is a great shame if people constrain what they  teach  or
what  they  ask  students  to  do because of grades.  Grades should be a
means not an end.  Grades should not impede  learning.   When  they  do,
they are being used in the wrong way.

3. How I Wish People Would Teach

3a. Making Students' Obligations Clear

    Although  I am an undergraduate, and I have yet to teach a course, I
strongly agree with people who say that a professor or instructor should
make  clear  at the beginning what the students' obligations are and how
success in meeting those  obligations  will  be  judged.   Let  me  also
suggest  that  any assignment should always be attempted by a professor,
instructor,  or  TA,  before  being  given  to  the  students,  so  that
impossible  or  unexpectedly  difficult assignments are avoided, and the
assignment does not have to be  changed  shortly  before  the  due-date,
because of questions raised by the students who are trying to finish it.
(This  last  problem  is  the  reason  why  most  students  don't  start
assignments until the last minute.)

3b. How Much Work Should Students Do?

    Another thing that I wish people  teaching  courses  would  consider
more  often  when decided what students are obligated to do is that most
students take 4 or 5 courses at once.  Students must divide  their  time
between  courses.   Thus, if each of five courses has an assignment each
week which takes as little as eight hours to complete (and that is not a
large  assignment  by  the  standards of most courses I have taken) this
means that students already have the equivalent of a  full-time  job  in
addition  to  the  lectures,  labs, and whatever part-time work they may
have to do to pay for their  education.   When  considering  whether  an
assignment  is  too  much  to  finish, find out how long it will take by
attempting it yourself (or getting someone else  to  attempt  it),  then
allow  for  the lack of expertise on the part of students, then multiply
by the average number of courses which students take.  It is unrealistic
to  assume  that  students  will fail all of their other courses for the
sake of yours, and yet, many people teach as though  their  course  were
the only one.

    Think about it.

        Mary-Anne Wolf (decvax!yale-comix!mwolf)

spuhler@hplabs.UUCP (Tom Spuhler) (01/04/84)

Lance Norskog was correct in that UCSC( University of California, Santa
Cruz ) had no "grades", but that is not to say that there wasn't an
evaluation system.  UCSC has NES ( Narrative Evaluation System ) which
means that, as Bill McKeeman described, that  an English language
descriptive evaluation was given.  These tended to be from a few lines
to a half page of text that 1) described the course content, 2) the
student's performance in that course.  I found these much more useful
then a simple letter grade. ( I had both in college ).

There were the effects that Dr. McKeeman did describe however, a major
problem was that there was no Grade Point Average for graduate schools
and prospective employers to plug into their formulas, and reading all
those evaluations was too much work.  Another problem was the sometimes
uneven quality of the evaluations; they went from a 1 line "average, good
excellent" type, to a half page of prose, describing in detail ones
performance.  Dr. McKeeman, as I recall, wrote reasonable evaluations.



-- 

__
                         Tom Spuhler
                         UUCP:  hplabs!spuhler
                         CSNET:  spuhler@hp-labs
                         HPMAIL:  tom spuhler/hp1900/01

myunive@nsc.UUCP (Jay Zelitzky) (01/05/84)

........
	Like Lance I attended UC Santa Cruz.
I want to clear up one point Lance made.  UCSC does indeed not
have grades.  They do have written evaluations of your performance
in a class.  You get an evaluation that is anywhere from a paragraph
to several pages long of what you did in that class.

	Computer Science classes there are otherwise fairly similar to
those elsewhere complete with exams and projects.  There is a greater
tendency of professors there to assign group projects.  There is also
a very strong group spirit there.  Students in all classes tend to
work together since there is very little competition.

	Students from Santa Cruz rarely have trouble finding jobs as
most employers find evaluations much more valuable than grades in finding
out the suitability of people to work there.

	For those who wonder how UC Santa Cruz can continue to
exist in this society without grades.  It has never given grades since
it first opened in the early 60's.  It is a school of 7000 students in the
University of California system.   It was a haven for me and remains a haven
for people like me who value learning for its own sake.
				Jay Zelitzky
				hplabs!menlo70!nsc!myunive

bcn@mit-eddie.UUCP (Clifford Neuman) (01/07/84)

At MIT, a student's Freshman year is pass/fail.  This is done in an
attempt to aid students in transitioning to an environment that is more
difficult than most of them are used to.  I think it also tends to
reduce competition, even in future years when the students are on
grades.  In place of grades, the students get evaluations in the middle,
and at the end of each term.  The evaluation has two parts: the
student's own evaluation of his performance, and the instructor's.
Students are expected to evaluate their own performance, and submit the
evaluation to their instructor, who then makes an evaluation, usually
emphasizing the areas that the student brought up.  It is also possible
for a student to find out how he would have done in a course if he did
receive grades.  I made it a point to do so in order to have some idea
of how I would do in future years.  (I wish they weren't pass/fail, as
that was my best year).

In one course I took (not pass/fail anymore) the grade was based on two
take home exams, and a paper.  When the first exam was handed back, the
professor told the students that if they felt any grade was not
indicative of there understanding of the material that they should speak
to him, and that emphasis in grading could be shifted to the other
assignments.

In another course, exams presented a problem AND a step by step method
to solve the problem.  This made it possible to give problems that were
much harder (well, they would have been if the method were not
included).  Additionally, the exams were open book which took emphasis
off memorization.  I found the exams to be very good and one would
actually learn a fair amount of material just by taking the exam.  For
an exam like this, cramming the night before the test did not do much
good, and it was necessary to have kept up with the material as it was
presented.

Clifford Neuman
BCN@MIT-MC.ARPA
{decvax!genrad,ihnp4}!mit-eddie!bcn.UUCP