Re: what's a grade, anyway?

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

At 09:41 AM 1/5/00 -0600, Jack Uretsky wrote in parg:
> The grades that you give your students should ultimately represent
> you warranty as to their ability to "perform physics".

Wow.  What a provocative statement.

First of all, note that the statement says "...should ultimately..." and we
agree that there is a big distinction between what a grade should ideally
represent and what real-world grades actually represent.

At one level, the sentiment Jack expressed is tempting, but the more I
think about it the more reservations I have.

At 12:18 PM 1/5/00 -0600, Jack Uretsky wrote:
>        Can I reliably assign a grade that predicts
>        a student's future performance in an engineering environment
>        or in an advanced academic setting?  Yes.  People have been
>        doing it for generations.

If you can do it reliably, you are a tremendous exception.  Most other
people have been doing it unreliably for generations.

IMHO the question as to what a grade ideally should represent is a horribly
ill-conditioned and ill-posed problem, overdetermined in some ways and
underdetermined in other ways.  For starters, it might represent
 a) how much physics the student learned in this course
 b) how much physics the student knew at the end of the course
 c) attendance, spelling, penmanship, et cetera
 d) an estimate of how well the student would do in a real-world job
 e) a way of motivating the students to study harder

As someone who makes hiring decisions, I implore you *not* to go overboard
in the direction of option (d).  It is very unlikely that you know what I'm
looking for.  (Maybe jobs here are not typical "real world" jobs, but they
are pretty nice jobs.)

At 01:01 PM 1/5/00 -0500, Richard W. Tarara wrote in part:
> It doesn't matter how much one knows, or how skilled one is, if one doesn't
> demonstrate the knowledge and/or skill in some 'useful' way.

I agree 100% --- although depending on context that might be misleading.
Some of the best students recognize at an early stage that grades _per se_
are less important than understanding the material, and many of the things
necessary to get really good grades are not 'useful' in the large sense.

> The student above (and
> those 'bright' students from another note) who don't have the discipline to
> do the assignments are poor risks for employers (whether in research,
> business, or academia).  Yes, some will rise above the immaturity that leads
> to 'blowing off' assignments, but perhaps the very impetus to do so will be
> the lower grade received in the course.  In other words--you NEED to
> penalize the person who does not turn in a report.

Well, I am reconciled to the fact that students who don't turn in homework
will be penalized.  I recognize that there are certain real-world jobs for
which such behavior would be disqualifying.

OTOH there are many different kinds of jobs in the world, so please don't
go too far in the direction Jack suggested, i.e. don't let yourself think
that GPA (let alone the grade in a single course) is a serious predictor of
success in a particular career.

> Grades are only one indication of ability.

True!

> That's why schools and employers
> usually want two or three evaluations.  Grades, standardized tests, letters
> of recommendation are common.

Yes!  And there may also be
 -- an application essay and/or
 -- a job interview, and/or
 -- (for higher-level candidates) a publication record...
... for a total of at least SIX data points.

Of these, grades are the *least* significant -- partly because of rampant
grade inflation, and partly because of the differences from course to
course and from school to school.

> The first often can indicate as much about
> work-ethic as raw ability, the second about retained knowledge/skills and
> sometimes raw intelligence, and the third can provide insight into the
> individual and can often explain away lower levels in the first two.  The
> recommendations are usually the most important piece of information for the
> evaluator and therefore it was EXTREMELY alarming to me to hear some of the
> stories (from a thread a few months ago) about how the letter of
> recommendation process has been subverted in some schools.

On numerous occasions I've ginven fellowships and/or job offers to people
who had lousy grades.

Think about it from the employer's point of view:
 *) One candidate got his parents to pay the expenses, took easy classes,
got straight As and mediocre test scores.
 *) Another candidate worked his way through school, took graduate-level
courses staring his junior year, earned Bs and Cs in all courses except for
As in the graduate courses, published a couple of papers, and got a 990 on
the GRE.

Whom do you think I'm going to hire?

==============

I say again, IMHO the whole concept of class grades raises unanswerable
questions.

Consider the following scenario:  It's the first day of class.  Student #1
already knows more about the subject than student #2 will ever know.  What
do you do?

I'm not making this up;  I've been part of that scenario -- sometimes as
student #1, sometimes as student #2, and sometimes as the teacher.  It's a
tough situation.

---

Grading the homework is particularly troubling.  Consider the following
three students:
  -- Student X attempts the homework, gets it wrong, and learns something
in the process.
  -- Student Y doesn't attempt the homework out of laziness and winds up
not understanding the material.
  -- Student Z takes one look at the homework, decides it is obvious and
dead boring, and decides not to work through the details because his time
would be better spent practicing for his upcoming piano recital.

If you grade the homework, there is no practical way to grade student Y
differently from student Z.

==========================================================

I think there is a big difference between grading and teaching, and having
the teacher do the grading tends to drive a wedge between teacher and
student.  There is an alternative:  At some well-known places they don't
give significant class grades, and rely instead on exams that are set by
third parties.  That tends to push the students and teachers together;
"let's work together to show those examiners a thing or two".