[From Bill Powers (931012.1115 MDT)]
Thomas Baines (931012) --
In replying to Chuck Tucker, you say
Standards for pass/fail should be a joint matter among those
who must deal with the consequences - teachers, business
people, those responsible for the administration of justice,
etc.
This is the "user-oriented" way of setting standards. The
"student-oriented" approach would be concerned with the education
of the students at least as much as the requirements of the big
wide world into which they will be going. To some extent the two
are related, but if the "user-oriented" approach is given too
much weight, what we get is a regression toward mediocrity and a
continual dumbing-down of the educational system.
The problem with pass-fail is that it's a sliding standard. The
passing mark is always set considerably lower than perfection, so
those who pass will have mastered only a fraction of the
material. When they get around to setting standards in their
turn, they will consider their own achievements to be at the high
end, and will set the break-point lower than the best they could
do (they don't know what they _might_ have learned -- only what
they _did_ learn). So each succeeding generation will learn less
and less and less, and still pass. What we get is what we've got:
a generation of ignoramuses who see little value in eduation.
They might be right.
I think Bill Glasser's idea of "Schools without failure" is the
right approach: never let a student out of a course until it is
perfectly clear that the student has maxed out. And then plan
what comes next on the basis of where that maximum is.
This, of course, places a burden on the teaching system. In the
first place, the course material has to be learnable. That is,
there has to be something to teach that is clearly defineable
enough that both teacher and student can know when it's been
learned. A course on music appreciation or poetry is pretty hard
to set up that way, because there are no clearly right or wrong
matters to be learned. In courses like that, the material to be
learned has to be defined at a higher level: how to approach a
subject, what principles apply. And testing has to be concerned
with that level more than with procedures and facts.
Mathematics and the sciences have it easier: right and wrong
answers are obvious. Even so, the higher levels are at least as
important as the lower; understanding principles is far more
useful than being able to execute procedures, because you can
always look up the procedures once you know what kind of
procedure is needed.
I think that pass-fail is really a declaration of no confidence
in students. It says that this poor dumb inattentive joker with a
ghetto accent and color couldn't possibly aim for perfection;
let's just give this person a low hurdle to jump and get it over
with. This leads to teachers spending the most time with good
students and ignoring bad ones, which is exactly the opposite of
what is needed.
The measure of a good teacher is the improvement in performance
by bad students; the good ones are no challenge. It's always
amused me that the "best" institutions of higher learning pride
themselves on their high standards of admission. The higher you
set those standards for incoming students, the less their
achievements say about your teaching. A real teacher goes for the
tough cases, not the easy ones.
I think we could go for a long time without considering the
requirements of educators, industry, or the legal system. If we
raised the standards and required more of teachers, we could
start turning out people who know _more_ than what is required by
the poorly-educated people who now set the standards.
···
--------------------------------------------------------------
Best,
Bill P.