Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

testing vs. teaching



The discussion of the recent PSAT question
6. METER : DISTANCE ::
(a) ounce : pound
(b) gram : weight
(c) container : liquid
(d) size : height
(e) boundary : periphery

has missed a couple of important points:

1) The ETS is in the testing business, not the teaching business. They
don't much care whether the students learn something, learn nothing, or
even "learn" misconceptions during the test.

2) Under the rules ETS sets for its testing game, the customer is supposed
to choose the *best* answer. The best answer need not be a perfect answer.
From a testing (not teaching) point of view, you can make a strong case
that you measure something important when you ask the customer to choose
the least of evils.

Specifically, on a test where you know there are no "least of evils"
questions, you can quickly plow through the test by crossing out all the
options that contain significant misconceptions. In contrast, if there are
"least of evils" questions, you have to work much more carefully.

===

One can contrive situations where testing is more important than teaching,
perhaps for instance when selecting participants for some grueling yet
critical military mission.

But fortunately, most of us live in an environment where teaching is the
primary goal. In such an environment, testing should be used in the
service of teaching, not vice versa.

It is troublesome how powerfully standardized tests distort the curriculum.
Students ask, if it's not on the test, why study it? For better or worse,
students sometimes remember test questions for a lifetime.

So one way to approach this is to try to enlist ETS in the teaching
business. We could ask them to minimize the use of "least of evils"
questions -- not because they are bad from a testing point of view, but
because they are bad from a teaching point of view.

Another way to approach this is related to Bob's insights about the
relationship of introductory/elementary thermo to
real-world/industrial-strength thermo. When students leave the classroom
and move into the real world, they will continually be faced with "least of
evils" choices. There comes a time in the growing-up process when students
must learn to deal with an imperfect world. Maybe having a few "least of
evils" questions is a good thing.

______________________________________________________________
copyright (C) 1999 John S. Denker jsd@monmouth.com