What should tests/exams test?

"William S. Verplanck" <[email protected]> Thu, 11 Apr 2002 17:31:48 -0400
Newsgroups gmane.psychology.behavior-analysis
Message-ID <[email protected]>
A while back, a chemistry prof bragged of a course he'd given.  It was so
good, he assured us, that nobody in the class got better than a "D."

We 'think' that we're evaluating how much each student has learned by
giving exams.

We are forgetting that the (mean or median) grades of a group of students
in a class, or in a school, _also_ enable us to evaluate (1) the curriculum
to which they have been exposed which indicates what each student "should
have" learned, (2) the instructor who does the teaching, and (3) the
specific methodologies of teaching.

It is damned easy to fail the student when it may be, in fact, the course,
the teacher, the educational system, or all three which have failed.

It is the money-and-power-hungry people who put together such tests as the
ACT and GRE who have systematically misevaluated _both_ individual students
and the several systems in which they've been taught!

No more "teaching to the test."  Now it must be "test to what's been
taught" (or  "_presumably_ taught").

Bill V.
_____________________________________________
    William S. Verplanck          [email protected]
  An award-winning site: http://web.utk.edu/~wverplan/
    http://web.utk.edu/~wverplan/kantor/kantor.html
_____________________________________________