Date: Aug 31, 2000 10:36 PM
Subject: Re: Ruth Parker in Mountain View
From: Ruth Parker <email@example.com>
Norm-referenced standardized tests, however, guarantee
that there will always be fully half of our students and roughly half of
schools at or below the 50th percentile and thus considered inadequate.
Now that is simply not true. Nothing guarantees that the school averages
will be statistically distinguishable. It might happen that the school
averages read like
50.001, 49.998, 50.02, .......
when the underlying individual scores have large standard deviations. Then
only the averages for very large schools would be statistically
significant. There is no guarantee of that half of the schools will be
considered inadequate. If it turns out that some school averages are 4
standard deviations apart, well then they are.
Now let's turn to the question of half of the students being considered
inadequate. When I was at MIT, I tried to take music appreciation. I
spent countless hours in the library listening to Wachet Auf (Johann
Sebastian Bach, Cantata No. 140). I never got it. I the choice of an F or
dropping the course. I dropped. I never have felt that I was inadequate
because of that. Just different. In Jr. HS, I couldn't throw the softball
OR run the mile OR high jump OR chin up OR..... for a grade better than D.
I did feel a bit inadequate over that. I SHOULD have been content with
'different'. Feelings of inadequacy have more to do with social attitudes
than the reality of differential abilities. The differential abilities
will still exist even if you don't test for them. In many schools, the
differential ability gap will grow large simply because people will deny
that they exist if the data don't hit them up side of the head.
High-stakes norm-referenced testing is detrimental to education and
as teachers experience increased pressures to just teach to the test. The
multi-billion dollar textbook/testing industry is driving educational
in this country to the detriment of mathematics education.
The teachers should teach beyond the test rather than to it. Tests and
standards are minimal criteria not end goals.