\doc\web\97\08\mintest2.txt
from Arthur Hu
to Opinion at Seattle Times


re:\clip\97\23\mintest.txt
http://www.seattletimes.com/extra/browse/html97/scor_101097.html
Seattle Times Company Friday, Oct. 10, 1997 Minority test scores
`devastating' by Jolayne Houtz Seattle Times staff reporter


I've got an email list of people who are opposed to the ideas
underlying this new test, and I'd like to volunteer to be a spokesman
for these people since you don't seem to have any other contacts with
any organized opposition to this juggernaut of reform


------------------------------


I saw the same low minority test results with the controversial CLAS
test in California.  But to your credit, at least this state and the
press have chosen to highlight low minority performance, rather than
sweep it under the rug. California largely ignored the minority
disparity issue, which proved that supposedly minority-friendly
features such as no multiple choice questions, multicultural content,
no single right answer, full credit for incorrect answers with good
explanations, and no credit for correct answers with no explanations
did not make up for an underlying poor grasp of the actual skills and
facts.


In general, the spread between the races increases with the level of
difficulty of the standard. Considering that the math pass level was
set at the 80th percentile, and that much of the content of the test
was 2 years beyond what is typically expected of 4th grade level
math, it should not be surpising to find that proportionally only 1/4
as many blacks met the standards as the mostly white state average
when Blacks generally lag grade level by 2 or more years. If multiple
choice tests are bad for minorities, written response questions are
even more difficult for poorly prepared minority students. 


That's a problem with the reforms in general which seem bent on
taking the performance of the top 10-20% of students, and then making
the bold assumption that all students can and will perform on that
level simply by setting expectations that high. In reality, such an
approach simply leaves the poorly prepared up the creek without a
paddle when such students haven't even mastered even the most basic
facts and skills expected under the old standards.



As for the common statements that middle class suburban minorites
score just fine, my study "Education and Race" from the Sept 15
National Review shows with figures from the California's CLAS test,
SAT's and the CTBS test broken down by race show that minorities
perform more poorly than Whites in even the most affluent and highest
scoring districts. This is true in San Francisco, Seattle, or
Washington DC, while Asians generally perform as well as Whites in
much more affluent districts. If you to do a similar study of the new
assessment, you should not be surpised to find blacks and Hispanics
score no better than state average even in the best suburbs, and
worse than urban Seattle in highly rated districts like Bellevue.  It
is a combination of class and race that really explains test score
outcomes.


We don't need a new series of standards, all we have to do is bring
up the number of students who can meet the normal standards we
already had.  The only groups of people outscoring average Americans
are the Asians and the Asian Americans, who leave other Americans in
the dust with the old fashioned tests with good old fashioned rote
drill and memorization, not these nutty new reforms.  Just bring the
bottom 25% up to average first. That's much more realistic than
assuming that every child be an honors student just by relabeling it
to be the "essential skill" level. You don't have to have
revolutionary reforms like Mao used to inflict upon China every 2 or
3 years. You just have to do a better job with the basics at hand. Is
that really so hard?


Arthur Hu