z39\doc\web\99\17\psat.txt


From: "acronin" 
Date: Mon, 27 Dec 1999 11:21:00 -0800


Mr. Bracey,


You are woefully uninformed about the College Board.  The PSAT has been
normed by regions for years and the reason for this is because mean scores
vary greatly by region.  The norming is used to determine the commended and
semi-finalist for National Merit.  Commended is distinquished by a score of
the top 5% in the region and the semi-finalist are the top 1.5% of the
region.  Thus a raw score that would qualify for the coveted semi in one
region would not qualify in another.  These regional variations are so
distinct that if the highest percent of qualifiers were used nation wide,
there would be few semis in certain areas of the country, the northwest
being one of them.  The northeast would dominate.  


As to the test itself, the PSAT now has an entire new section called
writing which was added to the verbal and math section in the fall of 97. 
The writing section has no match in the SAT.  It was added to equalize the
high math score trend and correct for the previous method of doubling the
verbal score and adding the math score for the National Merit. The test is
so different from its predecessor that comparisons are impossible.


The SAT is always undergoing changes, with the grid-ins being the most
dramatic recent change. Every SAT uses a completely new set of questions
for that test date. Re-norming makes previous comparisons impossible to any
scores before 96.  The test is tested, of course, but anyone knows that the
same test is never given twice.  The being tried out test sections are
easily identifiable to the testees because the  questions are always off
the wall. The fact that fewer students are taking the SAT hardly makes it
less reliable.  It should, in fact, be more reliable.  The College Board
has never billed the SAT as a test that does anything more than predict
with mediocrity the likelihood of performance in the freshman year of
college.  Beyond that, it is useless.



The College Board is a very profitable enterprise from where I am sitting. 
They sell products of a great variety to admissions offices while the
students taking their tests are paying for the sorting and classifying. 
The College Board is the original Big Brother and what it learns about each
of its testees would make your hair curl and all identifiable with your
SS#.  The newest rend is to profile the candidate most likely to
matriculate, as well as tailoring the marketing of the admissions office to
attract the best students and get them to matriculate.  The crunching is as
sophisticated as any corporation anywhere.  Visit their web site and get a
look at what they are up to.  They are the best example I can think of 
that represents the profit motive in driving the testing company decisions.
 The days of the non-biased, public service, testing group called the
College Board are long past.  Just as the episcopacy diminished as the SAT
ascended to its primacy, the meritocracy is in decline with the constant
pressure on the College Board to produce a more marketable and, therefore,
less reliable product, not that it was very reliable to begin with.  


Anne Cronin  


From: Gerald W. Bracey


I'm afraid, Ms. Cronin that your comments about the SAT and NAEP are
totally wrong.


SOME specific questions on both will change from year to year in the SAT
and assessment to assessment for NAEP, but for both common questions from
year to year and tryouts of new questions are carefully controlled.  Each
new form of the test is carefully equated to the previous forms--something
that the SAT-9 does not do.  The ITBS is equated because under Iowa law it
must be.  There are longitudinal data for the ITBS going back to 1955
(actually, farther back than that, but the ITBS was qualitatively changed
so much in that year the directors of the program think it best not to
extend the trendlines past 1955.


The SAT is not quite so good an indicator as NAEP because only 43% of the
seniors nationwide take it an the proportion varies from 4% in Utah to 80%
in CT and MA.  NAEP, though, is given to a representative sample of
students.  If we consider NAEP as the Rolls Royce of testing, then the
SAT-9 is not much more than a used Chevy.


I do not dispute that something devious is happening in the districts you
mention, but the data are contrary to what is normally seen.  Usually, when
a district adopts a new test, or a new form of the same test, scores
decline because students and teachers both are not familiar with the test.


That the kids did better on the SAT-9 than the ITBS is especially suspect
because here in Virginia, when the state changed from ITBS to SAT-9, the
results were virtually identical.  


Your comment that the PSAT is normed each year by region cannot be correct:
ETS and the College Board could not afford to norm a test each year.  Where
does this allegation come from?


You say a gain of 2 points on the SAT "does not meet most accepted
standards of significance."  I concur.  You also sya you'd put your money
on the ACT.  It doesn't matter.  While the verbal SAT rose only 2 points
from 1983 to 1998, the math rose 18 point.  The gains are larger for each
ethnic group considered separately.  From 1982 to 1990, the composite ACT
rose from 20.3 to 21.0.  Given the compressed ACT scale (compared to the
600 point scale of the SAT).


So I ask again, given that all of these indicators are rising, how can you
say, at least at the national level, things are declining?


You say that the number of students taking the SAT is down in your
district.  This strongly suggest that yours is a district undergoing a
demographic shift toward less educated, less affuent families.  This is
also backed up by your comment that the number of kids going on to college
at one school has been halved.  Nationally, a higher and higher proportion
of students are attending college (and they are taking tougher and tougher
high school courses--more math, more science).  If this school is losing
ground it has to be that it is changing demographically.