Re: It's the HIP-Hop

[Bob Spruill <BobSp@REVIEW.COM>]

Re: SAT as a predictor
Last spring, the California legislature held hearings concerning the
prospect of eliminating the SAT as an admissions criterion at UC schools;
these were held, as you might imagine, in the wake of the demise of
affirmative action in that state. At the hearings it was indicated that the
correlation between college GPA and an aggregate of SAT score and high
school GPA was about one half. I'm not much of a statistician, but to
explain roughly, a correlation coefficient of 1 would indicate a perfect
predictor, while a correlation coefficient of 0 would indicate as good a
predictor as rolling a die or flipping a coin. Out of this hearing also came
the information, from UC admissions personnel, that the correlation of SAT
alone is higher than that of HS GPA alone, although the two considered
together proved a much better predictor than either alone.

Obviously, a correlation of .5 isn't stellar, but it's not 0. I am not aware
that any predictor with a higher correlation was propsed, much less found. I
suspect not, since they would likely be using it if it had been. Still, even
if a better predictor were found, it would likely be combined with SAT and
HS GPA to produce one with a still higher predictive value. Admissions
committees are enamored of the results these tests provide, and since
they're ETS's customers, ETS is likely to keep its "nonprofit" operations
grossing $100 million+ for many fiscal years to come, since after all their
customers who believe in the exams have the control required to keep
students paying ETS its monopoly-sized fees.

The fact that admissions committees' arguments hinge on college GPA's,
however, is probably the characteristic that's most important to anyone who
wants to examine the cultural underpinnings of the admissions system. You
shouldn't be surprised to know that most of the large courses at UC
Berkeley, for example, are themselves graded on a curve designed to produce
a similar bell-shaped distribution to the one that ETS makes its millions
selling. This is yet another manifestation of the "usefulness" argument
testing companies find it necessary to make. The MCAT's predictive value,
for instance, is often expressed in terms of USMLE (licensing exam) scores,
another assessment that is primarily (surprise!) multiple-choice and
norm-referenced.

These usefulness arguments are more sophisitcated renditions, I think, of
the "rich people are smarter" pseudo-argument described below. The only
possible information such lines of inquiry can yield is to what extent a
particular assessment is in step with the educational assessment system at
large, which it has been quite rightly pointed out on this list is a primary
mechanism of solidifying and protecting class differences in this country.
The perfect circularity of the educational barrier is what I find most
impressive: those who get the "best" educational opportunities are the
children of the affluent; the "best" educational opportunities produce
candidates who can (among other things) score well on the SAT; those who
score well on the SAT are given access to the "best" university
opportunities; the "best" university opportunities put students through
curricula designed to develop the skills of those who can (among other
things) score well on the SAT; those whose skills are so developed are more
likely to become affluent, and thus produce children of the affluent... What
surprises me most is not that the SAT is complicit in the process, but that
the portion of the barrier it constitutes is as porous as it is. After all,
if the exam's goal really were to prevent all social mobility, the
correlation between its results and college GPA (for instance) would be 1,
not .5.

But then again, if its goal really were to prevent all social mobility, it
would be much easier to create righteous indignation toward the exams in
this "land of opportunity" than it currently is.

And as for the success of prep courses, there may not have been many studies
that the testing companies are willing to accept, but I am willing to accept
the evidence of my senses. Good ones work, and those who claim otherwise are
protecting the tests' legitimacy.

Bob Spruill
The Princeton Review of North Texas
email: bobsp@review.com



> ----------
> From:         James Powell
> Reply To:     Assessment Reform Network Mailing List
> Sent:         Monday, October 26, 1998 10:20 PM
> To:   ARN-L@LISTS.CUA.EDU
> Subject:      Re: It's the HIP-Hop
>
> Response to Nancy Patterson -  Jim Powell:
> I would also like to know how well the ACT and SAT predict anything. How
> many false
> positives and negatives?How much advantage do the cram courses on the test
> provide?  The standard error of measurement on the SAT is a bit less that
> 150
> points. That doesn't seem to me to be an instrument that finely
> discriminates.
> The highest correlation with the SAT is not high school GPA or parent
> level of
> education.  This highest correlation's is with parent income. I brought
> this up in
> an article in a local newspaper and an attorney, CPA wrote a letter to the
> editor
> in which he stated that he was not surprised because parents who have
> higher
> incomes are more intelligent.  Wow. I think it would be a good thing if
> students
> would refuse to take the SAT and ACT, but, since the universities and the
> Educational Testing Service are joined at the hip, I don't think that can
> happen
> now.  I will read the two books you suggested.
>
> Jim Powell
> powell@ns.net
>
> Nancy Patterson wrote:
>
> > An amusing little trend is happening in Michigan.  Parents are
> requesting
> > that their students be exempted from taking the high stakes tests that
> > determine whether a student gets a state endorsed diploma.  By law,
> parents
> > have a right to get their kid exempted from the test.
> >
> > And this doesn't seem to hurt the kid at all.  In fact, I recently
> learned
> > that state university admissions officers do not look at state tests.
> They
> > do look at SAT's and ACT's, however.  And they look at grade point
> average.
> >
> > There is some speculation that the Michigan Education Assessment Program
> > will drop by the wayside when our "education" governor John Engler is no
> > longer in office.  As it is right now, school districts around the
> Detroit
> > suburbs, where most of the parent exemptions are coming from, are being
> > asked to offer the tests during an additional testing window because so
> many
> > parents are exempting their children.
> >
> > I'd be interesting in knowing the correlation between SAT and ACT scores
> and
> > college completion.  A number of years ago I read that there did not
> seem to
> > be much correlation between whatever test it is that prospective law
> > students have to take and completion of the law degree.
> >
> > What we need is parents protesting the SAT and ACT and demanding that
> > students submit a portfolio for college admission.
> >
> > Nancy
> >
> > At 12:19 PM 10/26/98 EST, you wrote:
> > >Dear Nathan ,  First off, I have to say how impressed I am with your
> thoughts
> > >on the subject! Wish you were with us in Ohio! We could use your
> energy! The
> > >Ohio Proficiency Test or OPT, written into law with a 10 year phase-in
> period,
> > >if not stopped will bring public education to it's knees. We kicked of
> our
> > >first protest a year ago by not allowing our fourth graders to take the
> test.
> > >In 2002, fourth graders will have to pass the reading portion to be
> promoted
> > >to fifth grade. The 10th grade test in 2005 , replacing the 9th grade
> test, is
> > >the old 12th grade test made tougher, to be passed to graduate. Scores
> are
> > >used to compare districts , take over districts, grade districts, give
> raises
> > >and or bonuses to administrators and teachers , and whatever else the
> > >legislature can dream up to tie the test to. The teachers are afraid to
> speak
> > >out publicly. That leaves the job to parents like Jenny, Teri, and
> myself in
> > >Ohio for the moment. It won't happen overnight. I would love to see
> kids
> > >protest these tests but as a parent, I don't want kids to be penalized
> .
> > >Absolutely the reality of your generation does not have to be the same
> as
> > >ours. You may have found your life's work!              Warmly, Mary
> > >
> >
> >-------------------------------------------------------------------------
> -
> > >To unsubscribe from the ARN-L list, send command SIGNOFF ARN-L
> > >to LISTSERV@LISTS.CUA.EDU.
> > >
> > >
> > Nancy G. Patterson
> > patter@voyager.net
> > <http://www.msu.edu/user/patter90/opening.htm>
> > |  \  /  |
> > |   \/   |__  _   _
> > |       / __/| | | |
> > |  |\/| \___\| |_| |
> > |__|  |_/___/|_____|
> >
> >
> --------------------------------------------------------------------------
> > To unsubscribe from the ARN-L list, send command SIGNOFF ARN-L
> > to LISTSERV@LISTS.CUA.EDU.
>
> --------------------------------------------------------------------------
> To unsubscribe from the ARN-L list, send command SIGNOFF ARN-L
> to LISTSERV@LISTS.CUA.EDU.
>

--------------------------------------------------------------------------
To unsubscribe from the ARN-L list, send command SIGNOFF ARN-L
to LISTSERV@LISTS.CUA.EDU.