Professor Matthew Malkan is a former chair of the Committee on
Undergraduate Admissions and Relations with Schools. Â
By Professor Matthew Malkan
University of California President Richard Atkinson has proposed
that UC drop the SAT-I, which measures reading comprehension and
mathematical and verbal reasoning, and replace it with a
hypothetical new test which would be more closely based on
high-school courses (“UC Regents propose new test,”
News, March). This would occur in the same year that he has brought
about another enormous change in the role of academics in
admissions: UC has just eliminated its (very large) category of
highly ranked students who were admitted purely on the strength of
their academics, substituting, instead, a system of
“holistic” review where non-academic and subjective
factors play a greatly increased role.
At UCLA, for example, applicants not in the top 1 percent or 2
percent of high school graduates now stand little chance of
acceptance unless they can report substantial life hardships. For
most UC applicants, it now appears that disadvantage (in family and
school) has become more important than academics.
Many believe that this vaguely defined holistic review may
result in a gradual erosion of academic standards. The SAT-I is the
only standardized test taken by most college-bound students for the
last 30 years. It is the only practical way to compare the academic
preparation of high school students across the country. Its
elimination will make it extremely difficult to measure any decline
in standards.
The arguments for eliminating the SAT are weak and confused. One
popular myth asserts that the SAT fails to predict how well
students will do when they get to college. All of the data,
including that obtained by UC itself, prove the opposite. The SAT-I
has, in fact, gradually caught up with the other successful
predictor of college performance, namely high school grades,
because of the rampant inflation in the latter.
Many applicants to the more competitive UC campuses have
straight-A high school grades, which no longer guarantee their
admission. Thus, since 1999 the SAT-I has been better correlated
with success at UC than high school grades. Furthermore, several
well-understood statistical fallacies caused the President’s
Office report on the SAT-I to underestimate its predictive power.
Professor Berk demonstrated this at a UCLA forum in April
(“UCOP methods in SAT study flawed, says UCLA
professor,” News, April 4). In any case, the committee that
called for the end of the SAT-I explicitly admitted that the exam
was not being rejected for any statistical inadequacy as a
predictor of college grades.
Instead, the SAT is accused of more subjective
“problems.” It is said to “send the wrong
message” to high school students ““ that, rather than
working hard in their courses, they should pay for an expensive
private SAT cram course. All controlled studies of SAT prep courses
show that they offer little more gain (30 to 40 points out of 1600
total) beyond the improvement that occurs simply by taking the test
a second time. Students are already getting a loud and clear
message that they must study hard and get top grades in their high
school courses, which count more heavily in admissions than their
SAT-I score anyway.
Some SAT critics feel that standardized tests stand in the way
of increasing under-represented minority enrollments at competitive
colleges. Their feelings are contradicted by all the UC and
national studies (including the one by the President’s
Office), which demonstrate that no reshuffling of academic
measurements (high school grades, SATs, or other exam scores) would
result in any minority gain in admissions.
The College Board is now making substantial improvements to the
SAT-I. They are adding a writing sample and more advanced math
questions and eliminating the verbal analogies disliked by
Atkinson. It is therefore unwise for UC to commit itself this July
to replacing the SAT-I with a hypothetical test which some people
hope may, later in the decade, more closely follow the contents of
particular high school courses.
Because this wished-for “achievement” exam is
required, by definition, to be very different from the SAT-I, it
will not be acceptable as an SAT-1 substitute at any of the other
colleges in the country. Many have shown no interest in dropping
the SAT-I. In effect, Atkinson proposes a segregation of UC from
the rest of American higher education.
The UCLA faculty has been invited to hear administration
advocates make their case in a meeting this Tuesday, May 14. The
faculty should recommend that UC support the work of the College
Board to preserve the SAT-I.