<p>This is an example how using partial data can be misleading:</p>
<p>Golden writes, "In 1990, federal investigators concluded that UCLA's graduate department in mathematics had discriminated against Asian applicants."</p>
<p>Well, that was true in ... 1990. But was it true upon further examination?</p>
<p>
[quote]
University of California at Los Angeles Graduate Schools </p>
<p>In January 1988, OCR regional staff began a compliance review of
admissions practices of all 84 departments with graduate programs at the
UCLA. UCLA was targeted because preliminary information indicated that
although UCLA had a large number of Asian-American applicants, the
overall admission rate for Asian-Americans was lower than the overall rate
for whites in many programs and because the Department of Justice had
received a number of inquiries concerning the University of California
system.</p>
<p>In its LOF of October 1, 1990, OCR found UCLA in violation of title VI of the
Civil Rights Act of 1964 because of its admissions practices for the
graduate Mathematics Department. In particular, OCR found that the
department had discriminated against five Asian-American applicants
who, if provided equal treatment under admissions standards articulated
by the department, should have been accepted. OCR deemed UCLA’s three
different explanations of admissions decisions given over more than a year
to be pretext for discrimination.</p>
<p>UCLA disagreed with OCR’s findings. UCLA asserted that OCR (l)
misunderstood the department’s initial evaluation rating system, which
was just a recommendation to the vice-chair, and (2) failed to interview
the vice-chair who actually made the admissions decisions but was on
sabbatical when OCR first visited the Mathematics Department in 1989 and
1990. UCLA expanded the statistical analysis and produced statistics
showing no difference in admission rates for whites and Asian-Americans
for numerical applications when they were grouped with ratings of “3.0
and above” and “below 3.0.” OCR had limited its comparison to a group of
whites who had been admitted and a group of Asian-Americans who had
been denied admission. In UCLA’s expanded group comparison, UCLA
showed that there were 22 white applicants in the same rating range (that
is, ratings of 2.4 and above) as the three OCR-identified Asian-Americans
who were denied admission based on the use of the same criteria. UCLA
maintained that three admitted whites in that group had substantially
higher academic qualifications than the three rejected Asian-Americans
OCR identified.</p>
<p>*The supplemental investigation showed that OCR had not fully understood
the criteria it was given by UCLA officials in September 1990. *
The regional office found at the outset that it had received the wrong
information from university and Mathematics Department officials. In
reexamining files and expanding the examination to files of lower ranked
Asian-Americans admitted, *OCR found that lower ranked Asian-Americans
also benefited from the application of subjective admissions criteria. *</p>
<p>Further review showed only two possible examples of discrimination. Both of these involved students within the range of white applicants
admitted and white applicants rejected. Both cases of possible
discrimination were vulnerable to being rejected, one because the
applicant had a lower quantitative Graduate Record Examination (GRE)
score by a substantial degree than anyone admitted and the other because
the applicant had a combination of low GRE scores, a degree from an
unknown school, and a stated interest in obtaining a certified public
accountant license, a career goal outside mathematics.</p>
<p>The regional office submitted a revised investigative report to
headquarters on July 23, 1991, in which it concluded that UCLA’s
Mathematics Department was not in violation of title VI and recommended
the withdrawal of the violation LOF. On December 26, 1991, the Deputy
Assistant Secretary for Policy concurred and suggested revisions of the
draft investigative report to the regional office. The next 20 months were
spent by the regional office and headquarters exchanging drafts of the
revised LOF.</p>
<p>*On August 8, 1993, OCR issued a revised LOF concerning the Mathematics
Department. It stated that because of new evidence, OCR had revised its
original findings and no violation had been found to have occurred. *
[/quote]
</p>