"When I use a word," Humpty Dumpty said, in rather a scornful tone, "it means just what I choose it to mean--neither more nor less."
No doubt Mr. Dumpty would have liked the study on affirmative action in last week's Journal of the American Medical Association. Its authors feel the same way about their research results: They get to choose their meaning.
University of California, Davis medical school Profs. Robert C. Davidson and Ernest L. Lewis have been touting their study of 20 years of racial preferences at their institution as proof that such admissions policies are harmless--that they do not affect the quality of the doctors who graduate. There is just one small problem: Their study's data support the opposite conclusion.
That clearly wasn't the authors' intent. Their examination was obviously designed to provide ammunition for pro-preference forces. For example, the authors declined to compare the obvious groups--beneficiaries of racial preferences and nonbeneficiaries. Instead, they lumped beneficiaries of racial preferences into the broader category of "special admissions"--which makes up 20% of admissions at Davis medical school. For example, an MIT physics graduate with perfect MCAT scores--but, due to the extraordinary competition at MIT, a less-than-stellar grade point average--would be a special admission. Since only 42.7% of special admissions were "underrepresented minorities," this strategy seriously diluted the data.
Even so, this study turned around and bit its authors. Once in medical school, the special-admissions students performed strikingly less well than did the control group. In both academic and clinical courses, their average grades were lower--much lower. Regular admitees also qualified to join the medical honors society almost three times more often than did the specials. And, although it's not easy to flunk out of medical school or even to fail a course, the specials were three times more likely not to graduate.
Most distressing were the two groups' National Board of Medical Examiners tests. The Part I exam, measuring minimum competence in the core of medical science, includes basic anatomy, pharmacology and pathology; no one who fails it may call himself a doctor. Davis's special admitees were eight times more likely to fail this test than were regular admitees. Results for the Part II exam, which measures clinical knowledge, were similar.
This is consistent with the results of a 1994 JAMA study, which reported that an astonishing 51.1% of black medical students failed the Part I exam, vs. only 12.3% of white medical students. Racial preferences were almost wholly to blame for this gap: When black medical students competed against white exam takers with similar academic credentials, they fared about the same.
Preference supporters argue that with multiple attempts almost all students eventually pass. But that's not much comfort. Advances in modern medicine come at a breathtaking pace. Even the most capable doctors find it difficult to keep up. The medical student who, after intensive tutoring, finally demonstrates minimum competence on his third attempt is unlikely to stay current, no matter what specialty he chooses.
The new JAMA study nonetheless tries to persuade us that special admitees miraculously become "just as good" as other medical students after graduation. Yet the authors' evidence for that conclusion is sketchy at best.
Only one item is offered: results of a questionnaire asking residency program directors to rate the study subjects as residents. We aren't told what questions were asked. Were the directors asked to rate these former residents--some of whom had come through their programs more than 15 years ago--on a scale of one to 10? Or simply pass/fail? All we are told is that the responses were extremely close for the two groups. This is supposed to overcome the mountain of objective evidence--some of it generated by this very study--showing relatively poor medical school performance by beneficiaries of racial preferences.
Another questionnaire directed to the study subjects themselves was evidently designed to demonstrate that preferences are needed because minority doctors are more likely to practice in minority communities. Subjects were asked about their practices, including the proportion of their patients who are minorities, but the study found remarkably little difference between the two groups. If that's true, why do we need preferences?
The editors of JAMA wrote an editorial that appears with the study, calling for racial preferences as a means of restoring trust to the medical profession. JAMA's editors are right to be concerned about trust. When doctors are torturing data to make a political point, whom can you trust?
Ms. Heriot is a professor of law at the University of San Diego.
Material presented on this home page constitutes opinion of the author.
Copyright © 1997 Steven J. Milloy. All rights reserved. Site developed and hosted by WestLake Solutions, Inc.