Expert: Aptitude Tests Are Biased Expert: Aptitude Tests Are Biased
He Says U-M's Policy Can Correct Imbalance
February 7, 2001
BY ERIK LORDS
DETROIT FREE PRESS EDUCATION WRITER
Standardized tests for college and law school admission are inherently biased against women and minority applicants, an expert witness said Tuesday during the federal trial to decide whether the University of Michigan's Law School admissions policy is constitutional. Martin Shapiro, a testing specialist and psychology professor at Emory University in Atlanta, testified that using race as one of many factors in admissions is the only way colleges can offset biases in standardized tests.
Shapiro's testimony came as the trial resumed before U.S. District Judge Bernard Friedman in Detroit after a week's hiatus. At issue is whether the law school discriminated against Barbara Grutter, 47, of Plymouth Township, a white applicant who claims the university failed to admit her while accepting minority students with lower test scores and grade-point averages.
Shapiro said biases were built into standardized tests from the start. The Scholastic Aptitude Test, or SAT, which has been used for about 75 years, was designed to benefit certain groups and hurt others, he said. According to Shapiro, college officials decided during the mid-1920s that there were too many Jewish students in undergraduate Ivy League classes, including at Columbia University in New York, where 40 percent of all students were Jewish.
Test makers developed an exam with questions modeled after curricula at elite prep schools, where few if any Jewish students attended at the time, Shapiro said. "The number of Jewish students was reduced to 20 percent" at Columbia, Shapiro said. "There's no denying it. That is the beginning of the SAT." Shapiro said the bias that existed then is prevalent today, even as test makers devise new versions of the SAT. To determine what questions will be on future tests, test makers pretest groups of students, he said.
To maintain consistency with past questions, if the group that historically scores well performs well on a pretest, the question is kept. If a group that historically scores poorly does well on a pretest question, that item is tossed out, Shapiro said. Data show that women and minorities traditionally have scored lower than white men on the pretests, Shapiro said.
"The test is designed to preserve the status quo," he said. "Whoever did better on the test before will now be carried along" and likely do well on future tests. In his cross examination, Larry Purdy, a lawyer for Grutter, pounced on Shapiro's historical reference. Purdy asked now that Jewish students are so well-represented at Ivy League colleges, wouldn't that indicate that biases have been removed? Not exactly, Shapiro answered. "Maybe it's that more Jewish students started going to prep school."
Purdy also asked whether it would make sense for colleges to stop asking an applicant's race. "That's a fairy tale," Shapiro said. "You cannot remove race as a piece of information in the United States. It's too ascertainable." Purdy then asked about not using standardized tests in the admissions equation. Standardized tests "are not the devil," Shapiro said.
But "race and sex in our society are characteristics that affect how we perform. They are not irrelevant." In most reverse discrimination cases, the plaintiff focuses on test scores and grades. Shapiro said such attacks on affirmative-action policies are flawed because schools consider a wide range of factors, including letters of recommendation and what role race might have played in a person's life.
"The real issue is, do you want to be bound" by test scores, "when in fact you have all this other information if you just read the application?" Shapiro asked.
Contact ERIK LORDS at 313-222-6513 or lords@freepress.com.
News and Announcements | AAD Home Page