Report Shows LSAT Score Gap

A report released Tuesday by Testing for the Public, a Berkeley, Calif.-based think-tank showed large differences between the LSAT scores of minority and white students with identical GPAs.

The study revealed that, on average, minority students from five prestigious colleges-Harvard included-who applied to the University of California, Berkeley's Boalt Hall School of Law fared worse on the LSAT than their white counterparts who received nearly identical grades for their undergraduate coursework.

"Students who have achieved the same success in college should theoretically earn approximately the same scores on the LSAT," said David M. White, director of the institute.

"Unfortunately, the results of this study demonstrate that this is far from true," he said.

In order to conduct the study, Testing for the Public took a sample of students from the five schools who took the LSAT from 1996-1998 and subsequently applied to Boalt.

William C. Kidder, a first-year student at Boalt, actually compiled the data this summer.

On average, black students scored 9.30 points lower on the LSAT than white students; Latino students, 6.87 points lower; Native American students, 3.77 points lower; and Asian-American students, 2.48 points behind.

The possible range of scores on the LSAT is between 120-180 points.

"Since the study controls for college performance," Kidder said, "there is some combination of cultural bias on the test and atmospheric bias."

Kidder explained that stereotypes and other influences in the testing milieu may artificially depress the performance of minority students.

But many say the results of the study do notnecessarily indicate bias and are not necessarilysurprising.

Winthrop Professor of History Stephan A.Thernstrom said, "The percentage ofAfrican-American students at the top of the LSATcurve is traditionally low."

"The question you have to ask is, were thestudents taking essentially the same courses?" hesaid.

Thernstrom also said grade inflation has to betaken into account when considering students fromdifferent schools.

"At Stanford the median grade is said to be anA-. At Harvard it's a B+. So grades are not a verydiscriminating indicator of excellence in academicperformance," he said.

David Murray, director of research andstatistics at Washington, D.C. think-tankStatistical Assessment Service agrees that gradesalone are not always reliable.

"One sees that grades are not that strong anindex, because, one, there is more variability ingrades than there are in test scores, and two,more compelling, is that the grading scale is atbest, from 1 to 5. This varies from field tofield, so, basically, the [LSAT] can distinguishmore subtle differences than a 5-point scale can."

Officials from the Law School AdmissionsCouncil (LSAC), which is responsible for writingthe LSAT, also say Testing for the Public's studyis somewhat misleading, but concede it does revealsignificant ethnic gaps within the testingpopulation.

LSAC Associate Council Jim Vaseleck said "thefaulty assumption in the study's analysis is thatstudents with the same GPA from the same schoolshould have the same test scores."

The LSAT is a test of skills which may or notbe attained through mastery of college coursework,he said.

The LSAC could not offer any reasons for whyethnic minorities score lower than their whiteclassmates.

"What we do know is that the test performs itsprediction function equally for different ethnicgroups."

Thernstrom also agreed that LSAT scores were agood predictor of law school performance.

"43 percent of black students [who took theLSAT] either failed to complete law school orgraduated but were unable to pass the bar examwithin three years," said Thernstrom, who recentlycontributed a related article to ConstitutionalCommentary.

"All of the law schools are perfectly free todiscard the LSAT if they think it isdiscriminatory. But none of them have," he said.

Students were not generally surprised byTesting for the Public's findings, but felt thedifferences in scores reflect economic, notethnic, gaps.

Mark E. McIntosh '99, a black student who tookthe LSAT on Saturday, did not see the test asculturally biased, but offered possible reasonsfor the discrepancies.

Some students may have to work to pay for theireducation and not have the time or money for testpreparation services, McIntosh said.

"It is a test of preparation. It would help tocontrol this study on the basis of money, and notjust grades," he said.

Shannon K. Manigault '99 echoed thesestatements.

"I have not seen studies of the LSAT, but whenlooking at the SAT, scores have a correlation withincome," she said.

"My guess would be that black students [atthese schools] would have lower family incomesthan white students, which would affect themtaking courses a Princeton Review, Kaplan, etc.

But many say the results of the study do notnecessarily indicate bias and are not necessarilysurprising.

Winthrop Professor of History Stephan A.Thernstrom said, "The percentage ofAfrican-American students at the top of the LSATcurve is traditionally low."

"The question you have to ask is, were thestudents taking essentially the same courses?" hesaid.

Thernstrom also said grade inflation has to betaken into account when considering students fromdifferent schools.

"At Stanford the median grade is said to be anA-. At Harvard it's a B+. So grades are not a verydiscriminating indicator of excellence in academicperformance," he said.

David Murray, director of research andstatistics at Washington, D.C. think-tankStatistical Assessment Service agrees that gradesalone are not always reliable.

"One sees that grades are not that strong anindex, because, one, there is more variability ingrades than there are in test scores, and two,more compelling, is that the grading scale is atbest, from 1 to 5. This varies from field tofield, so, basically, the [LSAT] can distinguishmore subtle differences than a 5-point scale can."

Officials from the Law School AdmissionsCouncil (LSAC), which is responsible for writingthe LSAT, also say Testing for the Public's studyis somewhat misleading, but concede it does revealsignificant ethnic gaps within the testingpopulation.

LSAC Associate Council Jim Vaseleck said "thefaulty assumption in the study's analysis is thatstudents with the same GPA from the same schoolshould have the same test scores."

The LSAT is a test of skills which may or notbe attained through mastery of college coursework,he said.

The LSAC could not offer any reasons for whyethnic minorities score lower than their whiteclassmates.

"What we do know is that the test performs itsprediction function equally for different ethnicgroups."

Thernstrom also agreed that LSAT scores were agood predictor of law school performance.

"43 percent of black students [who took theLSAT] either failed to complete law school orgraduated but were unable to pass the bar examwithin three years," said Thernstrom, who recentlycontributed a related article to ConstitutionalCommentary.

"All of the law schools are perfectly free todiscard the LSAT if they think it isdiscriminatory. But none of them have," he said.

Students were not generally surprised byTesting for the Public's findings, but felt thedifferences in scores reflect economic, notethnic, gaps.

Mark E. McIntosh '99, a black student who tookthe LSAT on Saturday, did not see the test asculturally biased, but offered possible reasonsfor the discrepancies.

Some students may have to work to pay for theireducation and not have the time or money for testpreparation services, McIntosh said.

"It is a test of preparation. It would help tocontrol this study on the basis of money, and notjust grades," he said.

Shannon K. Manigault '99 echoed thesestatements.

"I have not seen studies of the LSAT, but whenlooking at the SAT, scores have a correlation withincome," she said.

"My guess would be that black students [atthese schools] would have lower family incomesthan white students, which would affect themtaking courses a Princeton Review, Kaplan, etc.