Study Raises Questions About GMAT Reliance

May 16, 2002
Print
Text Size:
A A
ANN ARBOR — If some executives are disappointed with the job performance of some new MBA employees having high GMAT scores, maybe they should give less weight to those scores.

That’s one suggestion arising out of a recent study conducted at the University of Michigan Business School indicating that the GMAT (Graduate Management Admission Test) isn’t necessarily a good predictor of success on the job.

And now the U-M Business School is thinking of supplementing its GMAT program with something that may turn out to be more reliable. 

What has happened is that a research project conducted in 1999 and 2000 among new MBA students at U-M shows — according to its authors — that tests which assess students’ practical abilities can be just as valuable as standardized intelligence devices such as the GMAT, and maybe more so.

Conducting the research were Robert Sternberg and his associate, Jennifer Hedlund. Sternberg, whom U-M terms renowned, is a professor of psychology at Yale University and is the author of the book “Successful Intelligence.”

“The types of problems found on standardized intelligence or aptitudes tests are quite different from the types of problems found in the real world,” Sternberg indicated when releasing the results of the 2-year study.

“Individuals who perform well on academic problems do not necessarily perform well on poorly defined practical problems. Thus, individuals who are successful by GMAT standards may not be successful by business standards.”

On the other hand, Sternberg and Hedlund found that tests of managerial potential which stress practical abilities tend to predict success both in the academic world and in practical endeavors.

Their study set forth to build a better predictor of success based upon Sternberg’s theory of successful intelligence (SI). It’s his view that successful intelligence is an interactive exercise of analytic, creative and practical abilities.

The pair created a test framework of case-based judgment problems, which sought short-answer responses that then were rated for their quality. Their assessment avoided the right-or-wrong answer format of the standardized GMAT.

Sternberg and Hedlund found that the higher the scores on both GMAT and SI tests, the better the students’ first-year and final-year grade point averages.

Likewise, the higher a student’s SI scoring, the higher the student scored on an applied team-consulting project — a cornerstone of U-M’s MBA program.

And beyond that, higher SI scores correlated with higher participation in academic clubs and holding more leadership positions.

But while GMAT scores were accurate predictors of grade averages in the MBA program, they did not necessarily predict grades on the applied project. Neither was there a close correlation between students’ GMAT and SI assessment scores.

Sternberg said the SI scores accounted for unique variance in both academic and practical performance beyond that accounted for by GMAT and undergraduate grade point average. He said he believes the study’s findings measure unique abilities that existing standardized tests simply don’t reveal.

Curiously, men tended to perform better than women on the GMAT, but women out-performed their male counterparts in the SI measures. Moreover, he said, “African-Americans scored only slightly lower on practical intelligence measures, but substantially lower than whites on the GMAT.”

He said the GMAT puts both women and African-Americans at a disadvantage, but that the disadvantage isn’t nearly as significant in the SI test employed at U-M.

“Although SI measures some disparities of their own,” he added, “the results suggest that they do not exhibit the same pattern or degree of disparity found for the GMAT.”

Sternberg said he and Hedlund plan further research and development of their SI format assessment.

And it can’t come too soon for Jeanne M. Wilt, the assistant dean of admissions and career development at the U-M Business School. Wilt said she is excited about the  possible use of SI tests.

“We are always looking for better ways to identify and develop leadership talent,” she said.

“This type of assessment not only has potential use as a complement to the GMAT in admissions decisions for MBA programs, but it also can be used as a tool for teaching students practical problem-solving skills that help them become more effective business leaders.”

Editor's Picks

Comments powered by Disqus