Observations on the GMAT

GMAT - a summary statistic of questionable
validity and fairness.

The Graduate Management Admissions Test (GMAT) remains a necessary accouterment in MBA circles with admission to many of the top schools requiring a score on 650 or even 700. Stanford students averaged 729, with Harvard, New York University and the University of Chicago all averaging 720+.

We asked the question, "What do we know about GMAT scores," and turned to the Graduate Management Admission Council's own data for answers.  Each year the Council publishes a Profile of Graduate Management Admission Test Candidates (available online), explicitly pitched to universities as a way to "market your program more effectively,"  a form of bragging rights of sorts.  We offer the following observations for the most recent year, 2012-2013:

Two hundred, thirty eight thousand (238K) student candidates took the test in 2012-2013, 49K fewer than the previous year.

Only 121K (51%) of these students expressed an interest in pursuing an MBA; the total MBA student market is much larger.

Eighty-five percent on these students were under the age of thirty years.  Almost 90% of women taking the GMAT were under the age of thirty. 

The highest test scores (by age category) were posted by students of less than twenty years of age: kids.  Males of this age averaged 604, while females averaged 597 (though this demographic accounted for only 0.6 % of test takers).

Test scores drop precipitously for both males and females: i.e., inversely correlated with age and presumably attributes associated with age, such as experience, maturity and wisdom.  Females: age 20-21 (avg. 572); 24-25 (528); 31-34 (504); 40-49 (449)

The number of woman sitting for the GMAT dropped precipitously after the age of 23 years: age 22-23 (N=27K); 24-25 (17K); 26-27 (12K); ... 31-34 (6K).

Women over the age of 21 years averaged less than a 550 score and would likely be excluded from many top MBA programmes.

Women consistently posted lower average test scores within age categories, oftentimes with meaningful differences (e.g., age 26-27: male avg. 571 vs. female avg. 533).

Males and females over the age of forty years posted average scores of less than 500.

Three countries, the U.S., India and China, accounted for 71% of all GMAT tests taken in the world, with the U.S. accounting for over 90K test takers.  India (577) and China (582) significantly outperformed the U.S. (532) on average.

The highest average GMAT scores were posted in Singapore (605) and New Zealand (608), though these nations benefit from a small-N advantage.

We conclude that the GMAT may favor younger test takers; measure the individual's test-taking ability (i.e., a recency effect); disadvantage women and individuals with greater experience (seasoned professionals); and remains a largely U.S. convention taken elsewhere by students hoping to study in the U.S.  Much of the world may not value this summary statistic of a person's ability in the MBA classroom or potential to become a global business leader. We're reevaluating the weight we place on GMAT in our selection processes.

The GMAT reminds me of a popular U.S. reality television show, 

"Are you smarter than a 5th grader?"

David