[Poets & Quants] Why Do GMAT Scores Keep Rising? GMAC Has (Some) Answers
By all accounts, the Graduate Management Admission Test is no easier today than it was five or 10 years ago. Yet in recent years score averages for elite business schools have been on the rise — including some dramatic examples recently reported by Poets&Quants. The puzzle has drawn enough speculation that the Graduate Management Admission Council, which administers the GMAT, released a report today (June 29) to address the issue.
In The Fallacy of Score Increases and the Impact of Score Preview, GMAC analyzed GMAT scores for citizens of eight countries that account for about 80% of GMAT exams taken each year since 2011 (the United States, Canada, France, Germany, the United Kingdom, China, India, and South Korea). Rebecca Loades, report author and director of innovation and next generation systems at GMAC, concludes that GMAT scores are actually stable, remaining consistent or growing by a statistically insignificant amount. But she adds that given trends and persistent perceptions, the report — five months in the making — may be the first of an annual series.
“Say you have 10 candidates and their average score is 650, and all of them sent their scores to a school,” Loades tells Poet&Quants. “So the schools got this range of scores going from 600 to 700. Now let’s say that we’ve got the ability to cancel scores. So candidates who have 630, 640 say, ‘I’ve previewed my score, and I’m canceling it — I’m not going to send it to that school.’ Even though there are 10 people who have taken the GMAT and their average score is 650, a school would only see the 650, the 670, the 700. And so it would appear candidates are doing better on the GMAT, but actually they’re doing just the same — they’re just self-selecting out.”
UNADDRESSED IN THE REPORT IS THE ROLE OF RANKINGS
The ability to preview and cancel a low score, introduced in 2014, is one of two main factors GMAC identifies that contribute to perceptions of rising scores and actual rising averages among admitted classes at leading graduate business programs. The other: a demographic shift among test-takers toward a younger population composed of students who are entering pre-experience specialty master’s programs. Younger test-takers still in college or only recently graduated tend to score higher because they are more practiced at taking tests.
Unaddressed in the report is the role of rankings, particularly the highly influential U.S. News ranking of MBA programs in the U.S. That publication uses the average GMAT score of the latest incoming class of MBA students in its metrics to create a ranking. That has put pressure on business school admission officials to cherry pick candidates with the highest scores and award them scholarships to get them to enroll.
Some school administrators believe the influence of rankings and their linkage to GMAT scores is the ‘elephant in the room.’ “I can imagine the added focus and pressure that American programs place on this particular component of a candidate’s profile knowing the direct impact it has on one of the most influential rankings in the U.S.,” says Niki da Silva, managing director of the full-time MBA program at the University of Toronto’s Rotman School of Management. “The reality is that the difference between a 680 and a 710 score comes down to luck, getting more or less of a question type you do well on, but is often the difference between an admit or refusal at many top programs despite the fact that there may be no significant difference in the aptitude of these candidates.”
SCORE PREVIEW AND WHAT IT HAS MEANT FOR PERCEPTIONS OF THE GMAT
Nonetheless, most applicants to the elite programs have a near obsession with scoring well on the text, a anxiety that has led many to cancel their scores and try for better results. GMAT’s score preview feature means lower test scores are removed from the pool through student cancellation. Cancelled scores don’t appear on score reports. One effect, as GMAC’s report posits, is that “a greater number of higher scores are reported to schools, a phenomenon common across all program groups analyzed in this study.” Since score preview was introduced, score cancellations have jumped from 2% the first year to 19% in 2016 and 27% this year, according to Brian Carlidge, Kaplan Test Prep’s executive director of pre-business and pre-graduate programs.
In looking at the effect of score “preview,” GMAC examined GMAT scores sent to 10 undisclosed schools (five “global programs ranked highly around the world” in the U.S., one in Canada, two in Europe, and two in Asia) and found that the average GMAT score was 651 — a score in the 76th percentile and an average 100 points higher than the overall average for the test. Only two years earlier, the average GMAT score for the group applying to the top 10 global programs was 22 points lower, at 629.
Carlidge tells Poets&Quants that Kaplan prepares thousands of students for the GMAT every year, and “we haven’t heard from them that they think the exam has gotten any easier over the years.” Indeed, he believes the GMAT remains a “rigorous” test — more rigorous than the GRE, which is increasingly finding favor in business school admissions — and aspiring MBAs share his view, a fact borne out by Kaplan research.
“We speak with many graduate school candidates who hold the perception that the GMAT is more rigorous and challenging than the GRE,” Carlidge says. “But many of those same aspiring MBAs still take the GMAT because some business schools, according to our own research, give an admissions advantage to GMAT takers, all other factors being equal.”
REPORT: SCHOOLS MUST BREAK CYCLE OF TURNING AWAY LOWER-SCORING CANDIDATES
GMAC says that roughly one in six GMAT scores, about 19%, were cancelled in 2016, with the average score being cancelled at 603. Globally ranked programs in the last two years have attracted more scores from test takers scoring 690 or higher — 13,262 in 2016 versus 12,022 in 2014 — despite there being fewer exams with this score in GMAC’s dataset. In 2016, candidates scoring 660 or higher accounted for 57% of the scores received by the highly ranked global schools.
One of the chief consequences of score preview, Loades writes, is that “as the number of higher-scoring candidates sending scores to Group A (globally elite) programs increases, these programs have more high-scoring candidates to choose from and so reported program averages rise, which, in turn, may drive lower-scoring candidates away (or encourage them to retake). Unless schools intervene, the cycle will continue until it breaks.” It’s a question of perception, she adds in conversation with P&Q: “Things like percentiles and things like average scores are calculated across the entire examinee base. So whether you keep or you cancel your score is irrespective. … Effectively, the same demographic groups are performing on a consistent basis year after year.”
Carlidge, however, sees a correlation. “Allowing test takers to cancel low GMAT scores and retake the test after more preparation is having the anticipated impact of raising scores. … Because the lower scores are no longer included in the equation, it makes sense that scores overall are higher. The test-taking year ends in a few days, and we’re looking forward to seeing some more data about this.”