Poets&Quants | Don’t Throw The Baby Out With The Bathwater

Sangeet Chowfla, former CEO of the Graduate Management Admissions Council, reflects on the business of business education

To test or not to test? Test requiring, test optional, test blind. There is an active debate about the use of standardized testing as an admissions instrument with valid arguments for and against. As the former CEO of the largest testing organization for graduate management education, I could be accused of bias on this issue. On the other hand, I have had a decade-long ringside view of this debate and would like to offer, what I believe is an objective commentary on the subject.

Standardized testing in GME has been around for three-quarters of a century. Why then the debate? The pandemic caused test
center shutdowns and forced schools to be test optional (understandably). The racial justice movement opened a second discussion about diversity in higher education and whether tests created a barrier to access. At the same time, declining applications in the U.S. has put pressure on admissions and recruitment teams leading some to conclude that the answer lay in not requiring tests and thereby removing a perceived barrier to entry. The first argument is no longer valid, but it is worth discussing the other two.

STANDARDIZED TESTS ARE NOT BIASED & DROPPING THEM WILL NOT INCREASE DIVERSITY

There are generally three arguments put forward by the proponents of the no-test or test-optional movements: 1) that standardized tests are biased against historically underrepresented communities, undermining our societal goals of equality; 2) that dropping the use of standardized testing would decrease racial and gender disparities in admitted cohorts; and 3) that standardized tests lack validity and predictive ability correlated to real-world performance in academic programs and that undergraduate GPAs (UGPA) combined with a holistic admissions process are an effective substitute.

It is true that underrepresented communities have median scores that are lower than their White (or Asian) counterparts. This is true for traditional higher education tests such as the SAT, ACT, GRE, and GMAT. It is also true for pre-SAT and PISA tests. Is it the tests? I would argue – and know from personal experience – that all tests of quality have extensive internal mechanisms to weed out bias in test questions through pretesting and differential item analysis. While the tests may be free from bias, we cannot ignore that there are inequalities in our educational and preparation environments and that certain communities may not have access to the same foundational education and means of preparation as others.

Test scores tell us something, even if it is the uncomfortable fact that we live in unequal societies and need to make adjustments for that fact. Ignoring them tells us nothing and is even less likely to drive change. The thermometer (or weather app these days) tells you it’s cold outside. Deleting the app will not make it warmer – but you will be more likely to dress incorrectly.

A related argument is that tests require preparation and that certain students do not have the time or means to adequately prepare. The argument may be true, but the implications bear some discussion. We are essentially saying that a particular student does not have the time and means to spend a few hundred dollars and 50 hours to prepare for a test but that the same student will have the time and means to devote years of nights and weekends and pay tuition in the tens of thousands to participate in a part-time or online program? Therein lies the problem of completion that I will come back to later.

Lastly, the argument that an undergraduate GPA is an effective substitute is not supported by empirical data. On the contrary, studies have shown the UPGA to be an unreliable indicator, largely due to the wide variability of GPA scoring in the U.S. and the inherent differences that exist across the globe. This was why standardized testing was invented in the first place. As schools expanded their reach, nationally and then globally, it became harder to rely on GPAs and the need for a standardized norming instrument arose. Removing test scores will have the unintended effect of admissions teams preferring GPAs from institutions they are familiar with, decreasing diversity.

If we accept, as shown, that tests are efficient predictors of success in the classroom, we must also accept that low-scoring candidates, regardless of background, will have a lower probability of success in that class. We cannot therefore simply waive test requirements and admit candidates who are not adequately prepared – having them expend their precious time and financial resources – but must use the diagnostic data that we gain to build additional support mechanisms. This is, after all, an important part of holistic admissions – the ability to see the diamond, even though all the facets are not shining through, and then polish these facets so that they can stand out, and shine, amongst their peers.

Bill Clinton talked about “the soft bigotry of lower expectations”. We must avoid this trap. Let’s not throw out test scores and ignore reality. Let’s instead use test scores to identify gaps and provide a helping hand.

DO TEST REQUIREMENTS REALLY DEPRESS APPLICATION VOLUME?

At first look, there is face value to this argument. Preparing for and taking a test is a commitment of time and money. Reducing it will remove a point of friction and reduced friction will increase velocity (application volumes). It’s a valid argument if you are a recruitment professional charged with application volumes or a dean looking at higher selectivity ratios.

Does it really help in the long run? There are three reasons why it does not.

One, as I have consistently argued, is about messaging. GME is a luxury good and must be managed as one. The lower our perceived admission standards, the lower the perceived value of the degree. A graduate business degree is for a lifetime, and we know that a degree from a particular school is not about what a student learned while at that school but about the potential of that individual. It’s a signaling mechanism and that signal is maintained, not by the individual but by the cohorts that came before and will come after. Students pay a lot of attention to admission standards and, while they may be perfectly happy to get a pass themselves, they are not eager to join a school that lets anybody in.

Groucho Marx famously joked that he would “not join a club that let him be a member.” It seems that students are equally not willing to join a school that would let him be a student.

Secondly, it should give us pause that over the last decade or so, we have seen admission standards (tests, essays, recommendations) lowered with no secular uptick in applications. Is this a vicious cycle where we respond to declining applications with more open standards which, in turn, reduces the perceived value of the degree and reduces applications?

I recently had a discussion with a senior business school administrator who told me that no-test policies must be working because more than half of his recent applications came without one. I asked him if his overall application volumes had increased, and he responded that they had actually declined. This makes my point. More no-test applications are not necessarily increasing the pipeline; they are just giving prospective students an easier path to application.

The third is the completion problem, something that is becoming more acute in part-time/online programs. I have always thought about preparing for the GMAT as taking your first class in business school. It brushes up foundational skills and, more importantly, tells me if my life circumstances (time more than money) allow for the commitment that a graduate program requires. Take that away and we take away another form of signaling. This one to the student about whether they are ready for what it takes. Without that signal we are (forgive me for putting it this way) exploitative, happy to fill the seat and take the tuition without ensuring that the means to complete the journey exists.

ADMITTING STUDENTS REQUIRES SCIENCE AND ART

We are having this debate because, over time, the use of standardized testing got ahead of itself. Rankings exaggerated its value and school websites took pride in promoting average test scores. This has led to the perception that tests are the ultimate goal of admissions. Too much is made of inconsequential single-digit changes in the scores of admitted classes as an example. I don’t believe that we should eliminate testing, but we do need to change how we think about them.

The test publishers need to sharpen their focus on the triple pillars of validity, security and accessibility. The test should measure what is required (and no more), there needs to be confidence that the score matches the actual person who took the test and it should be accessible. At the risk of showing my personal bias, the new GMAT focus is a good start. It is slimmer, more relevant and drops some of the artifacts of the past.

Admission professionals must use tests as an anchoring data point in a holistic process without making it the only data point. A test tells you a lot about the reasoning and critical thinking capabilities of a candidate but not enough about grit, drive, teamwork, and communication skills. This is the subjective element that needs to be bought into play.

Marketing and recruitment should think about how they talk about test scores. Schools are often eager to tout their test averages – particularly when they have gone up – giving a false impression about the centrality of such scores. One option is for schools to drop the use of test averages in their marketing communications entirely, replacing them with the 20^th and 80^th percentile ranges only. This would give prospective students the ability to see where they fit without a single average number becoming a point of focus. It would also demonstrate that an individual with a lower test score, but with a compelling life story, has the possibility to be admitted.

Standardized tests are a useful tool in the admission process. The benefit is in the word – standardized – that provides an empirical data point that is free from bias and interpretation. A GMAT 700 is a GMAT 700. You cannot wish it away if you are unfamiliar with the candidate’s background or uncomfortable with their color, origin, orientation, or political beliefs. They are not without issues – in structure and in usage. We should strive to make them better, not throw the baby out with the bathwater of the politics of the day.

Author and former GMAC CEO Sangeet Chowfla

Sangeet Chowfla led the Graduate Management Admissions Council as president and CEO for nearly ten years from 2014 to 2022. A globally recognized and respected executive with deep experience in the technology, telecommunications, and venture capital sectors, he began his career in New Delhi with IBM/IDM. Chowfla went on to spend 18 years with Hewlett-Packard Co. in Europe, the Middle East, Asia Pacific, and the United States. He culminated his tenure with the company as vice president and general manager of the Inkjet Media Division from 1995-2001. He then moved to Timeline Ventures as a partner in the venture capital partnership. In 2007, Chowfla became the chief strategy officer and executive vice president of the Mobile Services and Global Market Units of Comviva Technologies, a leading Indian telecommunications software company. Chowfla joined GMAC during a period of disruption in the organization and industry. During the last three years of his tenure, he helped to stabilize the candidate pipeline, renewed GMAT exam growth, diversified GMAC’s footprint and ensured a strong financial foundation to enable future investment.