Is Educational Testing Service (ETS) Headed for Bankruptcy? (Part 2)

Emerging Factors Accelerating the Collapse in Demand for Standardized Testing Among MBA and Master’s Applicants

As if the news wasn’t bad enough for testing services like GMAC and ETS, new factors are now emerging that could intensify that industry’s “perfect storm” of adversity into a “perfect hurricane.” We next review two of these factors, concluding with our speculation about the long-term future of ETS as an ongoing concern.

The GRE Predictive Validity Meta-Analysis

In February 2023, a team of professors from Utah State University and China’s Binzhou Medical University published a comprehensive meta-analysis of the GRE’s predictive validity. Titled “The Predictive Validity of the GRE Across Graduate Outcomes: A Meta-Analysis of Trends Over Time,” the report was published by the Journal of Higher Education.

This meta-analysis screened 3,036 scholarly journal articles about the GRE published between 1943 and 2022. The researchers eventually narrowed down the sample to the best 201 studies of the GRE’s predictive validity published between 1957 and 2022, with an average sample size of 362 examinees. The research team then performed a modified multiple regression analysis to determine correlations between GRE scores and various graduate student outcomes.

The results were devastating for ETS.

• The meta-analysis found that 62.3 percent of the reported effects were nonsignificant. That means in almost two-thirds of the studies, the GRE scores had no value in terms of predicting student outcomes.
• The magnitude of the observed predictive relationships didn’t improve over time as one would expect. Instead, the magnitude decreased significantly over time. In other words, the GRE got worse at predicting student outcomes over time, not better.
• Across all 201 studies and outcomes, the aggregate mean effect was statistically significant and positive—but extraordinarily small. For example, a GRE score only predicted about three percent of the variance across all measured outcomes. When strong data relationships exist in a multiple regression analysis, social scientists can find variables that explain as much as 30 percent or more of the variance in outcomes.
• Moreover, in terms of explaining grades, a GRE score only predicted 4 percent of the variance in overall graduate GPA and a mere 2.56 percent of the variance in first-year graduate GPA. Incredibly, the GRE only predicted a microscopic 0.16 percent of the variance in degree completion outcomes, and 0.64 percent of the variance in making progress towards a degree. The professors further explain:
  

  Thus, in the majority of comparisons the GRE held no predictive value and, on average, predictive value when observed was minimal (e.g., 96 percent of variance in overall GPA not predicted by GRE scores).

• Ideally, first-year GPA would be crucial to predict accurately for admissions decisions, because some graduate schools link a student’s GPA at the end of their first year with the school’s authorization for the student to continue in their MBA, master’s or PhD program and earn their degree. But according to this meta-analysis, the GRE is not capable of predicting 97.44 percent of the variance in first-year grades.
• Composition effects correlating GRE scores with race or ethnicity were not statistically significant. In fact, GRE scores demonstrated worse predictive ability for GPAs as the proportions of people of color within a sample increased—a clear equity and fairness issue for these folks. In other words, the authors conclude that “the predictive validity of the GRE is diminishing over time as the graduate population diversifies.”

In short, the professors conclude that the best data, going back as far as 1957, demonstrates that the GRE lacks validity, fairness, and suitability for helping admissions committees make decisions about which applicants are likely to succeed in graduate school.

Given the billions of dollars spent on developing and administering the GRE since 1943 and the money students spend on test prep courses and tutoring to try to boost their scores, one has to wonder why universities continue to require such an obviously defective product—an “examination” that purports to be fit for purpose when it clearly is not.

It’s inappropriate for us to comment about such a question in a report like this one. However, FairTest’s astute executive director Harry Feder offers a provocative theory in this June 2024 podcast.

Performance-based Admissions

Fortunately, MBA and other graduate school applicants who don’t want to take the GRE or the GMAT have plenty of options among the business schools we’ve covered here on BSchools since 2018.

As Poets’ John Byrne points out in Part One of this report, most of the newer online graduate degree programs, including MBA and specialized business master’s degree programs, are these days either test-free or test-optional. The few online MBA programs that still technically require admission tests almost always offer to grant waivers to candidates in exchange for years of work experience or high undergraduate GPAs.

But there’s another solution for online candidates who want nothing to do with tests or waivers. One of the most disruptive innovations ever to hit higher education is performance-based admissions or PBA. This alternate admissions paradigm grants provisional entry to all applicants. Meanwhile, PBA waives admission tests and all the other extensive documentation of a traditional application in exchange for earning good grades for graduate credit in one or two actual online courses within the degree program to which a candidate has applied.

When Northeastern’s College of Engineering in Boston offered performance admissions for its first two online degree programs, the university also offered a traditional admissions process for applicants who wanted that option—and the traditional process required GRE scores.

In a May 2024 interview with EAB’s higher education strategist Jennie Bailey, Robert Towner, the assistant vice president for new ventures and business operations at Northeastern, said that:

What we saw in our first term of launching this with engineering is over 99 percent of the students who came in chose the PBA process versus the traditional process.

In other words, given a choice between traditional admissions that require the GRE and an alternate process that did not, essentially 100 percent of the applicants chose the performance admissions process that obviated the need for the GRE.

One could argue that’s a small sample from a single university during a single semester—and that might be true. However, since December 2023, our team has studied the introduction of performance admissions at several institutions, such as the University of Colorado at Boulder, America’s first college to introduce performance admissions, back in 2019. And we have yet to learn of a single case of a student offered both traditional and PBA admissions who didn’t choose the performance admissions pathway.

Performance admissions represent a dangerous threat to testing firms like ETS and GMAC. Because there’s never an application with performance admissions, there are never any admission tests. And every student admitted to a degree program through a performance admission pathway represents hundreds of dollars worth of lost revenue to ETS or GMAC.

Why? That’s because in the United States, for most examinees, the GRE General Test costs $220, and the online GMAT costs $300. What’s more, 2015 ETS data indicated that nearly 25 percent of examinees retake the GRE each year, and 2016 data indicated that more than 20 percent of every year’s examinees sit for the GMAT more than once. Plus both companies offer a broad array of additional options—such as rescheduling services and additional score reports that all require separate fees ranging from $35 to $180—during examinees’ long and winding “testing journeys.”

So, is ETS Headed for Bankruptcy?

Large organizations can frequently operate at a temporary loss in the short run, and ETS is no exception. Plus, we aren’t aware of an industry analyst or commentator who believes that ETS might stop offering its tests or suspend operations during the next 12 to 18 months.

However, the probability of an eventual ETS bankruptcy amounts to the one possibility that everybody in the education industry is probably thinking about—but nobody is talking about. At least not yet.

ETS just lost a third of its annual revenue from the College Board’s firing, amounting to about $300 million. Moreover, the decline in standardized testing is a pervasive, long-term structural trend that will not reverse. In other words, the nearly universal standardized testing admission requirements that almost all universities demanded before 2018 are gone for good—and they’re never coming back. That fact means ETS is stuck selling defective products like the GRE that people with better options no longer want and are no longer forced by colleges to pay hundreds of dollars to obtain—along with buying thousands of dollars worth of test prep courses and tutoring services.

Under these circumstances, it is difficult to envision a scenario by which ETS, as we know it, will still exist as a going concern in the long run five years from now. There doesn’t seem to be a viable line of business with sufficient margins readily available to the firm that can replace both the College Board’s lost annual revenue, plus offset all the years worth of accumulated losses because of the declines in standardized testing since 2018.

A potential employer market for tests to assess employee skills acquired through credential programs has been proposed as a possibility by some within the industry. But currently the demand for such testing is way too small to drive a viable business model that could provide ETS with what it needs right away—additional annual revenue between $300 million and half a billion dollars a year, just to survive in its current form.

And if ETS can’t quickly find a replacement line of business with sufficient margins or raise a lot more capital without delay, the company will probably need to cut at least 1,000 more U.S. employees and scale back its operations all over the world—and fast. Whether the firm should pursue that strategy through a formal bankruptcy and reorganization procedure or some other mechanisms poses a question for the company’s legal counsel.