In a year unlike any in recent history, one thing will surely happen this winter as it always has. Students will finalize and submit applications to colleges and graduate schools across the nation. But this year, some students won’t submit any test scores at all. And after they receive their admissions decisions, they’ll never know how they were or were not selected.
Prior to the Covid pandemic, I’d heard fierce debate on the value of standardized testing, including questions about what place the tests should have in graduate admissions. While many schools have since temporarily waived test requirements at the graduate level, some universities and programs are making more permanent and monumental decisions about their admissions practices that will have an impact for many years to come.
I’m hoping what is best for the student remains at the center of these debates and decisions. And there is one thing we can be certain of, from decades of research: The tests are not the problem. How they are used can be.
It is critical that we continue to look at how assessments are being used, to ensure the most equitable and transparent processes possible. This is as true in graduate admissions as it is in K-12 assessments.
At ETS, of course we are going to defend testing — but not for the reason people may think. We defend testing because we are a nonprofit organization founded on a mission to advance equity in education. We believe in the power of assessment to do just that. For more than 70 years, we have seen what measurement data and policy research can do to propel people forward, regardless of income, social status, gender, ethnicity or race.
We believe every student is so much more than a test score. In graduate admissions, we think GRE scores should be included as an objective source of information among multiple criteria. A holistic approach ensures balances to the limitations of any single measure of knowledge, skills or abilities. This is because how scores are used is as important as the scores themselves.
GRE scores provide a standardized measure of graduate-level academic readiness. But what does this mean for individuals? It means the test can make a difference for students whose scores earn them a second look when they had to work at a job instead of doing undergraduate research. It can do the same for students whose scores demonstrate they have the right skills for a program, but who might otherwise be cut based on preferences for students from better-known or higher-rated institutions.
We know that many students who could succeed in graduate education have test scores that do not reflect all their skills. That’s why we’ve always advocated against the use of cut scores that institutions set and use to exclude all applicants whose scores falls below the cut.
For more than 70 years, we have seen what measurement data and policy research can do to propel people forward, regardless of income, social status, gender, ethnicity or race.
If a student received high marks from a top undergraduate program accompanied by strong reference letters and relevant research experience, a low GRE score should carry less weight.
Yet for many reasons — structural stigma, deeply segregated school districts, unequal school funding — students from underrepresented backgrounds often don’t have the same opportunities and experiences as their more privileged peers.
Students come to graduate admissions with varying experiences relevant to the path of study that lies ahead. That’s why we believe there should be a transparent and holistic review of all the materials they can provide to show what they know and can do.
A holistic approach cannot erase the inequities in educational and life experiences that are associated with income and opportunity, but it can ensure that scores are considered in the context of all the information candidates can submit to reflect their strengths.
We know the opportunities that high scores can represent, and it drives us to continue investing in measurements that help students demonstrate their academic readiness. So, for students whose scores show strength not evident in other parts of their graduate applications, we’ll continue to provide data that lets them show what they know and can do.
For institutions that value the role GRE scores play in presenting a student’s total picture, we offer a range of resources about holistic admissions practices, increasing access, diversity and inclusion.
For institutions that do not use GRE scores and have removed standardized testing from admissions, we are also working on solutions. We firmly believe in the valuable role of a standardized, objective measure, and we understand that programs are seeking more than GRE scores to achieve their goals.
As a research-based organization, we are uniquely positioned to join the higher education community in bringing forward new ways of meeting the evolving needs and values of today’s graduate programs — and we intend to do just that.
There is always more work to be done. We acknowledge that, but let’s not stand in the way of providing meaningful testing data that is critical for so many students, programs and institutions. After all, test scores are the only objective component in graduate admissions. Without them, we’re left making life-changing decisions with less data.
Walt MacDonald is the president and CEO of ETS.
This story about testing was produced by The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education. Sign up for Hechinger’s newsletter.
Why we should discourage holistic test use
In response to the recent opinion article by Walt McDonald (president of ETS), we agree that tests are not the problem, and that how they are used can be. We also strongly agree with the importance of transparency. However, adopting a holistic approach to admissions, as proposed in the article, only aggravates the problems regarding equity in education.
Holistic admissions typically not only imply utilizing multiple sources of information about applicants, but also weighting that information implicitly and inconsistently. Due to this implicit, ‘black box’ nature, the holistic approach and transparency are mutually exclusive. Standardization is not only important to ensure equity when it comes to the content and administration of tests, but is equally important when it comes to collecting other relevant information about applicants, and in combining that information to make admission decisions. Therefore, equity is not achieved through a holistic approach, but by determining what information is deemed relevant, how much weight each piece of information should get, applying that rule to all applicants equally, and collecting the same information from all applicants in a structured way.
In other words, drawing on an example provided by Dr. McDonald, if a student receives high marks from a top undergraduate program accompanied by strong reference letters and relevant research experience, the GRE score should not carry less weight. Instead, we should determine how much weight it should get for all applicants compared to all other factors deemed relevant, so that all applicants can compensate for lower GRE scores with high marks (or vice versa) to the same degree.
There is an enormous amount of evidence arguing against the use of the holistic approach. Therefore, as a research-based organization, ETS should not advocate its use.
Dr. A. Susan M. Niessen (Assistant professor in Psychometrics and Statistics, University of Groningen)
Prof. dr. Rob R. Meijer (Full professor of Psychometric and Statistical Techniques, University of Groningen)