Change the Discussion on Standardized Tests

How college admissions offices make decisions about whom to admit and deny is garnering increased concerns. The role of standardized admissions test and test-optional approaches to admissions decisions are also receiving a great deal of attention. We need only look at the previous two weeks of Inside Higher Ed to find two stories that focus on test-optional admissions and the role of test scores in admissions. At the annual January conference held by the University of Southern California Center for Enrollment Research, Policy and Practice, there was a lively conversation that critiqued both test-optional admissions schemes and the utility of using admissions tests to make decisions.

Much of the discussion was fueled by a new book, Measuring Success: Testing, Grades and the Future of College Admissions. Some at the conference voiced concerns that the book was not balanced and was too supportive of admissions tests. At the same time, many in attendance promoted the merits of test-optional admissions. When discussions like these arise, too often campus enrollment management and admissions professionals, along with academics and journalists, are often polarized at the extremes. They either gravitate toward a full endorsement of test-optional admissions along with a critique of standardized tests -- or they voice criticisms of test-optional admissions and focus on the benefits of standardized admissions testing.

Many chapters in Measuring Success support some of the benefits of admissions tests, but one chapter in the book, authored by Eric Maguire, provides a thoughtful case study examining how test-optional admissions was effective at his institution. Though not a part of the book, we would also note that William Hiss and Valerie Franks have documented two decades of the success of test-optional policies at Bates College. These results, along with Maguire’s case study, suggest that test-optional admissions enable colleges and universities to enroll a more diverse student body while at the same time maintaining high standards in the classroom and high graduation rates. As well, one of us wrote a chapter in the book that provides a rationale for both approaches to admissions. Ultimately the decision for test-optional admission should be made depending upon the mission of the institution alongside the implementation of evidence-based admission policy and practice.

In keeping with Jerry Lucido’s chapter, there are some very practical uses of the ACT and SAT. For underfunded public and private colleges, test scores can be a cost-effective way to make admissions decisions. For example, if an analysis of admissions decisions at an underfunded regional public university reveals that students with an SAT score of 1000 are admitted 90 percent of the time, and once admitted they graduate at rates equal to or higher than the campus average, for efficiency reasons, admissions professionals may to decide to rely heavily on these SAT scores when making admissions decisions for students in this range.

Advocates for using standardized test scores also assert that they are useful tools in making admissions decisions in a world where many high schools no longer rank students, where rankings helps differentiate what it means to have a 3.9 versus 3.4 grade point average from any given high school. In these situations, advocates of admissions testing suggest that test scores can be one more useful piece of information when making admissions decisions. Mainly, supporters of the use of admissions tests never advocate using only test scores to make admissions decisions; rather, advocates suggest that test scores can be used along with grade point averages and other factors when making decisions.

Standardized test scores can also serve an important societal function prior to a student’s matriculation. Ironically, it is common for many test-optional schools to use test scores (which they often criticize) to identify and recruit first-generation and low/moderate-income students, as well as to attract a more diverse entering class. Purchasing names from ACT and the College Board to identify low-income students is standard practice at many colleges and universities -- including test-optional campuses.

Returning to the case for test optional, we suggest that the truth with respect to the results of test-optional admissions schemes is more complicated than just being test optional. Hiss and Franks, in their study of Bates, find that test-optional practices have resulted in more ethnic diversity. However, this study is based on the results of a single institution, and not all studies have reached the same conclusion. We also have a successful case study from Maguire. On the other hand, a study by Andrew Belasco, Kelly Ochs Rosinger and James Hearn that examined enrollment patterns, using a sample of 80 private institutions rated as most competitive by Barron’s Profile of American Colleges, found that test-optional policies had no impact on the ethnic diversity of the student body.

All of this conflicting information demonstrates the complexity and nuances surrounding the use of standardized tests. We come down on the side of using admissions tests -- but as Lucido suggests, using them thoughtfully and within the constraints of institutional missions. This does not mean all colleges and universities should use admissions tests to make all admissions decisions. We want to bookend the Hiss and Franks study with the work of Hearn and colleagues, which arrive at different findings on the use of test-optional decisions. Drawing upon our own experience as enrollment managers, we offer that test-optional policies, when combined with financial aid and campus policies that create a welcoming campus environment to a diverse student body, can result in a more diverse student body. In other words, test-optional admissions policies alone are insufficient to enroll a more diverse student body. The unanswered question is how many test-optional schools are making all of these commitments?

We would be remiss, however, if we stopped here. At a more cynical level, test-optional admissions has resulted in conundrums for various institutional rankings schemes and potentially for the students, families and counselors that use them. Most students who do not submit test scores have lower scores than those who submitted them. The work of Belasco, Rosinger and Hearn, using a sample of 80 private selective institutions, found that test-optional policies had no impact on diversity but did increase average ACT/SAT scores and perceived selectivity. Thus using only the submitted scores can create the impression that average test scores are higher, possibly buoying an institution’s rankings. U.S. News & World Report has attempted to address this by using statistical techniques to give different weightings for test-optional schools.

College Niche, for example, rates test-optional colleges and universities separately from institutions that use admissions tests. These examples reveal what can be a pernicious strategy to increase institutional prestige and rankings. We suggest that the most transparent practice for test-optional schools is to follow the example of DePaul University, which requires all students to submit their test scores, with the option for students to indicate that they want their application to be reviewed without their scores. In these cases, students’ test scores are not used in the admissions review, although all test scores are reported for institutional rankings purposes.

It is clear to us that admissions test scores can positively affect the ability of institutions to identify and recruit prospective low-income students. In addition, test scores can be used by organizations to identify and mentor first-generation, low-income and often ethnically diverse students to pursue a college education, which in turn achieves both important institutional and societal goals.

Thus, in closing, we offer the following observations to those looking at test-optional admissions practices and on the broader topic of the use of standardized tests. Like many issues, it’s complicated. Test scores can be used to help increase the number of first-generation and low/moderate-income students who attend college. Test scores can also be used proactively to help national and community-based organizations target these students in high school and offer postsecondary guidance programs that will help increase socioeconomic and racial diversity.

However, at the institutional level, test-optional policies alone are insufficient to have a substantive impact on student diversity. Test-optional processes along with the use of financial aid, and efforts to make the campus culture more welcoming for a diverse student body, can work.

And to double down on our call for more research on test-optional policies, we want to add one more caveat that reflects upon how frequently institutions assess the effects of their admissions policies. In a recent study of college admissions practices, we found that two out of 10 case studies had undertaken any systematic research as to whether their admissions practices produce the desired results. In addition, two more were talking about undertaking studies. To the extent these 10 colleges are representative -- and we think they are -- it is possible that many institutions cannot document the impact of their admissions policies or the effects of any changes they may have implemented.

And given the increasing calls for higher education institutions to be accountable, what time is better than now to start?