Standardized Testing and College Admissions

The Issue:

Standardized testing has been widely used by colleges as an input to admissions decisions in the United States since the 1950s. These tests offered a way for colleges to recognize and admit talented students who they might otherwise overlook because they do not come from traditional “feeder” high schools. But the use of these tests has been extensively criticized for favoring children from wealthier families because these applicants have greater opportunity for test preparation, can afford to take the test multiple times, and, possibly, the test questions are biased in their favor. During COVID, many colleges made applicants’ submission of standardized test scores optional, or even adopted test-blind admissions. But now many top-tier private colleges have shifted back to requiring students to provide test scores. This memo looks at some recent research on whether standardized testing is a useful and fair tool in admissions decisions and discusses what this means for efforts to ensure that higher education contributes to a just and productive society.

Standardized testing can be a valuable tool in helping selective colleges identify talented students from disadvantaged backgrounds.

The Facts:

• Standardized test scores have traditionally been an important consideration in college admissions decisions. These tests were first used in the 1930s by Ivy League and other elite colleges as a way to broaden their applicant pools beyond students coming from privileged East Coast boarding schools.  An aptitude test was viewed as a way of comparing students across the country from very different income levels, social backgrounds, and educational opportunities in the absence of national curricula or nation-wide achievement testing found in some other countries, and thereby giving opportunities to talented students who might otherwise be overlooked. These tests became an important factor in admissions decisions for most selective public and private colleges in the 1960s and were used widely after that. 
• Over the years, the SAT and the ACT (another standardized test similar to the SAT) have been extensively criticized as favoring children from more privileged backgrounds. One concern was that these tests were culturally biased by favoring students from families more familiar with the language and cultural assumptions used in the tests. Over the years, the tests have been modified to try to reduce this bias. There has also been concern about the advantages for children from wealthier families who could enroll in expensive test preparation programs and could afford to take the tests multiple times (only the highest score from each section of each test is reported to colleges rather than all scores from all times the test was taken). A more fundamental issue is that the test scores of children from wealthier families benefit from much greater access to economic resources for their education. Wealthier families send their children to better funded schools, often after attending pre-K schooling not provided to poorer kids. The problem is increasing as neighborhoods have become more segregated by income level.
• Reflecting such concerns, in recent decades reliance on standardized test scores has been reduced and more attention paid to other indicators in an effort to identify talented kids from low income and disadvantaged backgrounds. Such indicators in include high school grade point average and class rank, teacher recommendations, and college essays. Some selective liberal arts colleges, mostly in the Northeast, became test-optional starting in the 1970s–for example, Maine’s Bowdoin College has been so for more than 50 years. The number of colleges that made standardized tests optional, or would not consider test scores at all, increased dramatically to two-thirds of colleges during the COVID pandemic.  According to Compass Education Group, this trend has increased further since the pandemic to 90 percent of colleges for students enrolling in 2024.
•  Most recently, some top-tier private schools, including Dartmouth, Georgetown, MIT, Harvard, and Yale, have reversed course and will again require submission of standardized test scores.  Extending the initial arguments in favor of standardized test scores, these colleges say standardized tests can be an important source of information for identifying talented but less advantaged applicants when taken together with other sources of information.  By contrast, however, the University of California system recently decided to extend their “test-blind” regime post pandemic. 
• Careful research can provide valuable evidence on whether standardized testing does offer useful information for selecting talented students from disadvantaged backgrounds and has, in fact, been used to address this issue. Scholars at Dartmouth College undertook a study of relevant predictors of college success at their institution.  Their analysis found higher tests scores were correlated with better first year college grades, much more so than high-school GPA, for students from both advantaged and disadvantaged high schools. This research was important in Dartmouth’s decision to require standardized test scores once again.  Another analysis looking at a somewhat broader group of highly selective colleges found that higher test scores were associated with better college performance while GPA was not as predictive. 
• This research based on experiences at top-tier colleges also suggests that test-optional admissions can actually have the inadvertent effect of reducing admissions of students from less privileged backgrounds. A test-optional admissions policy allows students to decide whether to submit test scores. The Dartmouth study found that less advantaged applicants who are high-achieving but have low scores compared to the average for the college (a statistic that is readily available) tend to not submit test scores. This reduces their likelihood of successfully applying to the college even though, in fact, admissions officers take into account their circumstances and look favorably at test performance that may be below the average score for the college but still impressive given the students’ background. This research also found that, after an initial surge in less advantaged applicants after dropping the testing requirement, there was not a sustained increase in the proportion of these applicants in the overall applicant pool.
• However, research based on broader high school populations beyond students applying to elite private schools is more favorable to optional test admissions policies. Studies of the Chicago  and California  high school student populations suggest that high school GPA scores provide a better predictor of college grades and graduation rates than standardized test scores after controlling for demographic factors. Studies of California and Texas public colleges that emphasize high school GPA rankings suggest that this information is useful for identifying talented students from low-income communities in the respective states.  In addition, a recent study  of 100 private institutions that shifted to test-optional admissions policies between 2005 and 2016 found that the shift in policy increased the diversity of enrollment by raising the share of Pell Grant recipients and first generation to college students from under-represented racial and ethnic backgrounds.