Standardized Testing, College Success, and Diversity: Planning a Predictive Education

Policymakers and educators have a vested interest in ensuring that each space at universities be filled by the most capable candidates, as spaces are limited. This is especially true at public institutions. Because the majority of funds come from the government, each pupil represents a significant investment of the state’s resources, and universities want to make sure that those investments are sound. However, predicting the performance of high school applicants in a college setting is no easy task. There are myriad methods by which schools evaluate applicants. The most common include standardized testing—such as the CollegeBoard’s SAT and the Smarter Balanced Assessment Consortium’s (SBAC) Smarter Balanced Assessment System—and high school grade point average (GPA).

In a survey of the 2014-2015 entering class at California public universities, Kurlaender, Kramer and Jackson aimed to determine which of these three criteria was most effective at predicting undergraduate performance, accounting for various socioeconomic factors. The researchers evaluated students at schools within two tiers: the California State University (CSU) system and the University of California (UC) system. Overall, researchers sampled 117,706 applicants at one of the 23 campuses of the CSU system, 36,388 of whom were accepted and decided to enroll. At UC Davis, within the University of California system, they sampled 34,439 applicants, 3,544 of whom enrolled.

The researchers chose to evaluate student performance in two ways. First, they measured academic performance through first year GPA. Second, they calculated persistence through retention rate, or the percentage of students finishing their freshman year and remaining at the school as sophomores. Within the CSU system, Kurlaender, Kramer, and Jackson found that high school GPA most reliably predicted both persistence and performance. At UC Davis, standardized testing—both the SAT and the SBAC—more strongly predicted both metrics of success.

In an effort to account for socioeconomic factors, researchers also analyzed their data along racial and economic lines. The researchers found that the overall trends of first-year GPA and first-year persistence they identified held true across all social and economic groups. However, they noted that the relationships varied across ethnic groups; for example, the relationship between high school and college performance was weaker for Latino students.

The researchers presented their findings with several caveats. First, while the SAT and SBAC measure many of the same skills, they have different formats. Most notably, the SAT can be retaken indefinitely, while the SBAC can only be taken once. Secondly, the College Board has since redesigned the SAT. The version that the sampled students took is no longer in use.

Nevertheless, educators and policymakers can draw some conclusions from these results. For one, the SBAC seems to be about the same as the SAT (or in some cases, slightly worse) at predicting college performance. Secondly, there is a stark difference as to which variable best predicts student success within CSU and within the University of California system. High school GPA best predicted success among CSU students, while standardized testing best predicted success at UC Davis. Finally, metrics to predict college success did not vary by race or socioeconomic status, though the reasons are not yet clear.

These implications should prove useful for college admissions officers hoping to strengthen their institutions, as well as for educators and policymakers striving to identify students most likely to succeed. Considering their applicability across all demographic groups, universal predictors of academic success both standardized testing and high school GPA can assist educators in creating diverse, high-performing student bodies.

Article source: Kurlaender, Michal, K.A. Kramer, and Erika Jackson. “Predicting College Success: How Do Different High School Assessments Measure Up?” Policy Analysis for California Education (2018).

Featured photo: cc/(monkeybusinessimages, photo ID: 607927560, from iStock by Getty Images)