公立大学系统中的标准化考试成绩与学业表现

Recent studies of Ivy-Plus institutions suggest that standardized test scores (SAT/ACT) are better predictors of college success than high school grade point average (HSGPA). In response, many elite colleges have re-instituted the requirement that test scores be submitted for admission. We ask whether re-establishing the SAT requirement for admission at a large urban public university system (hereafter the System) would improve the predictability of academic outcomes using administrative data for the 2010–2019 first-year cohorts. We also assess the use of a state-wide exam of English Language Arts (ELA) taken by students in their junior year of high school as a substitute for the SAT. Unobserved variables that determine who is admitted and where applicants matriculate can confound estimates of the effects of SAT and HSGPA on academic outcomes. While selection bias remains unavoidable, we estimate effects for subsamples of students whose choice of colleges outside the System, and campuses within the System, are likely limited. We also apply recent sensitivity tests that assess how strong a confounder would have to be to eliminate the effect of HSGPA or SAT on the outcomes. We find that a one-standard-deviation increase in HSGPA is at least six times more predictive of six-year graduation than a comparable increase in SAT or the ELA. The estimated effect of HSGPA is robust to the inclusion of student characteristics, whereas the SAT and ELA are not. The adoption of a test-optional policy should not undermine the System’s ability to assess applicants’ potential for success in college.