Test Scores, Class Rank and College Performance: Lessons for Broadening Access and Promoting Success.
ABSTRACT Using administrative data for five Texas universities that differ in selectivity, this study evaluates the relative influence of two key indicators for college success-high school class rank and standardized tests. Empirical results show that class rank is the superior predictor of college performance and that test score advantages do not insulate lower ranked students from academic underperformance. Using the UT-Austin campus as a test case, we conduct a simulation to evaluate the consequences of capping students admitted automatically using both achievement metrics. We find that using class rank to cap the number of students eligible for automatic admission would have roughly uniform impacts across high schools, but imposing a minimum test score threshold on all students would have highly unequal consequences by greatly reduce the admission eligibility of the highest performing students who attend poor high schools while not jeopardizing admissibility of students who attend affluent high schools. We discuss the implications of the Texas admissions experiment for higher education in Europe.
- [Show abstract] [Hide abstract]
ABSTRACT: UT-Austin administrative data between 1990 and 2003 are used to evaluate claims that students granted automatic admission based on top 10% class rank underperform academically relative to lower ranked students who graduate from highly competitive high schools. Compared with white students ranked at or below the third decile, top 10% black and Hispanic enrollees arrive with lower average standardized test scores, yet consistently performed as well or better in grades, first year persistence, and four-year graduation likelihood. A similar story obtains for top 10% graduates from Longhorn high schools verses lower-ranked students who graduated from highly competitive feeder high schools. Multivariate results reveal that high school attended rather than test scores is largely responsible for racial differences in college performance.Educational Evaluation and Policy Analysis 03/2010; 32(1):44-69. · 1.38 Impact Factor