Decomposing High-School GPA as a Predictor of First-Year GPA

Abstract

ABSTRACT Despite lack of standardization, high school grade-point average (HSGPA) is frequently the best single predictor of first-year postsecondary GPA (FYGPA). This study evaluated the extent to which HSGPA reflected standardized test scores, social-emotional learning (SEL) competencies, demographics, and high-school characteristics. Dominance analysis indicated that the variance in HSGPA accounted for by the available predictors reflected mainly academic achievement, followed by SEL, student demographics, and school characteristics. In structural equation models to predict FYGPA, HSGPA accounted for some of the predictive power of achievement test scores, SEL, demographics, and high-school characteristics. Overall, results were consistent with the interpretation of HSGPA as a measure of much more than knowledge and skills learned in the classroom. The predictive value of HSGPA likely arises from similarity in the degrees to which HSGPA and FYGPA reflect a combination of academic achievement, SEL, and other personal and school characteristics.