Do Standardized Tests Predict Graduate School Performance?

Abstract

This short paper analyses whether standardised tests predict graduate school performance by empirically examining the academic performance of over 300 students at a prominent school of public management in Manila, spanning its student cohorts from 2006 to 2013. In examining student performance in this management program, this study considers most of the observable student characteristics including age, nationality, and years of work experience, while also factoring the possible influence of cohort-specific effects. It finds initial evidence to support the claim that standardizes tests (in this case the GMAT) has some predictive power on student performance. However, the empirical results show that the magnitude of this predictive power dramatically declines once other variables are considered.

Keywords: standardised test scores; graduate school performance

JEL Classification: A23; I21