Choosing Among Multiple Achievement Measures: Applying Multitrait–Multimethod Confirmatory Factor Analysis to State Assessment, ACT, and Student GPA Data

Abstract

Practitioners and researchers interested in understanding student achievement, its predictors, and how it relates to other student outcomes are likely unaware of how the source information about achievement may offer subtly different pictures. This study applies multitrait–multimethod (MTMM) confirmatory factor analysis (CFA) within a structural equation modeling (SEM) framework to student achievement data to demonstrate empirically how commonly used measures of student achievement may reflect different information about student performance. Using student population-level data from a single state, this study presents a robust demonstration of the similarities and differences among three commonly used achievement measures—American College Testing (ACT) scores, state test scores, and grade point average (GPA). Results show that state assessment scores and ACT scores measured a similar achievement construct, whereas student grades reflected less of the achievement construct and a higher level of method effects. Possible sources of the similarities and differences among different achievement measures are discussed, along with implications for measurement among gifted students.