Confirmatory factor analysis of the indicators of basic early math skills

Abstract

A study was conducted to analyze the factorial structure and measurement invariance of the Curriculum-Based Measurement (CBM) Indicadores de Progreso de Aprendizaje en Matematicas (IPAM [Indicators of Basic Early Math Skills]), in 2nd grade Spanish students. The model proposed is a one-factor model in which the five IPAM tasks (i.e., number comparison, missing number, single-digit computation, multi-digit computation, and place value) serve as observable indicators for a single underlying factor (i.e., number sense). The IPAM, composed of three parallel forms (i.e., A, B, and C), was administered to 252 Spanish second graders, three times throughout the school year (i.e., fall, winter, and spring). Consequently, the goodness of fit of the proposed model was analyzed for each measurement time. Furthermore, longitudinal measurement invariance was explored to analyze whether the measurement model would remain stable throughout the three-time points of measurement. Discriminant, predictive, and concurrent validity were also tested. The results support that the number sense latent factor explains the common variance of the observable indicators throughout the school year. Each latent factor was found highly related to the next. Moreover, the IPAM showed adequate indices for discriminant, concurrent and predictive validity. We conclude that the IPAM is an appropriate measure to assess number sense competence in second grade.