Domain general and specific contributions to algebra: A sequenced longitudinal path model

Abstract

Algebra I is critical for future STEM aspirations. However, current understanding of key predictors has focused on large-scale big data studies (which often lack a strong set of potential cognitive and mathematical predictors) or on individual differences or experimental studies (with relatively small sample sizes). The present study was designed with a strong set of theoretically and empirically motivated mathematical and cognitive predictors of algebra performance and large sample size: 532 students taking Algebra I in Grades 8 or 9. Students were evaluated at the beginning and end of their course. Results revealed direct and indirect effects for an array of domain-general and domain-specific factors, with the most robust effects for language, whole number arithmetic, arithmetic concept understanding, and fractions competency. General patterns held in the context of a variety of demographic covariates though more generally for students taking Algebra I at the standard time relative to those taking it earlier. To prepare students for algebra, and while considering a host of contextual factors, results support the need to build direct and indirect supports for domain-general resources at younger ages, and to provide practice with domain-specific skills.