Multiple models regarding the etiology of externalizing behaviors have been proposed, each suggesting differing patterns of interrelations between externalizing factors longitudinally. Because prior studies investigating the longitudinal relations between externalizing factors have relied upon correlated-factor models, it is unclear if the longitudinal relations between externalizing factors are due to variance shared across factors, variance unique to specific factors, or non-externalizing factor variance. The goals of this study were to determine if an S − 1 Bifactor model provided the best fit across this early childhood sample, determine if the longitudinal relations between externalizing factors remained consistent across the primary groups in this sample (i.e., two-, three-, and four/five-year-old children, Black children, White children, Hispanic children, non-Hispanic children), and examine these longitudinal relations within the context of prominent etiological models. In this study, teachers rated the externalizing behaviors of 2,390 children (mean age = 42.80 Months; SD = 11.56) in early childhood settings at two points within the same year. Results indicated that an S − 1 Bifactor model provided the best fit for the overall sample and each subgroup examined. Results also supported measurement and structural invariance across age, race (i.e., Black vs. White children, and ethnicity (i.e., Hispanic vs. non-Hispanic) separately. Contrary to our expectations, there were no significant cross-factor longitudinal relations between assessment periods. Implications for the understanding and development of externalizing behaviors across early childhood are discussed
