In recent years, methodological advances for analyzing developmental data are coming thick and fast. Two of the most popular and rapidly developing frameworks are (i) longitudinal structural equation modeling and (ii) network modeling. The present paper outlines the incremental gain in what we can learn from data about co-developing skills and challenges when using these two frameworks in tandem. First, we discuss the proposed analytic paradigm in the context of fundamental questions in developmental psychology. Second, we present two different paths to formalize such questions, introducing, first, a recently developed network model for longitudinal panel data and, second, the notion of growth parameter networks based on latent growth curve models. Used in tandem, they can provide new insights into the longitudinal co-development of developmental domains. Specifically, we focus on integrating growth parameters from latent growth curve models into networks and analyzing them as such. Third, we illustrate these analytic steps with an empirical example using longitudinal data from the Millenium Cohort Study (N=7623). As illustrated and discussed in the real data example, the proposed approach offers a magnifying glass to the study of coupled developmental changes. Teasing apart the processes underlying the heterogeneity of childhood development can, in turn, add to substantive developmental theory.
Latent Growth Curve Models for Biomarkers of the Stress Response
