A Rosetta Stone for Modeling Change

This chapter focuses on the intersection of three families of models for developmental data analysis: multilevel modeling (MLM), single-level structural equation modeling (SEM), and multilevel structural equation modeling (M-SEM). The chapter links notation, diagrams, and terminology across these three frameworks within a running example, and it also describes the similarities and differences in their capabilities more generally. This chapter presents a sequence of models for the analysis of longitudinal change (aka, growth curve models) that examine how time-varying hours spent video gaming predict depression in children over time. To help readers in the practical translation of these strategies to their own research, an online supplement including data, syntax, and output using R, SAS, SPSS, Stata, and M plus for all of the example models (that are available within each program) is available at the author's website: https://www.lesahoffman.com/Research/Chapter21_Supplement.zip .