ABSTRACT
Change happens for people in places, and it is useful to consider how person-level and environmental characteristics may influence that change. Multilevel models (MLMs) offer an advantageous statistical modeling framework with which to examine change in an outcome, especially when considering both between-person and between-context influences on change. In this chapter, we offer a brief overview of MLM, discussing a number of advantages that the framework offers. We then demonstrate two useful MLMs for modeling change. First, we demonstrate the use of MLM to model trajectories for change over multiple measurement occasions via latent growth models. Second, we use an application of dyad-discrepancy MLM to model a difference in scores when assessing change over two measurement occasions. For both models, we discuss person-level and context-level variance and covariates. We close with recommendations for further readings.
