Abstract

This entry describes and illustrates techniques of modeling discontinuous change with hierarchical linear models (also known as multilevel models). To this aim, six models of increasing complexity are presented in a stepwise fashion. The entry discusses discontinuous change models with phase-specific elevation shift, discontinuous change models with phase-specific elevation shift and time as continuous Level-1 predictor, discontinuous change models with phase-specific elevation shift and slopes, the introduction and centering of Level-2 predictors, and discontinuous nonlinear change models using higher order polynomials. The models are exemplified by means of real-world data extracted from the German Socioeconomic Panel. The motivating research question was to assess the impact of the event “divorce” on the life satisfaction trajectories of men and women. The parameters of the various models are interpreted in detail, and the topic of model comparison is discussed.