Structural Equation Modeling and the Hidden Markov Models: a comparative and integrative approach. Evidence from life course studies.

The thesis is focused on the longitudinal Structural Equation Models (SEM) and Hidden Markov Model. The main goal is to provide a comparative approach to analyze the complementarity of the two models in order to benet from the joint advantages of the each single approach paying great attention to the modeling of that latent process who could influence and explain the observed phenomena. The purpose of the thesis is to contribute to the discussion on the integration of SEMs and Hidden Markov models both developing new statistical tools for the analysis of longitudinal data and comparing empirically the two approaches.