Events and Meetings of Italian Statistical Society, Advances in Latent Variables - Methods, Models and Applications

Font Size: 
Alternative solutions to the initial conditions problem in dynamic binary panel data models with time-dependent unobserved heterogeneity
Antonello Maruotti

Last modified: 2013-06-16


We illustrate a class of models for the analysis of longitudinal data suffering from an initial conditions problem in a latent class framework. A discrete latent process, following a time-dependent process, is introduced. This process accounts for unobserved heterogeneity between individuals in a dynamic fashion and for dependence between the observed process and the initial data mechanism. To deal with the topic we introduce two (conditional to initial conditions) models. To estimate the resulting models, we rely on the conditional maximum likelihood approach and for this aim we outline an EM algorithm.

Full Text: PDF