ABSTRACT
This chapter provides the role of time-dependent input variables for Latent Markov (LM) models. The LM models are closely related to latent class models. Jeroen Vermunt, R. Langeheine and U. Bockenholt and Vermunt and J. Magidson presented a logit regression approach that allows to regress the latent states occupied at the various points in time on time-varying covariates. Once the equation modeling estimates of the model parameters have been computed, their variances may be found from the information matrix. This is the inverse of the matrix of second order derivatives of the loglikelihood function toward all independent parameters. The research goal and the dependency between the repeated measures are different for approach and that of the growth model approach. The loglikelihood ratio statistic is asymptotically chi-square distributed with degrees of freedom equal to the number of different response patterns of the observed input and output variables, minus the number of independent model parameters.
