Abstract
We discuss the differences between several intensive longitudinal data models. The dynamic structural equation model (DSEM), the residual dynamic structural equation model (RDSEM) and the repeated measures longitudinal model are compared in several simulation studies. We show that the DIC can be used to select the correct modeling framework. We discuss the consequences of incomplete or incorrect modeling for the predictors in multilevel time series models. We also illustrate the advantages of the Bayesian estimation over the REML estimation for models with categorical data, subject-specific autocorrelations, and subject-specific residual variances. Dynamic factor analysis models are discussed where autoregressive relations occur not only for the factors but also for the residuals of the measurement variables. The models are also illustrated with an empirical example.
Notes
1 The RDSEM estimation can treat the initial conditions as missing variables as well using the option InitialCondition = sample.
2 The Mplus scripts are available at statmodel.com/download/rdsem.zip.