Trend vector models for the analysis of change in continuous time for multiple groups

A problem with the modeling of repeated multinomial response data is the dimensionality of the response variable. For reducing this dimensionality and enhancing interpretability multidimensional scaling techniques are utilized. The resulting trend vector model provides an easily interpretable graphical display with trajectories of different groups over time. A generalized estimating equations scheme is employed for obtaining estimates of the parameters. Model selection is based on the Bayesian Information Criterion and the bootstrap. For illustration, the model is applied to a data set.