ABSTRACT
In Chapters 4 through 7 of this book, we have illustrated various exploratory and inferential methods associated with antedependence models using four data sets introduced in Chapter 1: the cattle growth, 100-km race, speech recognition, and fruit fly mortality data. However, due to the specific focus on a particular method in each instance, the analysis of each data set was presented in a rather piecemeal fashion. Moreover, for obvious reasons only antedependence models were considered and we did not fit any alternatives often used for longitudinal data, such as random coefficient models or vanishing correlation models. In this chapter we present a concise summary of our previous analyses of each data set, adding to the analysis where it seems appropriate. Included among the supplemental analyses are fits and comparisons of alternative models. In each case we attempt to follow a coherent, data-driven approach to parametric modeling of the data’s mean and covariance structure. We begin the chapter with a description of the components of this approach, and we close it out with a discussion of what the case studies tell us about the relative merits of antedependence models and other models for longitudinal data.
