Exploratory Data Analysis
Most books on longitudinal data discuss exploratory analysis. See, for example, Diggle, Liang, and Zeger (1994). However, most effort is spent to model building and formal aspects of inference. In this section, we present a selected set of techniques to underpin the model building. We distinguish between two modes of display. In Section 4.2, the marginal distribution of the responses in the Vorozole study is explored, that is, we explore the observed profiles averaged over (sub)populations. Three aspects of the data will be looked at in turn: the average evolution, the variance function, and the correlation structure. Afterward, in Section 4.3, we will discuss some procedures for exploring the observed profiles in a subject-specific way.
KeywordsData Analysis Stochastic Process Probability Theory Statistical Theory Longitudinal Data
Unable to display preview. Download preview PDF.