STRUCTURED ANTEDEPENDENCE MODELS FOR LONGITUDINAL DATA
Dale Zimmerman
University of Iowa
Antedependence (AD) models are a useful, though not widely known,
class of models for the covariance structure of longitudinal data.
Like stationary autoregressive models, AD models allow for serial
correlation within subjects, but they are more general in the sense
that they do not stipulate that the variance is constant over time
nor that correlations between measurements equidistant in time are
equal. Thus, AD models provide a more parsimonious approach to
the analysis of nonstationary data than the completely unstructured
classical multivariate approach.
For some nonstationary longitudinal data, however, a highly
structured AD model may be more useful than an unstructured AD
model. For example, if the variances increase over time, as is
common in growth studies, or if measurements equidistant in time
become more highly correlated as the study progresses, then a
model that incorporates these structural forms of nonstationarity
is likely to be more useful. In this talk I review AD models in
general and then I introduce some structured AD models.
Properties of the models and estimation of model parameters by
maximum likelihood are considered. A new graphical diagnostic,
the PRISM, is introduced, which can help in determining the most
appropriate order and type of AD model.