ANALYZING INCOMPLETE LONGITUDINAL BINARY RESPONSES - A LIKELIHOOD-BASED APPROACH

In this paper, we describe a likelihood-based method for analysing bal anced but incomplete longitudinal binary responses that are assumed to be missing at random. Following the approach outlined in Zhao and Pre ntice (1990, Biometrika 77, 642-648), we focus on ''marginal models'' in which the marginal expectation of the response variable is related to a set of covariates. The association between binary responses is mo delled in terms of conditional log odds-ratios. We describe a set of s coring equations for jointly estimating both the marginal parameters a nd the conditional association parameters. An outline of the EM algori thm used to obtain the maximum likelihood estimates is presented. This approach yields valid and efficient estimates when the responses are missing at random, but not necessarily missing completely at random. A n example, using data from the Muscatine Coronary Risk Factor Study, i s presented to illustrate this methodology.