GMM Estimator Covariance Structure for Time-Dependent Covariates with Unbalanced Replication

• Activity Number: 482
• Type: Contributed
• Date/Time: Wednesday, August 7, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Trent L. Lalonde*+
• Companies: University of Northern Colorado
• Keywords: Generalized Linear Model ; Longitudinal Data ; Generalized Method of Moments ; Unbalanced

Abstract

When analyzing longitudinal data it is essential to model both the correlation inherent from the repeated measures of the responses as well as the correlation among varying values of predictors. The Generalized Method of Moments (GMM) for the coefficients in longitudinal data provides substantial gains in efficiency over generalized estimating equations (GEE) with the independent working correlation. It is common to have misbalance in longitudinal data sets, i. e. an unequal number of repeated observations for different subjects. However, current analysis techniques suggest the estimator covariance structure should be estimated using only complete-case subjects.

This paper presents a method of analyzing unbalanced longitudinal data such that all subjects are accounted for in the estimator covariance structure. This paper presents examples and programs intended for binary and for count response data.