Approaches for the utilization of multiple criteria to select a working correlation structure for use within generalized estimating equations

Abstract

Generalized estimating equations (GEE) incorporate a working correlation structure that is important because the more accurately this structure reflects the true structure, the more efficiently regression parameters may be estimated. Numerous criteria have therefore been proposed to select a working structure, although no criterion will always work better than all other criteria. In practice, it will be unknown which criterion will work best. Therefore, in this manuscript we propose how to utilize information from multiple criteria. We demonstrate the benefits of our proposed approach via a simulation study in a variety of settings and then in an application example.

Keywords:

• Correlation structure
• covariance matrix
• efficiency
• empirical longitudinal data

Acknowledgements

We would like to thank Dr. Richard J. Kryscio, Dr. Frederick A. Schmitt, and Dr. Erin Abner for allowing us to use data from the PREADViSE trial, which was supported by a grant from the National Institute on Aging (R01 AG019241). We also thank two anonymous reviewers for their helpful comments that improved this manuscript.