|Title||Analysis of Longitudinal Data of Repeated Observations Using Generalized Estimating Equations Methodology|
|Publication Type||Journal Article|
|Year of Publication||1998|
|Authors||Li, F, Maddalozzo, GF, Harmer, P, Duncan, TE|
|Journal||Measurement in Physical Education and Exercise Science|
|Pagination||93 - 113|
Data collected from longitudinal designs often have correlated observations arising from repeated measurements on the same subject. Conventional statistical methods (e.g., ordinary least squares linear regression, univariate analysis of variance with repeated measures) that assume independence of observations can be inadequate in providing test statistics. Liang and Zeger (1986) proposed a class of generalized estimating equations (GEE) to account for dependence among observations. The article provides a descriptive introduction to GEE and an example illustrating its use in continuous response data with a multivariate normal distribution. A discussion on the utility of the GEE procedure in longitudinal data analysis is provided.