POPULATION-AVERAGED GENERALIZED ESTIMATING EQUATION MODELS (GEEs)
Zeger and Liang [1986] describe a class of estimators that address correlated panel data. The user must specify both a generalized linear model specification valid for independent data and the correlation structure of the panel data.
Although fixed-effects estimators and random-effects estimators are referred to as subject-specific estimators, the GEEs available through PROC GENMOD in SAS or xtgee in Stata, are called population-averaged estimators. This label refers to the interpretation of the fitted regression coefficients. Subject-specific estimators are interpreted in terms of an effect for a given panel, whereas population-averaged estimators are interpreted in terms of an affect averaged over panels. When and whether to draw inference for average sampling units is considered in the next section.
The average human has one breast and one testicle.—Des McHale
Get Common Errors in Statistics (and How to Avoid Them), 4th Edition now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
