ABSTRACT
Chapter 20 uses the example data set to illustrate key features and technical details for inverse probability weighted analyses with generalized estimating equations (wGEEs). Illustrations include weightings the subject level, with one weight per subject reflecting the inverse probability of observing the dropout pattern, and weighting at the observation level, with one weight per subject per visit that reflects the changing probability of dropout as outcomes evolve over time.
