Missing Data and Complex Survey Designs

In this chaptermissingness mechanisms at two levels; what maximum likelihood does with missing outcomes; missing predictors and the congeniality of imputation and analysis models; multilevel multiple imputation by joint modeling, chained equations, and substantive-model-compatible methods, with a TIMSS example and a simulation; sensitivity to data missing not at random; complex survey designs; sampling weights in multilevel models, with a simulation; plausible values; reporting.