missing data

Pooling clubSandwich results across multiple imputations

A colleague recently asked me about how to apply cluster-robust hypothesis tests and confidence intervals, as calculated with the clubSandwich package, when dealing with multiply imputed datasets. Standard methods (i.