This article considers the estimation of power and sample size in experimental and quasi-experimental intervention studies, where there is clustering of subjects within one or both intervention arms, for both continuous and binary outcomes. A new command, clsampsi, which has a wide range of...