Due to the asymptotic nature of the sandwich estimators, simulations were conducted to assess their performance under varying small sample and rare event conditions. The sandwich estimators compared included the traditional Liang-Zeger (VLZ), Mancl and DeRouen (VMD), Pan (VP), Morel (VM _ T), a version of Morel inflated by the determinant rather than the trace, and the Rogers (VR) sandwich estimators. A model with one continuous covariate was used for simulation study. The number of clusters, prevalence, and correlation of the outcome variable were varied.
