lmiratrix/blkvar: ATE and Treatment Variation Estimation for Blocked and Multisite RCTs

This package contains code for four primary purposes: (1) estimating average treatment effects (ATEs) in block-randomized experiments, even those with singleton treated or control units, (2) estimating ATEs for multisite trials, (3) estimating cross-site variation in site-level ATEs using a variety of methods, and (4) providing code for conducting simulations of blocked or multisite randomized trials. There is also some additional code implementing power calculation functions for these contexts. See Pashley & Miratrix, 2020 <arXiv:1710.10342> for further details on (1).

install.packages("remotes")
remotes::install_github("lmiratrix/blkvar")