Using group mean centering for computing adjusted means by site in a randomized experimental design - The case of California's work pays demonstration project

When analyzing data from a randomized experiment that is replicated across multiple sites and includes covariates, the covariates can adjust for diffe rences from either the grand mean or the group (site) mean. The analysis st rategy determines the reference point, Pooling the sites and using a standa rd analysis of covariance (ANCOVA) adjusts for differences around the grand mean, whereas analyzing each site separately adjusts for differences aroun d each group (site) mean. This article demonstrates that group mean centeri ng permits pooling data from multiple sites into a single analysis while st ill using the group mean as a reference point for evaluating the covariate.