A two-stage approach to recovery of inter-block information and shrinkage of block effect estimates

Abstract

Linear mixed models often comprise several effects, and the focus is usually only on one or a few of them, while the other effects need to be fitted merely to adjust for all sources of variation. A typical example is the analysis of a blocked experiment, Show more