Tests When Errors are Correlated in a Randomized Block Design

摘要

Consider a randomized block design where the errors are correlated within a block but are independent from block to block. The theory for testing the significance for the treatment effects was done by Box (1954) and Geisser and Greenhouse (1958), and a partial solution was given by Gray ill (1954). A more general solution to this problem is now presented and several test procedures are derived. (Author)