A-efficient block designs for multiple parallel line assays

Abstract

A-optimality aspects of block designs for multiple parallel line assays for comparing odd number of test preparations with a single standard preparation have been studied. A general method of construction of A-optimal/efficient block designs for three major contrasts of interests, namley preparation, combined regression and parallelism contrasts have been obtained. A catalogue of 58 A-efficient block designs is also provided for comparing three test preparations with one standard preparation with ( ) where m is the number of doses of each preparation, b the number of blocks and k is block size along with the lower bounds to their A-efficiency. A-Optimality of block designs for multiple parallel line assays that allow estimation of three contrasts of major importance but do not necessarily allow the estimability of other treatment contrasts has also been studied and a method to obtain such designs has also been developed. A catalogue of 23 A-optimal block designs for ( ) has been prepared for one standard and three test preparatioins.