Optimal row-column designs for treatment control comparisons

Abstract

Problem of comparing a set of test treatments with a control or standard treatment arises in many applications. In the literature, there exist a number of alternative design settings available for the situation being considered. Problem is to choose one among several alternatives which is best in some sense. In this thesis, we considered two-way elimination of heterogeneity model with simultaneous confidence coefficient criterion. A procedure for making exact joint confidence statements for multiple comparisons with a control was described and some methods for the construction of the Balanced Treatment Row-Column Designs (BTRCD ’s) were given. Finally, tables of optimal BTRCD’s were provided for practical range of parameter values.