Department of Statistics, University of Waikato

2003-06-16

Room I1.01 (I Block, 1st floor), University of Waikato

Finding efficient resolvable row-column designs can be computationally expensive, especially for designs with a large number of treatments. Bailey and Patterson (1991) show that two-replicate resolvable row-column designs are combinatorially equivalent to a single-replicate row-column design for two factors. This two-factor design is called the contraction. In this talk we shall explore the relationship between a resolvable row-column with more than two-replicates and its contraction. In particular, we shall show that an optimal contraction leads to an optimal row-column design.