ABSTRACT
Lattice Square designs are two-way designs where both rows and columns are incomplete. Lattice Squares are a useful class of two-way designs because they have sufficient flexibility, have incomplete rows and columns, and eliminate two sources of variation, providing more efficient and economical designs for estimating treatment comparisons and thereby enhancing the quality of an experiment. In a tuberculin assay, a Lattice Square design can be used by making the rows correspond to the sites on the skins of the guinea pigs and the columns correspond to the days on which the experiments were performed. Complex experimental designs are important in many scientific investigations, such as bioassays for determining the relative potency of a drug. However, it can so happen that litters are too small or the “plates” are unable to take all the doses to be tested, and then an incomplete two-way design is useful even if the analysis is a little more complicated.
