A Series of Factorial Row-Column Designs with Incomplete Rows and Columns

Abstract

Row-column designs are advantageously used in experimental situations when the heterogeneity in the experimental material is due to two cross classified sources. If the experimenter wants to study the effect of two or more factors simultaneously under a row-column set up, a statistically more efficient and informative method is to use a factorial row-column design (RCD), than to run separate experiments for each factor. Factorial RCDs not only need fewer experimental units but can also address the issue of interactions between the factors. As the number of factors or the number of levels of the factors increases, the number of treatment combinations increases considerably and, hence, it may be practical to adopt factorial RCDs in incomplete rows and columns. A series of symmetric factorial RCDs with incomplete rows and columns has been obtained here and the effects confounded in rows and columns have been identified.