ABSTRACT
In many plant improvement programs, a considerable number of genotypes are grown in a wide range of environments. A crucial matter of strategy is whether the data-analytic procedure is thought of as a ‘meat-grinder’ that accepts 2-way and/or 3-way tables without any additional information, or whether this procedure looks for additional information, even if incomplete, which may illuminate the investigator’s logic in collecting the data, and can be useful in structuring the analysis. Comparison of the patterns of genotype response across environments is more useful than the traditional comparison of individual responses. In such experiments, the outcomes will be described by attributes measured on each genotype within a particular environment, producing a two-way genotype by attribute matrix. Alternatively, they can be characterized by the performance values for a single attribute measured in a number of environments, producing a genotype by environment matrix.
