Raúl Barnier B.1 y Surendra P. Sinha2
Various solutions for the analysis of the interaction variety x locality in a series of experiments are presented. A method is proposed which is equivalent to the cited procedures in the literature in this subject; based on orthogonal contrasts obtained from totals for each variety and locality.
The methods described were applied to an experiment in Potato which included 12 varieties and 4 localities.
A procedure is proposed for combining the observed means throughout all the localities and the components of interaction to adjust the observed means according to their higher or lower consistency.
1Ing. Agr. Jefe de Proyecto Estadística. Instituto de Investigaciones Agropecuarias, Casilla 5427, Santiago, Chile. Proyecto de Estadística y Diseño Experimental, Facultad de Agronomía, Universidad Católica de Valparaíso.
2Ph. D. Profesor de Estadística y Diseño Experimental, Universidad de Chile, Facultad de Agronomía. Consultor Estadística, Instituto de Investigaciones Agropecuarias.