ABSTRACT
The obviously prohibitive cost of dealing with every unit of the farm population in agricultural economic investigation and planning has stimulated the search for a reliable basis for generalization. Since work is carried out within obvious natural differentials of climate and soil, a great deal of confusion has arisen from the general-purpose nature of usual data collections. American workers in supply-response research have made important contributions on the necessary conditions for grouping. Other workers have made comparisons to show that homogeneity in the limiting resource, at least over some range of price possibilities, is the only necessary condition to minimize aggregation bias, when compared with other traditional criteria for classification. Tribal background differentiates characteristics secondary to natural conditions which might have created peculiarities in the system of agriculture. Usually, because the tribe has evolved as a community, its members have common objectives.
