Heterogeneous Samples in Organizational Research

Abstract

The utility of heterogeneous versus homogeneous samples in comparative organizational research is appraised through a path analytic comparison of the relationships between organizational size, spatial, vertical and horizontal differentiation established with samples representing different degrees of homo-heterogeneity. A causal model is estimated on data of 115 business and government organizations. The estimated parameters of the model are compared to those of Blau and Schoenherr based on a homogeneous sample of employment agencies, and Mileti, Gillespie, and Haas's sample of mixed government agencies. The findings suggest the potential utility of increasing our use of heterogeneous samples in comparative organizational research.