PROBLEMS OF GENERALIZATION IN CROSS-NATIONAL STUDIES OF ORGANIZATIONS

This article addresses the issue of how to generalise the results of case studies, which are the main instrument of most cross-national studies on organisations. Drawing on his own studies and those of others, the author reconstructs three ways of dealing with this problem: an ‘experimental’ procedure resting on the classical approach of comparing ‘matched’ pairs; a ‘holistic’, usually interpretive procedure based on the ‘cultural’ or ‘systemic’ reconstruction of social totalities and patterns of interdependence; and, lastly, an ideal-type procedure that is grounded in the construction of society-plant constellations that can stimulate an inductive-exploratory way of developing and verifying more general statements. The third procedure is partially illustrated by a Franco-German study previously conducted by the author. © 1991, SAGE Publications. All rights reserved.