Search results

In spite of an official record of success, much uncertainty surrounds the mean effect of sociotechnical interventions on organizational effectiveness. A new technique, a meta-analysis, is used to estimate both this mean effect and its variation across 17 sociotechnical studies. Several contingency variables which may have moderated this mean effect are also examined. Results indicate that interventions that (1) involved the formation of autonomous instead of semi- or non-autonomous workgroups, (2) included an increase in monetary incentives, and (3) occurred in a non-U.S. country rather than in the U.S. led to greater increases in productivity. Conversely, sociotechnical interventions that also included a change in technology resulted in a smaller improvement in productivity and a smaller drop in escape behavior than interventions without a technological change. More importantly, a curvilinear relationship was uncovered between the duration of an intervention and productivity. This curvilinear relationship rejects the hypothesis that the positive outcomes of sociotechnical interventions are temporary and indicative of a "Hawthorne Effect."

AbstractOrganizational theories frequently rely on notions of sharing and dependence among organizational participants, but researchers usually focus on characteristics of the actors themselves instead of the relational patterns among the actors. Loose coupling is one conceptual tool that emphasizes relational patterns. Loose coupling, however, is an abstract metaphor that is simultaneously fertile and ambiguous. This paper develops a rigorous and comprehensive framework that sharpens the theoretical contributions of loose coupling to our understanding of structural relationships. Characteristics of loose coupling capture some important and underexplored features of multidimensional fit and interdependence in organizations. The proposed framework clarifies these theoretical contributions of loose coupling with concepts and equations modified from network analysis. Testable hypotheses are proposed with respect to three key independent variables that may affect patterns of coupling: organization strategy, technology, and environmental turbulence. Additional hypotheses are advanced with respect to the use of the multidimensional approach to loose coupling in studying new organizational forms. Initial psychometric and empirical evidence are presented.