Organizationally-relevant configurations: the value of modeling local dependence

Configurations are important across all levels of organizations. Despite the interest in and importance of configurations in research for organizations, the empirical methods for assessing and classifying configurations has not kept pace with the theoretical advancements. Theory suggests that configurations must include aligned elements that have local dependence. Local dependence is defined as the interrelationships among variables necessary to form an internally consistent configuration. We explain how latent class cluster analysis (LCCA) enables modeling for local dependence and provides theoretical and methodological value for configurations researchers. Using primary data from two samples, we demonstrate that LCCA with local dependence outperforms traditional cluster analysis-based approaches. Our method can be used for detecting configurations at a variety of organizational levels (e.g. nation, industry, firm, and group). Copyright Springer Science+Business Media B.V. 2013