Hypothesized Interdependence, Assumed Independence

Abstract
Many theories hypothesize interdependence of observations among individuals within organizations, yet researchers frequently assume complete independence of observations in statistical models applied to their data. The resultant interpretational errors frequently appear in the organizational sciences literature. This paper illustrates the magnitude and direction of bias because of inconsistency of assumptions of interdependence. Both theorists and researchers are encouraged to state their assumptions of interdependence with mathematical models.