In 2003, in The Journal of Applied Behavioral Science, Cathleen McGrath and David Krackhardt published a study comparing how three alternative, and to some extent conflicting, models of network structure would support organizational change.
The density model is based on the notion that strong ties between members of different departments within an organization will facilitate acceptance of change. The existence of these sorts of ties is often difficult to achieve, in part because people tend to associate more with those in close proximity, and in a work environment there is usually a high level of task dependence among people in the same unit. A series of experiments was run to test the impact of strong cross-unit ties and the results were in agreement with what the model predicted.
The viscosity model takes an opposite starting point from the density model, suggesting that the implementation of change will be facilitated more when subunits within an organization are relatively isolated from one another. One of the assumptions behind this model is the notion that one department or group within an organization must institute and accept a change before that change can be diffused through the organization as a whole. Computer simulations were used to test this model and the findings suggested that change was more likely to be accepted throughout an organization, if it was first adopted by a relatively isolated or peripheral unit, in part because controversial activity within such a unit was less likely to generate broad opposition or backlash.
The structural leverage model is rooted more directly in the mathematics of networks and builds on the sociological phenomenon known as the friendship paradox. According to this paradox, our friends always appear to have more friends than we do. The question then becomes one of how to select a limited number of individuals within an organization through which to diffuse change. The paradox suggests that you randomly select an initial set of well-connected individuals and then ask them to nominate one of their friends. It is members of this second group that take on the role of change agent. While no direct evidence was collected to support this model, the mathematics suggests that diffusion of change will take place almost twice as fast through secondary contacts.
In discussing their findings, the authors conclude that the three models should be viewed as complementary, because they are based on different assumptions about how networks actually work. So, for managers charged with implementing change, their knowledge of the relationships among individuals within their organization will help them to assess what sort of network structure exists and therefore which method would likely give the best results.
The application of social network analysis in the study of organizations, as well as in other settings, is still in its infancy. The article discussed here is almost a decade old and, while tremendous advances have been made in our understanding of network dynamics, the complexity and uncertainty associated with social network research that this article demonstrates has, if anything, increased in the intervening years.
Take up the challenge. There are countless opportunities to use social network analysis in organizational studies as a way to make your mark as a researcher and become an authority in the field.