Network models of the diffusion of innovations.
Innovations, such as ideas, products, or opinions, spread or diffuse through society at a rate and specificity which can be understood by analyzing the pattern of communication (the social network) which exists between individuals in a social system. Analysis of such network models of diffusion reveals tipping points in the process that are studied through threshold models, which focus on individuals, and critical mass models, which describe social systems. Together, these models provide a comprehensive picture of how social systems determine social change. This book opens with an introduction which reviews the theory of diffusion of innovations, network analysis, and the three diffusion datasets used as examples. The concept of "contagion," or the specific process of innovation diffusion (also known as the "diffusion effect"), is defined. Chapter 2 provides a framework for understanding threshold and critical mass models by describing prior research on their effects. Chapter 3 describes relational diffusion network models, which maintain that individuals adopt innovations based on their direct relations with others in their social system. Structural diffusion network models, presented in chapter 4, hold that individuals adopt innovations based on their position in their social system, regardless of their direct ties to others. Chapter 5 covers threshold models of diffusion and introduces the notion that individuals may be innovative with respect to their personal network as well as to the social system. Chapter 6 deals with critical mass models of diffusion and points out that competing definitions of critical mass and a lack of clarity in critical mass research has hindered the theoretical development of these models. This chapter tests alternative models and shows how centralness and radiality of personal networks contribute to the critical mass. Chapter 7 develops a network threshold model which can be conceptualized in both relational and structural terms and which allows individual innovativeness to be measured relative to an individual's personal network or relative to a whole social system. This model can be used to predict diffusion, identify opinion leaders, understand the two-step flow model of opinion formulation, and determine the critical mass. Chapter 8 discusses other possible methods which are useful for understanding network models. The final chapter discusses applications, contrasts network thresholds with the classic diffusion model, and concludes that network characteristics are associated with adoption behavior at both the individual and the system level of analysis. The shortcomings of the modeling systems presented in this book and the limitations of the research are discussed, and indications for future research are given.