This project studies the effects of network topology on the spread of rumor and makes three major intellectual contributions. First, it provides a theoretical framework for explaining the relationship between networks and the spread of rumor. Second, it provides an empirical investigation of the diffusion of rumor through networks, including a comparison with the transmission of valid information. Finally, this project offers an empirical test of hypotheses generated by leading theoretical research on network models of social contagion. The project applies network models of cascades and social contagion to analyzing the spread of rumor. These models differentiate between simple and complex contagions, and predict that simple contagions will spread regardless of network topology while complex contagions will spread less effectively on small-world and scale-free networks than on cohesive graphs. Valid information, which an individual is likely to pass along even after hearing it from one source, is an example of a simple contagion, while rumor, which an individual will be reluctant to pass along without some corroboration, is an example of a complex contagion.
The project will use Netscan, a vast database of messages collected from Usenet, an Internet-based discussion tool, to empirically examine the spread of rumor and network models of social contagion. By providing both a detailed, systematic description of the spread of rumor and direct empirical tests of the hypotheses generated by the theoretical models, this study will advance our understanding of social contagion in general, and rumor transmission in particular.
Broader Impact: The results of this study will have implications on research in areas as diverse as computer science and organizational communication. Computer security experts, corporate executives, and political leaders all contend with separating rumor from verified information, and would welcome a systematic comparison of their diffusion processes. This research will offer opportunities for undergraduates and members of underrepresented minorities to get involved in academic research, providing cross-disciplinary research experience for students of sociology and information science.
