Information and attitude diffusion in networks

The availability of diverse information does not guarantee that a person's views will be equally diverse. Research has consistently shown that attitudinal positions on an issue will lead to selective reception and dissemination of information, which in turn have reciprocal effects on those attitudes. The current paper aims at understanding how the diffusion of information and attitudes in networks are dynamically related to each other from a computational perspective. Simulations from an agent-based model show that selective exposure of information and social reinforcement in active dissemination of information can lead to polarization of attitudes in the network. Network structures are shown to have significant effects on information and attitude diffusion. While simple contagion models of information diffusion predicts that hub nodes in a small-world network can facilitate propagation of information, our model shows that hub nodes can induce stronger polarization of attitudes when information and attitude diffusion can mutually influence each other. Results highlight the importance of incorporating social science research in network models to better establish the micro-to-macro links.