This paper investigates opinion dynamics and social influence in directed communication networks. We study the properties of a generalized boundedly rational model of opinion formation in which individuals aggregate the information they receive by using weights that are a function of their neighbors' degree. We then present an experiment designed to test the predictions of the model. We find that both Bayesian updating and boundedly rational updating a la DeMarzo et al. (2003) are rejected by the data. Consistent with our theoretical predictions, the social influence of an agent is positively and significantly a ected by the number of individuals she listens to. When forming their opinions, agents do take into account the structure of the communication network, although in a sub-optimal way.
Authors: Pietro Battiston and Luca Stanca
http://dipeco.economia.unimib.it/repec/pdf/mibwpaper267.pdf
No comments:
Post a Comment