Biased learning in a network setting

This program of research investigates the impacts of important and well-documented departures from perfect rationality on learning and information aggregation in a network setting. Existing work has gathered empirical evidence of these biases, and has built models to explain their effects on agents who learn from some fixed (external) source. This research extends analysis to settings where there is more than one agent (person) in a society, and these agents can learn from each other.