Abstract
In this work, I extend the normal form cognitive hierarchy model (Camerer et al. (2004)) to a class ofnite two-person extensive form games. I study two versions of such a model: therst is as faithful as possible to the normal form assumptions, while the second modies them slightly. In both cases I consider the possibility of bayesian updating. I apply the two versions to the centipede game, testing empiri- cally the relevant predictions. The results show that the cognitive hierarchy model outperforms backward induction in predicting experimental behavior. Moreover,
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have