Does Cascade Behavior Reflect Bayesian Updating? –An Experimental Study--

We examine the explanatory power of cascade models by implementing the BDM-mechanism in a simple cascade experiment in which subject have to decide on the prediction of a randomly chosen urn. Assigned price limits to participate in the prediction game are used as indicators of subjective probabilities. We are thus able to test the explanatory power of the standard BHW model (Bickchandani, Hirshleifer and Welch, 1992) in comparison to cascade models which incorporate individual decision errors. Focusing on stated price limits corresponding to urn predictions in line with both, standard and error BHW models our data indicate that the inclusion of errors does not significantly improve the explanatory power of the standard approach.