Some New (Bayesian) Light Through Some Old (Cognitive Modeling) Windows
Michael Lee · IMBS Colloquium · 2007
We consider four case studies from cognitive science where non-Bayesian methods for relating a cognitive model to empirical data failed to help answer important theoretical questions. These case studies involve the nature of stimulus representation, of category learning, of inductive and deductive reasoning, and of recognition memory. In each case, we show how the application of Bayesian methods, using graphical models as a formalism, usefully addresses the original theoretical question.