Behavioral Signatures of Cognitive Noise

The cognitive noise hypothesis is an increasingly popular explanation for a wide range of puzzling regularities in human behavior. Using simple, familiar tools and minimal assumptions drawn from neuroscience, cognitive noise models provide a parsimonious, unified explanation for a number of the most important anomalies documented in behavioral economics. The hypothesis proposes that limited cognitive resources force the brain to represent information imprecisely, generating coarser decisions, stochastic behavior and subjective uncertainty. Because brains deal with this kind of imprecision in a broadly Bayesian manner, imprecision leads to a number of well-known behavioral biases that may in fact be optimal given that imprecision. Because brains also allocate scarce cognitive resources adaptively, these biases and their severity are impacted by seemingly-irrelevant aspects of the choice environment, providing an explanation for a number of well-known context and experience effects that are otherwise difficult to explain. We review the basic structure of these models, their behavioral consequences, and the growing body of empirical evidence supporting them.