A Neuronal Theory of Human Economic Choice

We develop and test a model that provides a unified account of the neural processes underlying behavior in the classical economic choice task. The model portrays brain processes engaged in evaluating information in the experimental stimuli. This portrayal produces a consistent account of several important features of the decision process in different environments (e.g., when the probability is specified or not): these features include the choices made, the time to decide, the error rate in choice, and the patterns of brain activation. Complex information describing two economic options is represented on the retinas of the subject as a collection of photons. A collection of photons is converted (via a sequence of neuronal processes) to a measure indicating which option has the higher utility magnitude. Data are processed by the brain until evidence sufficient to favor one option is reached. The model predicts that the further two stimuli are from each other in utility space, the faster the reaction time will be, fewer errors in choice will be made, and less brain activation will be required to make the choice; the model also predicts that choices with ambiguity can be made quicker and will require less brain activation in the horizontal intraparietal sulcus than for choices with risk. Also, we demonstrate how, ceteris paribus, with a larger certainty option in the choice, there is more brain activation, and furthermore, with less experience on the part of the subject making choices, there is more activation.