Cognitive Processes In Choice And Decision Behavior PDF Download

A fundamental question in decision neuroscience is how cognitive processes underlying decision-making give rise to efficient decisions, especially in complex situations with multiple choice options, and each option consisting of multiple attributes such as outcome, cost, and uncertainty. These cognitive processes include attention, learning, and memory. While much is known about each process individually, less is known about how they interact: that is, the underlying cognitive dynamics. The present thesis aims to provide an integrated framework for understanding the cognitive dynamics at the behavioral, computational, and neural levels. Behavioral studies presented in this thesis investigated the influences of attention, memory, and reward on decision-making by quantitatively characterizing dynamic modulations of two critical components of choice behavior: bias (i.e., the skew of the choice distribution across options) and persistence (i.e. repeated choices). These studies provide evidence that attentional and motivational control guide multi-attribute decisions by selecting attributes to maximize payoff or minimize effort, and by generating robust persistent actions. In addition, memory related to recollection and familiarity influences choice bias and persistence, respectively. A computational model that accounts for attentional and motivational control based on the multi-attribute choice context is then presented. The model captured behavioral data, and suggests that (a) choices are selectively biased toward more distinct attributes; and (b) persistence results from memory-dependent learning rates. Finally, using single-unit recordings, this thesis directly explored neuronal representations of subjective utility during decisions under uncertainty in the lateral prefrontal cortex, a high-level cognition area of the brain. This study provides evidence that lateral prefrontal neurons code expected value and uncertainty of choices. By combining approaches from psychology, economics, and computational and systems neuroscience, this thesis provides a significantly improved characterization of the cognitive dynamics that, when in harmony, contribute to better decisions.