Kwangjun Lee: Hierarchical Drift-Diffusion Model on reward-based two-choice decision making process under perceptual and economic uncertainty

BCCN Berlin / TU Berlin

In the naturalistic and ever-changing environment, many decisions are made on the basis of noisy sensory evidence (perceptual uncertainty) and uncertain outcomes (economic uncertainty). Despite growing interests in decision making under uncertainties across a wide range of disciplines, the two types of uncertainty have been usually investigated separately. In a novel attempt to develop an integrated computational framework for human choice behaviors that incorporates the two types of uncertainty, we designed a simple two-alternative forced choice (2AFC) reward-based perceptual discrimination experiment with the well-known random dot motion (RDM) stimulus. Behavioral data collected across task conditions of with varying stimulus coherence and the reward probabilities were fitted to drift-diffusion model (DDM) using hierarchical Bayesian methods. The model fit was evaluated by both deviance information criterion values (DIC) and posterior predictive checks (PPC). Our results suggest that the hierarchical DDM adequately accounts for the patterns observed in choice proportions and reaction time (RT) distributions across the different task conditions. Moreover, we show that perceptual and/or economic uncertainty induce dynamic changes on the speed of decision-related information processing, response caution, a priori bias, and non-decision time during the decision making process.

Additional Information

Master thesis defense in the International Master Program Computational Neuroscience.

Organized by

Klaus Obermayer / Robert Martin