The Department of Cognitive Sciences Colloquium Series presents

"The Computational and Neural Bases of Goal-Directed and Habitual Decision Strategies"
with Mimi Liljeholm, Ph.D., Postdoctoral Scholar, Computation and Neural Systems, California Institute of Technology

Wednesday, January 30, 2013
12:00 p.m.
Social & Behavioral Sciences Gateway (SBSG), Room 1517

Two distinct strategies have been proposed to support action selection in humans and other animals: 1) a goal-directed strategy that generates decisions based on deliberate evaluation of the consequences of actions, and 2) a habitual strategy that relies on a more reflexive elicitation of actions by environmental stimuli. Although considerable evidence has substantiated this theoretical distinction, and in spite of its far-reaching implications, ranging from the structuring of economic policies to the treatment of compulsive pathology, very little is still known about what factors induce the use of one strategy over the other and how the two strategies are respectively implemented by the human brain. In this talk, Liljeholm will present behavioral and functional neuroimaging (fMRI) data in support of the outcome divergence hypothesis – the notion that shifts between decision strategies are driven by the degree to which alternative actions yield distinct outcomes. She will show that striatal and cortical BOLD activity discriminates between conditions of high and low outcome divergence, predicts individual differences in habitual performance, and does so differentially across the development versus deployment of goal-directed and habitual behavioral control. Moreover, she will show that many of these effects are preserved across experiential and vicarious learning domains. Taken together, her results indicate a broad division between neural systems mediating the processing of stimulus-response and action-outcome relationships.

For further information, please contact Clara Schultheiss, 949-824-7569 or clara@uci.edu.