Adolescence is widely considered to be a period of increased risk of addiction. Developmental differences in reward-related learning may contribute to this heightened vulnerability. The transition to drug addiction is thought to involve an imbalance between two dissociable instrumental learning systems. A goal-directed learning system represents the consequences of potential actions, enabling flexible selection of a response likely to obtain a desired outcome. In contrast, a habitual learning system forms strong links between previously rewarded actions and the cues or contexts with which they were associated. Compulsive drug use is thought to stem in part from an overreliance on habitual learning. Despite the hypothesized role of action selection biases in the etiology of addiction, there has been little study of goal-directed and habitual learning prior to adulthood, when addictive disorders typically originate. We have recently begun to examine developmental changes in action selection biases using a reinforcement-learning task that can distinguish the extent to which goal-directed or habitual cognitive representations guide one's choices. We find that whereas habitual learning is apparent from childhood onwards, evidence of goal-directed learning only emerges during adolescence. This suggests that during early adolescence when goal-directed learning processes may still be maturing, or for individuals in whom this maturation is delayed or disrupted, a propensity toward habit-based action may heighten vulnerability to addiction. Using computational modeling and functional and structural neuroimaging, the proposed research will 1) examine maturational changes in the neural systems supporting goal-directed and habitual learning and 2) clarify whether individual differences known to modulate addiction risk also bias the reliance upon these learning processes. This project will begin to elucidate the neural and cognitive maturational trajectory of basic instrumental learning processes that may increase vulnerability to drug addiction early in development.

Public Health Relevance

Drug use at an early age markedly increases one's risk of addiction. Our recent work suggests that individuals exhibit greater reliance upon habitual over goal-directed learning early in development, which might facilitate the transition from drug experimentation to compulsive use. This project will elucidate the neurocognitive mechanisms underlying this learning bias, examining how developmental changes in brain circuitry, as well as affective and cognitive individual differences, influence the recruitment of habitual versus goal-directed learning.

National Institute of Health (NIH)
National Institute on Drug Abuse (NIDA)
Small Research Grants (R03)
Project #
Application #
Study Section
Special Emphasis Panel (ZDA1-MXL-F (03))
Program Officer
Sirocco, Karen
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Weill Medical College of Cornell University
Schools of Medicine
New York
United States
Zip Code
Potter, Tracey C S; Bryce, Nessa V; Hartley, Catherine A (2017) Cognitive components underpinning the development of model-based learning. Dev Cogn Neurosci 25:272-280
Decker, Johannes H; Otto, A Ross; Daw, Nathaniel D et al. (2016) From Creatures of Habit to Goal-Directed Learners: Tracking the Developmental Emergence of Model-Based Reinforcement Learning. Psychol Sci 27:848-58
Hartley, Catherine A; Somerville, Leah H (2015) The neuroscience of adolescent decision-making. Curr Opin Behav Sci 5:108-115