Keynotes | K&Ts | GACs | Talks | Posters | Search

Poster B in Poster Session B: Tuesday, August 4, 2:00 – 3:45 pm, Kimmel Center, Shorin & Rosenthal Rooms

Early-life environments shape reinforcement learning computations across development

Nora C. Harhen¹, Noam Goldway², Jennifer Fielder³, María Alejandra Martínez Ortiz¹, Yvette Ma¹, Ethan McCormick⁴, Catherine Hartley¹; ¹New York University, ²King's College London, ³University College London, ⁴University of Delaware

Presenter: Nora C. Harhen

A central goal of developmental research is to understand how early-life environments shape individual differences. However, real-world environments vary along countless dimensions. Which ones are relevant? Theoretical work from reinforcement learning (RL) specifies which environmental dimensions should shape learning and decision making and how. Here, we test these normative predictions over developmental timescales. We recruited a large sample of 10- to 25-year-olds to complete a battery of RL tasks and a self-report measure of early-life experience. Our findings largely align with RL's predictions. Younger adolescents from less enriched environments learned more slowly from positive outcomes; those with less control over their environment exhibited stronger Pavlovian influence on learning; and those from less predictable environments planned more using a mental model of the environment. This alignment with normative predictions suggests that individual differences in RL reflect, in part, rational adaptation to developmental environments.

Topic Area: Development, Individual Differences & Clinical Populations