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

Poster C in Poster Session C: Wednesday, August 5, 9:30 – 11:15 am, Kimmel Center, Shorin & Rosenthal Rooms

Do you know what you can do? Measuring self-efficacy in sequential decision making

Jing Li¹, Mark K Ho², Angela Radulescu¹; ¹Icahn School of Medicine at Mount Sinai, ²New York University

Presenter: Jing Li

Effective planning requires not just a model of the environment, but a model of self: how precisely can I execute the actions my plan requires? This metacognitive dimension of planning has deep roots in psychology under the construct of self-efficacy , yet we lack formal tools for measuring it and understanding how it shapes sequential decision-making. We introduce the Climbing Challenge, a behavioral paradigm in which participants navigate two-stage decision trees by aiming at holds that vary in motor difficulty and reward. Because motor noise is the only source of uncertainty, path choice reflects how participants integrate reward structure with beliefs about their own capability. We formalize self-efficacy as a perceived motor noise parameter (σ̂) within a reinforcement learning planning model, and by comparing it to empirical motor noise (σ), obtain a measure of self-efficacy calibration. Preliminary results (N=74) reveal systematic individual differences in self-efficacy calibration, with planning strategies further modulated by reward structure.

Topic Area: Decision-Making, Cognitive Control & Event Cognition