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

Sensorimotor context modulates reinforcement learning

Elliot Huang¹, William Xu¹, Robert Wilson¹; ¹Georgia Institute of Technology

Presenter: Elliot Huang

Reinforcement learning, a framework for human learning and decision-making, is influenced by a wide array of factors. One underexplored domain is sensorimotor context. This study examined behavior on a two-choice reversal learning task under two conditions: stationary (arm movements between computers) and active (walking between computers). Participants chose repeatedly between two options rewarding at 70% and 30% probability, with the advantageous option reversing with 20% probability every 15 trials. A mixed-effects model revealed that sensorimotor context modulated reinforcement learning, but the effect is order-dependent. Participants who completed the active condition first are more likely to repeat their choices in the stationary trials. No such effect emerged in the stationary-first group. Subjective reports, computational modeling, and individual differences will reveal mechanistic and computational psychiatry implications.

Topic Area: Memory, Learning & Knowledge Structures