Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

Adaptive and Maladaptive Behavioural Flexibility after Reward and Punishment: Evidence Across Dynamic Learning and Goal Contexts

Poster Session E - Monday, March 9, 2026, 2:30 – 4:30 pm PDT, Fairview/Kitsilano Ballroom

Yajing Zhang¹ (), Benjamin James Dyson¹; ¹University of Alberta

Reward and punishment are central concepts in neurocognitive models of adaptive learning, shaping neural circuitry and behavioural responses. Yet, how early experiences with reward or punishment shape flexible or inflexible future performance in dynamic environments remains not well understood. Across four experiments (n = 370), participants performed multi-phase Matching Pennies tasks with an early phase featuring reward or punishment manipulated by fixing win rates (Far/Close Win, Far/Close Loss). Later phases featured computerized opponents that either followed an item bias (exploitable) or played randomly (unexploitable). Our analyses revealed robust asymmetries where early consistent reward (Far Win) promoted the development and persistence of exploitative win-stay and lose-stay strategies, leading to better strategy-performance links in late phases. In contrast, early punishment (Far Loss) generated elevated lose-shift behaviour but impaired the consolidation of effective win-based and stay-based strategies, even during late phase when environments became exploitable. Across Experiments 1-3, these effects appeared regardless of overall performance, with early punishment undermining reward-driven learning stability. However, in Experiment 4, using the same target score continuously across both phases, but with only the later phase allowing strategic exploitation, enabled participants to restore strategy-performance links despite early punishment. These findings demonstrate that early reward and punishment fundamentally alters the trajectory of behavioural adaptation. Even when later environments allow for exploitation, early punishment typically disrupts effective strategy development and persistence. However, this disruption is attenuated when the same target score is maintained across task phases, highlighting a boundary condition for neurocognitive mechanisms that support resilient strategy use.

Topic Area: THINKING: Decision making