When control slips away: temporal dynamics of learned helplessness and cognitive flexibility under reward uncertainty

Abstract Learned helplessness theory suggests that exposure to uncontrollable stress impairs motivation and cognitive flexibility, but how these effects evolve over time remains unclear. This study used a voluntary task-switching paradigm to examine the dynamic impact of controllability on cognitive flexibility under changing reward conditions. A total of 115 undergraduates were randomly assigned to one of three groups: controllable noise, uncontrollable noise, or no-noise control. All completed a baseline and a reward-based phase of the voluntary task-switching task. The reward phase featured four within-subject reward components: increasing, stable-high, decreasing, and stable-low. At baseline, both the uncontrollable and controllable noise groups showed elevated switch rates relative to the control group. However, hidden Markov modeling revealed divergent switching dynamics: participants in the controllable group more frequently transitioned into a high-switching, fast-response state, whereas those in the uncontrollable group predominantly remained in a low-switching state. During the reward phase, only the controllable and control groups modulated their switching in response to reward changes, whereas the uncontrollable group’s behavior remained flat and reward insensitive. Spectral clustering further showed that participants under uncontrollable stress increasingly adopted reward-insensitive behavioral profiles as trials progressed. Although both stress groups initially increased switching, only the controllable group sustained adaptive flexibility. In contrast, uncontrollable stress reduced reward sensitivity and fostered rigid behavior—hallmarks of learned helplessness. These effects emerged gradually, underscoring the importance of perceived control and the value of trial-level modeling in capturing the development of helplessness over time.