Heightened adaptability to environmental volatility in adolescence
Year: 2026
Type: preprint
Abstract: To learn to make good choices, learners must calibrate how quickly they update their beliefs based on the statistics of their environments: When environments change rapidly, people should base their choices on the most recent outcomes they have experienced, whereas when environments are more stable, people should use a longer history of experienced outcomes to guide their decisions. Prior research has revealed that adults dynamically adapt their learning computations in this way, and that such adaptation is a hallmark of healthy psychological functioning. To date, however, it is unclear how this adaptive calibration of reinforcement learning computations changes across adolescence. Here, 105 adolescents and adults (ages 12 - 25) completed a probabilistic learning task in which they learned to avoid aversive outcomes through trial and error, across environments with both stable and volatile reward contingencies. Computational modeling of participant choices revealed that younger participants demonstrated greater adaptation of learning computations across task contexts, placing more weight on recent outcomes and making noisier choices in the volatile versus stable environment. However, concurrent pupil size recordings during the task revealed that participants across age demonstrated similar sensitivity to reward uncertainty and outcome surprise during learning. Together, these findings suggest that adolescence may be characterized by heightened responsivity to environmental volatility, and lay the groundwork for future studies investigating how deviations in trajectories of adaptive learning may relate to the emergence of psychopathology.
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7164997682