Adolescents flexibly estimate hidden-agent intervention probability from experience, less tied to their explicitly reported beliefs than adults¶

Claim · All claims
Evidence · 1 study · causal r2 · q2
Currency · newest study 2020 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Adolescents were best fit by the adaptive Bayesian model that estimates intervention probability on each trial, whereas adults were best fit by the empirical Bayesian model that incorporates participants' own attribution judgments. → Alexandra O. Cohen 2020

Evidence¶

Alexandra O. Cohen 2020¶

Alexandra O. Cohen, Kate Nussenbaum, Hayley M. Dorfman, Samuel J. Gershman and Catherine A. Hartley. (2020). The rational use of causal inference to guide reinforcement learning strengthens with age. npj Science of Learning 5:16. https://doi.org/10.1038/s41539-020-00075-3

q2 · i? · causal · r2

Computational modeling comparison across age groups suggests adolescents flexibly updated their representation of the environment's structure during the task; the authors interpret this as adolescent learning being less tied to reported beliefs than adults'.

"Adolescents were best fit by the adaptive Bayesian model, which estimated the intervention probability on each trial, whereas adults were best fit by the empirical Bayesian model, which incorporates participant 's own attribution judgments into the model"

Discussion¶

Related Claims¶

• Children's reinforcement learning is best fit by a one-learning-rate model, whereas adolescents and adults are best fit by Bayesian models incorporating hidden-agent beliefs — related
• Participants aged 7–25 attribute positive and negative outcomes to hidden agents in line with the true causal structure of each environment — related
• Individuals can explicitly understand the causal structure of an environment yet fail to use that knowledge to guide action selection — related