Advanced search
2 files | 4.22 MB Add to list

Humans adaptively select different computational strategies in different learning environments

Pieter Verbeke (UGent) and Tom Verguts (UGent)
(2025) PSYCHOLOGICAL REVIEW. 132(3). p.581-602
Author
Organization
Project
Abstract
The Rescorla-Wagner rule remains the most popular tool to describe human behavior in reinforcement learning tasks. Nevertheless, it cannot fit human learning in complex environments. Previous work proposed several hierarchical extensions of this learning rule. However, it remains unclear when a flat (nonhierarchical) versus a hierarchical strategy is adaptive, or when it is implemented by humans. To address this question, current work applies a nested modeling approach to evaluate multiple models in multiple reinforcement learning environments both computationally (which approach performs best) and empirically (which approach fits human data best). We consider 10 empirical data sets (N = 407) divided over three reinforcement learning environments. Our results demonstrate that different environments are best solved with different learning strategies; and that humans adaptively select the learning strategy that allows best performance. Specifically, while flat learning fitted best in less complex stable learning environments, humans employed more hierarchically complex models in more complex environments.
Keywords
adaptive model selection, hierarchical learning, reinforcement learning, cognitive flexibility, FRONTAL-CORTEX, HIERARCHICAL CONTROL, PREFRONTAL CORTEX, COGNITIVE CONTROL, REINFORCEMENT, PSYCHOLOGY, MODEL, INFORMATION, MECHANISMS, DOPAMINE

Downloads

  • (...).pdf
    • supplementary material
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.80 MB
  • Verbeke Psych Rev AAM.pdf
    • full text (Accepted manuscript)
    • |
    • open access
    • |
    • PDF
    • |
    • 2.41 MB

Citation

Please use this url to cite or link to this publication:

MLA
Verbeke, Pieter, and Tom Verguts. “Humans Adaptively Select Different Computational Strategies in Different Learning Environments.” PSYCHOLOGICAL REVIEW, vol. 132, no. 3, 2025, pp. 581–602, doi:10.1037/rev0000474.
APA
Verbeke, P., & Verguts, T. (2025). Humans adaptively select different computational strategies in different learning environments. PSYCHOLOGICAL REVIEW, 132(3), 581–602. https://doi.org/10.1037/rev0000474
Chicago author-date
Verbeke, Pieter, and Tom Verguts. 2025. “Humans Adaptively Select Different Computational Strategies in Different Learning Environments.” PSYCHOLOGICAL REVIEW 132 (3): 581–602. https://doi.org/10.1037/rev0000474.
Chicago author-date (all authors)
Verbeke, Pieter, and Tom Verguts. 2025. “Humans Adaptively Select Different Computational Strategies in Different Learning Environments.” PSYCHOLOGICAL REVIEW 132 (3): 581–602. doi:10.1037/rev0000474.
Vancouver
1.
Verbeke P, Verguts T. Humans adaptively select different computational strategies in different learning environments. PSYCHOLOGICAL REVIEW. 2025;132(3):581–602.
IEEE
[1]
P. Verbeke and T. Verguts, “Humans adaptively select different computational strategies in different learning environments,” PSYCHOLOGICAL REVIEW, vol. 132, no. 3, pp. 581–602, 2025.
@article{01HNA7HEGNWPEHY32FJ2G5MMCK,
  abstract     = {{The Rescorla-Wagner rule remains the most popular tool to describe human behavior in reinforcement learning tasks. Nevertheless, it cannot fit human learning in complex environments. Previous work proposed several hierarchical extensions of this learning rule. However, it remains unclear when a flat (nonhierarchical) versus a hierarchical strategy is adaptive, or when it is implemented by humans. To address this question, current work applies a nested modeling approach to evaluate multiple models in multiple reinforcement learning environments both computationally (which approach performs best) and empirically (which approach fits human data best). We consider 10 empirical data sets (N = 407) divided over three reinforcement learning environments. Our results demonstrate that different environments are best solved with different learning strategies; and that humans adaptively select the learning strategy that allows best performance. Specifically, while flat learning fitted best in less complex stable learning environments, humans employed more hierarchically complex models in more complex environments.}},
  author       = {{Verbeke, Pieter and Verguts, Tom}},
  issn         = {{0033-295X}},
  journal      = {{PSYCHOLOGICAL REVIEW}},
  keywords     = {{adaptive model selection,hierarchical learning,reinforcement learning,cognitive flexibility,FRONTAL-CORTEX,HIERARCHICAL CONTROL,PREFRONTAL CORTEX,COGNITIVE CONTROL,REINFORCEMENT,PSYCHOLOGY,MODEL,INFORMATION,MECHANISMS,DOPAMINE}},
  language     = {{eng}},
  number       = {{3}},
  pages        = {{581--602}},
  title        = {{Humans adaptively select different computational strategies in different learning environments}},
  url          = {{http://doi.org/10.1037/rev0000474}},
  volume       = {{132}},
  year         = {{2025}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: