Feature-based reward learning shapes human social learning strategies.

Area of Science:

• Cognitive Science
• Evolutionary Psychology
• Behavioral Economics

Background:

• Human adaptation relies on social learning, with strategies like conforming to majorities or copying successful individuals identified.
• Existing theories often view social learning strategies as fixed heuristics, failing to account for observed flexibility and individual variation.

Purpose of the Study:

• To propose a domain-general reward learning framework that mechanistically explains key social learning strategies.
• To investigate how individuals learn to associate social features with experienced rewards.
• To demonstrate how this learning process generates adaptive social learning strategies.

Main Methods:

• Formalized a reward learning framework for associating social cues (e.g., behavior, success) with reward.
• Conducted six experiments with 1,941 participants to test flexible social learning adjustments based on experienced rewards.
• Utilized agent-based simulations to model the emergence of social learning strategies.

Main Results:

• Participants demonstrated flexible adjustments in their social learning strategies based on experienced rewards.
• The reward learning framework successfully predicted and explained the emergence of various social learning strategies.
• Individual variability in social learning is explained by differential reward associations.

Conclusions:

• Individuals learn how to learn from others through a reward-based mechanism.
• This flexible learning process enables adaptive knowledge transmission and societal adaptation.
• The findings challenge fixed-heuristic models and offer a unified account of social learning strategies.