Group level effects of social versus individual learning

Area of Science:

• Evolutionary Game Theory
• Behavioral Economics
• Computational Social Science

Background:

• Individual learning dynamics in iterated games can lead to emergent population-level behaviors.
• Previous research has explored various learning mechanisms within game theory frameworks.
• Understanding social learning is crucial for explaining cooperation and strategy adoption.

Purpose of the Study:

• To investigate the impact of social learning strategies within a two-population iterated game model.
• To compare the performance of populations with individual learners versus mixed populations with social learners.
• To identify the most effective social learning strategy for maximizing fitness.

Main Methods:

• Utilizing an iterated game model with random pairing between two complementary populations.
• Implementing and comparing individual learning versus social learning strategies.
• Analyzing the spread and frequency of imitators across generations.
• Evaluating different social learning rules, focusing on payoff-based imitation.

Main Results:

• Mixed populations with both individual and social learners outperform populations with only individual learners.
• The frequency of imitators can vary significantly over generational time.
• The strategy of imitating the neighbor with the highest payoff consistently yields superior results.
• Findings align with experimental and simulation studies across various contexts.

Conclusions:

• Social learning, particularly imitating high-payoff individuals, enhances population performance in iterated games.
• Mixed learning strategies (individual and social) offer advantages over homogeneous individual learning.
• The study provides a robust framework for understanding the evolution of social learning strategies.