Social reinforcement with weighted interactions.

Area of Science:

• Social Network Analysis
• Computational Social Science
• Behavioral Economics

Background:

• Behavior diffusion speed and extent depend on network topology and individual preferences.
• Confirmation bias influences how individuals process information, prioritizing conforming evidence.
• Online social networks often exhibit scale-free topologies with high degree asymmetry.

Purpose of the Study:

• To introduce weighted interactions based on preference similarity to model social reinforcement.
• To analyze behavior diffusion on scale-free networks alongside clustered networks.
• To determine the impact of realistic agent interaction assumptions on diffusion dynamics.

Main Methods:

• Agent-based modeling of behavior diffusion.
• Introduction of preference-weighted interactions to simulate confirmation bias.
• Comparative analysis of diffusion rates on clustered versus scale-free network topologies.

Main Results:

• Weighted interactions, reflecting confirmation bias, reduce diffusion rates in clustered networks.
• Scale-free networks demonstrate superior diffusion performance compared to clustered networks under weighted interactions.
• Degree distribution emerges as a more critical factor than clustering for efficient behavior adoption.

Conclusions:

• Realistic assumptions about agent interactions, including confirmation bias, significantly alter diffusion dynamics.
• Scale-free network properties are more conducive to rapid and widespread behavior adoption than clustered structures.
• Future research on social network diffusion should prioritize degree distribution in network topology analysis.