How social reinforcement learning can lead to metastable polarisation and the voter model
2024

How Social Reinforcement Learning Affects Opinion Dynamics

Sample size: 100 publication 10 minutes Evidence: high

Author Information

Author(s): Meylahn Benedikt V., Meylahn Janusz M.

Primary Institution: Korteweg-de Vries Institute for Mathematics, University of Amsterdam

Hypothesis

Can social reinforcement learning lead to metastable polarization in opinion dynamics?

Conclusion

The study concludes that polarization in opinion dynamics is metastable and consensus is eventually reached.

Supporting Evidence

  • The model shows that consensus is reached with probability one in the long run.
  • Simulations indicate that the time to consensus is influenced by network structure.
  • Heavy-tailed distributions in time to consensus suggest metastable states.

Takeaway

The way people learn from each other can make opinions stick together for a while, but eventually, everyone tends to agree.

Methodology

The study used simulations of an asymmetric reinforcement learning model to analyze opinion dynamics on a network.

Potential Biases

The model may not accurately reflect real-world dynamics due to its simplifications.

Limitations

The model assumes fixed agents and does not account for external influences or memory in agents.

Participant Demographics

The study involved a theoretical model with 100 agents on a geometric network.

Digital Object Identifier (DOI)

10.1371/journal.pone.0313951

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication