Rational Irrationality: Modeling Climate Change Belief Polarization Using Bayesian Networks.

Area of Science:

• Cognitive Science
• Social Psychology
• Computational Modeling

Background:

• Belief polarization, where individuals react to evidence by updating beliefs in opposing directions, appears irrational and contradicts normative models like Bayes' theorem.
• Despite evidence of rational behavior, belief polarization remains a puzzling exception in human cognition.

Purpose of the Study:

• To investigate belief updating in response to scientific consensus information on anthropogenic global warming (AGW).
• To identify factors contributing to belief polarization using Bayesian network (Bayes net) models.

Main Methods:

• Utilized representative samples from Australia and the U.S. to study belief updating.
• Employed Bayesian network models to simulate rational belief updating and analyze experimental data.
• Assessed the influence of worldview, specifically free-market support, on climate change beliefs.

Main Results:

• In Australia, consensus information partially mitigated worldview influence on climate change acceptance.
• In the U.S., while consensus information increased perceived consensus, strong free-market supporters showed reduced acceptance and perceived consensus.
• Bayes net analysis confirmed free-market support as a driver of climate change beliefs and trust in scientists.
• Active distrust in climate scientists among a subset of U.S. conservatives was identified as a key factor in contrary updating.

Conclusions:

• Belief polarization, particularly concerning anthropogenic global warming, can be driven by factors such as distrust in scientific authorities.
• Bayesian network models provide a framework for understanding and simulating seemingly irrational belief updating processes.
• Understanding the drivers of contrary updating is crucial for addressing societal challenges related to scientific information dissemination.