Identifying social partners through indirect prosociality: A computational account

Area of Science:

• Cognitive Science
• Social Psychology
• Computational Modeling

Background:

• Identifying prosocial individuals is crucial for forming beneficial social partnerships.
• Previous research focused on direct social interactions for inferring social tendencies.
• Indirect evidence from personal choices also reveals an agent's social considerations.

Purpose of the Study:

• To present a computational model for inferring social preferences from indirect actions.
• To investigate how people use indirect evidence to understand others' prosociality.
• To explore the bidirectional inference between social preferences and knowledge about others.

Main Methods:

• Developed a Bayesian computational model for action understanding.
• Conducted four experiments to test the model's predictions.
• Analyzed human inferences based on agents' choices with indirect social consequences.

Main Results:

• The model accurately captured how people infer social preferences from indirect actions.
• Demonstrated that people infer others' knowledge from their social preferences.
• Showed that simpler alternative explanations did not account for observed inference patterns.

Conclusions:

• People effectively discern potential social partners using indirect evidence of prosociality.
• This work advances understanding of partner detection and social cognition.
• The findings highlight the sophisticated inferential abilities underlying social interaction.