Learn how to be effective: neural and computational insights into charity efficiency learning

Area of Science:

• Neuroscience
• Behavioral Economics
• Computational Modeling

Background:

• Effective altruism (EA) seeks to maximize social benefit from donations.
• It is unclear if individuals learn efficiency information from feedback to guide charitable choices.

Purpose of the Study:

• To investigate if donors learn efficiency and amount information through feedback.
• To determine if learned representations guide subsequent charitable decisions.

Main Methods:

• Reinforcement learning (RL) computational modeling.
• Electroencephalography (EEG) to measure neural activity.
• A self-prosocial contrast framework (reward vs. charity tasks).

Main Results:

• Donors prioritized efficiency information in the charity task, updating it slower.
• Donation preferences were driven by efficiency differences, not just amount.
• EEG showed distinct neural responses to amount information, weaker in charity tasks.

Conclusions:

• Individuals can learn and utilize efficiency information for charitable decisions.
• Findings support the practical feasibility of EA's rational altruism framework.
• This research bridges computational modeling, neuroscience, and behavioral economics.