Neural arbitration between social and individual learning systems.

Area of Science:

• Neuroscience
• Cognitive Science
• Decision Science

Background:

• Decision-making integrates personal experience and social advice.
• Neural mechanisms of arbitrating between information sources are not fully understood.

Purpose of the Study:

• To formalize and investigate the neural basis of arbitration between individual and social learning systems.
• To examine how relative precision of predictions influences arbitration.

Main Methods:

• Hierarchical Bayesian modeling to define arbitration based on prediction precision.
• Probabilistic learning task involving predictions from self-sampled outcomes and/or an advisor.
• Functional neuroimaging (fMRI) to identify brain regions involved in arbitration.
• Decision confidence measured by points wagered.

Main Results:

• Arbitration, defined as a ratio of precisions, correlated with decision confidence.
• Neural arbitration signals were independent of decision confidence.
• Arbitrating for self-information activated dorsolateral prefrontal cortex and midbrain.
• Arbitrating for social information engaged ventromedial prefrontal cortex and amygdala.

Conclusions:

• Relative precision effectively captures behavioral and neural arbitration between social and individual learning.
• Distinct neural pathways support arbitration favoring self versus social information.