P300 amplitude variations, prior probabilities, and likelihoods: A Bayesian ERP study.

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Human Brain Function

Background:

• Bayesian inference provides a normative framework for rational decision-making under uncertainty.
• Understanding the neural mechanisms underlying Bayesian inference in the human brain is a key challenge in cognitive neuroscience.
• Event-related potentials (ERPs) offer high temporal resolution for investigating rapid neural processes.

Purpose of the Study:

• To investigate the neural correlates of Bayesian inference in the human brain.
• To determine if the brain differentially processes prior probabilities and likelihoods during probabilistic reasoning.
• To examine the relationship between manipulations of probabilistic contingencies and event-related potential (ERP) amplitudes.

Main Methods:

• Participants performed an urn-ball task with manipulated probabilistic contingencies (prior probabilities and likelihoods).
• Event-related potentials (ERPs) were recorded in response to stimuli with varying frequencies.
• Bayesian model comparison was used to analyze the association between probabilistic manipulations and ERP amplitude variations.

Main Results:

• Manipulations of prior probabilities and likelihoods elicited distinct and separately modifiable ERP responses.
• P3a amplitudes correlated with prior certainty, showing larger frontal P3a waves with higher prior probabilities.
• P3b amplitudes correlated with likelihood certainty, exhibiting larger parietal P3b waves with lower likelihoods.

Conclusions:

• The human brain appears to code prior probabilities and likelihoods through distinct neural mechanisms.
• Frontal P3a waves reflect the processing of prior certainty, while parietal P3b waves reflect likelihood certainty.
• These findings provide electrophysiological evidence for the neural basis of Bayesian inference in the brain.