Context effects on probability estimation

Area of Science:

• Cognitive Neuroscience
• Decision Science
• Neuroeconomics

Background:

• Decision-making involves evaluating outcomes and estimating probabilities.
• Outcome evaluation is subjective and context-dependent.
• Probability estimation relies on environmental statistics and presents unique challenges.

Purpose of the Study:

• To investigate context effects on probability estimation.
• To explore the neural basis of these context effects.
• To identify shared and unique neural substrates for probability estimation and outcome valuation.

Main Methods:

• Behavioral experiments measuring probability estimates under varying contexts.
• Blood-oxygen-level-dependent (BOLD) functional magnetic resonance imaging (fMRI).
• Multivoxel pattern analysis (MVPA) to link brain activity to behavioral effects.

Main Results:

• Probability estimation is biased by contextual events, similar to outcome evaluation.
• Context effects on probability estimates are scaled by estimated uncertainty, peaking at intermediate probabilities.
• Activity patterns in the dorsal anterior cingulate cortex (dACC), ventromedial prefrontal cortex (VMPFC), and intraparietal sulcus (IPS) predicted individual differences in these biases.

Conclusions:

• The ventromedial prefrontal cortex (VMPFC) is a shared neural substrate for both outcome valuation and probability estimation.
• The dorsal anterior cingulate cortex (dACC) and intraparietal sulcus (IPS) are uniquely involved in probability estimation.
• Context-dependent biases in probability estimation may impact a broad range of cognitive computations, from sensory inference to higher cognition.