A parallel accumulator model accounts for decision randomness when deciding on risky prospects with different...

Area of Science:

• Cognitive Neuroscience
• Computational Psychiatry
• Decision Science

Background:

• Individuals often exhibit decisional randomness, choosing different options with identical information, a phenomenon poorly understood in psychiatric disorders.
• Computational models propose decisions arise from noisy evidence accumulation in the brain, explaining random choices.
• The linear ballistic accumulator (LBA) model posits decisions are made when evidence for an option reaches a threshold.

Purpose of the Study:

• To investigate the mechanisms underlying decisional randomness using the LBA model.
• To apply the LBA to a decision-making task with explicit risk and expected value (EV) signals.
• To predict choice behavior based on fitted LBA parameters.

Main Methods:

• Applied the LBA model to a task where risk and EV were signaled before choice.
• Estimated separate drift rates for stimuli representing high/low EV and high/low risk.
• Used fitted LBA parameters to predict choices on unseen trials for stimulus pairs.

Main Results:

• LBA model predictions showed significant correlation with actual subject choices across all stimulus pairs.
• The model successfully accounted for decisional randomness in an explicit probabilistic task.
• Estimated drift rates varied based on task stimuli, reflecting differences in EV and risk.

Conclusions:

• Sequential sampling models, like the LBA, can explain decisional randomness in probabilistic tasks.
• Findings suggest implications for understanding decision-making in both healthy and psychiatric populations.
• The LBA provides a framework for analyzing the neural implementation of choice under uncertainty.