Decision-making components and times revealed by the single-trial electroencephalogram

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience

Background:

• Decision-making involves complex information processing between stimulus onset and response.
• The precise cognitive sequence underlying reaction time remains unclear despite extensive research.

Purpose of the Study:

• To model the sequence of cognitive events during decision-making using electroencephalogram (EEG) data.
• To identify distinct stages of information processing from stimulus to response.

Main Methods:

• Applied the hidden multivariate pattern method to single-trial EEG data.
• Modeled EEG as a sequence of trial-recurrent, time-varying, stable topographies.
• Estimated the number and nature of cognitive events during a decision task.

Main Results:

• Identified five distinct cognitive events: two for visual encoding, and one each for attention orientation, decision, and motor execution.
• Observed adherence to Piéron's law (stimulus intensity vs. reaction time) between encoding and attention.
• Observed adherence to Fechner's law (stimulus intensity vs. decision) between attention and decision commitment.
• Characterized the final decision event as a ramping signal in parietal areas.

Conclusions:

• The study elucidates a five-stage cognitive sequence in decision-making.
• The timing, amplitude, and build-up of the decision-related parietal signal predict accuracy.
• This provides a neurophysiological framework for understanding decision-making processes and reaction time.