Parsing a cognitive task: a characterization of the mind's bottleneck

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Psychology

Background:

• Understanding the temporal dynamics of cognitive processes is crucial in neuroscience.
• Existing models like psychological refractory period and random-walk models offer insights into response time.
• A unified model is needed to explain how task components contribute to response time and interference effects.

Purpose of the Study:

• To present a theoretical model that parses mental operations into distinct stages.
• To predict how response time distributions change when tasks are performed concurrently.
• To test the model's predictions using a carefully designed number-comparison task.

Main Methods:

• Developed a theoretical model assuming three consecutive stages: perception, decision (evidence integration), and response.
• Proposed that perceptual and motor stages can be parallel, while the decision stage is a serial bottleneck.
• Used a number-comparison task with variations in notation, numerical distance, response complexity, and temporal asynchrony with a probe task.

Main Results:

• The model successfully parsed the number-comparison task, with variables affecting distinct stages.
• Numerical distance uniquely influenced the evidence integration stage, identified as the serial bottleneck.
• Perceptual and motor stages were found to operate in parallel with an interfering task, with their durations not significantly affecting response time variance.

Conclusions:

• The proposed three-stage model offers a parsimonious explanation for response time dynamics and dual-task interference.
• The central decision process, involving evidence integration, is the primary source of response time variability.
• Cognitive tasks can be decomposed into parallelizable perceptual/motor stages and a serial bottleneck decision stage.