Early versus late noise differentially enhances or degrades context-dependent choice

Area of Science:

• Neuroscience
• Computational Neuroscience
• Decision Science

Background:

• Noise is a fundamental challenge in neural information processing.
• Stochastic errors in choice are attributed to noise in decision-making.
• The differential impact of noise sources on value coding and choice remains unclear.

Purpose of the Study:

• To investigate how noise from different stages of decision-making differentially affects context-dependent choice behavior.
• To reconcile conflicting findings in the literature regarding context effects in choice.

Main Methods:

• Model simulations to predict the impact of early versus late noise on choice accuracy.
• Experimental manipulation of early noise (option value uncertainty) and late noise (time pressure) in human participants.
• Analysis of choice behavior under varying noise conditions and contextual influences.

Main Results:

• Model simulations predicted that early noise enhances choice accuracy with context, whereas late noise degrades it.
• Human experimental data confirmed these opposing predictions, showing dissociable positive and negative context effects.
• Early and late noise sources were experimentally manipulated to produce distinct behavioral outcomes.

Conclusions:

• The timing of noise in the decision process critically modulates the influence of context on choice.
• A unified mechanism for context-dependent choice is proposed, accounting for different noise sources.
• Understanding noise sources is crucial for comprehending how neural computations interact with behavior.