A two-stage diffusion modeling approach to the compelled-response task.

Area of Science:

• Cognitive Neuroscience
• Psychology
• Decision Science

Background:

• Understanding the interaction between perception and motor planning is crucial for explaining action selection.
• The compelled-response task involves making an eye movement before knowing the correct target, forcing choices based on limited information.
• Salinas' race-to-threshold model explains variability in choices by the initial state of information accumulation.

Purpose of the Study:

• To propose and validate a novel non-time-homogeneous two-stage-diffusion model for decision-making.
• To account for response time distributions and choice probabilities in the compelled-response task.
• To connect the compelled-response paradigm with established decision-making theories.

Main Methods:

• Development of a two-stage-diffusion model.
• Modeling of response time distributions and choice probabilities.
• Analysis of behavioral data from the compelled-response task.

Main Results:

• The proposed model successfully predicts both response time distributions and choice probabilities.
• The model utilizes interpretable parameters and does not require cross-trial parameter variability.
• The model's predictions, such as bimodal reaction time distributions, are falsifiable.

Conclusions:

• The two-stage-diffusion model offers a robust alternative to existing models for understanding decision-making under uncertainty.
• This model provides a framework that aligns with both behavioral and neural data in various experimental contexts.
• The findings advance our understanding of the interplay between sensory information processing and motor output generation.