Action biases perceptual decisions toward expected outcomes.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Sensory Perception

Background:

• Action control theories propose suppressing perception of expected outcomes to prioritize surprising events.
• Bayesian models in sensory cognition suggest increased perception of expected events due to their higher likelihood.

Purpose of the Study:

• To investigate how expectations influence perceptual decisions regarding action outcomes.
• To adjudicate between action control and Bayesian models of sensory perception.

Main Methods:

• A signal detection paradigm was employed across three experiments.
• Participants performed manual actions followed by visual outcomes, varying in expectation and relevance.
• Computational modeling analyzed the source of expectation effects on perception.

Main Results:

• Participants showed a bias to report expected action outcomes, supporting Bayesian accounts.
• No significant effects of expectation were found on perceptual sensitivity.
• Computational modeling indicated the bias originated from sensory evidence accumulation, not decision circuits.

Conclusions:

• Findings contradict dominant action control 'cancellation' models.
• Results align with Bayesian perspectives, suggesting perception is enhanced for expected outcomes.
• This bias may facilitate veridical representation of actions and consequences in uncertain environments.