Updating perceptual expectations as certainty diminishes

Area of Science:

• Cognitive Psychology
• Neuroscience
• Perception Science

Background:

• Perception is often facilitated by expectations derived from environmental statistical regularities.
• The dynamic nature of the environment can lead to degradation or disappearance of these statistical relationships.
• The impact of altered environmental statistics on perceptual expectations remains poorly understood.

Purpose of the Study:

• To investigate how changes in statistical relationships influence perceptual expectations.
• To examine the mechanisms underlying the updating of perceptual expectations under diminishing statistical certainty.
• To understand how learning and perception interact to reflect a changing environment.

Main Methods:

• Two studies involved training participants on perfect action-outcome relationships (finger abduction and grating orientation).
• The predictive relationship was then degraded during a test phase (50-75% expected vs. remaining unexpected events).
• Drift diffusion modelling was employed to analyze perceptual decision-making processes.

Main Results:

• Perceptual decisions were faster and less error-prone for expected versus unexpected events when relationships were strong.
• This advantage diminished as the predictive relationship weakened.
• Drift diffusion modelling revealed shifts in evidence accumulation starting points and rates, reflecting training and test session statistics.

Conclusions:

• Perceptual expectations are updated dynamically as statistical certainty decreases.
• Interacting influences, potentially related to learning consolidation, modulate these updates.
• Underlying mechanisms optimize the interplay between learning and perception in a constantly changing world.