Related Experiment Video
Updated: Nov 28, 2025

10:39
The "Motor" in Implicit Motor Sequence Learning: A Foot-stepping Serial Reaction Time Task
Published on: May 3, 2018
8.9K
The effect of visual uncertainty on implicit motor adaptation.
Jonathan S Tsay1,2, Guy Avraham1,2, Hyosub E Kim3
1Department of Psychology, University of California, Berkeley, California.
Journal of Neurophysiology
|November 25, 2020
Summary
Visual uncertainty hinders sensorimotor adaptation, but only with small errors. This finding challenges existing models of how the brain learns and adapts movements based on feedback.
Area of Science:
- Neuroscience
- Motor Control
- Human Movement Science
Background:
- Sensorimotor adaptation is crucial for adjusting movements based on sensory prediction errors.
- Existing models suggest that increased feedback uncertainty should universally reduce motor learning.
- Previous research faced confounds between error size and other learning processes.
Purpose of the Study:
- To investigate the joint effect of visual uncertainty and error size on sensorimotor adaptation.
- To test predictions of optimal sensory integration models in a novel visuomotor task.
- To isolate implicit adaptation independent of strategy use.
Main Methods:
- A novel visuomotor task was employed to avoid limitations of standard tasks.
- Visual uncertainty and error size were manipulated factorially across two experiments.
- Implicit adaptation was measured independently of explicit strategy use.
Main Results:
- Visual uncertainty attenuated sensorimotor adaptation exclusively when the error size was small.
- Uncertainty had no significant effect on adaptation when the error size was large.
- This interaction contradicts the basic predictions of optimal integration theory.
Conclusions:
- Sensorimotor adaptation is modulated by a complex interaction between error size and feedback uncertainty.
- Novel constraints are highlighted for computational models of sensorimotor adaptation.
- The findings suggest that the motor system may dynamically adjust learning rates based on error relevance and uncertainty.

