Working memory constraints for visuomotor retrieval strategies

Area of Science:

• Cognitive Neuroscience
• Human Motor Control
• Perception and Action

Background:

• Visuomotor adaptation relies on cognitive strategies, with algorithmic approaches being flexible but computationally intensive.
• Memory retrieval offers a less demanding alternative for visuomotor adaptation but is constrained by working memory (WM).
• Understanding the capacity and precision limits of WM in visuomotor adaptation is crucial.

Purpose of the Study:

• To estimate the capacity and precision constraints of working memory (WM) retrieval in visuomotor adaptation.
• To differentiate between WM retrieval and algorithmic strategies like mental rotation.
• To investigate how long-term memory (LTM) retrieval influences these constraints.

Main Methods:

• Five experiments involving visuomotor rotation and item-recognition tasks with varying set sizes (number of rotation-target associations).
• Manipulation of item recency to assess memory decay.
• Application of a Bayesian latent-mixture model to analyze performance decrements and estimate WM capacity.

Main Results:

• Performance (accuracy and speed) decreased with larger set sizes and less recent items, indicating WM retrieval signatures.
• A Bayesian model revealed that performance decline resulted from increased guessing and reduced memory precision.
• Estimated WM capacity was limited to 2-5 items before guessing behavior significantly increased.

Conclusions:

• Visuomotor adaptation can be achieved through working memory retrieval, but it is subject to capacity and precision limitations.
• Performance decrements in WM are attributable to both increased guessing and less precise memory representations.
• Long-term memory retrieval, developed through extensive practice, can overcome the limitations of working memory for visuomotor adaptation.