Variability in motor learning: relocating, channeling and reducing noise.

Area of Science:

• Motor control and learning
• Biomechanics
• Statistical analysis of movement

Background:

• Motor performance variability persists despite practice due to inherent motor noise.
• Understanding how performers manage variability is key to improving motor skills.

Purpose of the Study:

• Develop a statistical method to quantify motor variability components (tolerance, noise, covariation).
• Analyze how these variability components change during motor learning.

Main Methods:

• Introduced a novel statistical approach quantifying tolerance cost (T-Cost), noise cost (N-Cost), and covariation cost (C-Cost).
• Applied the TNC-Cost analysis to a virtual throwing task with 15 participants over 6 or 15 days of practice.
• Measured variability in release angle and velocity to assess trajectory accuracy.

Main Results:

• All three cost components (T-Cost, N-Cost, C-Cost) were present and decreased with practice.
• T-Cost reductions were significant early in practice; C-Cost and N-Cost decreased more gradually.
• N-Cost remained the highest component throughout practice, indicating persistent noise effects.

Conclusions:

• Motor performance variability can be reduced by optimizing tolerance, covariation, and noise in execution.
• Learners may exploit T-Cost and C-Cost to mitigate the impact of intrinsic motor noise.
• The TNC-Cost framework provides a quantitative tool for analyzing motor learning and adaptation.