Characterizing perceptual learning using regression statistics: development of a perceptual calibration index

Area of Science:

• Cognitive Psychology
• Perception Science
• Motor Learning

Background:

• Perceptual learning enhances skill through practice, improving accuracy and consistency.
• Regression statistics can quantify calibration of perceptual properties.
• Existing methods obscure simultaneous changes in accuracy and consistency.

Purpose of the Study:

• To introduce a perceptual calibration index for quantifying perceptual learning.
• To visualize conjoint changes in regression statistics.
• To demonstrate the index's utility in a hand-held object wielding experiment.

Main Methods:

• Utilizing regression statistics (r2, slope, intercept) to measure perceptual performance.
• Developing a perceptual calibration state space diagram.
• Calculating a perceptual calibration index based on distance to an optimal performance attractor.

Main Results:

• The perceptual calibration index quantifies improvements in both accuracy and consistency.
• State space diagrams reveal conjoint changes in regression statistics.
• The index effectively illustrates perceptual learning in a motor task.

Conclusions:

• The perceptual calibration index offers a unified measure of perceptual learning.
• This method provides simultaneous insight into accuracy and consistency improvements.
• The index is a valuable tool for studying perceptual skill development.