Identifying Qualitative Between-Subject and Within-Subject Variability: A Method for Clustering Regime-Switching...

Area of Science:

• Cognitive Science
• Data Science
• Human-Computer Interaction

Background:

• Modern technology generates vast amounts of high-frequency human dynamic process data.
• Analyzing variability within and between subjects in this data is crucial for understanding complex behaviors.

Purpose of the Study:

• To introduce a novel statistical approach for characterizing qualitative variability from quantitative changes in multi-subject time-series data.
• To evaluate the strengths and limitations of this approach through simulations and real-world application.

Main Methods:

• Development of a statistical model to analyze multi-subject time-series data.
• Monte Carlo simulations to examine the model's performance and limitations.
• Application to real-time hand movement data from an embodied learning platform.

Main Results:

• The proposed approach effectively characterizes between- and within-subject variability.
• Demonstrated ability to identify clusters of dynamics, including phase transitions.
• Successful application in analyzing hand movement data from a mathematical learning context.

Conclusions:

• The introduced statistical method offers a robust way to analyze complex human dynamic processes.
• It provides valuable insights into behavioral variability and learning dynamics.
• Potential applications span various fields requiring analysis of high-frequency human movement data.