A SIMPLICIAL DESIGN FOR THE ANALYSIS OF CORRELATIONAL LEARNING DATA

Area of Science:

• Cognitive Psychology
• Educational Psychology
• Learning Sciences

Background:

• Traditional factor analysis of learning trials often yields uninterpretable superdiagonal matrices.
• These matrices result in generalized factors (early, middle, late trials) lacking specific insights into the learning process.

Purpose of the Study:

• To investigate a novel method for obtaining a more meaningful factor structure in learning studies.
• To demonstrate how incorporating diverse learning measures can enhance factor analysis outcomes.

Main Methods:

• The study introduces several independent measures of learning for each trial.
• Factor analysis is applied to the intercorrelation matrix derived from these enhanced trial measures.

Main Results:

• The proposed method yields a more interpretable factor structure compared to traditional approaches.
• Factors derived from multiple measures offer deeper insights into the learning process dynamics.

Conclusions:

• Employing multiple independent learning measures per trial significantly improves the meaningfulness of factor analysis in learning research.
• This methodology provides a more nuanced understanding of learning dynamics beyond simple temporal factors.