Proportion Explained Component Variance in Second-Order Scales: A Note on a Latent Variable Modeling Approach

Area of Science:

• Psychometrics
• Behavioral Science
• Statistical Modeling

Background:

• Evaluating the variance explained by underlying traits in behavioral scales is crucial for psychometric assessment.
• Existing methods like omega-hierarchical coefficients have limitations in fully capturing explained variance.
• Second-order factor structures are common in complex behavioral scales.

Purpose of the Study:

• To outline a procedure for evaluating the proportion of explained component variance by the underlying trait in behavioral scales with second-order structure.
• To introduce a novel index as a complement to conventional psychometric coefficients.
• To describe a point and interval estimation method for this new index.

Main Methods:

• Utilizes confirmatory factor analysis (CFA) within latent variable modeling.
• Develops a procedure for calculating the proportion of explained variance across scale components.
• Employs a point and interval estimation technique for the proposed index.

Main Results:

• The proposed index effectively quantifies the proportion of variance explained by the underlying trait.
• This index serves as an informative complement to omega-hierarchical coefficients and explained component correlation.
• The estimation method is practical and can be implemented using standard statistical software.

Conclusions:

• The developed procedure provides a valuable tool for assessing the psychometric properties of behavioral scales.
• The new index enhances the understanding of how well underlying traits account for variance in scale components.
• This method supports rigorous evaluation of scale reliability and validity in research.