An External Validity Approach for Assessing Essential Unidimensionality in Correlated-Factor Models

Area of Science:

• Psychometrics
• Psychological Measurement
• Factor Analysis

Background:

• Psychometric measures often support both unidimensional and correlated factor analysis solutions.
• Choosing the most appropriate factor structure is crucial for data interpretation.
• Current decision procedures rely solely on internal item score information.

Purpose of the Study:

• To propose an external auxiliary procedure for deciding between psychometric data structures.
• To introduce differential and incremental validity procedures using external variables.
• To enhance the methodological framework for structural model analysis in psychometrics.

Main Methods:

• Development of two groups of external procedures: differential validity and incremental validity.
• Utilizing a second-order structural model with latent variables.
• Assessment through a simulation study and a real-data example in personality research.

Main Results:

• The proposed external procedures provide new methodological insights into factor structure assessment.
• Differential validity procedures assess how primary factor scores relate to external variables.
• Incremental validity procedures evaluate the added predictive value of primary factor scores over general factors.

Conclusions:

• The proposed external procedures offer a valuable alternative to internal-only methods for psychometric structure decisions.
• These methods enhance the understanding of factor score relationships with external criteria.
• The approach is demonstrated to be functional and useful in practical psychological research.