Partial confirmatory factor analysis: described and illustrated on the NEO-PI-R

Area of Science:

• Psychometrics
• Quantitative Psychology
• Personality Psychology

Background:

• Exploratory Factor Analysis (EFA) is commonly used to identify underlying factor structures.
• Researchers often recommend Confirmatory Factor Analysis (CFA) to validate EFA findings.
• A gap exists in methods to pre-emptively assess the suitability of EFA models for CFA.

Purpose of the Study:

• Introduce Partial Confirmatory Factor Analysis (PCFA) as a technique to evaluate EFA-derived models.
• Justify the recommendation for subsequent CFA testing of EFA models.
• Assess the appropriateness of the Five-factor model for CFA using PCFA.

Main Methods:

• Described Partial Confirmatory Factor Analysis (PCFA) methodology.
• Applied PCFA to a Revised NEO Personality Inventory facet correlation matrix.
• Examined the Five-factor model using the PCFA technique.

Main Results:

• The Five-factor model, derived from EFA, was found not suitable for Confirmatory Factor Analysis (CFA) testing.
• Partial Confirmatory Factor Analysis (PCFA) provided evidence against the direct application of CFA.
• The analysis highlighted limitations in directly confirming the EFA solution via CFA.

Conclusions:

• Researchers should supplement Exploratory Factor Analysis (EFA) with Partial Confirmatory Factor Analysis (PCFA).
• PCFA serves as a crucial step to justify the transition from EFA to CFA.
• This approach enhances the rigor of factor model validation in personality research.