The bifactor structure of the Self-Compassion Scale: Bayesian approaches to overcome exploratory structural equation...

Area of Science:

• Psychology
• Psychometrics
• Clinical Measurement

Background:

• Self-compassion research heavily relies on Neff's Self-Compassion Scale (SCS).
• Debate exists regarding the global factor structure of the SCS (one vs. two factors).
• Previous models (ESEM + 2CFA) had methodological limitations and yielded inconsistent interpretations.

Purpose of the Study:

• To test a more appropriate bifactor model with two global factors for the SCS using Bayesian SEM.
• To address the debate on the SCS's global structure and its implications for research and practice.
• To provide a more accurate understanding of self-compassion measurement.

Main Methods:

• Applied recent advances in Bayesian SEM frameworks and fit indices.
• Tested a bifactor model with two global factors (6CFA + 2GlbBF).
• Analyzed correlations between compassionate self-responding (CS) and uncompassionate self-responding (RUS) factors.

Main Results:

• The two-global-factor bifactor model (6CFA + 2GlbBF) demonstrated good fit.
• Correlations between CS and RUS factors were significantly less than 1.0, refuting a single bipolar factor.
• The previously used 6ESEM + 2GlbCFA model was found to be internally inconsistent and discredited.

Conclusions:

• The SCS is better represented by two global factors (CS and RUS) rather than a single bipolar factor.
• Clinical applications and scoring of the SCS should consider these distinct components.
• The methodology offers a framework for evaluating other clinical measurement instruments.