Factor Structure of the PANAS With Bayesian Structural Equation Modeling in a Chinese Sample

Area of Science:

• Psychological assessment
• Quantitative psychology
• Structural equation modeling

Background:

• The Positive and Negative Affect Schedule (PANAS) is a widely used self-report measure for assessing affect.
• Inconsistent findings exist regarding the factor structure of the PANAS, necessitating further investigation.

Purpose of the Study:

• To investigate the factor structure of the PANAS using Bayesian structural equation modeling (BSEM).
• To compare the fit of four different structural models for the PANAS.

Main Methods:

• Bayesian structural equation modeling (BSEM) was applied to data from 893 Chinese middle and high school students.
• Four models were tested: orthogonal two-factor, oblique two-factor, three-factor, and bi-factor models.
• Prior specifications included approximately zero cross-loadings and residual covariances.

Main Results:

• The orthogonal two-factor model, with informative priors for cross-loadings and residual correlations, demonstrated the best model fit.
• Confirmatory factor analysis (ML-CFA) based on BSEM modifications showed improved fit compared to frequentist analysis.

Conclusions:

• The orthogonal two-factor model provides the optimal structure for the PANAS in this sample.
• BSEM is a valuable tool for addressing model misspecifications in psychological measurement.