The aims of this study of oncology outpatients with cancer pain were to perform an exploratory factor analysis (EFA) of the 48-item Coping Strategy Questionnaire (CSQ) and a confirmatory factor analysis of the 6-factor solution of the Coping Strategy Questionnaire-Revised (CSQ-R) suggested by Riley and Robinson in 1997. In addition, differences in latent factor means and in the CSQ-R subscale scores between inpatients and outpatients were evaluated.
Data from oncology outpatients (n=217) and inpatients (n=225) with pain were used. The Mplus program was used to perform both the EFA and confirmatory factor analysis treating the items as ordinal, and using robust maximum likelihood estimation. Quartimin oblique rotation was used for the EFA. Model fit was evaluated with the χ2 test, the comparative fit index, the root mean square error of approximation, and the standardized root mean square residual, as well as by substantive evaluation of the solutions.
The EFA of the original 48-item CSQ did not reproduce the factor structure defined by Rosentiel and Keefe. The 6-factor structure from the 27-item CSQ-R fit both the inpatient and outpatient data well with strong factorial invariance, as well as the combined data, allowing some correlated errors among items. Differences were found between the 2 samples for the ignoring, catastrophizing, and praying latent factor means, and for the catastrophizing and praying subscale means.
The 27-item CSQ-R is recommended for use as a clinical instrument. However, further research of the 6-factor structure is recommended to identify reasons for the correlated errors.