Rasch and Confirmatory Factor Analyses of the Arabic Version of the Diabetes Self-Management Scale (DSMS): An Intercultural Approach

Abstract

The current study was designed to validate the Arabic version of the Diabetes Self-Management Scale (DSMS) using Rasch and confirmatory factor analyses. This included person and item fit, separation, and reliability; rating scale functionality to evidence substantive validity; unidimensional structure to evidence structural validity; and item technical quality to evidence content validity. The study was conducted between September 2021 and March 2022. Utilizing AMOS-based confirmatory factor analysis (CFA), the study also assured the dimensionality of the DSMS. The participants were 103 diabetic patients in Saudi Arabia with a mean age of 44.72 years (standard deviation = 17.35). The analysis was performed using a trichotomous rating scale, and only one item exhibited a misfit (DSMS14). The item difficulty range was −1.0 to +1.0 logits, while the person’s ability range was −3.0 to +3.0 logits. The first construct proved one Rasch dimension, which was explained and further analyzed using AMOS-CFA for the one-factor model. The DSMS was shown to be beneficial as a screening instrument for patient-reported diabetes self-management, despite several flaws that need to be addressed to improve the scale further.

Keywords

• diabetes mellitus
• compliance
• diabetes self-management scale
• Rasch model
• psychometric testing
• AMOS