Refinement and revalidation of the demoralization scale: The DS‐II—internal validity

BACKGROUND The Demoralization Scale (DS) was initially validated in 2004 to enable the measurement of demoralization in patients with advanced cancer. Subsequent shortcomings indicated the need for psychometric strengthening. Here, the authors report on the refinement and revalidation of the DS to form the DS‐II, specifically reporting the scale's internal validity. METHODS Patients with cancer or other progressive diseases who were receiving palliative care (n = 211) completed a revised version of the 24‐item DS and a measure of symptom burden (the Memorial Symptom Assessment Scale). Exploratory factor analysis and Rasch modeling were used to evaluate, modify, and revalidate the scale, providing information about dimensionality, suitability of response format, item fit, item bias, and item difficulty. Test‐retest reliability was examined for 58 symptomatically stable patients at a 5‐day follow‐up. RESULTS Exploratory factor analysis supported a 22‐item, 2‐component model. Separate Rasch modeling of each component resulted in collapsing the response option categories and removing 3 items from each component. Both final 8‐item subscales met Rasch model expectations and were appropriate to sum as a 16‐item total score. The DS‐II demonstrated internal consistency and test‐retest reliability (Meaning and Purpose subscale: α = .84; intraclass correlation [ICC] = 0.68; Distress and Coping Ability subscale: α = .82; ICC = 0.82; total DS: α = .89; ICC = 0.80). CONCLUSIONS The DS‐II is a 3‐point response, self‐report scale comprising 16 items and 2 subscales. Given its revalidation, psychometric strengthening, and simplification, the DS‐II is an improved and more practical measure of demoralization for research and clinical use. External validation of the DS‐II will be reported subsequently. Cancer 2016;122:2251–9 . © 2016 American Cancer Society .