Rasch analysis suggested three unidimensional domains for Affiliate Stigma Scale: additional psychometric evaluation

• Published in: Journal of clinical epidemiology 2015-06, Vol.68 (6), p.674-683
• Main Authors: Chang, Chih-Cheng, Su, Jian-An, Tsai, Ching-Shu, Yen, Cheng-Fang, Liu, Jiun-Horng, Lin, Chung-Ying
• Format: Article
• Language: English
• Subjects: Affiliate stigma; Age of Onset; Anxiety - epidemiology; Caregiver; Caregivers - psychology; Caregivers - statistics & numerical data; Classical test theory; Comorbidity; Epidemiology; Family; Female; Humans; Internal Medicine; Male; Mental disorders; Mental Disorders - epidemiology; Mental Disorders - psychology; Methods; Middle Aged; Psychometrics - instrumentation; Psychometrics - methods; Rasch analysis; Self Concept; Severe mental illness; Social Stigma; Statistical analysis; Studies; Surveys and Questionnaires; Taiwan - epidemiology; Validity
• ISSN: 0895-4356; 1878-5921
• DOI: 10.1016/j.jclinepi.2015.01.018

Abstract Objectives To examine the psychometrics of the Affiliate Stigma Scale using rigorous psychometric analysis: classical test theory (CTT) (traditional) and Rasch analysis (modern). Differential item functioning (DIF) items were also tested using Rasch analysis. Study Design and Setting Caregivers of relatives with mental illness ( n  = 453; mean age: 53.29 ± 13.50 years) were recruited from southern Taiwan. Each participant filled out four questionnaires: Affiliate Stigma Scale, Rosenberg Self-Esteem Scale, Beck Anxiety Inventory, and one background information sheet. Results CTT analyses showed that the Affiliate Stigma Scale had satisfactory internal consistency ( α  = 0.85–0.94) and concurrent validity (Rosenberg Self-Esteem Scale: r  = −0.52 to −0.46; Beck Anxiety Inventory: r  = 0.27–0.34). Rasch analyses supported the unidimensionality of three domains in the Affiliate Stigma Scale and indicated four DIF items (affect domain: 1; cognitive domain: 3) across gender. Conclusion Our findings, based on rigorous statistical analysis, verified the psychometrics of the Affiliate Stigma Scale and reported its DIF items. We conclude that the three domains of the Affiliate Stigma Scale can be separately used and are suitable for measuring the affiliate stigma of caregivers of relatives with mental illness.