Evaluation of the psychometrics of the Social Impact Scale: a measure of stigmatization.

As stigmatization has a large impact on patients, therapists need a measure of this impact to provide patients with adequate services. This study, therefore, examined the reliability and validity of the Social Impact Scale (SIS) when applied to three groups of individuals diagnosed with major depression, schizophrenia, or HIV/AIDS. The study sample (N=580) included 237 patients with depressive disorder, 119 with schizophrenia, and 224 with HIV/AIDS. Of these, 56% were men, 45.5% had an elementary school education or less, 48% were employed, and 56% were single. The Rasch measurement model, an item-response theory, was used to analyze the SIS structure and quality. The Rasch model solves several statistical problems of traditional measurement theory, such as misuse of ordinal data as interval data and sample dependence. Rasch analysis indicated that the 24 items of the SIS fit the measurement model. The match between item difficulties and person abilities was adequate. All items showed acceptable rating scale structure. The separation reliability of the scale reached 0.99. The SIS had acceptable psychometric qualities in terms of internal consistency, item validity, person validity, sensitivity, and concurrent validity when applied to patients with depression, schizophrenia, and HIV/AIDS in Taiwan.