Validating Three Common Occupational Therapy Assessments in Participants With Schizophrenia

Social dysfunction is one of the main characteristics of participants with schizophrenia and influences participants' daily functions and quality of life. The Social Functioning Scale-Taiwan short form (SFST) is commonly used to measure social function in participants with schizophrenia. However, no study has examined psychometric properties of the SFST in participants with schizophrenia exclusively. Therefore, the study aimed to examine missing rate, ceiling effect/floor effect, internal consistency, and convergent validity of the SFST in participants with schizophrenia. Investigators obtained schizophrenic self-report data (the SFST and the ADLRS-3) from occupational therapy records. The Chinese version of the Activities of Daily Living Rating Scale (ADLRS-3), which has 10 domains, is commonly used for assessing activities of daily living (ADL) in participants with schizophrenia. However, construct validity (i.e., unidimensionality) for each domain of the ADLRS-3 is unknown, limiting the explanations of the test results. From occupational therapy records, investigators obtained self-report data of the ADLRS-3. Confirmatory factor analysis (CFA) was conducted to examine the 10 one-factor structures. If a domain showed an insufficient model fit, exploratory factor analysis (EFA) was performed to investigate the factor structure and choose one factor representing the original construct. Internal consistency was examined using Cronbach's alpha. Ceiling and floor effects were determined by the percentage of participants with the maximum and minimum scores in each domain, respectively. To evaluate construct validity (i.e., unidimensionality and convergent validity) and Rasch reliability of the Comprehensive Occupational Therapy Evaluation Scale (COTES) in participants with schizophrenia. From occupational therapy records, investigators collected COTES data of participants with schizophrenia. The 20-item COTES contains three subscales. For construct validity, investigators first examined the unidimensionality of each subscale using Rasch analysis. After unidimensionality was supported, investigators examined convergent validity using Pearson's r and Rasch reliability of the individual subscales in unidimensional and multidimensional approaches.