Anders Kottorp1, Ingela Petersson. 1. Karolinska Institutet, Department of Neurobiology, Care Sciences and Society, Division of Occupational Therapy, Stockholm, Sweden.
Abstract
OBJECTIVE: The aim of this study was to evaluate the psychometric properties of the Assessment of Awareness of Disability (AAD), applied to a sample of clients with a variety of diagnoses, using two different Rasch models. METHODS: The sample consisted of 245 clients admitted to occupational therapy services in Sweden. The psychometric properties of the AAD measures were analysed and compared using both Rasch partial credit (PCM) and rating scale (RSM) models. RESULTS: In relation to rating scale properties and person response validity, the RSM yielded better results. However, the PCM demonstrated better item goodness-of-fit statistics than the RSM, supporting the internal scale validity of the AAD. Some systematic indications of sample-dependent misfit were also found in the data. CONCLUSION: The results support the validity of the AAD in a sample of clients with various diagnoses, although some limitations of the tool are addressed. The outcomes also demonstrate the importance of making cautious choices when selecting a Rasch model for investigating psychometric properties, as the choice has a substantial impact on the interpretation of the findings.
OBJECTIVE: The aim of this study was to evaluate the psychometric properties of the Assessment of Awareness of Disability (AAD), applied to a sample of clients with a variety of diagnoses, using two different Rasch models. METHODS: The sample consisted of 245 clients admitted to occupational therapy services in Sweden. The psychometric properties of the AAD measures were analysed and compared using both Rasch partial credit (PCM) and rating scale (RSM) models. RESULTS: In relation to rating scale properties and person response validity, the RSM yielded better results. However, the PCM demonstrated better item goodness-of-fit statistics than the RSM, supporting the internal scale validity of the AAD. Some systematic indications of sample-dependent misfit were also found in the data. CONCLUSION: The results support the validity of the AAD in a sample of clients with various diagnoses, although some limitations of the tool are addressed. The outcomes also demonstrate the importance of making cautious choices when selecting a Rasch model for investigating psychometric properties, as the choice has a substantial impact on the interpretation of the findings.