Rasch analysis of the Chedoke–McMaster Attitudes towards Children with Handicaps scale
Horton, Mike C.
Disability and Rehabilitation
Taylor & Francis
Reason for embargo
Aim To assess whether the Chedoke–McMaster Attitudes towards Children with Handicaps (CATCH) 36-item total scale and subscales fit the unidimensional Rasch model. Method The CATCH was administered to 1881 children, aged 7–16 years in a cross-sectional survey. Data were used from a random sample of 416 for the initial Rasch analysis. The analysis was performed on the 36-item scale and then separately for each subscale. The analysis explored fit to the Rasch model in terms of overall scale fit, individual item fit, item response categories, and unidimensionality. Item bias for gender and school level was also assessed. Revised scales were then tested on an independent second random sample of 415 children. Results Analyses indicated that the 36-item overall scale was not unidimensional and did not fit the Rasch model. Two scales of affective attitudes and behavioural intention were retained after four items were removed from each due to misfit to the Rasch model. Additionally, the scaling was improved when the two most negative response categories were aggregated. There was no item bias by gender or school level on the revised scales. Items assessing cognitive attitudes did not fit the Rasch model and had low internal consistency as a scale. Conclusion Affective attitudes and behavioural intention CATCH sub-scales should be treated separately. Caution should be exercised when using the cognitive subscale.
National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care of the South West Peninsula (PenCLAHRC)
Copyright © 2016 Taylor & Francis This is an Accepted Manuscript of an article published by Taylor & Francis in Disability and Rehabilitation on 16 February 2016, available online: http://www.tandfonline.com/10.3109/09638288.2016.1140833
Published online: 16 February 2016