Free
Research Article  |   January 2008
Rasch Analysis of the ADL Scale of the A-ONE
Author Affiliations
  • Guðrún Árnadóttir, MA, BOT, is Associate Professor, Division of Occupational Therapy, Faculty of Health, University of Akureyri, Iceland; Coordinator of Occupational Therapy Research and Development Projects, Occupational Therapy, Grensás, Landspítali University Hospital, Reykjavík, Iceland; and doctoral student, Division of Occupational Therapy, Department of Community Medicine and Rehabilitation, Umeå University, Umeå, Sweden. Mail to: Sörlaskjól 52, 107 Reykjavík, Iceland; a-one@islandia.is
  • Anne G. Fisher, ScD, OTR, is Professor, Division of Occupational Therapy, Department of Community Medicine and Rehabilitation, Umeå University, Umeå SE-901 87, Sweden
Article Information
Neurologic Conditions / Work and Industry / Rehabilitation, Disability, and Participation
Research Article   |   January 2008
Rasch Analysis of the ADL Scale of the A-ONE
American Journal of Occupational Therapy, January/February 2008, Vol. 62, 51-60. doi:10.5014/ajot.62.1.51
American Journal of Occupational Therapy, January/February 2008, Vol. 62, 51-60. doi:10.5014/ajot.62.1.51
Abstract

The ADL-focused Occupation-based Neurobehavioral Evaluation (A-ONE; Árnadóttir, 1990) can be used to evaluate both performance of activities of daily living (ADL) tasks and neurobehavioral problems that interfere with ADL task performance among clients with neurological disorders. This study examined the rating scale structure and aspects of validity and reliability of the A-ONE's ordinal ADL scale by applying Rasch analysis methods (Bond & Fox, 2001). Rasch analysis of 209 clients’ A-ONE assessments indicated that misfit of items to the ADL scale could be reduced by removing the two communication items. Threshold disordering could be corrected by combining two adjacent scoring categories (supervision and verbal assistance), thus supporting four response categories. Separation reliability for item calibrations (.98) was high and acceptable for people (.90). Finally, principal components analysis of the residuals supported unidimensionality. The study provided support for converting the ordinal ADL scale to an interval scale that has potential to be used to measure changes in ADL task performance over time.