Rasch Analysis of Disability Measures

Rasch analysis is a psychometric method, based on Georg Rasch's probabilistic measurement model, used to test and refine the disability, function, and participation scales that pervade disability and rehabilitation research. As set out for clinicians by Alan Tennant and Philip Conaghan in 2007, fitting the Rasch model checks whether a scale's items genuinely measure a single underlying trait at interval level, so that summing item scores into a total is justified. Because so many disability outcome measures simply add ordinal item ratings — assuming items are equally difficult and that ordinal categories behave like interval data — Rasch analysis provides the rigorous test of whether that common practice is actually valid.