To make scores from tests designed for special populations exchangeable, the tests must first be equated on the same scale. This study examined the utility of a Rasch model in equating motor function tasks. Using an existing gross motor function data set and a semisimulation design, an artificial equating and cross-validation sample, as well as two artificial tests, were created. Based on these samples and tests, the accuracy and stability of Rasch equating was empirically determined using a standardized difference statistic. It was found that Rasch equating could accurately equate tests and was generalizable when applied to a cross-validation sample. After equating, tests can be compared on the same scale, and interpretation of cross-test scores becomes possible. In addition, with the conversion table and graph generated from Rasch equating, the application of test equating was demonstrated as simple and practical.
ASJC Scopus subject areas
- Physical Therapy, Sports Therapy and Rehabilitation