Most popular tests of motor proficiency have been developed using classical test theory. Recently, item response theory (IRT) has been used to address some of the difficulties inherent with these traditional psychometric approaches. This paper focuses on the application of IRT in the psychomotor domain where analyses involving both dichotomous (Cole, Wood, & Dunn, 1991; Werder & Bruininks, 1988) and polytomous criterion referenced measurements (Puderbaugh & Fisher, 1992) have met with equivocal results. The application of the method to quantitative data has to date proved challenging. An example reported here is based on outcome measures used in the development of a test of movement skills for children aged 5 and 6 years. The results appear very promising, but some caution is required in interpreting results.
