A measure of revised token test ordinality and intervality

Abstract

Abstract Multidimensional scoring of behaviours associated with aphasia came to the fore initially with the publication of the Porch Index ofcommunicative Ability (PICA) (Porch 1967, 1971). Before this test was standardized and proposed for clinical use, it was designed to determine the effectiveness of a 16-point multidimensional scoring system for the assessment of aphasia. Multidimensional scoring was developed in order to minimize the weaknesses of plus-minus scoring, descriptive notes, rating scales and simple category scales. While multidimensional scoring does appear to be a successful method for attenuating the weakness present in other scoring systems, the data derived from this method must meet certain statistical requirements. When mean Scores are used to evaluate patient performance, the scoring system must be at least ordinal. If parametric statistics are used to evaluate the data, intervality of the scoring system is required (Labowitz 1970). Studies by McNcil, Prescott and Chang (1975) and Duffy and Dale (1977) have provided methods and data which suggest that while the PICA scoring system may not be exactly ordinal nor interval, ifused in the context of the PICA, it is robust enough to function as an ordinal and interval scale.