A qualitative study on token test performance in aphasic and non-aphasic brain damaged patients

Abstract

The performance of 100 aphasic and 100 non-aphasic brain damaged patients in part V of the TT was subjected to a qualitative analysis. The items were ranked according to the frequency of errors occurring in each of the subgroups. The distribution of rank order of errors was compared across aphasic and non-aphasic patients. An attempt was made to determine whether the items most critical for either group were the same as those identified by Whitaker and Noll [9] in a study of normal children. Finally, the linguistic or psychological properties determining the ease or difficulty of a given item for either group were considered. The performance characteristics of aphasic adults and normal children on the Token Test were different. In the aphasic subgroup, the Whitaker and Noll items were, in fact, clustered in the lower half of the rank-ordered distribution. However they accounted for only 37 per cent of the total errors. Therefore, the linguistic hypothesis suggested by Whitaker and Noll did only partly explain the aphasic patient's failure.