Limitations of the brief form of the Halstead Category Test

This study attempted to cross-validate a short form of the Category Test (subtests 1-4) first used by Calsyn, O'Leary, and Chaney (1980). One hundred and sixty-eight subjects were assigned to one of six neurological categories, and Category Test scores predicted (via regression analysis) from performance on the short form were compared with actual scores obtained from the original long form. While there was a high correlation (r = .91) between the two sets of scores, a large number of normal subjects were misclassified as brain-damaged. Additionally, subjects with focal right lesions could not be differentiated from normals either on the basis of error scores predicted from short form performance or the absolute number of errors made on subtests 1-4. These results suggested that shortening the Category Test may limit its applicability with certain populations.