Application of a short form of the Category Test for individuals with a traumatic brain injury: a cautionary note

The Category Test, notable for its length, has been the subject of many proposed short-form versions. In our present managed care era, the pressure to utilize short forms of various tests with clinical populations has increased, yet the appropriateness of using short forms with various populations must be demonstrated not assumed. The present study examined the appropriateness of applying the short form of the Category Test proposed by Charter and colleagues (1997) with a sample of 100 outpatients with traumatic brain injuries referred for neuropsychological testing. Both variance accounted for by the short-form Category Test and the residuals or errors of prediction resulting from the use of the short-form Category Test were examined. While the short form of the Category Test accounted for 93% of the variance of full-form Category Test scores, examination of residuals indicated clinically significant (approximately 10 points or greater) estimation errors for 25% of the subjects. In light of the clinical significance of such extreme estimation errors, it was not recommended that this short form of the Category Test be used for individuals with a traumatic brain injury.