Double cross-validation of the booklet Category Test in detecting malingered traumatic brain injury

Two studies were conducted to determine whether the Category Test (CT), in booklet form, can effectively discriminate between groups of brain-injured, malingering, and normal subjects. Discriminant function analyses and clinical decision rules were used to identify malingering on the CT and the relative effectiveness of these two methods was evaluated. Because each of these methods achieved similar classification rates with both samples of subjects, optimal discriminant functions and clinical decision rules were developed combining the subjects from both studies. Discriminant functions correctly classified from 72.5% to 92.2% of subjects and optimal cut-off scores correctly classified from 64.1% to 91.5% of subjects. Effective identification of malingerers was obtained with both higher than expected base rate (29%) and a much lower, more realistic base rate (4%). Thus, the CT performances of individuals who attempt to malinger sequelae of traumatic brain injury appear to be discriminable from genuine brain-injured patients.