Abstract
The detection of malingering in neuropsychological assessment requires valid measures of incomplete effort. The Computerized Assessment of Response Bias (CARB) is a digit-recognition procedure designed to detect poor effort during a neuropsychological evaluation. We examined the CARB performance of a consecutive series of litigating patients (N = 119) with a full range of head-injury severity, from trivial to severe. Patients with trivial or mild head injuries performed more poorly than did patients with frank brain injuries. Specifically, patients with moderate to severe traumatic brain injuries obtained higher total scores and demonstrated briefer response latencies than did patients with trivial or mild head injuries.