Hypothesis

We investigated the effects of faking bad on the Wisconsin Card Sorting Test, comparing the perfor-mance of simulating malingerers (M; n = 24) to controls (C; n = 21), closed head-injured patients (CHI; n = 70), and patients with mixed CNS pathology other than CHI alone (CNS; n = 89). Step-wise discriminant functions achieved good accuracy (91-96%), sensitivity (58-100%), and speci-ficity (92-100%) in differentiating simulating malingerers from these groups. The Categories core was a consistent significant independent discriminating variable. Categories alone could differenti-ate between the M and C groups. However, in discriminating between simulating malingerers and both CHI or CNS patients, more complex patterns of performance merged, consisting of relatively poorer performance ratios on Categories compared to Perseverative Errors. This supported the Pattern of Performance theory of the effects of simulating malingering on ncuropsychological tests, which holds that people simulating malingering do more poorly on obvious vs. subtle tusks com-pared to people with verified brain damage. Due to the difficulty of tracking one's performance on several different indices simultaneously, known patterns of performance for patients with verified brain dysfunction shauM be very difficult o fake, even with coaching