Scaled P300 scalp distribution correlates of verbal deception in an autobiographical oddball paradigm: Control for task demand

Subjects participated in two trial blocks of an autobiographical oddball paradigm. In the first block, 14% of the stimuli were the subject's phone number, 86% were other, meaningless numbers. Subjects responded yes or no, verbally and truthfully. In the second block, stimuli were dates, and the oddball was the subject's birthdate. Subjects again responded verbally, but dishonestly on about 50% of the trials and truthfully on the other 50%. Reaction times differed between the first and the second blocks, but not between the honest and dishonest trials of the second block. P300 amplitude was reduced in dishonest trials of the second block. Honest trials of both blocks had similar P300 amplitude. Scaled scalp distributions were the same for honest trials of both blocks, but differed between honest and dishonest trials of the second block. There were no latency effects. The results are discussed from the viewpoint that task demand effects do not mediate the P300 differences between honest and dishonest responses.