Abstract The P300 component of the event-related potential was recorded during two blocks of an autobiographical oddball task. All participants performed honestly during the first block (Phone), i.e., the oddball stimuli were phone numbers. During the second block (Birthday), in which the oddball stimuli were participants' birthdays, a Truth group (N = 13) performed honestly and a Malinger group (N = 14) simulated amnesia. Amnesia simulation significantly reduced P300 amplitudes, both between groups and within the Malinger group (Phone vs. Birthday), possibly because of an increase in task difficulty in the Malinger condition. Analysis of scaled amplitudes also indicated a trend for a feigning-related alteration in P300 topography. Bootstrapping of peak-to-peak amplitudes detected significantly more (93%) Malinger individuals than bootstrapping of baseline-to-peak amplitudes (64%). Bootstrapping also provided evidence of a feigning-related amplitude difference between oddball stimuli (i.e., Phone > Birthday) in 71% of Malinger group individuals. In this comparison, the peak-to-peak measure also performed significantly better in intraindividual diagnostics.
