Witnesses sometimes refuse to identify a culprit to police (due to fear or
allegiance) or may be physically unable. Lefebvre et al. (2007, 2009)
applied the P300-based Guilty Knowledge Test to eyewitness identification.
We tested a replication of Lefebvre and colleagues findings with our own
materials. Participants were shown a crime video twice to maximize
exposure. Then, they studied a new face to a criterial recognition
performance. These two faces were later repeatedly presented intermixed
with ten new faces in an oddball paradigm. We instructed half of
participants to lie by not identifying the culprit overtly. Groupwise
P300 amplitudes to the criminal were compared to amplitudes to the studied
face and to the new faces to classify amplitudes to the criminals face as
identifications or rejections.