1.) Is it possible to develop a P300-based deception detection test which is accurate and resistant to countermeasures?
2.) Does the profile or pattern of event-related brain wave (ERP) amplitudes across the scalp vary from deceptive to honest mind states? We mostly look at the P300 ERP in response to autobiographical information. There is an obvious application here to the field of detection of deception, but there are other more theoretical concerns: Is there a profile specific to deception (a "Pinnochio" effect)? Does the brain work in a unique way during deception? Does altruistic deception show a different P300 profile than selfish deception? Do the brains of psychopaths produce different P300 profiles than the brains of normals during deception? As far as P300 is concerned, we don't see differences. We have recently see that there is a difference in normals between the late slow wave responses accompanying lies versus truths, but we do not see this effect in people with psychopathic traits.
3.) What is the difference in brain function during retrieval of a) real memories vs. b) honestly believed, but false memories vs. c) malingered false memories (which the subject knows are not real, but dishonestly claims are real)? This question is addressed also by comparing P300 scalp profiles associated with the three kinds of memories. We also look at the latency of the P300 wave (time from stimulus to wave peak) and have so far found it is the best discriminator of a) and b) above. This suggests that P300 latency is a correlate of unconscious recognition, as discussed in the some of the papers posted on the publications page.
It is noted that a recent review paper covering all our work in these areas under 1) and 2) is the 2002 book chapter entitled "Event-related Potentials in Detection of Deception."
These studies all use EEG and ERP recording methods, as illustrated in the lab scenes to which there are links on my lab pictures page and the Discovery Channel taping page. You can also view the Discovery Channel's coverage of our deception detection studies by clicking here. (It is recommended that any plugin used to view the video be set to have at least a 30 second buffer. Otherwise, the clip may not stream well. Unfortunately, it is not possible to provide universally-applicable instructions for increasing the size of a video plugin buffer, but one should be able to do so by right clicking on the video window and selecting options. After changing the buffer, reload the page. The clip will begin playing automatically after refreshing the page, but it may take awhile since 30 seconds of the clip must be downloaded before playback can begin.)