DETECTION OF DECEPTION WITH EVOKED POTENTIAL P300 EMERGING ON FIRST STIMULUS AT CPZ: NOT AFFECTED BY

1
DETECTION OF DECEPTION WITH EVOKED POTENTIAL
P300 EMERGING ON FIRST STIMULUS AT CPZ NOT
AFFECTED BY SENSATION SEEKING STATE OF
PARTICIPANTS C.
Tanguay, J.D. Cross, J. A. Dale Program
660.66 Neuroscience Psychology
Departments Allegheny College, Meadville PA
16335, USA
Figures
Abstract
Sample Stimuli
Detection of deception in a mock theft was
studied in 16 undergraduate (8 high
sensation-seeking and eight low-sensation
seeking-5 men and 11women) participants using a
modified control-question technique. In an
isolated participant room each person was asked
to steal and hide one of three objects (a cell
phone, a CD player or a wrist watch). The stolen
object was to be revealed only after the
experiment. Evoked potentials from electrode
positions FZ, FCZ, CZ, CPZ, and PZ were recorded
using a Neuroscan Labs, Nuscan model 7181 and a
Neuromedical Supplies electrode cap with linked
mastoid references. Four visual views of each of
the items were presented 25 times on a computer
screen for two seconds separated by a blank white
screen also presented for 2 sec so that each item
was presented 100 times. After the experiment,
the control item for the analysis was selected by
a coin toss. In an analysis of the first stimulus
presentation of the hundred, the mean positive
potential from an average of 100 to 0 for
300-400 msec after the presentation onset was
significantly higher (p. lt.05) for the "stolen
item than for the control item at CPZ but no
differences were detected at other sites. In an
additional analysis of average ERP of the full
100 presentations, the effect remained at CPZ but
was stronger for males than females (p lt.05).
Although the results were statistically
significant they were not forensically
significant in that some participants deceptions
could not be detected. Support Contributed By
W.M. Keck Foundation..

A participant whose deception would not be
detected averaging all trials
A participant whose deception would be
detectable with the average ERP of all trials.
Images were captured digitally with a pinhole
(lensless) digital camera.
Four different views of each object were
presented twenty five times
.
Equipment
Fig 3 Microvolts at CPZ from 300 to 400 msec
for the first presentation of the Not Stolen
stimulus versus the Stolen stimulus corrected
for baseline. Error bars are standard errors of
the mean. Note the p300, on the average, was
absent and ,in fact, negative when the object
was not stolen. this effect on occurred at CPZ
and was not seen at FZ,FCZ CZ or PZ.
Fig 4 Microvolts at CPZ of averaged ERP for 30
to 400 msec corrected for baseline shown as a
function of gender. Error bars are standard
errors of the mean. Note that the detection of
deception occurred mostly in males.
Method
Participants

Results
A large sample of 119 participants (39 males and
80 females filled out Zuckeman and Lunbins
(1965) Multiple Affect Adjective Check list and
this was scored for sensation seeking. Eight
participants from the high end (mean 10) and 8
from the low end (mean 2.75) were selected and
invited to return for the study. These included 5
males and 11 females. Subjects were encouraged to
return by the offer of a chance to win a raffle
for 50.00.

In an analysis of the first stimulus presentation
of the hundred, the mean positive potential from
an average of 100 to 0 for 300-400 msec after
the presentation onset was significantly higher
(p. lt.05) for the "stolen item than for the
control item at CPZ but no differences were
detected at other sites. In an additional
analysis of average ERP of the full 100
presentations, the effect approached significance
(plt.10) at CPZ but was stronger for males than
females (p lt.05). Averaged ERP analyses for other
points may yield additional interesting findings.
At left, the Neuroscan Labs NuAmps 7181,
18-electrode recording system and at right, a
Neuroscan Quik-Cap, were used to record cortical
EEG at 32 electrode locations Compumedics
Neuroscan? (2003). http//www.neuroscan.com/produ
ct.sstg?id31pic1
Purpose
Conclusions
Procedure
It was suggested by Farewell and Donchin in
1991that the ERP may be a reliable means of
detecting deception. The current study is an
attempt to replicate their findings and see if
the effect is moderated by the personality
variable of sensation seeking or gender.
In an isolated participant room each person was
asked to steal and hide one of three objects (a
cell phone, a CD player or a wrist watch). The
stolen object was to be revealed only after the
experiment. Evoked potentials from electrode
positions FZ, FCZ, CZ, CPZ, and PZ were recorded
using a Neuroscan Labs, Nuscan model 7181 and a
Neuromedical Supplies electrode cap with linked
mastoid references. Four visual views of each of
the items were presented 25 times on a computer
screen for two seconds separated by a blank white
screen also presented for 2 sec so that each item
was presented 100 times. After the experiment,
the control item for the analysis was selected by
a coin toss.
Although the results were statistically
significant they were not forensically
significant in that some participants deceptions
could not be detected. Certainly more needs to be
done to ensure that the task is relevant and
engaging for the participant. Experimenters of
both genders should replicate this methodology to
see if the reason for detection in males was
related to the female experimenter.
Hypotheses
It was hypothesized that participants would react
with greater amplitude P300 responses to images
of objects that they had stolen versus objects
that they merely seen. We also predicted that
High Sensation seekers would differ in
reactivity on this measure when compared to Low
Sensation seekers.