Countermeasures to P300-based Guilty Knowledge Tests of Deception

1
Countermeasures to P300-based Guilty Knowledge
Tests of Deception

• J.Peter Rosenfeld, Matt Soskins,Joanna Blackburn,
  Ann Mary Robertson
• Northwestern University.
• Supported by DoDPI

2
Countermeasure issues

• Among the problems with both the ANS-based CQT
  and CIT raised by the report of the National
  Research Council of the National Academy of
  sciences (National Research Council, 2003) is the
  potential susceptibility of all ANS-based methods
  to countermeasures (CMs). As stated by (Honts,
  Devitt, Winbush, Kircher, 1996, p. 84),
  Countermeasures are anything that an individual
  might do in an effort to defeat or distort a
  polygraph test. The National Research Council
  report went on to state that Countermeasures
  pose a serious threat to the performance of
  polygraph testing because all the physiological
  indicators measured by the polygraph can be
  altered by conscious efforts through cognitive or
  physical means (National Research Council,
  2003, p. 4).

3
ERPs to the rescue?

• Deception researchers all hoped and indeed
  expected that when the P300 Event-Related EEG
  Potential was introduced as the dependent index
  of recognition in a CIT (Farwell Donchin, 1991
  Rosenfeld, Angell, Johnson, Qian, 1991
  Rosenfeld et al., 1988), the CM issue would be
  resolved. For example, the eminent inventor of
  the GKT / CIT, (Lykken, 1998, p. 293), suggested
  about CMs to P300 CITs Because such potentials
  are derived from brain signals that occur only a
  few hundred ms after the GKT alternatives are
  presented it is unlikely that countermeasures
  could be used successfully to defeat a GKT
  derived from the recording of cerebral signals.
  (Ben-Shakhar Elaad, 2002, expressed a similar
  view.) All this optimism, as shown below, turned
  out to be misplaced.

4
Some History (earliest publications)

• Rosenfeld et al., 1987,1988,1991
• Farwell and Donchin, 1991
• Allen, Iacono, Danielson, 1992
• Johnson and Rosenfeld, 1992
• Since we were there at beginning, why do we
  challenge as late as 2003-4 with countermeasures?
  (1) Its about time.

5
2) Farwells web page, claiming 100 accuracy
6
Stimuli used in 3-SP

• (1)Probes (P or R in figures) Items which
  subject is suspected of knowing (e.g., murder
  weapons). Subject denies(lies by pressing NO ).
• (2)Irrelevants (I or W in figures) Items of
  which subject has no knowledge and denies,
  honestly, by pressing NO .
• (3) Targets (TR) Items Irrelevant Items to which
  subject presses YES . (Benchmark P300).

7
We ultimately knew we could beat the test..

• In the ordinary, un-countered 3-stimulus
  protocol, the subject is instructed to make
  unique responses to explicitly assigned targets
  which are readily executed with the typical
  result that large target P300s are evoked, since
  these targets are also rare and additionally,
  meaningful, due to their unique button
  requirement. (Rareness and meaningfulness are the
  major antecedents for P300 Johnson, 1986.)

8
IF.the subject can follow an experimenters
instruction to respond uniquely to an
experimenter-chosen irrelevant (an explicit
target) .. then the subject could also
covertly define some (or all) irrelevants for
himself as implicit targets to which he could
make unique responses. These originally
irrelevant but now secret targets would also
elicit large P300s so that one could no longer
depend on the probe P300 amplitude to reliably
exceed that of the irrelevant P300. The
larger probe P300 is, of course, what ordinarily
makes the diagnosis of possession of concealed
information.
9
How P300 amplitude is supposed to catch Liars
1)PgtI (BAD) 2)P-TR corr gtP-I corr(BC-AD)
1)PI 2)P-I corr gtP-TR corr
10
Whither R-TR correlation if there are latency
differences?
Probe P3 Target P3
Nothing should happen to bootstrapped amplitude
difference test (BAD) but bootstrapped
cross-correlation test (BC-AD) should fail.
11
Experiment 1, based on Farwell Donchin (1991)

• --6 Different Probes (multiple probe protocol)
• --Innocent, Guilty, and Countermeasure(CM) Groups
• --Countermeasure Associate various latent
  responses to different categories (jewelry type,
  drawer color, operation name, etc.), all
  irrelevant members of the category.

12
General Instructions.

• Mock crime scenario
• Press Yes to Targets (on list)
• Press No to all other stimuli (Possibly guilty
  probes and Irrelevants).

13
More simply.

• Probe Target I1 I2 I3 I4 ring
  bracelet necklace watch broach tiara
• pink brown yellow purple red
  blue
• donkey tiger lion cow pig
  horse
• etc., etc... (only half the matrix
  here.)
• All these are shuffled, presented in random
  order, involving 4 repetitions of each item.

14
What are the covert countermeasures for the 6
categories of 6 probes?

• 1) Jewelry category.micro right finger wiggle
• 2) drawer lining category. left
  
• 3) owners name category. toe
  
• 4) operation name category right
• 5) location of item category Imagine professor
  slaps you
• 6) desktop categoryDo Nothing
• I.e., make irrelevants into relevant targets.

15
Guilty group Probe(R) gt Irrelevant (W).
R gt W
16
Guilty Group TR vs R
Both have P300
17
Innocent Group R vs W
Both lack P300
18
Innocent Group TR vs R
TR towers over P (R)
19
CM Group R vs W
No difference P(R) vs I (W)
20
CM Group Tr vs R
Target gt Probe
21
Results, Exp. 1 CM works, and analysis method
matters
Diagnoses of Guilty
Amplitude Difference (BAD) method,p.1
Innocent Group
Guilty Group
CM Group
9/11(82)
1/11(9)
2/11(18)
Cross-Correlation(BC-AD) Method, p.1
6/11(54)
0/11(0)
6/11(54)
--with Off the street subjects see next 2
slides.

22
Guilty Subject, probe(R) vs irrelevant(W)
R gtgt W subject clearly guilty
23
but clear latency shift in TR and R P300s
..so BC-AD fails, BAD catches the S.
24
Experiment 2 (One probe protocol) Rosenfeld et
al., 1991).

• 100 simpler to conceptualize
• There are 6 items, repeated 30-40 times
  randomly, consisting of..
• A probe, subject maybe guilty,
• A target to force attention (not really necessary
  but allows us to test Farwell cross-correlation
  method vs. ours.)
• 4 irrelevants.

25
Standard autobiographical oddball paradigm.

• 6 dates are presented, one is the birth date.
• 5 other dates with no personal significance, one
  is a target, four are irrelevant.
• Subject is feigning head injury, the inability
  to recognize birth date 50 of the time, so he
  presses yes and no to all dates on 50-50 basis.

26
Design

• One group of advanced Ss run in 3 successive
  weeks. (Also Control group.)
• Week 1 All naïve(about CM) and guilty (of having
  birth date and recognizing it despite overt lies.
• Week 2 Use explicit countermeasure (to be
  explained).
• Week 3 Repeat first week, stop the CM.

27
Whats the CM?

• When you see the first Irrelevant, do micro left
  finger wiggle.
• When you see second Irrelevant, do micro right
  toe wiggle.
• When you see third irrelevant, imagine me
  slapping your face.
• When you see fourth Irrelevant, do nothing.
• I.e., make irrels. into rel. targets.

28
Results, week 1(guilty) Probe(R) vs.Irrelevant
(W)
As usual, RgtgtW
29
Week 1 Probe(R) and Target(TR)
Both have nice P3
30
Week 2 Explicit CM, R v W
Not so different anymore...
31
Week 3 5/12 test beaters(effortless), R v W
Surprise! R W
32
Week 3 5/12 test beaters, R v TR--classic
defeats
..and TR gtgt R as with innocents.
33
Exp. 2 quantitative.

• WK BAD BC-AD
• no CM 12/13(.92) 9/13(.69)
• CM 6/12(.50) 3/12(.25)
• no CM 7/12(.58) 3/12(.25)
• (Control group nothing much happened over 3
  weeks of repeating week 1.)

34
RTs for 3 weeks week 1 week 3, proving CM not
used in week 3.
35
Conclusions, bottom lines..

• 6-probe/ 3S protocol beat-able, and the 6 probe
  combination lacks a real rationale anyway.
  (Lykken wouldnt like?)
• 1-probe/3S protocol may be explicitly beat-able,
  but the very slow Irrelevant RT distribution will
  raise suspicions. 1 probe per run is more
  Lykkenable.
• BUT---1-probe paradigm after CM practice is
  beat-able, period.

36
Farwell (SPR 08) didnt agree
37
But at the meeting, his letter, not he, showed up
38
Cogn Neurodynamics DOI 10.1007/s11571-012-9230-0

• Brain fingerprinting field studies comparing
  P300-MERMER and P300 brainwave responses in the
  detection of concealed information
• Lawrence A. Farwell Drew C. Richardson Graham
  M. Richardson
• Pub. On line Dec 2012

39
Includes full 100,000 Reward (CM) Study.

• But How does he know the CMs are really done??!!
• There are NO (Zip) Reaction Time data.
• And the ERPs do not suggest CMs are being done.

40
Labkovsky Rosenfeld (2011) Real CM effects on
RT
41
(No Transcript)
42
Controls (???)
43
When CMs (2004) are really done
44
(No Transcript)
45
(No Transcript)
46
So we can forget Farwell. (Every one else has)

• The 3SP is vulnerable to CMs, no doubt about it.

47
What to do?

• Go to a new paradigmthe Complex Trial Protocol
  (Rosenfeld et al., 2008)