J. Peter Rosenfeld, John Meixner, Michael Winograd, Elena Labkovsky, Alex Haynes, Northwestern...

Post on 31-Mar-2015

222 views 2 download

Tags:

Transcript of J. Peter Rosenfeld, John Meixner, Michael Winograd, Elena Labkovsky, Alex Haynes, Northwestern...

J. Peter Rosenfeld, John Meixner, Michael Winograd, Elena Labkovsky, Alex Haynes, Northwestern University

AKA—”Guilty Knowledge” Test.

Physiological responses accompany recognition of information known only by guilty perps and authorities.

Responses are traditionally Autonomic –HR, GSR—but we use brain waves:

An Endogenous, Event-Related Potential (ERP)

Positive polarity (down in Illinois).

Amp = f(1/[stim. Probability], meaningfulness)

80% to 95% correct detection rates….but….

*Rosenfeld et al. (2004) and Mertens, Allen et al. (2008):These methods are vulnerable to Counter-measures (CMs)

via turning I’s into covert T’s.

When you see a specific irrelevant, SECRETLY make some specific response, mental/physical.

After all, if you can make special response to TARGET on instruction from operator, you can secretly instruct yourself.

Irrelevant becomes secret target. It makes big P300. If P = I, no diagnosis.

2 tasks for each trial: 1. implicit probe recognition vs. 2. explicit Target/Non-Target

discrimination

Hypothesized Result: Mutual Interference of 2 tasks more task demand reduced P300 to P. CMs hurt Old test.

Farwell’s web page, Farwell’s web page, claiming 100% accuracy:claiming 100% accuracy:

Results from Rosenfeld et al. (2004): Farwell-Donchin paradigm(BAD and BCAD are 2 analysis methods.)Diagnoses of Guilty

Guilty Group Innocent Group CM Group

9/11(82%) 1/11(9%) 2/11(18%)

Amplitude Difference (BAD) method,p=.1

Cross-Correlation(BC-AD) Method, p=.1

6/11(54%) 0/11(0%) 6/11(54%)

Week BAD* BC-AD*

1: no CM 12/13(.92) 9/13(.69)

2: CM 6/12(.50) 3/12(.25)

3: no CM 7/12(.58) 3/12(.25)

*Note: BC-AD and BAD are 2 kinds of analytic bootstrap procedures.

2 stimuli, separated by about 1 s, per trial,

S1; Either P or I…..then…..S2 ; either T or NT.

*There is no conflicting discrimination task when P is presented, so P300 to probe is expected to be as large as possible due to P’s salience, which should lead to good detection; 90-100 % in Rosenfeld et al.(2008) with autobiographical information. It is also CM resistant. (Delayed T/NT still holds attention.)

* “I saw it” response to S1. RT indexes CM use.

WEEK Hit Rate

Week 1 (no CM): 11/12 (92%)

Week 2 (CM): 10/11 (91%)

Week 3 (no CM): 11/12 (92%)

• 3 groups (n=12)• Simple Guilty (SG), Countermeasure (CM), Innocent Control (IC)

• All subjects first participated in a baseline reaction time (RT) test in which they chose a playing card.

• SG and CM subjects then committed a mock crime.• Subjects stole a ring out of an envelope in a professor’s mailbox.

• All subjects were then tested for knowledge of the item that was stolen. There were 1 P (the ring) and 6 I( necklace,watch,etc).

• CM subjects executed covert assigned responses to irrelevant stimuli in an attempt to evoke P300s to these stimuli to try and beat the Probe vs. Irrelevant P300 comparison.

Condition Detections Percentage

SG 10/12 83

CM 12/12 100

IC 1/12 8

• Unlike 3-SP, the CTP is highly sensitive at detecting incidentally acquired concealed knowledge in a mock-crime (as with autobiographical knowledge).

• Another advantage of the CTP vs 3-SP or polygraph CIT: resistance to CM use.

• CM use produces a large increase in RT between the baseline and test block, and within test block, probe vs irrelevant RT .

5-button ‘I SAW it’ box. The subject is instructed to press, at random*, one of the 5 buttons.

We hoped that this would make CMs harder to do. It didn’t, but we caught the CM users anyway.

Autobiographical information (birthdates): One P and 4 I (other, non-meaningful dates).

3 Groups as before: SG,CM, IC.

NEW: mental CMs to only 2 of the 4 Irrelevants: Say to yourself your first name OR your last name. These are assigned prior to run.

Only one block per group (no baseline).

Group BT/Iall.9 BT/Imax.9

SG 13/13 (100%) 13/13 (100%)

IC 1/13 (7.6%) 1/13 (7.6%)

CM 12/12 (100%) 10/12 (83%)

RT still nicely represents CM use within a block.

Rosenfeld & Labkovsky

John Meixner & Peter Rosenfeld

How do you catch bad guys before crimes are committed, and before you know what was done, where,

when?

+Experimental guilty subjects come to lab and study 3 brochures dealing with pros & cons of 1) What CITY to attack, 2) What METHOD to use, 3) What DATE to attack on for later 3 blocks of CTP tests.

+ Then they write letter to boss with recommendations.

+ Innocent controls study vacation brochures, write recommendation letter to parents/room-mates.

0

4

8

12

Guilty Innocent

Group

Am

pli

tud

e (

µv

)Probe

Iall

A Mock Terrorism Application of the P300-based Concealed Information TestDepartment of Psychology, Northwestern University, Evanston, IL

Iall Imax Blind Imax

Guilty Innocent Guilty Innocent Guilty Innocent

1000 648 985 287 985 603

1000 610 999 416 998 602

955 598 889 476 892 649

996 611 898 430 893 605

994 150 946 17 943 689

909 475 698 284 761 547

945 600 677 365 702 536

997 555 959 250 961 569

999 586 908 217 907 565985 690 888 382 886 706

912 390 667 129 698 650

903 644 837 215 842 702

966 546 863 289 872 619

12/12 0/12 12/12 0/12 10/12 0/12

AUC = 1.0 AUC = 1.0 AUC = .979

Table 1. Individual bootstrap detection rates. Numbers indicate the average number of iterations (across all three blocks) of the bootstrap process in which probe was greater than Iall or Imax. Blind Imax numbers indicate the average number of iterations in which the largest single item (probe or irrelevant) was greater than the second largest single item. Mean values for each column are displayed in bold above detection rates.

CTP is a promising, powerful paradigm, against any number of CMs, mental and/or physical and RT reliably indicates CM use. The new “P900” might also.

jp-rosenfeld@northwestern.edu

Instead of CM first, then I saw it response…

Do them simultaneously! Then Bye Bye RT index:

Probe

Iall