1
NOT FOR DISTRIBUTION

The Science of Deception: Integration of Theory and Practice
Arlington VA, July 17-18, 2003

PREFACE.......................................................................................................................... 2
SCENARIO #1: WALK-INS ARE IMPORTANT BECAUSE THE MOST VALUABLE PEOPLE ARE
NOT THOSE WHOM WE RECRUIT OR CAPTURE, BUT THOSE WHO APPROACH US. ................ 3
Current challenges: Operational constraints ............................................................. 3
View your interaction with the walk-in as a trade negotiation................................... 4
One lie begets another ................................................................................................ 4
Source evaluation........................................................................................................ 4
Know yourself: deception is 50% self-deception ........................................................ 6
Rules for how to talk with the walk-in ........................................................................ 8
Set up standard operating procedures........................................................................ 9
Implement evaluation mechanisms ........................................................................... 10
As much as possible, control the context of the interview ........................................ 11
General guide for deceiving an interviewer: You are given the task of deceiving an
interrogator. What should you do?.......................................................................... 13
Research challenges.................................................................................................. 15
SCENARIO #2 - LAW ENFORCEMENT THREAT ASSESSMENT POST 9/11 LACKS TRIAGING
MECHANISMS ................................................................................................................. 15
Categorize threats and prescript appropriate responses.......................................... 16
Research challenges.................................................................................................. 17
SCENARIO #3 - LAW ENFORCEMENT INTERROGATION AND DEBRIEFING ....................... 17
Research challenges: Personal characteristics of good deception detectors........... 17
Research challenges: characteristics of normal conversations................................ 18
Research challenges: the reliability of technological devices to detect deception... 19
Additional research challenges................................................................................. 20
SCENARIO #4 - INTELLIGENCE GATHERING ................................................................... 20
Systematic data collection: needs and challenges .................................................... 20
GENERAL RESEARCH CHALLENGES................................................................... 21
PARTICIPANT LIST (PARTIAL) ............................................................................... 23

 2

Preface
“Lying is a fact of everyday life. Studies in which people kept daily diaries all of their lives suggest that
people tell an average of one or two lies a day. . . . People lie most frequently about their feelings, their
preferences, and their attitudes and opinions. Less often, they lie about their actions, plans, and
whereabouts. Lies about achievements and failures are also commonplace.” (DePaulo et al., 2003, p. 76).
“The pervasive experience of self-deception within deception is a result of our ability to entertain models of
the world independent of our experience. Our capacity for understanding is particularly implicated in its
ability to allow us to create or imagine possible as well as actual worlds.” (Mitchell, 1996, p. 842)

A meeting of social and behavioral scientists and personnel from law enforcement and
intelligence agencies was held in Arlington, VA on July 17 and 18, 2003. The topics
were deception and deception detection in four situations: the evaluation of information
offered by a ‘walk-in’ to a U.S. embassy abroad, the evaluation of the veracity of threats
by telephone, letter and email to the FBI and U.S. Secret Service, the interrogation of
suspected criminals by law enforcement, and intelligence gathering by U.S. government
agencies. The conversations were focused around four scenarios that were provided by
the intelligence and law enforcement communities.
This report contains a summary of the conversations recorded at the meeting. It includes
some description of current standard operating procedures, recommendations and
suggestions offered to intelligence and law enforcement personnel in the situations
described by the scenarios, and a list of issues where further consideration and research
are needed.

Susan E. Brandon
January 12, 2004

 3

Scenario #1: Walk-ins are important because the most valuable people are not those
whom we recruit or capture, but those who approach us.
Throughout the world, individuals regularly walk into an American Embassy and state
that they have valuable information regarding threats to U.S. interests abroad or
sensitive intelligence information. These “walk-ins” or “volunteers” will also phone in
or send emails to U.S. embassies with offers of information.
Some walk-ins actually have valuable, sensitive information, some are mentally unstable,
and some are “provocations,” hoping to pique the interest of the CIA in order to identify
Agency employees and /or the methods used by the Agency to handle walk-ins. Others
simply wish to provide false and misleading information to the Agency, and still others
may have useful information but are trying to peddle the information to multiple
governments in the hope of reaping financial gain.
Determining who has legitimate information of value and who is engaging in deception is
clearly a critical issue. Constraints include limited time to interview the walk-in,
variability in interviewer skill (sometimes the interviewer is an Embassy Security Officer,
other times it is an Agency ops officer), or lack of face-to-face observation in the case of
individuals calling in or writing email.
Current challenges: Operational constraints
(i) No one, or many, individuals may interact with the volunteer before he sees the
interviewer.
(ii) Time is critical: “One hour is long and two is a gift.”
(iii) This may be the only visit to the embassy that the walk-in will make.
(iv) The embassy staff – likely to be the Regional Security Officer, who is responsible
for the physical security of the facility – is not likely to know the local language, and
the interview may be conducted with the help of a translator or translation cards.
(v) The walk-in may be from various backgrounds of culture, language, age, sex,
education, and previous experience with Americans.
(vi) The situation is highly likely to be viewed as high risk by both parties, but
particularly by the walk-in.
(vii) The FBI, DoD, and CIA may have different objectives and approaches to the
walk-in situation.

 4
Participants’ experience with many years of walk-in situations suggested that U.S.
embassy personnel are remarkably bad at walk-in evaluation. A worst case scenario is
telling a walk-in to “go away and write your story and then come back, because I don’t
have time for you right now.” In such instances, interrogators forfeit visual cues and
spontaneity, both of which are important to evaluation.
Empirical research and current collective experience suggests the following tactics and
strategies relevant to for the embassy employee (assumed to be an American male) who
processes a walk-in. The walk-in is assumed to be male and a resident of the local
population. Where empirical data are available, sources are noted.
View your interaction with the walk-in as a trade negotiation
Each of you can be assumed to have something that the other wants. It can be assumed
that the walk-in has information that is valuable to the U.S. It can be assumed that as an
agent of the U.S. government, you have something that is valuable to the walk-in, such as
money, security, or relocation. The probability that walk-ins will drop by and deliver a
significant item of information on a first visit for nothing is low, however. It is advisable
to view this first encounter as the beginning of a negotiation, where “we have money and
some services to offer,” and “they have information that may be useful to our national
interests.”
One lie begets another
The walk-in wants to be able to trust you, and he will not do so if he perceives that you
are lying to him. Be as truthful as you can.
Source evaluation
The first step in evaluating a walk-in is to establish whether or not this source is likely to
be legitimate or not.
Some evasiveness or reticence to divulge identity or all of the information that the
subject is wishing to sell should be expected. However, evasiveness is not the
same as inconsistency. Given reasonable concerns for exposure, most walk-in
sources are likely be initially hesitant about presenting a full, verifiable identity if
he is intent on remaining a resident in his current community. The common fate
of traitors is death in most of the world, and no amount of financial reward is
going to stack up against this. So, unless the subject is seeking relocation for
himself and his family as part of some compensation, he should make some
efforts to preserve his future by protecting his identity at the start.

 5
In addition, it should be expected that anyone selling information is not going to
give that information up at the start. If the information is perceived to be of
monetary value, why would he divulge it for free? It is reasonable to assume that
the source should be able to give details regarding the nature of the information
and with sufficient level of detail either about the information or about himself
that he can be identified as a potentially credible source. The walk-in who has
good information usually provides only partial information, and looks for
assurances of safety (e.g., for family).
If the walk-in is trying to deceive you as to the source or content of his information, it is
possible that he will provide some cues that indicate his deception:
(i) Research shows that, in some situations, liars tell their stories in a less
compelling and less immediate manner and with fewer relevant and irrelevant
details than truth-tellers. Lies also have fewer spontaneous corrections and liars
less often admit that they can’t remember. There is no difference in eye contact
between those who tell the truth and not, but there is some evidence that liars are
more likely to raise their chins while talking. Signs of mental effort may be
present with liars, such as pauses and disturbances in speech, more pupil dilation
and a higher pitched voice. (These signs of mental effort should increase to the
extent that a person is asked to offer long, rather than short, answers to
questions.)1 Such cues to deception are more pronounced when people are highly
motivated to deceive, especially when the motivations were not monetary or
material.2
Some of these cues may be more evident if the interrogator has an opportunity to
first observe the walk-in and establish as much of a “baseline of behavior” as
possible. Amarillo Slim, a renown and successful poker player, includes among
his list of the “top ten keys to poker success,” that one is “always observing at a
poker game. “The minute you’re there, you’re working. Watch the other players
for ‘tells’ before you look at your own cards.”3
(ii) The walk-in who deliberately is lying might engage in deliberate posture
mirroring. Behavioral mirroring or mimicry occurs when an individual imitates
the posture behaviors of another person. Mimicry is known to build affiliation,

1

DePaulo, B.M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues
to Deception. Psychological Bulletin, 129, 74-118. See also DePaulo, B. M., & Morris, W. L. (in press,
2003). Discerning lies from truths: Behavioral cues to deception and the indirect pathway of intuition. In
P. A. Granhag & L. Stomwall (Eds.), Deception detection in forensic contexts.
2
Forrest, J. A., & Feldman, R. S. (2000). Detecting deception and judge’s involvement: Lower task
involvement leads to better lie detection. Personality & Social Psychology Bulletin, 26, 118-125; DePaulo,
B. M., & Tang, J. (1994). Social anxiety and social judgment: The example of detecting deception.
Journal of Research in Personality, 28, 1420153.
3
Preston, A. S. (2003). Amarillo Slim in a World Full of Fat People. With Greg Dinkin. Harper Collins,
New York (p. 126).

 6
liking and rapport between two people.4 This may serve to enhance the walk-in’s
ability to deceive. Since this behavior occurs in normal (i.e., nondeceptive)
social interactions, as well, it simply is advisable to be aware of the effectiveness
of mirroring.
(iii) Research shows that most people are not good at detecting lies: across many
situations, accuracy at detecting deception is about 53%, which is little better than
chance.5 There is some evidence that people are better at reporting an accurate
sense of ambivalence about a deception situation, than they are about the specifics
of deception.6 In some instances, “implicit measures” of deception have been
shown to be significantly better than explicit measures.7
Know yourself: deception is 50% self-deception
Know your views, attitudes, and beliefs about your own culture, as well as your
views, attitudes, and beliefs about the culture in which you are working. This is
as important as knowing the views, attitudes, beliefs and customs of the locals –
and far more possible, given the relatively limited training that most embassy staff
receive.
Cognitive schemas are how people organize their knowledge of the world. These
schemas affect our perceptions and our expectations, so that we are more likely to
perceive and expect incoming information that is consistent with these schemas
than that which is not.8 Deliberate deception is most effective when the walk-in
knows what your expectations and views of the world are, so that he can build a
story that fits with these expectations, and prescript your imagination Such a
story also needs fewer explicit, external details, because it can be assumed that
you will fill these in. The walk-in then does not need to offer details that might be
unverifiable or contradictory.9 A good interrogator therefore works hard to avoid
providing the walk-in with a priori expectations.
4

Lakan, J. L., & Chartrand, T. L. (2003). Using nonconscious behavioral mimicry to create affiliation and
rapport. Research Article, Ohio State University; Chartrand, T. L., & Bargh, J. A. (1999). The chameleon
effect: The perception-behavior link and social interaction. Journal of Personality and Social Psychology,
76, 893-910.
5
Bond, C. F., & DePaulo, B. M. (2003). Accuracy and truth bias in the detection of deception. Manuscript
in preparation.
6
DePaulo, B. M., Rosenthal, R., Green, C. R., & Rosenkrantz, J. (1982). Diagnosing deceptive and mixed
messages from verbal and nonverbal cues. Journal of Experimental Social Psychology, 18, 433-446;
DePaulo, B. M., Charlton, K., Cooper, H., Lindsay, J. J., & Muhlenbruck, L. (1997). The accuracyconfidence correlation in thd detection of deception. Personality and Social Psychology Review, 1, 346357.
7
Vrij, A., Edward, K., & Bull, R. (2001). Police officers’ ability to detect deceit: The benefit of indirect
deception detection measures. Legal and Criminalogical Psychology, 6, 185-196.
8
Catrambone, R., & Markus, H. (1989). The role of self-schemas in going beyond the information given.
Social Cognition, 5, 349-368; Tesser, A., & Leone, C. (1978). Cognitive schemas and thought as
determinants of attitude change. Journal of Experimental Social Psychology, 13, 340-356.
9
Mitchell, R. W. (1996). The psychology of human deception. Journal of Social Research, 63, 819-861.

 7

(i) To get a sense of how to construct a good deception plan, the walk-in might
first engage in legitimate and truthful interactions. The deceiver might at this
point also provide “facts” that are unverifiable (e.g., claims of people who have
died), just to engage you in conversation.
(ii) The walk-in who is deliberately lying might take try to get you to offer up
some of your own judgments or perceptions of him, so that he can adapt his own
behavior in such a way that he will provide apparent confirmation of your views
of him. (That is, the walk-in may try to take advantage of people’s inherent
tendency to confirm their own biases.10) The walk-in might try to convince you
that you have figured him out. You are then more likely to think that you know
how to evaluate subsequent information that he provides you, and he can use this
as a means of safeguarding other information or as a platform for presenting
intentionally deceptive information. For example, an interrogator usually expects
that he is going to be lied to. If a walk-in allows himself to be "caught," the
theory is that once satisfied that he had the subject accurately assessed all along,
the interrogator will stop digging or will not attend to other data that suggest he
should look further. Therefore, if you strongly suspect that the walk-in is going to
lie, he may create an opportunity for you to catch him in a “lie” by designing an
unconvincing cover story that can be penetrated – so that your desire to catch a lie
is satisfied. This bogus deception might be the truth (which the walk-in would
want the interrogator to dismiss), or it might involve matters unrelated to his
intended goal.
You might want to try to see if this is happening by presenting a false early
characterization of the walk-in or situation that he does not actually hold and then
observe whether he makes any attempts to adapt his own stories accordingly.
(iii) A good interrogator knows what his own expectations are, and avoids using
conformity to those expectations as evidence. Do not seek out information that is
too closely tailored to your own expectations. It would be best to look for
deception in the situation without preconceptions, but this is not possible – we all
have preconceptions of one sort or another (including the preconception that we
do not have preconceptions). However, there are means that can be used to
counter this tendency.
(a) The interrogator can initially enter the situation with a minimum of
biasing information.
(b) The interrogator can keep focused on a particular task. Specifically, in
initial contacts, it may be more important to validate the source first before
attempting to validate the content of his information. Also, it may be
10

Mynatt, C. R., Doherty, M. E., & Tweney, R. D. (1978). Confirmation bias in simulated research
environment: An experimental study of scientific inference. Quarterly Journal of Experimental
Psychology, 29, 85-95.

 8
helpful to focus on particular aspects of information and try to validate
those first, to avoid getting lost in the larger picture/data set of information
being presented.
(c) As an interrogator, it might be helpful write out any expectations you
an honestly identify before the start of the interrogation series and then
consciously seek disconfirming evidence.
(d) Never offer real opinions, theories, ideas, personal information, etc., to
the walk-in that might give away expectations, biases, or beliefs that could
be used as tools for deception by the target.
(e) Review ahead of time the focus of your questions, and assess where
they potentially indicate a particular bias or intent.
(f) Assign determinations of whether a subject is being deceptive or not to
a second observer (live or video feed). This allows an agent who is not
directly involved in the interaction to focus specifically on the target's
behavior.11
Rules for how to talk with the walk-in
(i) Make questions open-ended. The initial goal of any interrogation is to get the
walk-in to do one thing: talk. A common error among novice interrogators is to
either ask yes-no type questions, or to move so rapidly into direct questions
related to the end goal of the interrogation that the subject's resistance peaks and
he stops talking. Thus, in most cases, you want open-ended questions.
(ii) Ask surprise questions. Much has been made about the assumptions of
interrogators and how they may be misled by the deceptive agent playing to these
assumptions or biases. The same is true with reference to the walk-in.
Interrogation is generally a novel experience for most people. One way most
people attempt to deal with this kind of situation is to make predictions about how
it will go. We know that establishing some predictability in the environment is
one method for exercising an element of control on what is otherwise a very
stressful, uncontrolled situation.12 When a walk-in is being interviewed and the
process is in line with expectations, the situation will be perceived as less
stressful. Doing something or asking something that serves to surprise him is a
way to keep him off balance and feeling out of control. Surprises serve to push
the person outside the range of his mental paradigm, and in that domain there is
likely to have been no rehearsal, no pat answers, no thought-out cover.

11

G. Hazlett, personal communication (2003).
Miller, S. M. (1989). Controllability and human stress: Method, evidence and theory. Behaviour
Research & Therapy, 17, 287-304.

12

 9
(iii) As the interrogation continues, it becomes more of a conversation between
two people and less of an interrogation. It is increasingly subject to the same
social contextual variables as normal conversations. Because interrogation
becomes more and more a social interaction, you should understand what your
social rank is and how it compares to the walk-in, and how the conversation will
drift across time because of the social nature of the interaction.
Set up standard operating procedures
(i) Set up procedures for sending the walk-in to multiple decision makers who are
at different agency/departmental levels but all are free to have input (in a
nonhierarchical manner). These additional interrogators should be ready to be
involved as soon as needed:
a) Include relatively junior level (rank) interrogators, because they may
offer less of a threat and encourage the walk-in to let down their guard.
(b) However, the initial contact should be someone of a higher rank.
Culturally, there are certainly occasions where a high level source might
be unwilling to cooperate with a perceived lower level staff person and
‘your source walks out the door.’13
Perhaps unfortunately, the reality is that lower level staff persons are most
likely to be whom the walk-in first encounters. It is these individuals who
should be provided with the necessary tools and training to at least make a
reasonable screening of a walk-in before recommending further follow-up.
(c) A good practice is to include two to three interrogators working across
shifts continuously but on an individual basis. The other interrogators
would be able to watch on video or review tapes and use the dynamics of
personality differences to get the walk-in to show different responses to
different interrogators across time. The interrogators can compare notes
and then simultaneously appear before the walk-in and make him discuss
any issues on which they have noted discrepancies.
(ii) Videotape the interview and seek independent assessment. Research in a
variety of settings about how quickly people make general judgments about
interviewees has shown that accuracy is greatest in the first minute of contact, is
not improved by additional minutes up to a certain point, and then gradually
degrades as the length of the interview increases.14 It is advisable to have several
independent raters observe a few minutes of tape and make a judgment rather than
to place the evaluation into the hands of the actual interrogator who sits with the
subject for up to several hours.
13
14

G. Hazlett, personal communication (2003).
G. Hazlett, personal communication (2003).

 10

(iii) Keep detailed records: Collect data about the antecedent and consequent
events to know rates of false alarms and misses, as well as successes.
Cues to deception vary with particular deceivers and particular situations. The
most promising way to learn about cues to deception or to train people to detect
deception is to study the particular context of interest.15 Interrogators should
learn about deception from videotapes of suspects/walk-ins known to have been
lying or telling the truth during actual interrogations. The particular training
technique of showing trainees a clip, asking them to guess whether the person on
the tape was lying or telling the truth, then telling them the answer has several
advantages. First, it has been shown to at least modestly improve the accuracy of
deception detection. Second, it can be very informative and sobering to
professionals who often have greater confidence in their own lie-detection
abilities than is warranted. Third, although the creation of such a collection of
tapes can be challenging, once they are in hand, the training procedures are easy
to implement.16, 17
Implement evaluation mechanisms
Set up routine methods for the interrogator to assess himself for operating
categorically in a manner that may obscure important details. Such assessment
should include issues of:
(i) How motivated is the interrogator? The embassy personnel may never
find someone who has useful information, despite many interviews. How
can the motivations of the interrogator be maintained under conditions of
low hit rates?

15 B. DePaulo, personal communication (2003).
16 Mann, S., Vrij, A., & Bull, R. (2002). Suspects, lies and videotapes: An analysis of authentic highstake liars. Law and Human Behavior, 26, 365-376; Vrij, A., & Mann, S. (2001). Telling and detecting
lies in a high-stake situation: the case of a convicted murderer. Applied Cognitive Psychology, 15, 187203; Mann, S., Vrij, A., & Bull, R. (in press 2003). Detecting true lies: Police officers’ ability to detect
suspects’ lies.
17
“The training studies I find most intriguing are ones in which the trainees are given no specific
instructions at all. Instead, they watch a clip of someone lying or telling the truth, indicate their judgment as
to whether the person was lying or telling the truth, then are told the correct answer. In some studies,
trainees show some improvement in accuracy after just 8 trials. (E.g., Zuckerman, Koestner, & Alton, 1984,
“Learning to detect deception,” Journal of Personality and Social Psychology, 46, 519-528.) I think it
might be promising to train people in this way with tapes of the kinds of deceptions that they face on the
job. For example, if there were a collection of videotapes of walk-ins, and it were known definitively which
people were lying and which were telling the truth, it could be useful to train people who deal with walkins using these tapes. “ (B. DePaulo, personal communication, 2003)

 11
(ii) How stressed is the interrogator (due to fatigue or job pressure, etc)?
Individuals under stress are more likely to perceive a threat than someone
who is not stressed.18
(iii) How long has the interrogator been conducting interrogations in the
absence of feedback? When there is no feedback loop, judges of
deception are likely to feel over-confident of their ability to detect
deception, and subsequently more likely to make erroneous judgments.
Interrogators are usually rewarded by getting “something,” and not by
getting “the right something.” Many mistakes are made and a lot of poor
work is done that goes unchallenged. It is not that interrogators don't want
to do well, it is just that there is very little follow-up for most.
(iv) As a caveat regarding feedback, it is important to understand the
powerful effects of hindsight bias. If errors in judgment are made,
reevaluate the facts as they were known before the error was discovered,
so that you can see what facts were ignored or misinterpreted.
As much as possible, control the context of the interview
(i) If the interview occurs in an environment that is under your control (e.g., a
room set aside for this purpose), you have the opportunity to set up the context to
your own advantage: in some situations, hearing a threat lurking outside the door
can be conducive to greater cooperation in the room. High ambient noise may
interfere with the capacity to carry on tasks that require high degrees of
concentration and attention.19 If the walk-in is obviously naïve, signs that
indicate you have much information already (folders, maps, etc.) and that you
have sophisticated technology available to you, may be effective. However, if
these are a ruse and the walk-in discovers this, this damages the interaction
significantly. (A few verified facts placed carefully to ensure that walk-in that he
is known and has little to gain from deception, is more effective.)
Alternatively, it usually is true that people are more comfortable in their own
surroundings, where there may be fewer cues to resistance. This is the kind of
issue which speaks to the need to conduct questioning/inquiry covertly in the
community in which the walk-in lives in vs. in a government holding facility.

18

Vila, B. J., Dennis, J. K., Morrison, G. B., & Reuland, M. (2000). Evaluating the effects of fatigue on
police patrol officers. Washington DC, U.S. Department of Justice; Vila, B., & Taiji, E. Y. (1999). Police
work hours, fatigue and officer performance. In D. J. Kenney & R. McNamara (Eds.), Police and Policing
(2nd Ed.). Westport CT: Praeger.
19
Military psychological operations conduct loud-speaker operations that, although grating to listen to and
appearing to be somewhat silly, have been shown to be effective in increasing the mental load on targets
(G. Hazlett, personal communication, 2003).

 12
(ii) If possible, take advantage of the fact that most people function better
cognitively during the daytime than during the nighttime, and that taking people
out of their normal sleep-wake cycle can degrade their cognitive performance.20
Sophisticated resistance to interrogation requires a lot of mental agility. Anything
that reduces that agility may increase the likelihood of the resistant subject
making errors. However, this kind of technique generally means that you have
the capacity to exercise that level of control over the walk-in's environment and
schedule.
(iii) Consider when and how to increase the anxiety of the walk-in:
Anxiety and fear are known to have enhancing effect on some memory and
cognitive tests when they are at a medium level, and a debilitating effect when
very high.21 However, when the cognitive challenge is complex (as deception
might be), anxiety generally degrades performance.22 If the goal is to break a
deceptive cover, enhancing anxiety might be an effective interrogation strategy.
An individual who arrives at an embassy and appears to have little anxiety may be
worth challenging once his initial pitch has been made. In such cases, one might
initially assume that either the walk-in is a foreign intelligence operative sowing
disinformation, or he is a merchant of information with something to sell and has
no belief that such a transaction constitutes any threat to him.
In formal interrogation settings, anxiety can have a role to play, particularly in the
early phases soon after capture when tactical intelligence is desired. Most
detained individuals will relatively quickly habituate to captivity and soon learn
that there are limited physical pressures that can be exercised against him.
Anxiety and fear are based on not knowing what may happen to you and can be a
strong motivator. Once that ignorance is removed, it is difficult to overcome
resistance.
It is not unreasonable to engage in deliberate deception operations against a
detained terrorist that may cause him some anxiety. However, individuals who
have been the most successful at interrogation are those who allowed the subjects
20

Englund, C. E., Ryman, D. H., Paul, N., & Hodgdon, J. A. (1985). Cognitive performance during
successive sustained physical work episodes. Behavior Research Methods, Instruments & Computers, 17,
75-85; Haslam, D. R. (1983). The incentive effecy and sleep deprivation. Sleep, 6, 362-368; Haslam, D.
R. (1985). Sustained operations and military performance. Behavior Research Methods, Instruments &
Computers, 17, 90-95; Mertens, H. W., & Collins, W. E. (1985). The effects of age, sleep deprivation, and
altitude on complex performance. FAA Office of Aviation Medicine Reports. Rpt. No. FAA-AM-85-3,
18, US: Aviation Medicine.
21
Broadhurst, P. (1957). Emotionality and the Yerkes-Dobson Law. Journal of Ecperimental Psychology,
54, 345-351.
22
Sarason, S. B., Hill, K. T., & Zimbardo, P. (1965). A longitudinal study of the relation of test anxiety to
performance on intelligene and achievement tests. Monographs of the Society for Research in Child
Development, 29, 1-51; Shephard, A. H., & Abbey, D. S. (1960). Manifest anxiety and performance on a
complex perceptual motor task. Perceptual & Motor Skills, 8, 327-330; Sinha, D., & Singh, T. R. (1960).
Manifest anxiety and performance on problem solving tasks. Journal of Consulting Psychology, 23, 469.

 13
to generate their own anxiety and did not rely upon aggressive/physical means to
extract information. A threat is only good if there is some basis for believing it.
Once it is discovered to be a false threat, the result is enhanced resistance from the
target. Intimidation also is a limited use phenomenon because it habituates
quickly when no harm is done. There is also the danger that if a threat is made,
the interrogator may at some point feel compelled to carry it out. Once a false
threat is identified or intimidation is perceived as a nuisance vs. a threat,
credibility of the interrogator is lost along with any efficacy in that role.
(iv) Consider when and how to decrease the anxiety of the walk-in:
The possibility should be considered that individuals who really have something
to offer in terms of information are likely to be generating anxiety on their own.
Increasing anxiety is only likely to scare such an individual away. Making
contact with Americans aversive is not likely to be an effective long-term strategy
for encouraging anyone to come in with information.
It would seem desirable to encourage rather than discourage individuals to contact
U.S. interests abroad. For example, if it is a regular occurrence for citizens in
Baghdad to come to the embassy to sell bogus information for $20, then anyone
who really has some valuable information always has a cover story about why he
is in the Embassy. Visiting the embassy may become a common practice in the
community, which may make it more difficult to ascribe responsibility to anyone
when there is suspicion that some real information was actually passed and acted
upon. This means more work for some embassy staff, maybe even some
dedicated staffers to this task. In the end it would be relatively cheap. Also,
among any pool of subjects seeking to lie for a few bucks, there is someone who
will tell the truth for a few more if properly recruited.
General guide for deceiving an interviewer: You are given the task of deceiving an
interrogator. What should you do?
(i) Minimize dishonesty. In process of an interview where the intent is to deceive,
the subject should as much as possible avoid telling lies, particularly
spontaneously developed ones that tend to fall apart very rapidly.
(ii) Control content. Stay on whatever line you have established as the
information you wish to relay. Avoid talking about subjects outside of that area.
(iii) Control quantity. Say very little more than what is necessary to communicate
the information you want to convey. The more you say, the more opportunities
there is for you to present potentially conflicting bits of data that can be used to
establish a lack of credibility.

 14
(iv) Volunteer nothing (source of effort). Credible information has to be drawn
out by an interviewer, he expects to work for it. What you volunteer may be
discarded or give away an obvious intent to push a particular message which
could arouse suspicion.
(v) Make your point and get out (limit time). If you are the walk-in, you should
make the point that you have information, validate yourself to a sufficient degree
and then get out so you can have greater control when you push the actual
deception. The longer you talk to the snake, the more likely you will be bitten.
(vi) Establish control (retain initiative). If you have sufficiently aroused the
interest of an interviewer or interrogator, it would be best to make the target (the
interrogator) work harder to get the information. The more work involved in
getting it, the more the information obtained will be valued and defended by the
target. A successful walk-in contact could be gauged by the willingness of the
target to follow up and seek an additional meeting(s). Setting the time and place
of subsequent contacts where the deceptive information is passed will allow the
source to minimize factors which might reveal the deception. Outside of the
embassy, a source can choreograph the remaining contacts for maximal
effectiveness. Consistent with this, never go back to an embassy once the initial
contact has been made.
(vii) Be realistic. Presentation of deceptive information should be framed in a
realistic context. Normal people are very uncomfortable with interrogation or
focused questioning of almost any kind. The display that accompanies any offer
of deceptive information should be consisted with what is normally seen by the
target. This is an area where target expectancy is in fact very important. If you
don't look like his usual customers, the poor fit with his existing schemas will
heighten his suspicion about your intent and the validity of your information.
Nonverbal cues may play a significant role in selling or invalidating a message.
(viii) Scenario training is essential. Some data regarding identity may have to be
divulged to validate you as a source. If you are to be successful, you should be
trained in significant depth on any line of deception that may be involved. You
must know the details of what you may talk about in depth and with multiple
associations so that the information resides in memory as part of a normal web of
knowledge as opposed to as a trivial and isolated fact. It is the third and fourth
question on a particular subject that tends to reveal the usually very shallow
nature of most attempts at deception.
(ix) Scenarios must involve layers of deception. If you can't control the situation
fully, it is necessary to provide people with a fall back position that allows them
to confirm interrogator suspicion while at the same time stopping further inquiry.
The longer one talks to a trained interrogator, the more likely he will make a
mistake and be challenged on it. If a deception effort is to be successful, the
target has to come away with some confidence in the data. Errors should be

 15
planned for in the execution of the effort and there should be engineered within
the scenario several fall back positions that will allow the source to be
"discovered" while at the same time retaining his validity as a source.23
Research challenges
What history of successes and failures do we have, and what can we learn from our
own history (avoiding hindsight bias, if possible)?
What use can we make of instances of deception among ourselves, such as the
experiences of undercover agents, spies, the selective sharing of information depending
on classification rating, from socially acceptable lies (e.g., about weight or age), and
from deception that occurs as a function of differential social status?24
How does the behavior of the U.S. embassy personnel (or whomever the officers are)
impact on how people offer information?
Might someone who wants to evade detection systems be able to fake a temporary
physical illness or mental illness in order to get past deception detection systems and
then once inside (the embassy, the hospital, the office, whatever) have access to secure
systems?
How important are differential power and status between informer and officer? What
role might social stigmatizing play?
How does motive affect social negotiation?
What are the American cultural stories (folk psychology) regarding truth and lies, and
how do we value these in different situations? What are other cultural stories?25
Scenario #2 - Law enforcement threat assessment post 9/11 lacks triaging mechanisms

23

G. Haslett (personal communication, 2003.)
Older Americans may be better than younger Americans at not revealing deception. See Parham, I. A.,
Feldman, R. S., Oster, G. D., & Popoola, O. (1981). Intergenerational differences in nonverbal disclosure
of deception. Journal of Social Psychology, 113, 261-269.
25
Research by Charles Bond and colleagues includes projects relevant where nationals from 60 countries
were surveyed about their beliefs concerning deception. The most commonly expressed opinion was that a
liar can’t look you in the eye. In general, he has shown that judgments of deception from behavior occurs
with about 55% accuracy across considerable situational variability, although sometimes this runs as high
as 75% in a situation-dependent manner. From Bond, C. F., Berry, D. S., & Omar, A. (1994). The kernel
of truth in judgments of deceptiveness. Basic and Applied Social Psychology, 13, 523-534; Bond, C. F., &
Atoum, A. O. (2000). International Deception. Personality & Social Psychology Bulletin, 26, 385-395;
Bond, C. F., Thomas, B. J., & Paulson, R. M. (in press). Maintaining lies: The multiple-audience problem.
Journal of Experimental Social Psychology.
24

 16
Law enforcement, especially the FBI and Secret Service, receive threats by phone, letters,
and email. Responding in terms of false positives and false negatives has significant impact
in terms of public safety and expending resources. How can these types of threats be
validated?
A current characteristic of this challenge is that there is (post 9/11) almost no triaging of
threats: law enforcement operates under the felt need to follow up on almost all threat
communications.
Another challenge is that this threat information is often very short – as short as a few
sentences. The brevity ensures that the message encapsulates very little information.
Law enforcement and intelligence agents also must avoid “enriching” the threatener by
making their reactions to a threat public, so that the threatener knows what the
police/intell action will be and can plan accordingly.
There is the need to acknowledge that people may have information that is false, but
which they believe to be true. In these instances, research shows that the usual cues to
deception are absent: there is no extra mental effort involved in maintaining the “lie,” and
no apparent behavioral indices of deception.26
Categorize threats and prescript appropriate responses
Determine a course of action depending on the specific combination of characteristics.
Threats can vary in terms of:
(i) Temporal (immediate/seconds vs longterm/years)
(ii) Type (domestic violence; bomb; WMD)
(iii) Intent of person making threat
(iv) Ability to carry out threat
(v) Vulnerability of target
(vi) Nature of target (President vs shopping mall)
(vii) Characteristics of threatening person
(viii) Characteristics of person to whom threat is being communicated
(ix) Characteristics of a likely deception

26

For an example of how assertions to false events can appear truthful, see Ceci, S. J. (2003). American
Psychologist, 58, 855-864.

 17
(a) Lies told for self-serving reasons
(b) Lies told for other reasons, or to be altruistic
(c Lies of omission and commission
(d) Bald-faced lies vs exaggerations
(e) Self-presentational lies (when liar cares about relationships with
others)
(f) Instrumental lies: for a particular purpose
(g) Lying as acceptable in business negotiations or among politicians, or
among the public generally.27
Analyses of narratives around threats have shown that there are predictors of the
imminence and severity of the threat. These analyses might be useful for some of the
threats and intelligence routinely received as described in this scenario.28
Research challenges
What is the impact of our felt need to act secretly? In haste? How would our
behavior be different if those needs were not present?
How can we analyze these situations using the procedures of decision-making
analysis that have been used elsewhere, for example, in the practice of medicine?
Scenario #3 - Law Enforcement Interrogation and Debriefing
Law enforcement routinely question witnesses and suspects regarding criminal activity. How
do you tell if the individual is telling the truth, lying, or something in between? Acts of
omission and acts of commission are both important to identify.
Research challenges: Personal characteristics of good deception detectors
Research with people who score well on at least three kinds of “lie detection” tests show
that (i) even these people did least well on those tests with which they had least
experience, indicating the operation of a significant effect of experience with the tests. It
was suggested that good deceivers are like Olympic champions or Grand Masters in
27

In the World Values Survey (http://wvs.isr.umich.edu/), people in 35 out of 35 countries responded that
they agreed with the statement that it sometimes justifiable to lie in one’s own self-interest. This had a
positive correlation with educational level. One problem with cross-cultural analyses is that lying means
different things in different cultures, and in some social situations, rudeness outweighs deception as a
prohibited behavior. In the U.S., out of more than 500 trait terms, “liar” is rated as one of the most
negative. (C. Bond, personal communication, 2003.)
28
Suedfeld, P., & Wallbaum, A. B. C. (1992). Modifying integrative complexity in political thought: Value
conflict and audience disagreement. International Journal of Psychology, 26, 19-36; Suedfeld, P., & Bluck,
S. (1988). Changes in integrative complexity prior to surprise attacks. Journal of Conflict Resolution, 32,
626-635.

 18
chess: they are good because they practice extensively.29 (ii) It also was found that about
80-90% of these individuals had unusual childhood experiences that might be understood
as giving them an outsider point of view, and aware of emotionality early in life. (iii)
There was some evidence that the ability to detect deception well correlated positively
with (a) high scores on the general SAT, and (b) an interest in psychological
phenomenon. Generally, detection of deception is better under lower motivational levels
than under higher motivational levels.30
It was suggested that such individuals might be identified also as Personality Type D –
that is, as individuals with a pervasively pessimistic view of life. This pessimism might
give them a predilection for setting their threshold for determination of the deceptive
content of a message at a relatively low level. It might also be considered that these
individuals show a lower threshold for other kinds of detection tasks, such as motion
detection tasks, and that the predilection for setting a low threshold, evident in those
tasks, might be useful as a screening mechanism for good lie detectors.31
It should be recognized, however, that a trait psychology view – where individuals good
at deception or deception detection are identified via personality characteristics and
personality testing -- may not be useful: in this scenario it is not desired to know how
hard it is for someone to lie, but whether someone is lying at that point in time.
We can be informed by research on self-persuasion, which attempts to understand what
makes us believe our own thoughts. The self-validation hypothesis is that attituderelevant thoughts that are perceived by oneself as valid have a strong impact on attitudes,
and those that are perceived as invalid, do not. In support of this hypothesis, behavior
that increases one’s confidence in one’s own thoughts has been shown to have a
favorable impact on one’s attitude as towards those thoughts. For example, (i) if a
thought comes to mind quickly, it is viewed as more credible (by our selves), and (ii) if
people write a message with either their dominant or nondominant hand, they are more
likely to believe what they wrote if it is written with the dominant hand.32
Research challenges: characteristics of normal conversations
Research has shown that everyday conversations have the following characteristics:33
(i) Maxim of quantity: all necessary information about a topic is provided
29

Maureen O’Sullivan, The Diogenes Project.
Forrest, J. A., & Feldman, R. S. (2000). Detecting deception and judge’s involvement: Lower task
involvement leads to better lie detection. Personality & Social Psychology Bulletin, 26, 118-125; DePaulo,
B. M., & Tang, J. (1994). Social anxiety and social judgment: The example of detecting deception.
Journal of Research in Personality, 28, 1420153.
31
Roy Freedle (personal communication, 2003).
32
Brinol, P., & Petty, R. E. (in press). Self-validation processes: The role of thought confidence in
persuasion. In G. Haddock & G. Mate (Eds.), Theoretical perspectives on attitudes for the 21st century.
Philadelphia, PA: Psychology Press.
33
Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and Semantics,
Vol. 3: Speech Acts (p. 41-58). New York: Academic Press
30

 19

(ii) Maxim of quality: the information given is true and not misleading
(iii) Maxim of relation: information given is relevant to the goal of the
conversation
(iv) Maxim of manner: the communication is clear, brief, and orderly, and
ambiguity and obscurity is avoided.
Research might consider the extent to which these rules are gender-dependent and
culture-dependent.
Research challenges: the reliability of technological devices to detect deception34
(i) There are no data to support the assertion that anxiety related to deception results
in measurable fluctuations in vocal recordings.
(ii) Thermal imaging – reflecting changes in blood flow, particularly around the eyes
– may be successfully combined with polygraph measures to enhance deception
detection; the method is under development.
(iii) Pupillometry, the measurement of changes in pupil size or movement, has been
shown to identify guilty participants in Concealed Information Test studies, but this
response can be manipulated voluntarily by the subject trying to be deceptive, and
this response occurs in situations not involving deception, such as in conditions of test
uncertainty.
(iv) Event-Related Potential (ERP): The act of deception is built upon several
cognitive processes, each of which is likely to be related to portions of the ERP
signal. Three components of ERP (elicited by exposure to stimuli that have some
task relevance35) most often have been investigated in deception research: two P300
components, P3b and P3a, and N4. Attention switching based on deception appears
to be related to changes in the P3a, an early waveform in the anterior region of the
brain; however, the P3a is coincident with general shifts in attention as well. Making
the decision to lie is related to suppression in the P3b, the most frequently-used ERP
measure of deception, but the P3b probably is involved in many types of higher
cortical functions. The N4 component is sensitive to semantic incongruity, assumed
to be part of the deception process, but obviously is part of other cognitive processing
as well. The P300 is heavily impacted by characteristics of the stimuli, and there is

34

Most of this summary is from J. M. C. Vendemia, “Alternative Technologies for the Detection of
Deception,” University of South Carolina.
35
Donchin, E., Smith, D. B. (1971). The contingent negative variation and the late positive wave of the
average evoked potential. Electroencephalography & Clinical Neurophysiology, 29, 201-203.

 20
evidence that certain individuals, such as those with PTSD36 or convicted
murderers,37 do not produce a P300 related to recognition memory.
(v) Functional magnetic resonance imaging is only beginning to be used in the
analysis of deception. The changes in cerebral blood flow are assumed to reflect
changes in cognitive load and demand.38 This technology is expensive and not
applicable to most operational settings, however.
Additional research challenges
How do we find out if the informant has knowledge of which s/he is not aware?
What pharmacological agents are known to affect apparent truth-telling behavior?
What are the mechanisms and processes of learning to lie? Can these be acquired
within relatively short periods of time (e.g., within a polygraph test session)?
What are sensory overloads on the maintenance of deceptive behaviors? How might
we overload the system or overwhelm the senses and see how it affects deceptive
behaviors?
What are the effects the personal characteristics of an examiner in polygraph test
situations?
Scenario #4 - Intelligence Gathering
Many government agencies, including but not limited to DOD, FBI, DOE, CIA, and the
Department of Treasury, gather “intelligence” of one type of another. A high priority is
how to evaluate the authenticity of both the source of information and the information
itself. Issues that need to be addressed include whether the person is who s/he says s/he
is, whether s/he has access to the information s/he purports to have, and whether the
information accurate.
Systematic data collection: needs and challenges

36

McFarlane, A. C., Seber, D. L., & Clark, C. R. (1994). Abnormal stimulus processing in posttraumatic
stress disorder. Biological Psychiatry, 34, 311-320.
37
Kiehl, K. A., Hare, R. D., Liddle, P. F., & McDonald, J. J. (1999). Reduced P300 responses in criminal
psychopaths during a visual oddball task. Biological Psychiatry, 45, 1498-1507.
38
Spence, S. A., et al. (2001). Behavioural and functional anatomical correlates of deception in humans.
Brain Imaging, 12, 2849-2853; Lee, T. M. C., et al. (2002). Lie detection by functional magnetic
resonance imaging. Human Brain Mapping, 15, 157-164; Ganis, G., Kosslyn, Sm M., Stose, S.,
Thompson, W. I., & Yurgelun-Todd, D. A. (2003). Neural correlates of different types of deception: An
fMRI investigation. Cerebral Cortex, 13, 830-836.

 21
The enterprise would be more effective if data were collected and systematically
reviewed on previous instances of the veracity of intelligence communications and their
sources. The challenges here are:
(i) Field officers or embassy personnel may become personally identified with a
source, and protect the source even when they are of little use or are being
deceptive.
(ii) When the field officer or embassy personnel is transferred out, their source
may become unavailable to others; it was the personal relationship of handler to
source that made the source useful.
(iii) There is more incentive for a source to provide some information than to
provide verifiable information.
(iv) If or when confirming evidence is available, it is likely to be swamped by
other data, so that the isolated initial data are not properly evaluated, and
considerable hindsight bias is operating.
(v) The creation of databases is time-consuming and expensive, and the products
are only as good as the data that are entered. These databases also must be made
available to the people who need them, which necessitates an understanding of the
local culture of these personnel (their needs, skills, and knowledge), and sufficient
training for use must be implemented.
General research challenges
What are the working definitions of deception and deception detection? Should
these include spinning, withholding, false memories, trickery, plausible denial
statements, and dissociative memories and personalities? What are the relationship of
these to error & lapses of memory, self-deception, Mass Psychogenic Illness, and
malingering?
Can we reasonably map types of lies, or conceptions of lies, to specific behaviors and
to the likelihood of their occurrence to specific situations (e.g., lies about bombs in
suitcases to airport security)?39 Once we do this, can we create categories and
relationships among categories, and then test the validity of such across samples?
Might it be possible to greatly increase the likelihood of success if we agree that
different strategies are needed to different types of interactions? That is, to propose
that the same methodology may not be appropriate for personnel selection as for
surveillance or law enforcement or airport screening?

39

There is a multitude of modeling techniques available for this kind of an approach; it was suggested that
connectionist models – e.g., J. L. McClelland & D. E. Rumelhart (1986), Parallel Distributed Processing.
MIT Press, Cambridge, MA. – might be considered.

 22
Consider that deception systems co-opt perceptual systems for all the senses – vision,
hearing, taste, smell, and touch. For example, when a primate signals the presence of a
predator to his or her cohorts, the other animals in the group fail to employ their own
perceptual systems (that is, they don’t wait to look for themselves, or smell for
themselves, or listen for danger). The normal sensory systems are short-circuited.
Might an individual’s ability to tune into signals from their own perceptual systems
correlate with their ability to detect deception?
Consider situations where the usual cues to deception are absent. Might we be more
receptive to other cues to deception if they are not overshadowed by the obvious cues?
For example, Secret Service personnel often are described as scoring better than the
average in deception detection tests. Might they be good at this because they have
learned to detect less obvious or more subtle cues – such as nonverbal cues that are
evident in the behavior of individuals in crowds? How might this issue relate to
“implicit measures” of deception detection, discussed earlier?
Further investigation of the cross-cultural variations in physiological measures of
emotionality. For example, it has been shown that psychopaths show less “fear
potentiated startle,” a common measure of anxiety, than do normals.40
Further investigation is needed into methods of inoculation against confirmation
biases.
What are the important general receiver characteristics? That is, how does how you
respond to me change the way I deceive?
We often assume deceit involves malice and intent to harm. But it is as often
motivated by compassion, dedication to good, or religious beliefs. How might
deception behaviors be different depending on the underlying motivational system?
If we are to construct a system for deception detection, how can we ensure that it
“fails well?” (I.e., that it might stretch and sag before breaking, each component
failure leaving the whole as unaffected as possible.)
Research is needed into how interrogators might take advantage of some of the
transference and counter-transference strategies used by psychotherapists.
Research is needed about the relevant ethical and personal privacy issues in deception
research that must be addressed?
Is it possible to determine deceit in an individual by a change in the pattern of social
networking and interactions that that individual displays, or in changes in patterns of
friends? Lying is damaging to social networks, so if someone engaged in more
40

Patrick, C. J. (1994). Emotion and psychopathy: Startling new insights. Psychophysiology, 31, 319-330.

 23
deception than usual, this should force a change in that network. This has been seen in
retrospective analyses: can it be used prospectively?
What are the implications of the apparent fact that lying is ubiquitous in some sense,
and yet about 90% of those who are tested for security clearances receive them?
What is the feasibility and practicality of using priming techniques and implicit
measures of cognition and affect to determine veracity? These are relatively easy to
administer and they provide immediate feedback.41 There is controversy about the
extent to which such measures reflect familiarity versus veracity, at present. If the
issue is one of whether people are who they say they are, there is a wide range of
techniques available to assess this described in self-system research, which investigates
how people conceive of themselves. Self can be accessed by others without awareness
of the person, for example, by assessing reaction time when confronted with one’s
name in a group of names.42

Participant list (partial)
Anderson, Bill
Band, Stephen
Bennett, Michael
Bond, Charlie
Brandon, Susan
Capps, Mike
Demaine, Linda
DePaulo, Bella
Freedle, Roy
Gerwehr, Scott
Griffin, Jim
Griffith, Doug
Happel, Mark
Hazlett, Gary
Helms, Hollis
Hubbard, Kirk
Jessen, Bruce
Kinscherff, Robert
Lang, Eric
Lassiter, Dan
Mericsko, Bob
41

For a description of the Implicit Attitude Test and demonstration of how to use the test, see
http://www.tolerance.org/hidden_bias/index.html.
42
Koole, S. L. & Pelham, B. W. (2003). On the nature of implicit self-esteem: The case of the name letter
effect. In S. J. Spencer & S. Fein (Eds.), Motivated social perception: The Ontario symposium, Vol. 9.
Ontario symposium on personality and social psychology. Mahwah, NJ, US: Lawrence Erlbaum Associates
(pp. 93-116).

 24
Mitchell, Bob
Mitchell, James
Morgan, Andy
Morris, Jon
Mumford, Geoff
O'Sullivan, Maureen
Pelham, Brett
Petty, Richard
Pinizzotto, Anthony
Povinelli, Daniel
Rosal, Carmel
Rosmarin, David
Strong, Gary
Vendemia, Jennifer
Zeffiro, Tom