Reflection

ORIGINAL ARTICLE

Increasing Cognitive Load to Facilitate Lie Detection: The Benefit of Recalling an Event in Reverse Order

Aldert Vrij Æ Samantha A. Mann Æ Ronald P. Fisher Æ Sharon Leal Æ Rebecca Milne Æ Ray Bull

Published online: 13 August 2007

� American Psychology-Law Society/Division 41 of the American Psychological Association 2007

Abstract In two experiments, we tested the hypotheses that

(a) the difference between liars and truth tellers will be greater

when interviewees report their stories in reverse order than in

chronological order, and (b) instructing interviewees to recall

their stories in reverse order will facilitate detecting decep-

tion. In Experiment 1, 80 mock suspects told the truth or lied

about a staged event and did or did not report their stories in

reverse order. The reverse order interviews contained many

more cues to deceit than the control interviews. In Experiment

2, 55 police officers watched a selection of the videotaped

interviews of Experiment 1 and made veracity judgements.

Requesting suspects to convey their stories in reverse order

improved police observers’ ability to detect deception and did

not result in a response bias.

Keywords Verbal and nonverbal cues to deception � Lie detection � Cognitive load

With current concerns over security, it is becoming

increasingly important to discriminate between suspects

who lie versus those who tell the truth. Nevertheless, a

substantial empirical base shows that laypeople and even

trained investigators (e.g. police) are often poor at dis-

criminating between liars and truth tellers (Vrij 2000, 2004,

in press). Obviously, liars’ behaviour often does not differ

much from truth tellers’ behaviour—at least not as we

currently measure them—and so the task of discriminating

between them is quite difficult. One reason for observers’

poor discriminatory performance, we believe, is that they

take a passive approach to the task of detecting deception.

That is, observers simply monitor liars’ and truth-tellers’

naturally occurring behaviours during an interview and

look for various non-verbal and verbal cues to distinguish

between them. Clearly, these discriminatory signs are not

so obvious. We suspect that observers could improve their

deceit-detection performance by taking a more active

approach to the task, and specifically by introducing a

manipulation that will magnify the differences between

liars and truth tellers. By definition, anything that magnifies

the difference between liars and truth tellers should

enhance our ability to discriminate between the

two—assuming, of course, that we can measure these

magnified differences. In this article we introduce such an

intervention that we expect to magnify the differences

between liars and truth tellers: Asking interviewees to re-

port their stories in reverse order. We argue that this will be

particularly debilitating for liars, because their cognitive

resources have already been partially depleted by the

cognitively demanding task of lying. As a result, we ex-

pected that more non-verbal and verbal differences would

emerge between liars and truth tellers in the reverse order

interviews than in the chronological interviews (Experi-

ment 1), which should facilitate the observers’ task of

discriminating between them (Experiment 2).

Cognitive Demand

One reason why differences in non-verbal and verbal cues

occur between liars and truth tellers is that lying can be

A. Vrij (&) � S. A. Mann � S. Leal � R. Milne

Psychology Department, University of Portsmouth, King Henry

Building, King Henry 1 Street, Portsmouth PO1 2DY, UK

e-mail: [email protected]

R. P. Fisher

Florida International University, North Miami, FL, USA

R. Bull

University of Leicester, Leicester, UK

123

Law Hum Behav (2008) 32:253–265

DOI 10.1007/s10979-007-9103-y

cognitively more demanding than truth telling (DePaulo

et al. 2003; Zuckerman et al. 1981). Several aspects of

lying contribute to this increased mental load. First, for-

mulating the lie itself may be cognitively taxing. Second,

liars are typically less likely than truth tellers to take their

credibility for granted (DePaulo et al. 2003; Gilovich et al.

1998; Kassin 2005; Kassin and Gudjonsson 2004; Kassin

and Norwick 2004; Vrij et al. 2006d). As such, liars will be

more inclined than truth tellers to monitor and control their

demeanour so that they will appear honest to the lie

detector (DePaulo and Kirkendol 1989), which should be

cognitively demanding. Third, because liars do not take

credibility for granted, they may monitor the interviewer’s

reactions more carefully in order to assess whether they are

getting away with their lie (Buller and Burgoon 1996;

Schweitzer et al. 2002). Carefully monitoring the inter-

viewer also imposes cognitive load. Fourth, liars may be

preoccupied by the task of reminding themselves to act and

role-play (DePaulo et al. 2003), which requires extra cog-

nitive effort. Fifth, liars have to suppress the truth while

they are lying and this is also cognitively demanding

(Spence et al. 2001). Finally, whereas activating the truth

often happens automatically, activating a lie is more

intentional and deliberate, and thus requires mental effort

(Gilbert 1991; Walczyk et al. 2003, 2005).

Obviously, lying is not always more cognitively

demanding than truth telling (McCornack 1997). Perhaps

the six reasons given as to why lying is more cognitively

demanding could give us insight into when it is more

cognitively demanding. That is, lying is more cognitively

demanding to the degree that these six principles are in

effect. For example, lying is likely to be more demanding

than truth telling only when interviewees are motivated to

be believed. Only under those circumstances can it be as-

sumed that liars take their credibility less for granted than

truth tellers and hence will be more inclined than truth

tellers to monitor their own behaviour and/or the inter-

viewer’s reactions. Second, for lying to be more cogni-

tively demanding than truth telling, liars must be able to

retrieve their truthful activity easily and have a clear image

of it. Only when liars’ knowledge of the truth is easily and

clearly accessed will it be difficult for them to suppress the

truth. On the other side of the equation, truth tellers also

need to have easy access to the truth for the task to be

relatively undemanding. If truth tellers have to think hard

to remember the target event (e.g. because it was not dis-

tinctive or it occurred long ago), their cognitive demands

may exceed the cognitive demands that liars require for

fabricating a story.

In experimental studies researchers ensure that inter-

viewees are motivated (typically by giving a reward for

making a credible impression) and that the target event is

easily retrieved (typically by interviewing the suspects

shortly after informing them about the target event). In

those experiments lying has been found to be more

demanding than truth telling in various settings. Partici-

pants who have directly assessed their own cognitive load

report that lying is more cognitively demanding than truth

telling. This occurred when lengthy, elaborative responses,

were required (Granhag and Strömwall 2002; Hartwig et

al. 2006; Strömwall et al. 2006; Vrij et al. 2001b, 2006d:

Vrij and Mann 2006; White and Burgoon 2001), and also

when short responses were sufficient (Caso et al. 2005;

Vrij et al. 1996, 2006b). In fMRI deception research, lying

and truth telling is differentiated only by the act of pressing

either a ‘‘lie’’ or ‘‘truth’’ button. Nevertheless, participants’

brain activity reveals that lying is more cognitively

demanding than truth telling (Spence et al. 2004).

In forensic settings, we can reasonably assume that

interviewees will be motivated, but we cannot assume that

interviewees will always be able to retrieve the target event

easily, as this should vary from one case to another.

Analyses of police interviews with real-life suspects ,

however, suggests that lying is often more cognitively

demanding than truth telling in the forensic setting. First, in

those police interviews, lies were accompanied by

increased pauses, decreased blinking, and decreased hand

and finger movements, all of which are signs of cognitive

load (Mann et al. 2002; Vrij and Mann 2003). Second,

police officers who saw videotapes of these suspect inter-

views reported that the suspects appeared to be thinking

harder when they lied than when they told the truth (Mann

and Vrij 2006).

Magnifying the Differences between Liars and Truth

Tellers

Although liars should experience more cognitive load than

truth tellers, the differences between liars and truth-tellers

may be relatively small, and perhaps not readily discern-

able by observers (DePaulo et al. 2003; Zuckerman et al.

1981). Our goal, therefore, was to magnify the differences

between liars and truth tellers. Vrij et al. (2006b, in press)

suggested that this might be accomplished by devising an

interview protocol that posed excessive cognitive demands

on the suspects (e.g. recalling events in reverse order). The

underlying assumption is that cognitively demanding

interviews (reverse order) will be particularly debilitating

for liars, whose cognitive resources have already been

partially depleted by the cognitively demanding task of

lying. This is analogous to the finding in the cognitive-

attention literature that information processing in the

primary task is slower in dual-task conditions than in

single-task conditions (Briggs et al. 1972).

Either of two mechanisms of attention might account for

the expected finding, that liars will be particularly debilitated

254 Law Hum Behav (2008) 32:253–265

123

in the demanding interview protocol (reverse order recall).

One possible mechanism is that the cognitive demands of

lying and describing an event in reverse order are drawn from

a common, limited pool of cognitive resources. When the

cognitive resources required to perform two demanding tasks

exceed the limit of attention, there will be a breakdown in

performance, the so-called divided-attention (e.g. Johnston

et al. 1970). Such a breakdown, or system overload, is more

likely to occur when the two tasks are each cognitively

demanding than when one of the tasks is relatively simple or

well practiced (recalling in chronological order). A second

possible attention mechanism suggests that performing two

tasks simultaneously entails shifting attention rapidly between

the two tasks (Broadbent 1957). The more cognitively

demanding a task is (reverse order recall), the more attentional

time it demands, thereby leaving less attention devoted to the

second task. Ultimately, the less attention that is devoted to a

task, the poorer performance will be (Kahneman 1973). We do

not attempt to discriminate between these two attentional

mechanisms, as both predict that liars should be particularly

debilitated in the more demanding reverse-order interview

protocol.

We selected reverse order recall as the interview pro-

tocol to increase cognitive demand because (a) it runs

counter to the natural forward-order coding of sequentially

occurring events (Gilbert and Fisher 2006; Kahana 1996)

and (b) it disrupts reconstructing events from a schema

(Geiselman and Callot 1990). Empirical support that re-

verse order recall is, in fact, resource-demanding derives

from a time-sharing study showing that performance on a

concurrent psycho-motor task declines when the memory

list is recalled in reverse order recall rather than in forward

order (Johnston et al. 1970). Although the reverse-order

strategy has been used for other purposes (as a memory

enhancing technique within the Cognitive Interview, Fisher

and Geiselman 1992), we used it here only because of its

high cognitive demand.

Verbal and Non-verbal Cues of Cognitive Load

We examined several verbal and non-verbal cues that are

associated with cognitive load.

Verbal Cues

Describing events in detail is typically more cognitively

challenging for liars than for truth tellers (Köhnken 1996,

2004; Vrij 2000, 2005). Liars may lack the imagination to

invent many details, or they may find it difficult to fabricate

a detailed story that sounds plausible. Verbal lie detection

tools that assess the number of details mentioned by

interviewees, such as Criteria-Based Content Analysis

(Vrij 2005) and Reality Monitoring (Masip et al. 2005),

distinguish between general categories of details and more

specific types of details. In the present experiment we

examined only general categories of details: Visual details

(details about what someone saw), auditory details (details

about what someone heard), and contextual embeddings

(details about locations, ‘‘the stool was underneath the ta-

ble’’ and details about time, ‘‘about one minute later...’’).

Since we assume that lying is more cognitively demanding

than truth telling, and that telling a story in reverse order is

more cognitively demanding than telling a story in chro-

nological order, we hypothesised that liars would include

fewer visual, auditory and contextual embedding details in

their stories than truth tellers, and that this would occur

mainly in the reverse order condition.

The final verbal cue we examined, a cue belonging to

the RM tool, was cognitive operations. This category varies

in its definition among Reality Monitoring researchers

(Masip et al. 2005). Here we define cognitive operations to

refer to evidence in the narratives of various cognitive

activities, such as thoughts or reasonings (‘‘I must have had

my coat on, as it was very cold that night’’) and cognitive

suppositions of sensory experiences , e.g. ‘‘She seemed

quite clever’’ (Vrij et al. 2007). Since cognitive operations

refer to cognitive activities, we could expect cognitive

operations to be more frequent in situations when people

carry out many cognitive activities, as in situations with

high cognitive load. We thus predicted that liars would

include more cognitive operations into their accounts

than truth tellers, and particularly in the reverse order

condition.

Vocal Cues

We examined several vocal cues that are associated with

cognitive load. Research has demonstrated that increased

latency (time lapse between question and beginning of the

answer), more pauses (between words or sentences), more

speech hesitations (use of speech fillers such as ‘‘um’’,

‘‘uh’’, ‘‘er’’ etc), more speech errors (grammatical errors,

stutters, false starts etc.) and a slower speech rate are all

associated with cognitive load (Goldman-Eisler 1968;

Smith and Clark 1993; Sporer and Schwandt 2006). We

therefore predicted that, in comparison to truth tellers, liars

would (a) display longer latency periods, (b) demonstrate

more pauses, speech hesitations and speech errors, and (c)

speak slower. Furthermore, these patterns would occur

particularly in the reverse order condition.

Visual Cues

Research has indicated that people tend to decrease several

kinds of movements when they have to think hard (Ekman

Law Hum Behav (2008) 32:253–265 255

123

1997; Ekman and Friesen 1972). This decrease occurs

because cognitive demand results in a neglect of body

language, reducing overall animation. We looked at several

types of movements: Illustrators (gestures that accompany

speech), hand/finger movements (movements of hand and

fingers without arms being moved), eye blinks, leg and foot

movements, and chair swivelling, and predicted that liars

would show fewer illustrators, hand/finger movements,

eye-blinks, leg/foot movements and chair swivels than

truth tellers, and particularly in the reverse order condition.

We looked at two more visual cues: Gaze aversion and

self-adaptors (scratching the head, wrists etc.). We did not

formulate hypotheses regarding those two cues because we

did not expect them to be associated with deception in our

experiment. Nevertheless we examined those two cues

because people typically believe that they are associated

with deception (Strömwall et al. 2004; Vrij et al. 2006a).

Gaze aversion is associated with cognitive load. People

tend to look away when they think hard, because looking

someone in the eye is too distracting when load is expe-

rienced (Doherty-Sneddon et al. 2002). Despite being

associated with cognitive load, gaze aversion does not

appear to be related to deception (DePaulo et al. 2003).

People typically believe that looking away makes a sus-

picious impression, and liars therefore avoid looking away

in an attempt to appear credible (Hocking and Leathers

1980; Vrij 2000). Self-adaptors are not associated with

cognitive load; rather, they are associated with experienc-

ing negative emotions (such as fear). Self-adaptors tend to

increase when certain negative emotions are experienced

(Ekman 1985/2001), but they are typically not associated

with deception (DePaulo et al. 2003). Perhaps they are not

related to deception for the same reason why gaze aversion

is not related to deception. People believe that fidgeting

makes a suspicious impression, and therefore liars avoid

fidgeting in order to appear credible.

Experiment 1

Method

Participants

A total of 80 undergraduate students participated: 40 males

and 40 females. Their average age was M = 20.88

(SD = 3.89) years.

Procedure

The experiment took place at a Students’ Union in a British

university. Undergraduates were recruited under the guise

of participating in an experiment about ‘telling a

convincing story’ with the possibility of earning £15. The

participants signed an informed consent form, and then

were randomly allocated to the truth telling condition or

the deception condition.

The 40 truth tellers participated in a staged event in

which they played a game of Connect 4 with a confederate

(who posed as another participant). (Connect 4 is a popular

two-player game where players drop counters into a slotted

grid to achieve, and simultaneously prevent their opponent

from achieving, four of their counters in a row). During the

game they were interrupted twice, first by another con-

federate who came in to wipe a blackboard and later by a

third confederate who entered looking for his or her wallet.

Upon finding the wallet, this latter confederate then

claimed that a £10 note had gone missing from it. The

participant was then told that s/he would be interviewed

about the missing money. We used the same event as Vrij

et al. (2006c, d) and Vrij et al. (2007).

The 40 liars did not participate in this staged event.

Instead, they were asked to take the £10 from the wallet,

but deny having taken this money in a subsequent inter-

view. They were told to tell the interviewer that they

played a game of Connect 4 just as the truth tellers had.

The liars were then presented with a document containing

the following information about the staged event that the

truth tellers had participated in.

‘You enter the room to find another participant,

‘Sam’, and the two of you play Connect 4 alone to-

gether for a while. You sat where you are sitting now

and the other participant sat opposite you. You had a

general conversation with the other participant as you

played, until the other participant’s mobile phone

rang and they excused themselves and left the room,

leaving you alone for a minute or so. When they

returned you continued playing the game. Then

someone else entered the room, made a comment

about you playing the game, wiped the mathematical

formulas that you can see off the board and then left.

You continued playing the game when someone else

entered the room looking for his/her wallet. The

wallet which you can see in front of you, is found

somewhere around the room (up to you to decide

where – it was varied in the scenario). You continue

playing the game when the experimenter came back

in, with the wallet-owner, and informs you both that

some money had gone missing from the wallet and

you are both to be interviewed.’

In summary, the liars did not engage in any of the activities

the truth tellers were engaged in (playing Connect 4, etc.).

Instead, the liars took the money out of the wallet, hid it

somewhere on their person, and pretended that they had

been engaged in the truth tellers’ activities. They therefore

256 Law Hum Behav (2008) 32:253–265

123

lied about the entire scenario, including taking £10 from

the wallet. The procedure reflects a situation where a liar is

familiar with the event s/he described but lacks the expe-

rience of true participation in that event.

Both liars and truth tellers were told that if they con-

vinced the interviewer that they did not take the money,

they would receive £10 for participating in this study. If

they did not convince the interviewer, they would instead

have to write a statement about what actually occurred. The

participants were then brought to the interview room where

they were interviewed by a uniformed, male, British police

officer. The interviewer was blind to the participants’

condition (truth telling or lying). The interviewer started

the interview by saying ‘‘£10 has gone missing from a

wallet in the room next door and I have to find out whether

or not it was you who took it’’. After several introductory

questions, the actual interview commenced. Participants

were asked to explain in as much detail as possible what

happened when they played Connect 4. The following

instructions were given by the interviewer to participants in

the non-instructed condition, and participants in the reverse

order condition. The parts denoted in italics were identical

for both and hence have not been repeated here. The

interviewer asked the 20 liars and 20 truth tellers in the

non-instructed condition:

‘‘Please tell me, in as much detail as possible, what

happened when you were in the room with Sam just

now. Mention all details, all conversations that took

place, and give as much information as you can about

everyone who entered the room, however irrelevant it

may seem. I will only be asking this one question.

You will have this one opportunity to give me as

much information as you can possibly remember.

Therefore, please tell me as much as you possibly

can, as I will use all the information you give me to

decide whether or not I think you are telling me the

truth. If you are unsure what I want you to do then tell

me now.’’

The 20 liars and 20 truth tellers in the reverse order con-

dition heard the following first two sentences in place of

the first sentence above:

‘‘I want you to tell me everything that happened in

the room just now with Sam, but in reverse order.

Therefore you should start your story with entering

this room, and end it with how you came to be in the

room next door with Sam....’’

After the interview the police officer gave each participant

a questionnaire, which he or she completed in another

room. Participants were asked the following three manip-

ulation checks: (i) to what extent they were motivated to

appear convincing during the interview, (ii) what they

thought the likelihood was of getting the £10, and (iii) what

they thought the likelihood was of being made to write a

statement. Answers were given on Likert scales ranging

from (1) very unlikely to (7) very likely. To ensure that all

participants were paid the same amount (£10), the experi-

menter told them that the police officer had been convinced

by their story.

Verbal and Non-verbal Coding

The interviews were videotaped and transcribed, and these

transcripts were the basis for all verbal coding. We had

previously trained two people to code visual cues, auditory

cues, contextual embeddings, cognitive operations, and

chronological productions. The two raters individually

coded the statements from the present study. They were

both blind to the hypotheses under investigation, to the

staged event, and to the experimental condition. One rater

coded all the statements and a second rater coded a random

sample of 40 statements (50% of the total). The two raters

coded per interview the frequency of occurrence of visual

details (e.g. ‘‘He walked over to the whiteboard’’ contains

three visual details); auditory details (e.g. ‘‘She said to sit

down’’ contains one auditory detail); contextual embed-

dings (e.g. ‘‘We started playing’’ is one temporal detail and

‘‘And then the pieces fell on to the floor’’ contains one

spatial detail); cognitive operations (e.g. ‘‘She seemed

quite clever’’ contains one cognitive operation); and chro-

nological production (the number of times the order of

event in the narrative differed from the chronological order

of the event, typically indicated by the participant by

saying ‘‘Before that...’’, ‘‘Prior to that...’’). The frequency

scores of the two raters correlated highly with each other

for each of the verbal cues (visual details, r = .97; auditory

details, r = .80; contextual embeddings, r = .72; cognitive

operations, r = .90; and chronological production, r = .92.

Coding of vocal cues and all visual cues occurred on the

basis of the videotapes, except for the coding of speech

hesitations and speech errors, which occurred on the basis

of the transcripts. We had used these coding schemes in

numerous experiments, including Mann et al. (2002), Vrij

(2006), Vrij et al. (1996, 1997, 2000, 2001a, b, 2004), and

Vrij and Winkel (1991, 1992). Another two raters indi-

vidually coded the videotapes. These raters were also blind

to the hypotheses under investigation, to the staged event,

and to the experimental condition. One rater coded all the

transcripts/videotapes and a second rated coded a random

sample of 16 transcripts/statements (20% of the total). The

following cues were coded: Latency period (a noticeable

pause of a second or more between the interviewer asking

the question and the interviewee responding with an

answer, bearing in mind that in this experiment only one

question was asked, r = .83); pauses (a noticeable pause of

Law Hum Behav (2008) 32:253–265 257

123

a second or more in the interviewee’s monologue, r = .91);

speech hesitations (frequency of saying ‘ah’ or ‘mm’ be-

tween words, r = 1.00); speech errors (frequency of word

or sentence repetition, sentence change, sentence incom-

pletion, and slips of the tongue, r = .87); illustrators (fre-

quency of arm and hand movements which were designed

to modify and supplement what was being said, r = .97);

hand and finger movements (frequency of movements of

the hands or fingers without moving the arms, r = .94); eye

blinks (where the eye shuts briefly but completely for a

blink, r = .97); leg and foot movements (frequency of

movements of feet or legs. Simultaneous movements of

feet and legs were scored as one movement, r = .98); chair-

swivels: (participants sat on a swivel chair; we counted

each time the chair was propelled one way or another by

the participant, r = .87); gaze aversion (number of seconds

for which the participant looked away from the interviewer,

r = .98); self-adaptors (frequency of scratching the head,

wrists, etc. Rubbing one’s hands together were not coded

as self-adaptors but as hand and finger movements,

r = .76). Speech rate was defined as the number of words

(calculated with the word count in Word divided by length

of answer in seconds). All visual and vocal cues except

speech hesitations and speech errors were adjusted for the

duration of the interview and were calculated per minute of

interview. Speech hesitations and speech errors were cal-

culated per 100 words.1

The average length of the interviews was M = 170.15 s

(SD = 80.1). To examine differences in length of interview

as a function of the experimental condition a 2 (Verac-

ity) · 2(Order) ANOVA was carried out. The analysis re-

vealed a significant main effect for Order, F(1, 76) = 5.93,

p < .05, g2 = .07 reflects that the Reverse Order interviews

lasted longer (M = 191.42, SD = 79.9) than the control

interviews (M = 148.88, SD = 76.4). None of the other

main or interaction effects were significant, all Fs < .89, all

ps > .35.

Results

Manipulation Checks

Four 2 (Veracity) · 2(Order) ANOVAs were conducted

with the four manipulation checks as the dependent vari-

ables. The ANOVA regarding chronological production

revealed one significant effect, a main effect for Order, F(1,

76) = 349.66, p < .01, g2 = .82. Participants in the Reverse

Order condition recalled their stories less chronologically

(M = 6.08, SD = 2.04) than participants in the control

condition (M = .13, SD = .33). In fact, all participants re-

called their stories in reverse order after being instructed to

do so. By comparison, participants in the control condition

hardly ever recalled non-chronologically.2

The experimental manipulations did not affect the par-

ticipants’ motivation (all Fs < .61, all ps > .43). The vast

majority of participants (80%) reported that they were

motivated to appear convincing during the interview (a

score of 5 or higher on the 7-point scale).

The ANOVA regarding the likelihood of receiving an

incentive of £10 resulted in main effects for Veracity, F(1,

76) = 14.05, p < .01, g2 = .16, and Order, F(1, 76) = 7.06,

p < .01, g2 = .09. Truth tellers were more convinced that

they would receive the incentive (M = 4.78, SD = 1.8) than

liars (M = 3.40, SD = 1.7) and participants in the control

condition were more convinced that they would receive the

incentive (M = 4.58, SD = 1.7) than the participants in the

Reverse Order condition (M = 3.60, SD = 1.8). Those re-

sults suggest that truth tellers thought that they performed

better than liars and that participants in the control condi-

tion thought that they performed better than participants in

the Reverse Order condition.

The ANOVA regarding receiving a penalty (writing a

statement) revealed a main effect for Order, F(1,

76) = 4.74, p < .05, g2 = .03. Participants in the Reverse

Order condition thought it more likely to receive a penalty

(M = 4.23, SD = 1.6) than participants in the control con-

dition (M = 3.48, SD = 1.6). This finding again suggests

that participants in the control condition thought that they

performed better than participants in the Reverse Order

condition. In summary, the participants were motivated to

be convincing, and participants in the Reverse Order con-

dition in particular thought that they performed worse than

participants in the control condition.

Hypotheses-Testing

A 2 (Veracity) · 2(Order) MANOVA was conducted with

the 16 verbal, vocal and visual cues as dependent variables.

The Veracity main effect, F(16, 61) = 4.20, p < .01,

g2 = .52, and Veracity · Order interaction effect, F(16,

61) = 2.46, p < .01, g2 = .39 were significant, whereas the

Order main effect was not, F(16, 61) = 1.22, p = .28.1 We did not calculate the verbal cues per minute of speech because

we believe that this changes the nature of the verbal cues. That is, the

number of details mentioned in a statement is different from the

number of details mentioned per 100 words, because the latter refers

to the conciseness of presenting information whereas the former does

not. When we included the duration of answer as a covariate in our

analysis, the results for the verbal cues showed the same pattern as

presented in the main text.

2 Perhaps one would expect liars to break up their stories in larger

chunks than truth tellers, as this is probably easier to do. Indeed, truth

tellers told their stories less chronologically (M = 6.65, SD = 2.0)

than liars (M = 5.50, SD 2.0), F(1, 38) = 3.36, p < .01, one-tailed,

g2 = .08. As such, the tendency to comply with the request to tell the

story in reverse order could be used as an indirect tool to detect deceit.

258 Law Hum Behav (2008) 32:253–265

123

Of interest for the hypotheses is the Veracity · Order

interaction effect. At a univariate level significant effects

emerged for auditory details, F(1, 76) = 11.07, p < .01,

g2 = .13; speech rate, F(1, 76) = 4.62, p < .05, g2 = .06;

hand/finger movements, F(1, 76) = 6.11, p < .05, g2 = .07;

and leg and foot movements, F(1, 76) = 7.52, p < .01,

g2 = .09. Marginally significant effects emerged for con-

textual embeddings, F(1, 76) = 3.75, p = .057, g2 = .05,

and speech hesitations, F(1, 76) = 3.77, p = .069,

g2 = .06.

Table 1 depicts the findings for the Reverse Order and

control conditions separately. In the Reverse Order con-

dition, liars mentioned fewer auditory details than truth

tellers. Liars also included more speech hesitations in their

statements, spoke at a slower speech rate and moved their

legs and feet more than truth tellers. In the control condi-

tion only one significant finding emerged: Liars moved

their hands and fingers less than truth tellers. Although the

Veracity · Order interaction effects for cognitive opera-

tions, speech errors and eye blinks were not significant,

they did significantly discriminate between truth tellers and

liars in the Reverse Order condition but not in the control

condition. In the Reverse Order condition, liars included

more cognitive operations and more speech errors in their

speech than truth tellers, as well as blinking more.

Discussion

The findings supported our prediction that liars would

display more signs of cognitive load than truth tellers, and

particularly in the Reverse Order condition. In the Reverse

Order condition liars included fewer auditory details and

contextual embedding details and more cognitive opera-

tions in their stories than truth tellers, three signs of cog-

nitive load. Furthermore, liars made more speech

hesitations, spoke with a slower speech rate, and made

more speech errors than truth tellers, which are three more

signs of cognitive load. Liars, however, did not just reveal

more signs of cognitive load than truth tellers. They also

made more leg and foot movements than truth tellers and

blinked more. These are signs of nervousness, rather than

signs of cognitive load. It thus appears that the instruction

to tell a story in reverse order not only made the partici-

pants have to think harder, but it also made them more

nervous. Participants’ perceptions that they perform worse

in the Reverse Order condition than in the control condition

(as indicated by their expectation of being less likely to

receive an incentive and more likely to receive a penalty)

may have contributed to their anxiety.

Participants in the control condition showed only one

cue to deceit: Liars moved their hands and fingers less than

truth tellers. This cue has emerged as a sign of deceit in

many of our previous studies (e.g. Akehurst and Vrij 1999;

Caso et al. 2006; Vrij 1995, 2006; Vrij and Mann 2001;

Vrij and Winkel 1991; Vrij et al. 1996, 1997, 2000, 2001a,

2004) and is, in our deception research, one of the most

consistent cues to deceit. Yet, very few other researchers

examine those movements (DePaulo et al. 2003). A

decrease in hand/finger movements could be a sign of

cognitive load, but could also be a sign of trying to make a

convincing impression on others. That is, liars will attempt

to control their behaviour and will avoid making move-

ments that they believe look suspicious, and making hand/

finger movements could be one cue that liars believe look

suspicious. It is also possible that the reduction in hand/

finger movements was caused by a combination of cogni-

tive load and attempts to control behaviour.

In summary, we examined 16 non-verbal and verbal

cues. Of these, we expected 14 cues (all except gaze and

self-adaptors) to emerge as cognitive cues to deceit,

Table 1 Verbal, vocal and visual cues as a function of veracity for the reverse order and control conditions separately

Reverse order Control condition

Truth Lie F (1, 38) g2 Truth Lie F (1, 38) g2

M SD M SD M SD M SD

Auditory details 9.00 4.6 5.35 3.6 7.62*** .17 6.25 4.2 8.85 4.2 3.80 .09

Contextual embeddings 18.95 7.3 12.15 5.9 10.51*** .22 17.05 10.3 17.10 7.4 .00 .00

Speech hesitations 4.54 2.6 5.96 2.0 3.86* .09 4.36 2.2 3.99 2.0 .32 .01

Speech rate 163.62 29.2 142.13 13.7 5.15** .12 162.84 21.0 169.02 32.9 .50 .01

Hand/finger 11.78 9.8 15.41 13.1 .98 .02 14.55 8.7 7.33 6.4 8.84*** .19

Leg/foot 5.21 5.3 13.80 16.9 4.68** .11 14.97 21.1 6.28 5.8 3.15 (.08) .08

Cognitive operations 1.75 2.0 4.35 4.6 5.30** .12 2.05 3.9 3.65 3.4 1.93 .05

Speech errors 1.26 0.7 1.91 1.2 3.87* .09 1.45 1.2 1.51 1.2 .02 .00

Eye blinks 17.38 9.7 27.07 12.7 7.34*** .16 19.56 9.1 24.75 12.1 2.34 .06

*p < .05, one-tailed test; **p < .05, two-tailed test; ***p < .01, two-tailed test

Law Hum Behav (2008) 32:253–265 259

123

particularly in the high cognitive load condition (Reverse

Order). In the high cognitive load condition six of the 14 cues

emerged as cognitive cues to deceit, whereas in the low

cognitive load condition (control) only one such cue

emerged. Six out of 14 cues (43%) may not be seen as strong

support for our hypothesis, but it is a high percentage com-

pared to other studies: In most studies where a substantial

number of cues have been examined (see DePaulo et al.

2003, and Vrij 2000, for reviews) fewer than 43% of the

investigated cues actually emerge as cues to deceit. Our

findings are even more striking considering the task we set

our liars. Because we told our liars what the truth tellers had

experienced during the staged event, we provided liars with a

wealth of visual details that they could incorporate into their

stories. As a result, differences in visual details between liars

and truth tellers did not arise. If we had not coached our liars

so thoroughly, the predicted difference that is typically found

(Vrij 2005) between liars and truth tellers in mentioning vi-

sual details may have occurred. Also, the extensive coaching

of liars may have made the task of lying somewhat easier for

them, and this may explain why the expected differences in

latency period, pauses and speech errors did not emerge. To

be sure, we cannot explain why the differences between liars

and truth tellers show up for some measures but not for

others. The real concern, however, is whether observers can

discriminate between liars and truth tellers more effectively

when stories are told in Reverse Order. Experiment 2

addresses this issue.

Experiment 2

The fact that reverse order interviews reveal more cues to

deception than control interviews does not automatically

imply that observers will be able to discriminate better

between truths and lies in reverse order interviews. Suc-

cessful discrimination depends on whether observers

interpret the diagnostic cues correctly. Lack of detail, in-

creases in speech hesitations and speech errors, slower

speech rate and an increase in movements (some of the

cues that differentiated liars from truth tellers in the reverse

order interviews), all create the impression of suspicion

(Strömwall et al. 2004; Vrij et al. 2006a). Therefore, we

were hopeful that observers could differentiate between

truths and lies in the reverse order interviews. In contrast,

the situation appears gloomier for the control interviews.

Only one cue, a decrease in hand/finger movements,

differentiated liars from truth tellers in these interviews,

and this cue is typically not associated with deception. In

fact, observers believe that liars make more movements

than truth tellers (Strömwall et al. 2004; Vrij et al. 2006a).

Observers might therefore be using the wrong decision-

making strategy when judging the control interviews and

might associate an increase in movements with deception

rather than truth telling. This would result in judging

truthful stories as deceptive and deceptive stories as

truthful.

We also examined whether interviewees gave the

impression that they were nervous or thinking hard. Since

several of the cues that differentiated liars from truth tellers

in the reverse order interviews were signs of cognitive load

(lack of detail, more speech hesitations, more speech

errors, and slower speech rate) or perhaps nervousness

(increase in leg/foot movements and eye blinks), we pre-

dicted that in the reverse order condition liars would give

the impression of thinking harder and being more nervous

than truth tellers. Given the lack of cues that differentiated

liars from truth tellers in the control interviews, we did not

expect liars and truth tellers to differ from each other in

terms of giving the impression of having to think hard or

being nervous in the control interviews.

Method

Participants

The participants were 55 British police officers: 33 males

and 22 females. The largest group (62%) were general

uniformed officers, with the remaining 38% being specia-

lised in CID (Criminal Investigations Department). None

of the participants had received training in lie detection

(such training does not exist in England and Wales). Their

average age was M = 30.60 years (SD = 8.4). All of the

police officers except one were Constables. The remaining

participant was a Sergeant. Their average length of service

in the police was M = 2.81 years (SD = 6.2). When asked

to indicate on a 5-point Likert scale how experienced they

considered themselves in interviewing (M = 1.76, SD =

1.2), 76% rated themselves as ‘inexperienced’ (a score of

1 or 2 on the 5-point Likert scale) whereas 12% declared

themselves as ‘experienced’ (a score of 4 or 5 on the 5-

point Likert scale). When asked to indicate on a 5-point

Likert scale how motivated they were to perform well on

the task, 77% reported themselves as fairly or highly

motivated (a score of 4 or 5 on the 5-point Likert scale,

M = 4.02, SD = .8).

Procedure

The study took place at training colleges with police con-

stabularies in the South of England. Between seven and

fifteen participants were tested simultaneously. This vari-

ation in group size reflected only the number of officers

that trainers were willing to release from class at that time.

It did not in any way affect the conduct of the experiment.

The videotaped interviews (‘clips’) were shown on a large

260 Law Hum Behav (2008) 32:253–265

123

screen (approximately 2m x 1m), in a large classroom

that would have enabled twenty participants to have

seen the screen clearly, sitting far enough apart so as

not to see each other’s answers. Participants were given

questionnaires and asked to complete the first section

relating to the details discussed in the Participants sec-

tion above. They were then informed that they were

about to see a selection of clips of students who were

either lying or telling the truth about a scenario that

involved the theft of money from a wallet. The scenario

involved their playing a game of Connect 4 with another

participant (actually a stooge) whilst various people

entered or exited the room. Truth tellers had actually

participated in this event, and truthfully had not taken

any money; liars were merely informed about the event,

and had actually taken the money from the wallet. The

experimenter did not tell the participants how many

clips they would see, or what percentage were truths or

lies, so as to avoid participants calculating how many

truths and lies they were actually being shown, and

hence deliberately trying to achieve a certain number of

truth/lie responses. Instead they were informed that

although they would not be told how many clips they

would see, there would not be as many clips as were in

their questionnaire. They were told that after each clip

the tape would be stopped, and when everybody had

completed all questions on the questionnaire relating to

that clip, the next clip would be shown. 31 officers saw

12 interviews told in reverse order and 24 officers saw

12 interviews told in chronological order. Those 24

interviews were a random sample of the interviews from

Experiment 1. Of the 12 interviews seen by each

observer, six interviewees lied and six told the truth.

In the experimental condition, prior to watching the

reverse order interviews the participants were informed

that the students that they were about to see have been

asked to explain everything that happened in reverse order,

so they will be reporting their story backwards.

After watching each clip the observers were asked to

answer three questions: (i) Do you think that the suspect is

telling .... (dichotomous answer, the truth/a lie), (ii) To

what extent does the person in the video look as if he/she is

having to think hard?, and (iii) To what extent does the

person in the video look nervous? Answers were given on 7

point Likert scales ranging from (1) not at all to (7)

extremely. The study took about 1 h to conduct.

Accuracy was measured by calculating the percentage of

correct veracity judgements given by each participant in

judging the truthful clips (truth accuracy) and deceptive

clips (lie accuracy). The impressions of having to think

hard and being nervous were measured by calculating the

average cognitive load and nervousness scores allocated to

liars and to truth tellers.

Results

Overall Accuracy, Lie Accuracy and Truth Accuracy

The overall accuracy scores ranged from a low of 17% (one

participant in the control condition) to a perfect 100% (one

participant in the Reverse Order condition) with an average

of M = .53 (SD = .17). This average percentage did not

differ significantly from the level of chance (.50),

t(54) = 1.22, p = .23. The truth accuracy (M = .54, SD =

.21) and lie accuracy (M = .52, SD = .21) were almost the

same and did not differ significantly from each other, F(1,

54) = .21, p = .64. Neither truth accuracy nor lie accuracy

differed significantly from chance, both ts < 1.28, both

ps > .20.

Hypothesis-Testing

A 2 (Veracity) · 2(Order) mixed ANOVA was conducted

with Veracity as the within-subjects factor, Order as the

between-subjects factor, and accuracy as the dependent

variable. The analysis revealed a significant Order effect,

F(1, 53) = 6.98, p < .05, g2 = .12. None of the other main

effects or interaction effects were significant, all Fs < 2.81,

all ps > .10. Accuracy in the Reverse Order condition

(M = .58, SD = .16) was superior to accuracy in the control

condition (M = .46, SD = .17). Accuracy in the Reverse

Order condition was above the level of chance,

t(30) = 2.73, p < .05, whereas accuracy in the control

condition did not differ from chance, t(23) = 1.12, p = .27.

The data were separated into truth and lie accuracy

scores. The Reverse Order condition resulted in M = .56

(SD = .22) truth accuracy and M = .60 (SD = .20) lie

accuracy, whereby both truth accuracy t(30) = 14.14,

p < .01, and lie accuracy score, t(30) = 2.68, p < .05 were

significantly above the level of chance. The control con-

dition resulted in M = .50 (SD = .20) truth accuracy and

M = .42 (SD = .19) lie accuracy, whereby the lie accuracy

score was significantly below the level of chance,

t(23) = 1.97, p < .05. Lie accuracy in the Reverse Order

condition (60%) was significantly higher than lie accuracy

in the control condition (42%), F(1, 53) = 10.53, p < .01,

g2 = .17; truth accuracy did not differ significantly between

the two conditions, F(1, 53) = 1.26, p = .26.

A 2 (Veracity) · 2(Order) mixed MANOVA was con-

ducted with Veracity as the within-subjects factor, Order

as the between-subjects factor, and the impressions of

having to think hard and nervousness as dependent vari-

ables. This analysis revealed a significant Veracity effect,

F(2, 52) = 15.61, p < .01, g2 = .38, a significant Order

effect, F(2, 52) = 6.37, p < .01, g2 = .20, and a signifi-

cant Veracity · Order interaction effect, F(2, 52) =

23.71, p < .05, g2 = .48. At a univariate level, both

Law Hum Behav (2008) 32:253–265 261

123

Veracity · Order interaction effects were significant:

Cognitive load (F(1, 53) = 38.02, p < .01, g2 = .42) and

nervousness (F(1, 53) = 18.03, p < .01, g2 = .26. Since

these two Veracity · Order interaction effects are more

informative than either of the two Veracity and Order main

effects, only the Veracity · Order interaction effects will

be discussed further.

In the control condition, liars and truth tellers did not

differ from each other in terms of giving the impression of

having to think hard or being nervous, but in the Reverse

Order condition they did, as can be seen in Table 2. Liars

gave the impression of having to think harder and being

more nervous than truth tellers.

Discussion

This experiment (Experiment 2) demonstrated that

instructing participants to tell their stories in reverse order

facilitates lie detection. Police officers who observed the

stories told in reverse order were better at detecting lies

(60% accuracy) than those who observed stories told in

chronological order (42%). Also, the 60% lie accuracy

obtained in the Reverse Order condition was significantly

better than could have been expected by chance, whereas

the 42% lie accuracy obtained in the control condition was

significantly below chance level.

The improvement in lie detection as a result of asking

interviewees to tell their stories in reverse order did not

occur at the expense of the ability to detect truths. Truth

detection accuracy was somewhat higher in the Reverse

Order condition (56%) than in the control condition (50%)

although the difference was not significant. However, total

accuracy (lie and truth accuracy combined) in the Reverse

Order condition (58%) was significantly higher than the

total accuracy obtained by observers in the control condi-

tion (46%). The total accuracy obtained in the Reverse

Order condition was also significantly above the level of

chance, whereas the total accuracy in the control condition

did not differ from chance.

We acknowledge that the accuracy scores are not high,

even in the Reverse Order condition. Note that we told our

liars (see Experiment 1) what truth tellers had actually

experienced during the staged event. This may make lying

somewhat easier than in situations where liars have to

invent all the details themselves. However, as in many

experimental studies, it is the statistical difference between

experimental conditions that is important rather than the

absolute level of accuracy scores.

We initially assumed that liars would have to think

harder than truth tellers, and in particular when recalling in

reverse order. Indeed, in this condition in particular liars

gave the impression of having to think harder than truth

tellers (Experiment 2). Liars also gave the impression of

being more nervous than truth tellers in the Reverse Order

condition (Experiment 2), which fits well with the findings

of Experiment 1, that liars made more movements than

truth tellers in the Reverse Order condition. We believe

that the unexpected challenge of having to tell their stories

in reverse order made liars nervous. Although we did not

predict that liars would be more nervous than truth tellers

in the Reverse Order condition (Experiment 1), the finding

that liars displayed more nervous behaviours in the Reverse

Order condition than truth tellers (Experiment 1) perhaps

facilitated lie detection in the Reverse Order condition in

Experiment 2 because observers typically think that liars

show more nervous behaviours.

General Discussion

We started this article by noting the paucity of attempts to

actively magnify the differences between liars and truth

tellers. In this article we attempted to fill this gap and

empirically tested an interview style designed to enlarge

the non-verbal and verbal differences between liars and

truth tellers: Instructing interviewees to recall their stories

in reverse order. We predicted that this would be particu-

larly debilitating for liars, because their cognitive resources

Table 2 impression liars and truth tellers made on the observers in the reverse order condition and the control condition

Reverse order condition Truth tellers Liars F (1, 30) g2

M SD M SD

Having to think hard 3.75 .7 4.89 .7 90.35** .70

Looking nervous 3.69 .6 4.48 .7 49.59** .62

Control condition Truth tellers Liars F (1, 23) g2

M SD M SD

Having to think hard 3.95 .9 3.72 .7 1.36 .05

Looking nervous 3.56 .7 3.59 .7 .04 .00

*p < .05; **p < .01

262 Law Hum Behav (2008) 32:253–265

123

have already been partially depleted by the cognitively

demanding task of lying. These predictions were supported.

The reverse order interviews contained many more cues to

deceit than the control interviews, and the instruction to

convey stories in reverse order improved police observers’

ability to detect deception.

We believe that the present findings are useful for pro-

fessional lie detectors, and can be adapted for police

interviews. Asking interviewees to report their stories in

reverse order is already practiced in police interviews in

several countries including the UK and US (Milne and Bull

1999, 2003), albeit in interviews with cooperative wit-

nesses rather than suspects. Cooperative witnesses are

requested to tell their stories in reverse order because it

facilitates recalling more accurate information than con-

ventional interview techniques (Geiselman et al. 1986;

Köhnken et al. 1999). The present findings demonstrate

that this technique could also be used in suspect interviews

to facilitate truth and lie detection.

One possible limitation of this study is that police offi-

cers may be uncomfortable in a real investigation to request

suspects to describe events in reverse order. It has been

noted, for example, that British police have generally not

used the reverse-order instruction of the Cognitive Inter-

view (Fisher and Geiselman, 2002) when interviewing

cooperative witnesses (Kebbell et al. 1999). We believe

that a reason for this reluctance is that the police typically

use a predetermined structure in which, in their view, the

reverse-order instruction does not really fit. However,

police may feel more comfortable using the technique

when interviewing suspects, where the interview is more

dynamic and is less likely to follow a predetermined

structure. Indeed, we have been told by several American

investigators who used the reverse-order instruction when

interviewing suspects, that suspects frequently gave

themselves away with obviously non-credible stories that

were replete with inconsistencies.

Our new approach to lie detection is not restricted to

asking interviewees to recall their stories in reverse order:

Numerous other requests that make the interview setting

cognitively more challenging should have a similar effect.

For example, interviewees could be instructed to look the

interviewer in the eye while reporting their activities.

Constantly maintaining eye contact while talking is cog-

nitively demanding (Beattie 1981), because it can be dis-

tracting (Doherty-Sneddon et al. 2002; Doherty-Sneddon

and Phelps 2005). As we found in the present experiments,

it could thus result in cues to deceit and facilitate lie

detection.

Whether our predictions about the effects of holding

constant eye contact on deception and lie detection will

hold true will need to be experimentally tested. In addition,

we have no doubt that there are other additional requests

that could be made in interview settings that would make

the settings more demanding for the liar. We hope that our

work inspires researchers to further develop the cognitively

based lie detection method introduced in this article.

Acknowledgements This project was sponsored by a grant from the

Economic and Social Research Council (RES-000-23-0292).

References

Akehurst, L., & Vrij, A. (1999). Creating suspects in police

interviews. Journal of Applied Social Psychology, 29, 192–210.

Beattie, G. W. (1981). A further investigation of the cognitive

interference hypothesis of gaze patterns during conversation.

British Journal of Social Psychology, 20, 243–248.

Briggs, G. E., Peters, G. L., & Fisher, R. P. (1972). On the locus of the

divided attention effects. Perception & Psychophysics, 11,

315–320.

Broadbent, D. E. (1957). A mechanical model for human attention

and immediate memory. Psychological Review, 64, 205–215.

Buller, D. B., & Burgoon, J. K. (1996). Interpersonal deception

theory. Communication Theory, 6, 203–242.

Caso, L., Gnisci, A., Vrij, A., & Mann, S. (2005). Processes

underlying deception: An empirical analysis of truths and lies

when manipulating the stakes. Journal of Interviewing and Offender Profiling, 2, 195–202.

Caso, L., Vrij, A., Mann, S., & DeLeo, G. (2006). Deceptive

responses: The impact of verbal and nonverbal countermeasures.

Legal and Criminological Psychology, 11, 99–111.

DePaulo, B. M., & Kirkendol, S. E. (1989). The motivational

impairment effect in the communication of deception. In J. C.

Yuille (Ed.), Credibility assessment (pp. 51–70). Dordrecht, the

Netherlands: Kluwer.

DePaulo, B. M., Lindsay, J. L., Malone, B. E., Muhlenbruck, L.,

Charlton, K., & Cooper, H. (2003). Cues to deception. Psycho- logical Bulletin, 129, 74-118.

Doherty-Sneddon, G., Bruce, V., Bonner, L., Longbotham, S., &

Doyle, C. (2002). Development of gaze aversion as disengage-

ment of visual information. Developmental Psychology, 38,

438–445.

Doherty-Sneddon, G., & Phelps, F. G. (2005). Gaze aversion: A

response to cognitive or social difficulty? Memory and Cogni- tion, 33, 727–733.

Ekman, P. (1985). Telling lies: Clues to deceit in the marketplace, politics and marriage. New York, NJ: W. W. Norton. (Reprinted

in 1992 and 2001).

Ekman, P. (1997). Deception, lying, and demeanor. In D. F. Halpern

& A. E. Voiskounsky (Eds.), States of mind: American and post- soviet perspectives on contemporary issues in psychology (pp. 93–105). New York, NJ: Oxford University Press.

Ekman, P., & Friesen, W. V. (1972). Hand movements. Journal of Communication, 22, 353–374.

Fisher, R., & Geiselman, R. E. (1992). Memory-enhancing techniques in investigative interviewing: The cognitive interview. Spring-

field, IL: C.C. Thomas.

Geiselman, R. E, & Callot, R. (1990). Reverse and forward order

recall of script based text. Journal of Applied Cognitive Psychology, 4, 141–144.

Geiselman, R. E., Fisher, R. P., MacKinnon, D. P., & Holland, H. L.

(1986). Enhancement of eyewitness memory with the cognitive

interview. American Journal of Psychology, 99, 385–401.

Gilbert, D. T. (1991). How mental systems believe. American Psychologist, 46, 107–119.

Law Hum Behav (2008) 32:253–265 263

123

Gilbert, J. A. E., & Fisher, R. P. (2006). The effects of varied retrieval

cues on reminiscence in eyewitness memory. Applied Cognitive Psychology, 20, 723–739.

Gilovich, T., Savitsky, K., & Medvec, V. H. (1998). The illusion of

transparency: Biased assessments of others’ ability to read one’s

emotional states. Journal of Personality and Social Psychology, 75, 332–346.

Goldman-Eisler, F. (1968). Psycholinguistics: Experiments in spon- taneous speech. New York, NJ: Double day.

Granhag, P. A., Strömwall, L. A. (2002). Repeated interrogations:

Verbal and nonverbal cues to deception. Applied Cognitive Psychology, 16, 243–257.

Hartwig, M., Granhag, P. A., Strömwall, L., & Kronkvist, O. (2006).

Strategic use of evidence during police interrogations: When

training to detect deception works. Law and Human Behavior, 30, 603–619.

Hocking, J. E., & Leathers, D. G. (1980). Nonverbal indicators of

deception: A new theoretical perspective. Communication Monographs, 47, 119–131.

Johnston, W. A., Greenberg, S. N., Fisher, R. P., & Martin, D. W.

(1970). Divided attention: A vehicle for monitoring memory

processes. Journal of Experimental Psychology, 83, 164–171.

Kahana, M. J. (1996). Associate retrieval processes in free recall.

Memory & Cognition, 24, 103–109.

Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ:

Prentice-Hall.

Kassin, S. M. (2005). On the psychology of confessions: Does

innocence put innocents at risk? American Psychologist, 60,

215–228.

Kassin, S. M., & Gudjonsson, G. H. (2004). The psychology of

confessions: A review of the literature and issues. Psychological Science in the Public Interest, 5, 33–67.

Kassin, S. M., & Norwick, R. J. (2004). Why people waive their

Miranda rights: The power of innocence. Law and Human Behavior, 28, 211–221.

Kebbell, M. R., Milne, R., & Wagstaff, G. F. (1999). The Cognitive

Interview: A survey of its forensic effectiveness. Psychology, Crime and Law, 5, 101–115.

Köhnken, G. (1996). Social psychology and the law. In G. R. Semin

& K. Fiedler (Eds.), Applied Social Psychology (pp. 257–282).

London, Great Britain: Sage Publications.

Köhnken, G. (2004). Statement Validity Analysis and the ‘detection

of the truth’. In P. A. Granhag & L. A. Strömwall (Eds.),

Deception detection in forensic contexts (pp. 41–63). Cam-

bridge, England: Cambridge University Press.

Köhnken, G., Milne, R., Memon, A., & Bull, R. (1999). The cognitive

interview: A meta-analysis. Psychology, Crime, & Law, 5, 3–27.

Mann, S., & Vrij, A. (2006). Police officers’ judgements of veracity,

tenseness, cognitive load and attempted behavioural control in

real life police interviews. Psychology, Crime, & Law, 12,

307–319.

Mann, S., Vrij, A., & Bull, R. (2002). Suspects, lies and videotape: An

analysis of authentic high-stakes liars. Law and Human Behav- ior, 26, 365–376.

Masip, J., Sporer, S., Garrido, E., & Herrero, C. (2005). The detection

of deception with the reality monitoring approach: A review of

the empirical evidence. Psychology, Crime, & Law, 11, 99–122.

McCornack, S. A. (1997). The generation of deceptive messages:

Laying the groundwork for a viable theory of interpersonal

deception. In J. O. Greene (Ed.), Message production: Advances in communication theory (pp. 91–126). Mahway, NJ: Lawrence

Erlbaum.

Milne, R., & Bull, R. (1999). Investigative interviewing: Psychology and practice. Chichester, England: John Wiley & Sons.

Milne, R., & Bull, R. (2003). Does the cognitive interview help

children to resist the effects of suggestive interviewing? Legal and Criminological Psychology, 8, 21–38.

Schweitzer, M. E., Brodt, S. E., & Croson, R. T. A. (2002). Seeing

and believing: Visual access and the strategic use of deception.

The International Journal of Conflict Management, 13, 258–275.

Smith, V. L., & Clark, H. H. (1993). On the course of answering

questions. Journal of Memory and Language, 32, 25–38.

Spence, S. A., Farrow, T. F. D., Herford, A. E., Wilkinson, I. D.,

Zheng, Y., & Woodruff, P. W. R. (2001). Behavioural and

functional anatomical correlates of deception in humans. Neu- roreport: For Rapid Communication of Neuroscience Research, 12, 2849–2853.

Spence, S. A., Hunter, M. D., Farrow, T. F. D., Green, R. D., Leung,

D. H., Hughes, C. J., & Ganesan, V. (2004). A cognitive

neurobiological account of deception: Evidence from functional

neuroimaging. Philosophical Transactions of the Royal Society of London, 359, 1755–1762.

Sporer, S. L., & Schwandt, B. (2006). Paraverbal indicators of

deception: A meta-analytic synthesis. Applied Cognitive Psy- chology, 20, 421–446.

Strömwall. L. A., Granhag, P. A., & Hartwig, M. (2004). Practitio-

ners’ beliefs about deception. In P. A. Granhag & L. A.

Strömwall (Eds.), Deception detection in forensic contexts (pp.

229-250). Cambridge, England: Cambridge University Press.

Strömwall, L. A., Hartwig, M., & Granhag, P. A. (2006). To act

truthfully: Nonverbal behaviour and strategies during a police

interrogation. Psychology, Crime, & Law, 12, 207–219.

Vrij, A. (1995). Behavioral correlates of deception in a simulated

police interview. Journal of Psychology: Interdisciplinary and Applied, 129, 15–29.

Vrij, A. (2000). Detecting lies and deceit: The psychology of lying and the implications for professional practice. Chichester: John

Wiley and sons.

Vrij, A. (2004). Invited article: Why professionals fail to catch liars

and how they can improve. Legal and Criminological Psychol- ogy, 9, 159–183.

Vrij, A. (2005). Criteria-Based Content Analysis: A qualitative

review of the first 37 studies. Psychology, Public Policy, and Law, 11, 3–41.

Vrij, A. (2006). Challenging interviewees during interviews: The

potential effects on lie detection. Psychology, Crime, & Law, 12,

193–206.

Vrij, A. (in press). Detecting lies and deceit: Pitfalls and opportu- nities (2nd ed.). Chichester: John Wiley and sons.

Vrij, A., Akehurst, L., & Knight, S. (2006). Police officers’, social

workers’, teachers’ and the general public’s beliefs about

deception in children, adolescents and adults. Legal and Criminological Psychology, 11, 297–312.

Vrij, A., Akehurst, L., & Morris, P. (1997). Individual differences in

hand movements during deception. Journal of Nonverbal Behavior, 21, 87–103.

Vrij, A., Akehurst, L., Soukara, S., & Bull, R. (2004). Detecting

deceit via analyses of verbal and nonverbal behavior in children

and adults. Human Communication Research, 30, 8–41.

Vrij, A., Edward, K., & Bull, R. (2001a). People’s insight into their

own behaviour and speech content while lying. British Journal of Psychology, 92, 373–389.

Vrij, A., Edward, K., & Bull, R. (2001b). Stereotypical verbal and

nonverbal responses while deceiving others. Personality and Social Psychology Bulletin, 27, 899–909.

Vrij, A., Edward, K., Roberts, K. P., & Bull, R. (2000). Detecting

deceit via analysis of verbal and nonverbal behavior. Journal of Nonverbal Behavior, 24, 239–263.

264 Law Hum Behav (2008) 32:253–265

123

Vrij, A., Fisher, R., Mann, S., & Leal, S. (2006). Detecting deception

by manipulating cognitive load. Trends in Cognitive Sciences, 10, 141–142.

Vrij, A., Fisher, R., Mann, S., & Leal, S. (in press). Increasing

cognitive load in interviews to detect deceit. In B. Milne, S.

Savage, & T. Williamson (Eds.), International developments in investigative interviewing. Uffculme: Willan Publishing.

Vrij, A., & Mann, S. (2001). Telling and detecting lies in a high-stake

situation: The case of a convicted murderer. Applied Cognitive Psychology, 15, 187–203.

Vrij, A., & Mann, S. (2003). Deception detection. In P. W. Halligan,

C. Bass, & D. A. Oakley (Eds.), Malingering and illness deception (pp. 348-362). Oxford University Press.

Vrij, A., & Mann, S. (2006). Criteria-Based Content Analysis: An

empirical test of its underlying processes. Psychology, Crime, & Law, 12, 337–349.

Vrij, A., Mann, S., & Fisher, R. (2006a). An empirical test of the

Behaviour Analysis Interview. Law and Human Behavior, 30,

329–345.

Vrij, A., Mann, S., & Fisher, R. (2006b). Information-gathering vs

accusatory interview style: Individual Differences in respon-

dents’ experiences. Personality and Individual Differences, 41,

589–599.

Vrij, A., Mann, S., Kristen, S., & Fisher, R. (2007). Cues to deception

and ability to detect lies as a function of police interview styles.

Law and Human Behavior, published online.

Vrij, A., Semin, G. R., & Bull, R. (1996). Insight into behaviour

during deception. Human Communication Research, 22, 544-

562.

Vrij, A., & Winkel, F. W. (1991). Cultural patterns in Dutch

and Surinam nonverbal behavior: An analysis of simulated

police/citizen encounters. Journal of Nonverbal Behavior, 15,

169–184.

Vrij, A., & Winkel, F. W. (1992). Cross-cultural police-citizen

interactions: The influence of race, beliefs and nonverbal

communication on impression formation. Journal of Applied Social Psychology, 22, 1546–1559.

Walczyk, J. J., Roper, K. S., Seemann, E., & Humphrey, A. M.

(2003). Cognitive mechanisms underlying lying to questions:

Response time as a cue to deception. Applied Cognitive Psychology, 17, 755–744.

Walczyk, J. J., Schwartz, J. P., Clifton, R., Adams, B., Wei, M., &

Zha, P. (2005). Lying person-to-person about live events: A

cognitive framework for lie detection. Personnel Psychology, 58,

141–170.

White, C. H., & Burgoon, J. K. (2001). Adaptation and communi-

cative design: Patterns of interaction in truthful and deceptive

conversations. Human Communication Research, 27, 9–37.

Zuckerman, M., DePaulo, B. M., & Rosenthal, R. (1981). Verbal and

nonverbal communication of deception. In L. Berkowitz (Ed.),

Advances in experimental social psychology (Vol. 14, pp. 1–57).

New York: Academic Press.

Law Hum Behav (2008) 32:253–265 265

123

<< /ASCII85EncodePages false /AllowTransparency false /AutoPositionEPSFiles true /AutoRotatePages /None /Binding /Left /CalGrayProfile (None) /CalRGBProfile (sRGB IEC61966-2.1) /CalCMYKProfile (ISO Coated v2 300% \050ECI\051) /sRGBProfile (sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Error /CompatibilityLevel 1.3 /CompressObjects /Off /CompressPages true /ConvertImagesToIndexed true /PassThroughJPEGImages true /CreateJDFFile false /CreateJobTicket false /DefaultRenderingIntent /Perceptual /DetectBlends true /ColorConversionStrategy /sRGB /DoThumbnails true /EmbedAllFonts true /EmbedJobOptions true /DSCReportingLevel 0 /SyntheticBoldness 1.00 /EmitDSCWarnings false /EndPage -1 /ImageMemory 524288 /LockDistillerParams true /MaxSubsetPct 100 /Optimize true /OPM 1 /ParseDSCComments true /ParseDSCCommentsForDocInfo true /PreserveCopyPage true /PreserveEPSInfo true /PreserveHalftoneInfo false /PreserveOPIComments false /PreserveOverprintSettings true /StartPage 1 /SubsetFonts false /TransferFunctionInfo /Apply /UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile () /AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages false /DownsampleColorImages true /ColorImageDownsampleType /Bicubic /ColorImageResolution 150 /ColorImageDepth -1 /ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true /ColorImageFilter /DCTEncode /AutoFilterColorImages false /ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /ColorImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /JPEG2000ColorACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict << /QFactor 0.76 /HSamples [2 1 1 2] /VSamples [2 1 1 2] >> /GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict << /TileWidth 256 /TileHeight 256 /Quality 30 >> /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict << /K -1 >> /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org?) /PDFXTrapped /False /Description << /ENU <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> /DEU <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> >> >> setdistillerparams << /HWResolution [2400 2400] /PageSize [5952.756 8418.897] >> setpagedevice