Reflection on Crocodile Tears | Custom PHD Thesis

Read the article, and write a reflection about it. What is the article about? An overview to be exact. Do you think it can work? can it

applied in the field for professionals to use? and any ideas that you may have. Pros/cons about the article.

A reflection paper is meant to illustrate understanding of the material and how it affects the student’s ideas and possible practice in

future.

The reflection paper should have structure with an opening paragraph, main body, and conclusion and be between 3 and 5 pages. It is written

in first person, no abstract, and references are required.

Crocodile Tears: Facial, Verbal and Body Language Behaviours Associated
With Genuine and Fabricated Remorse
Leanne ten Brinke

Sarah MacDonald

University of British Columbia

Memorial University of Newfoundland

Stephen Porter and Brian O’Connor
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

University of British Columbia

Emotional deception is a common behaviour that can have major consequences if undetected. For
example, the sincerity of an offender’s expressed remorse is an important factor in sentencing and parole
hearings. The present study was the first to investigate the nature of true and false remorse. We examined
facial, verbal and body language behaviours associated with emotional deception in videotaped accounts
of true personal transgressions accompanied by either genuine or falsified remorse. Analyses of nearly
300,000 frames indicated that descriptions of falsified remorse were associated with a greater range of
emotional expressions. Further, sequential analyses revealed that negative emotions were more commonly followed by other emotions—rather than

a return to neutral emotion—in falsified versus sincere
remorse. Participants also exhibited more speech hesitations while expressing deceptive relative to
genuine remorse. In general, the results suggest that falsified remorse may be conceived as an emotionally turbulent display of deliberate,

falsified expressions and involuntary, genuine, emotional leakage.
These findings are relevant to judges and parole board members who consider genuine remorse to be an
important factor in sentencing and release decisions.

duration and frequency of the [crimes], the criminal record of the
offender, the effects on the victim and the presence or absence of
. . . remorse” (R. v. B., 1990; R. v. W.W.M., 2006). In evaluating
remorse, judges often refer to the defendant’s demeanour: “The
appellant’s responses to questions posed him and his demeanour
showed that his main concern was for himself. . . I would have
expected some show of distress or anguish for having raped a
previously chaste young woman. The appellant showed no such
signs” (Balkissoon v. Canada (Citizenship and Immigration),
2001). On the other hand, defendants who portray remorse for their
actions are considered to be good candidates for treatment and
rehabilitation: “[The defendant’s] remorse, guilt, and shame
should provide him with a strong motivation to work at changes
that will prevent future acts of violence” (R. v. Struve, 2007). It is
clear that impressions of remorse, based on demeanour, can influence the length of an offender’s sentence.
Apparent level of remorse also is an important consideration
when an offender is eligible for release upon parole. A primary
purpose of a parole interview is to provide the offender with the
opportunity to show the parole board that his/her attitude has
changed since he/she first arrived in prison (Ruback & Hopper,
1986). If an offender displays remorse during the interview, parole
board members may take this as a positive indication that the
offender is prepared for reintegration into the community. However, it is clear that many decisions are not well informed and risky
releases can occur when offenders manipulate parole decisionmakers. Ruback and Hopper (1986) investigated judgments made
by parole board members regarding the offenders’ success on
parole both before and after an interview with the parole candidates. Before the interview, the board members had access to the

Deception is a common aspect of human social interaction.
People admit to using deceit in 14% of emails, 27% of face-to-face
interactions and 37% of phone calls (Hancock, 2007), and lie twice
a day, on average (DePaulo, Kashy, Kirkendol, Wyer, & Epstein,
1996). While these acts of deception typically are minor, white lies
of little consequence (DePaulo et al., 1996), lies accompanied by
powerful emotions and false emotional displays can have major
consequences at individual and societal levels. For example, in
sentencing and parole hearings, the presence of remorse is a major
factor in decision-making; the perceived credibility of a defendant’s emotion during his/her testimony inform ultimate decisions
pertaining to their future. In the absence of genuine remorse, an
offender may be highly motivated to fake regret for his/her actions,
and the judge, jury or parole board member may be similarly
motivated to detect “crocodile tears”. In sentencing decisions,
judges are instructed to consider “the age of the victim, the

This article was published Online First February 8, 2011.
Leanne ten Brinke, Stephen Porter, and Brian O’Connor, University of
British Columbia, Kelowna, BC, Canada; Sarah MacDonald, Memorial
University of Newfoundland, St. John’s, NL, Canada
The authors would like to acknowledge the assistance of Brendan
Wallace and Natasha Korva for their contributions to this research. This
project was supported by the Social Sciences and Humanities Research
Council of Canada (SSHRC) through scholarships awarded to the first and
second authors, as well as research grants awarded to the third and fourth
authors.
Correspondence concerning this article should be addressed to Leanne
ten Brinke, University of British Columbia, Kelowna, BC, Canada. E-mail:
lbrinke@interchange.ubc.ca
51

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

52
offenders’ files, detailing all relevant information required to make
a decision regarding the offenders’ release. During the interview,
parole board members gained information concerning the offender’s attitude and apparent level of remorse. Results indicated that
decisions made by board members became less accurate in predicting offenders’ success upon subsequent release, relative to
their judgments based on file information alone. Interestingly, this
indicates that the face-to-face contact with the offender impaired
the parole officers’ judgments, such that high-risk offenders were
more often released. Similarly, Porter, ten Brinke, and Wilson
(2009) explored the probability of offenders’ release as a function
of their level of psychopathy. Psychopathic offenders, known for
their remorselessness and ability to charm and manipulate others,
were approximately 2.5 times more likely to be granted conditional
release compared to non-psychopaths. As with Ruback and Hopper’s (1986) findings, one explanation is that the decision-makers
were duped by offenders during their interviews, persuaded by
false displays of remorse.
But what is remorse? Most of us— barring psychopaths—feel
remorseful for something we have done to another person (e.g.
cheating on our partner, lying to our parents). Yet, there has been
hardly any psychological research on the manifestation of emotional facial expressions during this common human experience
(cf. Keltner & Buswell, 1996). Remorse, or guilt, is defined as the
negative feeling resulting from the violation of one’s moral standards, and may arise from situations such as lying, neglecting a
friend or family member, cheating, or failure at a particular duty
(Keltner & Buswell, 1996; Tangney & Dearing, 2002). Feelings of
guilt occur particularly in situations where the regrettable action
was controllable by the transgressor (Tracy & Robins, 2006).
Although Izard (1977) and Keltner & Buswell (1996) suggest that
a single universal expression does not represent remorse, no empirical investigations to date have considered examining the combination of

universal emotional expressions associated with this
affective state. Further, no studies have examined the potential
emotional expression differences between genuine and fabricated
remorse, reflecting the paucity of research surrounding colloquial
ideas about emotional deception more generally (Porter & ten
Brinke, 2010). The issue of whether “crocodile tears” are identifiable has received no empirical attention.
Theory and research suggest that various behavioural cues
associated with deception do exist (see Porter & ten Brinke,
2010). In theory, the presentation of liars and truth-tellers
should differ because of the elevated arousal, cognitive load and
behavioural control involved in fabricating a personal account
or feelings (Vrij, 2008). The liar is forced to avoid betraying the
deception by controlling his/ her feelings of guilt or excitement,
monitoring his/her words and keeping the story straight, while
supplying sufficient detail to appear credible but not so much as
to lead to a failure in recalling and maintaining the falsified
details. While telling the story, the liar must monitor his/ her
facial expressions (sometimes having to conceal or falsify an
emotion) and body language. This necessary “multi-tasking”
should reduce the level of conscious control the liar has over
each behavioural channel, and increase the amount of relative
“leakage” from one or the other(s), depending on the where
he/she is devoting relative degrees of effort (Porter & ten
Brinke, 2010).

Facial Cues to Emotional Deception
Given that the face is the primary focus during interpersonal
interaction, a wise liar might devote particular attention to controlling his/her facial expression. We rapidly “read” faces to decide
whether a stranger is friend or foe, make personality assessments,
and inform the course of our interactions (Fowler, Lilienfeld &
Patrick, 2009; ten Brinke & Porter, 2009; Willis & Todorov,
2006). Thus, deceptive individuals must maintain their duplicity
by falsifying emotional expressions concordant with the lie, and
suppressing “leakage” of their true emotions (Ekman & Friesen,
1975). For example, a deceptive employee must convincingly
express sadness as he explains to his boss that he will need to miss
work to attend his aunt’s funeral out of town, simultaneously
suppressing excitement about his real plans to extend a vacation
with friends. Given the complex musculature of the face, and the
heavy cognitive load associated with such a task, it is not surprising that expression of genuine and falsified emotions can be
perceptibly different.
False facial expressions were first studied in the nineteenth
century by Duchenne (1862/1990) who examined the muscle actions associated with real and false smiles. Using electrical stimulation of facial

muscles, he noted that a genuine happiness expression involves not only the contraction of the zygomatic major
muscle which upturns the mouth corners into a smile, but also the
orbicularis oculi which creates crow’s feet around the eye. This
observation was later validated by Ekman, Davidson, and Friesen
(1990). A follower of Duchenne’s work, Darwin (1872) later
noted: “A man when moderately angry, or even when enraged,
may command the movements of his body, but. . .those muscles of
the face which are least obedient to the will, will sometimes alone
betray a slight and passing emotion” (p. 79). He proposed that
some facial actions associated with strong emotion cannot be
inhibited voluntarily and that the same muscle actions cannot be
engaged voluntarily during emotional simulation. Implicit in this
statement is the suggestion that leakage of one’s true emotion will
be proportional to the intensity of the felt emotion. Collectively,
these statements form the inhibition hypothesis (Ekman, 2003a).
Micro-expressions—a derivation of Darwin’s idea—are defined as
short-lived leakage of one’s true emotion manifested as an emotional expression lasting 1/25th to 1/5th of a second (Ekman, 1992;
see also Haggard & Isaacs, 1966). Despite the popularity of both
Darwin’s and Ekman’s hypotheses, little research has been conducted to substantiate these ideas.
In a direct examination of Darwin’s (1872) hypothesis, Porter
and ten Brinke (2008) investigated the nature of facial expressions
accompanying four types of falsified or concealed universal emotions: happiness, sadness, fear and disgust. Participants viewed
powerful emotional images, responding with a genuine or convincing but false expression while being judged by a naı¨ ve observer.
False expressions, in response to emotionally provocative images
were either simulated (an unfelt emotion was expressed) or
masked (a false expression replaced that of the felt emotion).
When the 697 videotaped expressions were exhaustively analyzed
(each 1/30th second frame for more than 100,000 frames), involuntary leakage was found to be ubiquitous. No participant was able
to falsify each emotion without such betrayals on at least one
occasion. Emotional leakage was significantly more likely to occur, and lasted longer, in masked expressions than genuine. How-

53
ever, the presence and duration of emotional leakage were not
significantly different across genuine and simulated emotional
expressions. In general, negative emotions (sadness, fear, disgust)
were more difficult to falsify than positive (happiness), and resulted in greater emotional leakage. Despite the presence of these
indicators, naı¨ ve judges performed only slightly above chance
when attempting to discriminate genuine and falsified expressions
(see also Hess & Kleck, 1994). Previous studies also have found
differences in the duration, onset, and offset times between genuine and false expressions of happiness and disgust (e.g. Frank,
Ekman, & Friesen, 2005; Hess & Kleck, 1990). Although no
research to date has examined potential sequential differences in
genuine versus deceptive emotional displays, it may be inferred
that emotions related to the act of deception could follow falsified
expressions. For example, Ekman (1992) noted that deceivers
might not anticipate the intensity of emotions related to the act of
lying, resulting in signs of shame or duping delight on the face
following falsified emotions.

Verbal and Body Language Cues to Deception
Empirical evidence suggests that there may be value in attending to verbal cues in catching emotional lies such as remorse
narratives (e.g. Porter & Yuille, 1995; Vrij, 2008). The broad
deception literature suggests that, relative to truth-tellers, liars are
more likely to speak slower and provide less detail (DePaulo et al.,
2003; Porter, Yuille, & Lehman, 1999). Liars also are more likely
to hesitate, including more ‘ums’ and ‘ers’ in their stories relative
to genuine individuals (DePaulo et al., 2003; Vrij, 2008). Further,
linguistic patterns can provide indications of dishonest speech
(Hancock, Curry, Goorha, & Woodworth, 2008). Research in
which participants wrote about their views on personally significant topics (e.g. abortion) indicated that deceptive and truthful
narratives were discriminated with a mean accuracy of 67% based
on linguistic characteristics alone (Newman, Pennebaker, Berry, &
Richards, 2003). Specifically, liars tended to use fewer first-person
pronouns and references to others.
The existing literature substantiates that there also are associations between body language and deception. In their meta-analysis,
DePaulo et al. (2003) found that a reduction in illustrator use was
among the most powerful indicators of deception. Other empirically supported indicators include an increase in selfmanipulations and blink

rate (specifically during emotional masking) (Porter, Doucette, Earle, & MacNeil, 2008; Porter & ten
Brinke, 2008). While gaze aversion is widely regarded as a reliable
cue to deceit, it has not found empirical support (Global Deception
Research Team, 2006). In fact, the opposite may be true; liars may
make greater eye contact than truth-tellers in a conscious attempt
to appear convincing (Vrij, 2008).

The Present Study
The present study investigated potential facial, verbal and body
language cues to deception in descriptions of true autobiographical
transgressions associated with genuine and falsified remorse. Differences in the presence and duration of emotional expressions
were expected to occur when participants described their genuine
versus their deceptive feelings of remorse (i.e. when participants
said, “I felt really guilty”, or something similar).

Hypothesis 1: The presence of sadness was expected to
dominate genuine, relative to deceptive descriptions of feelings. Deceptive descriptions of felt emotion, on the other
hand, were expected to be characterized by the presence of
various discordant emotional expressions, such as happiness
or anger.
Hypothesis 2: Similarly, genuine emotional descriptions were
expected to be accompanied by longer durations of sadness,
whereas deceptive descriptions would include longer durations of happiness and anger.
Such differences were not expected when emotional presence
and duration were collated across the entirety of each event description because most of each falsified remorse narrative was
comprised of truthful event details. Further, we examined the
presence of micro-expressions (Ekman, 1992) as a potential cue to
emotional deceit, but refrained from posing a hypothesis.
We also suspected that the sequences of facial expressions that
individuals display when they are being deceptive might differ
from the naturally occurring sequences of facial expressions that
occur during genuine remorse. The duping delight, embarrassment
or urge to express one’s genuine emotions may result in (genuine)
discordant emotions following (falsified) sadness expressions. To
our knowledge, this possibility has not been examined in previous
research. We therefore conducted sequential analyses on the facial
expressions data.
Hypothesis 3: It was expected that (falsified) negative expressions would be more often followed by positive expressions
(emotional leakage) during false remorse, compared to genuine regret.
In addition to emotional facial variables, verbal and non-verbal
cues to deception were examined.
Hypothesis 4: Verbal indicators, including slowed speech
rate, increased speech hesitations, and fewer self- and otherreferences, were expected to be associated with deceptive
remorse.
Hypothesis 5: Various body language behaviours (decreased
use of illustrators, increased use of self-manipulators, increased blink rate and potentially, decreased gaze aversion)
were examined as cues to falsified remorse.

Method
Participants
Undergraduate students (n ⫽ 31) in a Canadian university
participated in return for course credit points. Participants were 20
females and 11 males and had a mean age of 21.67 (SD ⫽ 4.34).
Three additional naı¨ve, undergraduate research assistants judged
the veracity of participant displays of remorse. The presence of an
observer was intended primarily to increase the realism of the task
and the motivation of the participant.

54
Apparatus
The testing room was arranged such that the participant was
seated in a chair facing video-recording equipment. Two digital
video cameras, recording at a rate of 30 frames per second, were
used to film the participant, capturing footage of the participant’s
entire body and a close-up of the participant’s face, respectively.
The experimenter and a naı¨ve judge were seated on either side (left
and right) of the video camera, facing the participant.

Procedure
Upon arriving for the study, all participants and judges completed an informed consent form and were provided with a definition of remorse,

as well as common situations associated with
this emotion, as outlined by Keltner and Buswell (1996). Participants were asked to tell the experimenter about a true non-criminal
event in their life that made them feel intensely and genuinely
remorseful, in as much detail as possible. Subsequently, the participants were asked to complete a filler-task questionnaire and a
short demographic survey (age, gender, ethnicity and education
level). Participants also were asked to describe a similar true event
(counterbalanced order) for which they felt no/little remorse, but
were asked to convincingly feign remorse for their actions. For
example, if the participant described an instance of cheating for
which he/she genuinely felt remorseful, he/she was asked to think
of another time he/she cheated but did not feel remorseful to
describe with falsified remorse. Participants were told to choose
events that occurred within 6 months of one another to reduce
differential effects of memory decay or emotional intensity on
narratives. Subsequently, participants were asked to indicate their
level of remorse for each transgression on a scale from 1 (not at all
remorseful) to 7 (highly remorseful). This served as a manipulation
check to ensure the two autobiographical events were indeed
associated with different levels of felt remorse. Naı¨ve judges
estimated the veracity of remorse in each description (genuine or
falsified) and rated their confidence in that decision. The presence
of a naı¨ ve judge during recording acted as social motivation,
encouraging participants to act convincingly remorseful. At the
completion of the study, all participants were fully debriefed about
its purposes.

relatively recent addition of contempt (Ekman & Friesen, 1975;
Ekman et al., 1987). A primary (graduate student) coder, blind to
veracity condition, examined all narratives. In order to examine
emotion in the upper and lower face separately, all 149,331 frames
were coded twice (once with attention to the upper face, and again
for the lower) for a total of 298,662 analyzed frames.
Body language variables examined included: rate of illustrators,
rate of self-manipulations, proportion of time spent averting gaze
and blink rate. Illustrators were defined as any movement/gesture
of the arms and/or hands, designed to illustrate what the participant
was saying. Self-manipulations were any instance where the participant touched/scratched his/her hand, head or body (Porter et al.,
2008). Verbal variables coded included: speech rate, rate of speech
hesitations (e.g. the use of “um” or “uh”, etc.), proportion of
self-references (e.g. “I”, “we”, “my”) and proportion of reference
to others (e.g. “him”, “she”, “they”).

Coding Reliability
Facial emotion variables. Sixteen (26%) of the narratives
were independently coded by a second individual (a trained undergraduate, blind to veracity and hypotheses) to assess inter-rater
reliability. The raters agreed on emotion codes for 84.07% of the
48,071 frames that were coded for the purposes of reliability. The
duration (measured in number of frames) of emotional expressions, as coded by the two individuals, were correlated at r (248) ⫽
.99, p ⬍ .001. Further, no mean difference between raters was
revealed, p ⬎ .05. Inter-rater reliability on the duration variable
was “excellent”, as defined by Cicchetti and Sparrow (1981) and
Fleiss (1981). Agreement on the dichotomously coded presence/
absence of emotional expressions also was high, Kappa ⫽ .70, p ⬍
.001, with agreement on 87.90% of all codes.
Verbal and non-verbal variables. A secondary coder reexamined 14 (23%) of the narratives for verbal and non-verbal
cues to assess inter-rater reliability. Verbal and non-verbal variables coded by the primary and secondary coder were highly
correlated, r ⫽ 0.60 – 0.99, ps ⬍ .05. Further, means were not
significantly different between coders, all ps ⬎ .05. Inter-rater
reliability was at least “good”, and more often “excellent” (as
defined by Cicchetti & Sparrow (1981) and Fleiss (1981)) on all
indices.

Results

Coding Procedures
Blind (to veracity) coding was conducted for all facial, verbal
and body language variables. Emotional facial expressions occurring during each narrative were coded using the procedure developed by Porter

and ten Brinke (2008). Training involves facial
musculature recognition, memorization of facial action units associated with universal emotions, and identification of universal
emotions. This training is based on the Pictures of Facial Affect
(universal emotions depicted in photographs; Ekman & Friesen,
1976) and the Facial Action Coding System (Ekman, Friesen, &
Hagar, 2002). See Porter and ten Brinke (2008) for further information on coding procedures and training. Each frame (1/30th of
a second) of the narratives was coded for the presence, duration
and sequence of universal emotional expressions in the upper and
lower portion of the face independently. Universal emotions include happiness, sadness, fear, disgust, anger, surprise and the

Manipulation Check
To ensure that each narrative was associated with different
levels of remorse (e.g. high remorse in the genuine condition and
low remorse in the fabricated condition), a paired samples t-test
was conducted for the participant ratings of each narrative. As
expected, participants reported significantly more remorse for their
genuine (M ⫽ 6.09, SD ⫽ 0.65) than fabricated (M ⫽ 2.06, SD ⫽
0.77) remorse event, t(30) ⫽ 24.61, p ⬍ .001, d ⫽ 6.29, 95% CI
[3.70, 4.37].

Facial Cues to Falsified Remorse
Descriptions of felt emotions. Although the narratives were
unstructured, the majority of participants included a description of

felt emotions associated with the event (e.g. “I felt so bad”; “I felt
really guilty”). Of the 31 participants, n ⫽ 23 included such a
statement in both of their (genuine and falsified) descriptions of
remorse. In order to test Hypotheses 1 and 2, the start and end
times of these descriptions were noted, and emotional facial variables occurring within these sections were calculated. Specifically,
the presence/absence, as well as the total duration, of each universal emotion served as dependent variables in subsequent analyses.
Hypothesis 1: The Generalized Linear Equations procedure in
SPSS was used to conduct a repeated measures logistic regression for the prediction of the presence/ absence of emotions. The durations of

the descriptions and the participant
ratings of remorse served as covariates. There was a significant effect for veracity (genuine vs. falsified remorse), Wald
Chi-Square ⫽ 117.1, p ⬍ .001. More of the seven universal
emotions were likely to be present during descriptions of
falsified remorse (M ⫽ .95, SE ⫽ .02), relative to genuine
remorse (M ⫽ .73, SE ⫽ .06). There were also effects for
facial hemisphere (upper vs. lower face), and emotion (happiness, sadness, fear disgust, anger, surprise and contempt),
including interactions. However, these effects are not relevant
to our hypotheses and thus, are not reported here. See Table
1 for descriptive statistics.
Hypothesis 2: The MIXED procedure in SPSS was used to
examine the effects of credibility on the duration of universal
emotional expressions during descriptions of remorse. The
durations of the descriptions and the participant ratings of
remorse once again served as covariates. However, there were
no significant effects involving credibility, ps ⬎ .05.
Complete description of event. A series of SPSS MIXED
model analyses was then conducted, to reveal potential differences
in facial emotion behaviour over the entirety of the narratives.

Veracity (genuine vs. falsified remorse), facial hemisphere (upper
vs. lower face) and emotion (each of the seven universal emotions)
were the within-subjects factors. The durations of the narrative and
the participant ratings of remorse served as covariates. In separate
analyses, the presence/absence and total duration (number of
1/30th second frames) of each of the seven universal emotions
served as dependent variables. No significant main effects or
interactions involving veracity emerged, ps ⬎ .05, suggesting that
emotional deception was restricted to descriptions of emotion,
specifically, as we hypothesized.
Micro-expressions.
Based on the definition of microexpressions offered by Ekman (1992), emotional expressions lasting one-fifth of a second or less occurring in

the upper or lower
face were isolated for further analysis. A total of 23 microexpressions were detected, including 12 upper face and 11 lower
face expressions. These affective displays were produced by 10
(32.3%) participants and occurred in 12 (19.4%) of the 62 narratives. Upper face micro-expressions in the fabricated condition
were predominantly of anger (5 of 9 micro-expressions) while the
majority of lower face micro-expressions were displays of sadness
(6 of 8 micro-expressions; refer to Table 2).
Sequential analyses of facial expressions. Sequential analyses and follow-up loglinear analyses were conducted to test
Hypothesis 3 and gain a more detailed understanding of the emotional processes involved in falsified remorse. In sequential analysis, a

stream of codes (in this case, the series of 905 codes for
facial expressions) are read to produce a transitional frequency
matrix, which is simply a table of frequency counts for the number
of times each code was followed by other codes in the coding
system. To simplify the analyses and to avoid dealing with large,
transitional frequency matrices, the facial emotion codes were
collapsed into three groups based on emotional valence: positive
(happiness), negative (sadness, fear, anger, contempt) or neutral
(neutral, surprise). Surprise was included in the “neutral” valence

Table 1
Descriptive Statistics of Facial Expression Variables During the Descriptions of Felt Emotions
Genuine

Upper face
Happy
Sad
Fear
Disgust
Anger
Surprise
Contempt
Neutral
Lower face
Happy
Sad
Fear
Disgust
Anger
Surprise
Contempt
Neutral

Deceptive

Mean (SD) duration
in frames

Presence
(% of narratives)

Mean (SD) duration
in frames

Presence
(% of narratives)

12.2 (23.8)
2.0 (9.4)
4.0 (19.2)
0.0 (0.0)
0.4 (1.9)
7.1 (16.2)
0.0 (0.0)
40.4 (28.6)

30.4
4.3
4.3
0.0
4.3
26.1
0.0
82.6

9.7 (18.5)
8.6 (24.5)
1.6 (7.5)
.6 (2.8)
5.6 (22.5)
9.1 (16.2)
.4 (1.7)
36.4 (30.2)

27.3
18.2
4.5
4.5
9.1
31.8
4.5
81.8

21.0 (26.0)
1.4 (6.3)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
43.6 (43.6)

47.8
8.7
0.0
0.0
0.0
0.0
0.0
65.2

28.6 (40.2)
4.0 (18.8)
0.0 (0.0)
.5 (2.1)
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
39.0 (30.2)

59.1
4.5
0.0
4.5
0.0
0.0
0.0
72.7

56
Table 2
Frequency of Emotional Micro-Expressions Revealed by the
Upper and Lower Face During Complete Descriptions of
Genuine and Fabricated Remorse

Table 4
Standardized Residuals for Antecedent-Consequent Transitions
in Emotional Facial Expressions

Lower face
Genuine

Fabricated

3
0
0
0
0
1
1
1

9
1
0
1
1
5
1
0

8
1
6
0
0
0
0
1

3
1
1
1
0
0
0
0

Upper face
Total
Happy
Sad
Fear
Disgust
Anger
Surprise
Contempt
Lower face
Total
Happy
Sad
Fear
Disgust
Anger
Surprise
Contempt

category because it is not easily characterized as positive or
negative. It should be noted, however, that while surprise did
(rarely) occur in the upper face, it never appeared in the lower face.
Thus, the “neutral” category in lower face sequential and loglinear
analyses are comprised of purely neutral expressions despite the
inclusion of surprise in this category. The software provided by
O’Connor (1999) was used to obtain four separate transitional
frequency matrices: (1) for the genuine upper face, (2) for the
genuine lower face, (3) for the deceptive upper face and (4) for the
deceptive lower face. Loglinear analyses were then conducted to
determine whether the transitional frequencies varied across the
genuine and deceptive conditions. The transitional frequencies
appear in Table 3, and the corresponding standardized residuals (z
values) for each cell are provided in Table 4.
For the upper face data, a hierarchal loglinear analysis with
backward elimination resulted in a model in which only the Lag 0
Table 3
Frequencies of Antecedent–Consequent Transitions in Facial
Emotions
Lower face

Upper face

Antecedent

Consequent

Genuine

Fabricated

Genuine

Fabricated

Positive
Positive
Positive
Negative
Negative
Negative
Neutral
Neutral
Neutral

Positive
Negative
Neutral
Positive
Negative
Neutral
Positive
Negative
Neutral

0
14
146
11
0
66
158
65
14

0
27
125
25
7
45
137
44
16

0
15
98
18
4
90
99
93
269

0
16
88
24
7
94
92
103
211

Note. Positive–positive transitions are structural zeros (i.e. zero by definition). Given that only happiness was included in the positive

valence
group, there was no opportunity for one positive emotion to follow another.

Upper face

Antecedent

Consequent

Genuine

Fabricated

Genuine

Fabricated

Positive
Positive
Positive
Negative
Negative
Negative
Neutral
Neutral
Neutral

Positive
Negative
Neutral
Positive
Negative
Neutral
Positive
Negative
Neutral

0
⫺4.26
2.52
⫺4.73
⫺3.04
7.72
2.70
6.88
⫺7.45

0
⫺2.67
1.73
⫺3.04
⫺.76
4.85
1.90
4.14
⫺5.41

0
⫺1.54
.76
⫺1.02
⫺3.23
2.18
.50
2.45
⫺1.50

0
⫺1.84
1.04
⫺.63
⫺3.43
2.33
.35
2.96
⫺1.89

Note. p ⬍ .05 for standardized residuals ⬎1.96 or ⬍⫺1.96. Positive–
positive transitions are structural zeros (i.e. zero by definition).

to Lag 1 (antecedent-to-consequent) term remained significant, ␹2
(4) ⫽ 16.5, p ⬍ .001. None of the terms involving veracity were
significant predictors of the transitional frequencies. These findings indicate that while there were significant antecedent-toconsequent

transitions, the patterns of transitional frequencies for
upper facial emotions were the same (i.e. not significantly different) across the genuine and deceptive conditions.
For the lower face data, a hierarchal loglinear analysis revealed
that elimination of the three-way interaction term (veracity⬘ antecedent⬘ consequent) resulted in a significant reduction in fit relative to

the saturated model, ␹2 (4) ⫽ 16.8, p ⫽ .002. This means
that the transitional frequencies varied depending on veracity. To
explore the interaction, loglinear analyses were conducted within
the genuine and falsified remorse conditions and the standardized
residuals from these analyses are provided in Table 4.
The patterns in the standardized residuals are easily summarized. Almost all of the lower face transitions were significant,
whereas few of the upper face transitions reached significance.
The same basic trends nevertheless existed in the transitions for
the upper and lower face. The directions of the lower face
transitions were simply stronger, or amplified. There is also a
distinct pattern in the transitions, which is perhaps most easily
discerned by focusing on the left-most column of numbers in
Table 4, which are the standardized residuals for the genuine
condition. The pattern has three components: (1) transitions
from positive to negative facial emotions, and from negative to
positive facial emotions, are notably rare; (2) transitions from
neutral emotions to positive and negative emotions are common
and (3) transitions from positive to neutral emotions, and from
negative to neutral emotions, are also common. In other words,
genuine participants did not display rapid or immediate transitions between positive and negative emotions, and their transitions in and out

of positive and negative emotions involved, or
occurred through, neutral emotions. This same three-component
pattern also occurred in deceptive remorse, but it was less
pronounced. Deception involved relatively more frequent direct
transitions between positive and negative emotions, and relatively fewer indirect sequences via neutral emotions.

Verbal and Body Language Cues to Falsified Remorse
Hypothesis 4: A repeated-measures MANOVA with veracity
(genuine vs. falsified remorse) as the within-participant factor
was conducted with speech rate, rate of speech hesitations,
proportion of self references, and proportion of other references as dependent variables. This analysis revealed a significant multivariate

effect of veracity, F(4, 26) ⫽ 2.84, p ⫽
.045, ␩2 ⫽ 0.10. Follow-up univariate analyses revealed that
participants exhibited a significantly higher rate of speech
hesitations during fabricated remorse (M ⫽ 0.06, SD ⫽ 0.05)
compared to genuine remorse (M ⫽ 0.04, SD ⫽ 0.04), F(1,
29) ⫽ 5.21, p ⫽ .03, ␩2 ⫽ 0.15, 95% CI [0.002, 0.029]. It also
is noteworthy that a second variable approached significance.
Participants mentioned more references to other people while
displaying fabricated remorse (M ⫽ 0.07, SD ⫽ 0.03) relative
to genuine remorse (M ⫽ 0.05, SD ⫽ 0.02), F(1, 29) ⫽ 3.53,
p ⫽ .07, ␩2 ⫽ 0.11, 95% CI [⫺0.001, 0.029].
Hypothesis 5: A repeated-measures MANOVA with veracity
(genuine vs. falsified remorse) as the within-participant factor
was conducted with rate of illustrators, proportion of time
spent averting gaze, and blink rate as the dependent variables.
The multivariate effect of veracity was non-significant, F(3,
27) ⫽ 0.47, p ⫽ .71.

Discussion
Feeling remorse is a basic and common human affective experience, often occurring in response to a personal failure or transgression (Keltner

& Buswell, 1996). Despite the potential implications of falsifying these feelings in forensic settings, there has
been hardly any research on the subject. To our knowledge, the
present study was the first to investigate genuine and falsified
remorse for behavioural cues that might be indicative of such
deception. It was expected that emotional presentations would
differ across levels of credibility specifically when participants
described their feelings associated with each transgression.

Emotional Facial Cues to Falsified Remorse
Description of felt emotion. One major finding was that
there were significant differences in the presence of universal
emotional facial expressions across real and false remorse. Falsified descriptions of remorseful feelings (e.g. “I felt so guilty”)
were associated with the presence of a greater range of emotions,
relative to genuine descriptions. Deceptive descriptions of remorseful emotions were commonly associated with the presence of
happiness and surprise. This may reflect the incomplete fabrication
of deceptive, negative emotions, and the leakage of genuine,
positive emotion in falsified accounts. Attempts to express falsified sadness involve complex and involuntary muscles in the
forehead. While most people can easily raise the eyebrows (frontalis muscle, Action Units 1 and 2; Ekman et al., 2002), it is
difficult for the untrained individual to engage the corrugator
muscle (Action Unit 4; Ekman et al., 2002) simultaneously, which
furrows the brows to mimic the upper face activation associated
with a genuine sad face (Ekman, 2003b). Thus, the liar activates
only the frontalis muscle, and appears surprised. Perhaps realizing

that their attempt at sadness is meagre at best, genuine happiness
associated with their embarrassing display may be leaked via a
smile with the lower face. Further, since genuine accounts were
accompanied by fewer emotions, the emotional presentation of the
liar may also reflect an over-compensatory strategy to the task,
resulting in a melodramatic emotional performance about a relatively minor, non-criminal past transgression that does not warrant
such a display.
Complete description of event. Given that both described
experiences had actually occurred (i.e. all event details were genuine) and that we exerted no control over the length or specific
content of each participant’s narratives, it was unsurprising that
differences in emotional facial presentation were not found when
the entirety of the narrative was examined. Any potential differences in presence or duration of each emotion may have been
diminished by individual differences in narrative content (i.e. level
of detail, description of emotion).
Micro-expressions. Micro-expressions also were examined
as a potential cue to emotional deceit and relative frequencies
suggested that they may reveal one’s true affective state. Microexpressions often signalled sadness during genuine remorse and
anger during fabricated guilt. While sadness is a component of
remorse, anger is generally considered to be discordant with feelings of regret (Smith, 2008). Thus, these very brief expressions
may indeed reveal covert (and unconcealed) feelings, as proposed
by Ekman and Friesen (1975). The finding that micro-expressions
(overall) were equally common among genuine and deceptive
expressions highlights the importance of considering the expressed
emotion in context rather than simply interpreting the presence of
a micro-expression as a signal of deceit. It also is interesting to
note that anger—an emotion singled out by Darwin (1872)—was
revealed by the upper face (involving Action Units 4, 5 and 7;
Ekman et al., 2002). The muscles underling these action units
should be of specific interest in future investigations as they may
be those which Darwin (1872) described as being “least obedient
to the will” (p. 79). Despite the (tenuous) support for microexpressions as a cue to deceit reported here, it should be noted that
micro-expressions occurred in less than 20% of all narratives and
were not an infallible cue to deception (or truth) in all cases. While
further research on this phenomenon certainly is warranted, empirical research to date suggests that over-reliance on microexpressions (e.g.

in security settings; Ekman, 2006) as an indicator
of credibility is likely to be ineffective (Weinberger, 2010).
Emotional sequences.
In addition to the presence of expressed emotion, we also found that the sequences of emotional
expressions for genuine remorse were different from those for
falsified remorse, but only in the lower face. Genuine participants
rarely displayed immediate transitions between positive and negative emotions, and their transitions in and out of positive and
negative emotions occurred through neutral emotions. This same
pattern occurred in deceptive remorse, but less strongly. In fact,
deception involved relatively more frequent direct transitions between positive and negative emotions, and relatively fewer indirect
sequences via neutral emotions. The internal states and preoccupations that are involved in deception apparently cause liars to
make more frequent, abrupt transitions between positive and negative facial emotions. To our knowledge, these sequential patterns
have not been previously reported.

58
The fact that emotional deception was revealed by the lower
face (i.e. muscles surrounding the mouth) replicates findings of
Porter and ten Brinke (2008) who found that masked emotions
were most often betrayed by leakage of inconsistent expressions in
the same area. This examination of emotional sequences, however,
is a novel approach to deception detection and offers new insights
into the processes involved in emotional deceit. Genuine negative
emotions expressed in the lower face almost always were followed
by a neutral expression. Thus, it might be concluded that genuine
expressions of remorse are experienced as a “linear series” of
universal emotions, separated by returns to a neutral baseline state.
In contrast, negative expressions (i.e. sadness) in fabricated remorse are more often followed by other, positive and negative,
emotions. This emotional turbulence may provide some indication
as to the liar’s actual affective state. For example, after an attempt
to appear sad, the liar— knowing that his/her expression does not
appear credible—may smile in genuine delight or embarrassment.
In general, these findings are consistent with the notion that
emotional deception may be betrayed by leakage of the liar’s true
emotion and/or the deceiver’s over-the-top performance, resulting
in a diverse emotional display.

Verbal and Body Language Cues to Falsified Remorse
Despite exhibiting non-significant mean differences for the majority of body language and verbal behaviours, participants did
show significantly more speech hesitations (e.g. um, uh, er) while
telling a story accompanied by fabricated remorse compared to
their genuine remorseful story. An increase in speech hesitations
generally is associated with an increase in cognitive complexity
(DePaulo et al., 2003; Vrij & Heaven, 1999). Given that the
increase in cognitive complexity in the fabricated remorse condition was minimal (i.e. emotion only), speech hesitations may be a
particularly sensitive measure of cognitive resources utilized.

Future Directions and Implications
The face-to-face contact with an offender during a parole interview has a large impact on decisions regarding release (e.g.
Ruback & Hopper, 1986). As such, identifying reliable behavioural cues that can differentiate genuine and fabricated remorse
could have considerable practical implications. The current findings can be of practical relevance to forensic psychologists, parole
officers and legal decision-makers when assessing the veracity of
remorseful displays. Although further research is necessary before
such cues inform aggravating or mitigating factors in the courtroom, clinicians may find this research useful in detecting and
confronting “crocodile tears”, potentially resulting in more honest
and effective treatment of offenders.
Further research should examine behavioural indicators of falsified remorse in a sample of criminal offenders and for more
severe transgressions. To strengthen the paradigm employed here,
future researchers should confirm the veracity of each event described by participants to ensure that findings are truly a result of
emotional, and not event, description. In addition to replicating the
current findings in a forensic sample, it would be interesting to
examine the predictive power of such cues. For example, one
could assess facial, verbal and body language cues to deceit, and
examine the correlation between the presence of such indicators

and the number of successful days on release. It is conceivable that
subtle facial cues associated with deception could be potent predictors of recidivism. In a similar line of research, Gottman and
Levenson (2000) found that a detailed examination of emotional
expressions during brief newlywed interactions could predict marriage longevity with 93% accuracy. Specifically, expressions of
contempt were predictive of early divorce, while a lack of positive
affect predicted later divorce and genuine displays of happiness
were a strong indicator of marriage stability. If similar relationships were found between offenders’ emotional expression in
parole hearings and success upon release, release decisions certainly could be enhanced for the safety of society. In general,
further research is necessary to gain a greater understanding of this
common and consequential human affective experience.

References
Balkissoon v. Canada (Citizenship and Immigration). (2001). CanLII
26712 (I.R.B.).
Cicchetti, D. V., & Sparrow, S. A. (1981). Developing criteria for establishing
interrater reliability of specific items: Applications to assessment of adaptive
behavior. American Journal of Mental Deficiency, 86, 127–137.
Darwin, C. (1872). The expression of the emotions in man and animals.
Chicago: University of Chicago Press.
DePaulo, B., Kashy, D., Kirkendol, S., Wyer, M., & Epstein, J. (1996).
Lying in everyday life. Journal of Personality and Social Psychology,
70, 979 –995. doi:10.1037/0022-3514.70.5.979
DePaulo, B., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K.,
& Cooper, H. (2003). Cues to deception. Psychological Bulletin, 129,
74 –118. doi:10.1037/0033-2909.129.1.74
Duchenne, G. B. (1862/1990). The mechanism of human facial expression.
NY: Cambridge University Press.
Ekman, P. (1992). Telling lies: Clues to deceit in the marketplace, politics,
and marriage. New York: W. W. Norton.
Ekman, P. (2003a). Darwin, deception and facial expression. In P. Ekman,
R. J. Davidson, & F. De Waals (Eds.), Annals of the New York Academy
of Sciences (Vol. 1000, pp. 205–221). New York: New York Academy
of Sciences. doi:10.1196/annals.1280.010
Ekman, P. (2003b). Emotions revealed: Recognizing faces and feelings to
improve communication and emotional life. New York: Owl Books.
Ekman, P. (2006, October 29). How to spot a terrorist on the fly. Retrieved
November 17, 2008, from washingtonpost.com.
Ekman, P., Davidson, R. J., & Friesen, W. V. (1990). The Duchenne smile:
Emotional expression and brain physiology: II. Journal of Personality
and Social Psychology, 58, 342–353. doi:10.1037/ 0022-3514.58.2.342
Ekman, P., & Friesen, W. V. (1975). Unmasking the face: A guide to
recognizing emotions from facial clues. New Jersey: Prentice-Hall.
Ekman, P., & Friesen, W. V. (1976). Pictures of facial affect. Palo Alto,
CA: Consulting Psychologists Press.
Ekman, P., Friesen, W. V., & Hagar, J. C. (1976/2002). Facial action
coding system: The manual on CD-ROM. Salt Lake City: Network
Information Research Corporation.
Ekman, P., Friesen, W. V., O’Sullivan, M., Chan, A., DiacoyanniTarlatzis, I., Heider, K., et al. (1987). Universals and cultural differences
in the judgment of facial expressions of emotion. Journal of Personality
and Social Psychology, 53, 712–717.
Fleiss, J. L. (1981). Statistical methods for rates and proportions (2nd ed.).
New York: John Wiley and Sons.
Fowler, K. A., Lilienfeld, S. O., & Patrick, C. J. (2009). Detecting psychopathy from thin slices of behavior. Psychological Assessment, 21,
68 –78. doi:10.1037/a0014938.
Frank, M. G., Ekman, P., & Friesen, W. V. (2005). Behavioral markers and
recognizability of the smile of enjoyment. In P. Ekman & E. L. Rosen-

59
berg (Eds.), What the face reveals (2nd ed., pp. 217–238). New York:
Oxford University Press.
Global Deception Research Team. (2006). A world of lies. Journal of
Cross-Cultural Psychology, 37, 60 –74. doi:10.1177/00220221
05282295
Gottman, J. M., & Levenson, R. W. (2000). The timing of divorce:
Predicting when a couple will divorce over a 14-year period. Journal of
Marriage and the Family, 62, 737–745. doi:10.1111/j.17413737.2000.00737.x
Haggard, E. A., & Isaacs, K. S. (1966). Micro-momentary facial expressions as indicators of ego mechanisms in psychotherapy. In L. A.
Gottschalk & A. H. Auerback (Eds.), Methods of research in psychotherapy (pp. 154 –165). New York: Appleton Century Crofts.
Hancock, J. T. (2007). Digital deception: When, where and how people lie
online. In K. McKenna, T. Postmes, U. Reips, & A. N. Joinson (Eds.),
Oxford handbook of internet psychology (pp. 287–301). Oxford: Oxford
University Press.
Hancock, J. T., Curry, L., Goorha, S., & Woodworth, M. T. (2008). On
lying and being lied to: A linguistic analysis of deception. Discourse
Processes, 45, 1–23. doi:10.1080/016385307017 39181
Hess, U., & Kleck, R. (1990). Differentiating emotional elicited and
deliberate emotional facial expressions. European Journal of Social
Psychology, 20, 369 –385. doi:10.1002/ejsp.2420200502
Hess, U., & Kleck, R. (1994). The cues decoders use in attempting to
differentiate emotion-elicited and posed facial expressions. European
Journal of Social Psychology, 24, 367–381. doi:10.1002/
ejsp.2420240306
Izard, C. E. (1977). Human emotions. New York, NY: Plenum Press.
Keltner, D., & Buswell, B. (1996). Evidence for the distinctness of embarrassment, shame, and guilt: A study of recalled antecedents and

facial
expressions of emotion. Cognition & Emotion, 10, 155–172. doi:
10.1080/026999396380312
Newman, M. L., Pennebaker, J. W., Berry, D. S., & Richards, J. M.
(2003). Lying words: Predicting deception from linguistic style.
Personality and Social Psychology Bulletin, 29, 665– 675. doi:10.1177/
0146167203029005010
O’Connor, B. (1999). Simple and flexible SAS and SPSS programs for
analyzing lag-sequential categorical data. Behavior Research Methods,
Instrumentation, and Computers, 31, 718 –726.
Porter, S., Doucette, N., Earle, J., & MacNeil, B. (2008). ‘Halfe the world
knowes not how the other halfe lies’: Investigation of cues to deception
exhibited by criminal offenders and non-offenders. Legal and Criminological Psychology, 13, 27–38. doi:10.1348/ 135532507X186653

Porter, S., & ten Brinke, L. (2008). Reading between the lies: Identifying concealed and falsified emotions in universal facial expressions.

Psychological Science, 19, 508 –514. doi:10.1111/j.14679280.2008.02116.x
Porter, S., & ten Brinke, L. (2010). The truth about lies: What works in
detecting high-stakes deception? Legal and Criminological Psychology,
15, 57–75. doi:10.1348/135532509X433151
Porter, S., ten Brinke, L., & Wilson, K. (2009). Crime profiles and
conditional release performance of psychopathic and non-psychopathic
sexual offenders. Legal and Criminological Psychology, 14, 109 –118.
doi:10.1348/135532508X284310
Porter, S., & Yuille, J. C. (1995). Credibility assessment of criminal
suspects through statement analysis. Psychology, Crime, and Law, 1,
319 –331. doi:10.1080/10683169508411969
Porter, S., Yuille, J. C., & Lehman, D. (1999). The nature of real, implanted, and fabricated memories for emotional childhood events:

Implications for the recovered memory debate. Law and Human Behavior,
23, 517–538. doi:10.1023/A:1022344128649
R. v. B. (1990). 36 O.A.C. 307.
R. v. W.W.M. (2006). 206 O.A.C. 342. R. v. Struve. (2007). BCSC 1316.
Ruback, B., & Hopper, H. (1986). Decision making by parole interviewers:
The effect of case and interview factors. Law and Human Behavior, 10,
203–214. doi:10.1007/BF01046210
Smith, N. (2008). I was wrong: The meanings of apologies. New York,
NY: Cambridge University Press.
Tangney, J. P., & Dearing, R. L. (2002). Shame and guilt. New York, NY:
Guildford Press. ten Brinke, L., & Porter, S. (2009). Friend or foe? The
role of intuition in interpersonal trustworthiness and vulnerability assessments. In A. Columbus (Ed.), Advances in psychology research
(Vol. 74). New York: Nova Science Publishers.
Tracy, J. L., & Robins, R. W. (2006). Appraisal antecedents of shame and
guilt: Support for a theoretical model. Personality and Social Psychology Bulletin, 32, 1339 –1351. doi:10.1177/014616720 6290212
Vrij, A. (2008). Detecting lies and deceit: Pitfalls and opportunities.
Chichester, England: Wiley.
Vrij, A., & Heaven, S. (1999). Vocal and verbal indicators of deception as
a function of lie complexity. Psychology, Crime & Law, 5, 203–215.
doi:10.1080/10683169908401767
Weinberger, S. (2010). Airport security: Intent to deceive? Nature, 465,
412– 415. doi:10.1038/465412a
Willis, J., & Todorov, A. (2006). First impressions: Making up your mind
after a 100-ms exposure to a face. Psychological Science, 17, 592–598.
doi:10.1111/j.1467-9280.2006.01750.x