Content uploaded by Christian A. Meissner
Author content
All content in this area was uploaded by Christian A. Meissner on Nov 02, 2018
Content may be subject to copyright.
© 2013 The Australian and New Zealand Association of Psychiatry, Psychology and Law
Psychiatry, Psychology and Law, 2013
Vol. 20, No. 4, 619–638, http://dx.doi.org/10.1080/13218719.2012.729018
Graves, B. (1994). When the abuser is an
educator. The School Administrator,51(9),
8–14, 16–18.
Green, S.B., & Salkind, N.J. (2003). Using
SPSS for Windows and Macintosh: Analyz-
ing and understanding data (4th edn). Upper
Saddle River, NJ: Prentice Hall.
Hughes, G. (2006, July 26). Female sex offenders
let off lightly? The Australian. Retrieved from
http://blogs.the Australian.news.com.au/gary
hughes/index.php/theaustralian/comments/
female-sex-offenders-let-off-lightly
Kelly, S.J. (1990). Responsibility and manage-
ment strategies in child sexual abuse: A
comparison of child protection workers,
nursers and police officers. Child Welfare,
69, 43–51.
Kennel, R.G., & Agresti, A.A. (1995). Effects of
gender and age on psychologists’
reporting of child sexual abuse. Professional
Psychology: Research and Practice,26, 612–
615.
Kite, D., & Tyson, G.A. (2004). The impact of
perpetrator gender on male and female
police officers’ perceptions of child sexual
abuse. Psychiatry, Psychology & Law,11(2),
308–318.
Maynard, C., & Wiederman, M. (1997). Under-
graduate students’ perceptions of child
sexual abuse: Effects of age, sex and
gender-role attitudes. Child Sexual Abuse
and Neglect,21, 833–844.
Saradjian, J., & Hanks, H. (1996). Women who
sexually abuse children: From research to
clinical practice. Chichester, UK: Wiley.
Tabachnick, B.G., & Fidell, L.S. (2001). Using
multivariate statistics (4th edn.). Boston:
Allyn & Bacon.
The Daily Telegraph. (2006, October 30).
Pervert teachers’ sex shame. The Daily Tele-
graph. Retrieved from http://www.news.
com.au/dailytelegraph/story/0,22049,20668
567-5001021,00.html
Timmerman, G. (2003). Sexual harassment of
adolescents perpetrated by teachers and by
peers: An exploration of the dynamics of
power, culture, and gender in secondary
schools. Sex Roles,48, 231–244.
Wellman, M.M. (1993). Child sexual abuse and
gender differences: Attitudes and preva-
lence. Child Abuse and Neglect,17, 539–547.
Generation and Detection of True and False Alibi Statements
1
Scott E. Culhane
a
, Andre Kehn
b
, Allyson J. Horgan
c
, Christian A. Meissner
d
,
Harmon M. Hosch
d
and Eric J. Wodahl
a
a
Department of Criminal Justice, University of Wyoming, Laramie, USA;
b
Department of Psychology,
University of North Dakota, Grand Forks, USA;
c
MRAC LLC, Cambridge, MA, USA;
d
Center for
Law and Human Behavior, Department of Psychology, University of Texas at El Paso, USA
This article reports two experiments focusing on two stages of the alibi process. In
Experiment 1, participants generated a true or false alibi for one of two dates (short or
long delay). Results showed that participants were most likely to report that they could
get alibi corroboration from a motivated alibi witnesses regardless of whether they were
being truthful or deceptive. Changes in details to the generated alibi were frequent for
both true and false statements. In Experiment 2, individuals were asked to discriminate
between true and false statements. The results indicated that participants were no better
than chance at detecting lies. As has been seen with in other domains (e.g. eyewitness
identification), confidence had no predictive power in distinguishing lies from true
statements.
Key words: alibi; alibi generation; autobiographical memory; deception detection.
The topic of alibis (Latin for ‘‘in another
place’’) has a more substantial history in
the criminal justice system than the realm
of research. However, it has received some
recent attention (Burke, Turtle, & Olson,
2007; Culhane & Hosch, 2004, 2012;
Dysart & Strange, 2012; Hosch, Culhane,
Jolly, Chavez, & Shaw, 2011; Olson &
Charman, 2012; Olson & Wells, 2004,
2012; Sommers & Douglas, 2007). There
are three primary domains of the criminal
justice process within which alibi claims
must be processed – namely, alibi genera-
tion (during which a suspect provides an
alibi for his or her whereabouts, including
any information which might be used to
validate the proposed alibi), alibi discrimi-
nation (during which time investigators
and prosecuting attorneys attempt to dis-
tinguish between truthful vs. deceptive alibi
claims), and finally alibi evaluation (in
which the alibi claim is evaluated in its
socially relevant, forensic context – in
particular, with regard to jurors’ evalua-
tions of the alibi in conjunction with other
direct and circumstantial evidence).
The majority of previous psychology
and law research has focused on the final
component: an individual’s evaluation of
alibi information. Alibi information can
contain pertinent details such as alibi
witnesses – individuals who are called
upon to support a suspect’s/defendant’s
alibi, and/or physical alibi evidence –
tangible items that substantiate an alibi
(e.g., ticket stubs, videotape surveillance,
Correspondence: Scott E. Culhane, Department of Criminal Justice, Department 3197, University of
Wyoming, Laramie, WY 82071, USA. Tel.: +1 307 766 2945, Fax: +1 307 766 3913;
Email: sculhane@uwyo.edu
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
620 S. E. Culhane et al.
timecard from work, etc.; Burke et al.,
2007). For example, Culhane and Hosch
(2004) and Olson and Wells (2004) each
examined issues of alibi witness’s motiva-
tion and its effect on evaluations of
defendant’s alibis. Each found that alibi
witnesses were only credible in the eyes of
the evaluator when the alibi witness had
no relationship to the defendant. This
work has been expanded by Hosch and
colleagues (2011), who found that within
subcategories of relationship types (biolo-
gical, affinal and social) differences in
evaluators’ scepticism exist, such that the
closer the alibi witness is to the suspect/
defendant, the greater the scepticism
displayed. This research provides a nice
start for understanding alibis, but there
are other aspects to an alibi. More
specifically, we know very little about
two steps in the alibi process: generation
and detection.
Experiment 1: Alibi Generation
According to Burke and colleagues
(2007), ‘‘there appears to be an implicit
assumption within the justice system that
people facing potential imprisonment . . .
should be able to recollect their actions
from an earlier date’’ (p. 12). Indeed,
researchers and scholars have often noted
the extent to which alibis are, on the one
hand, expected of suspects, but on the
other hand, sceptically dismissed as mis-
leading or false (Sanders, 1984). A history
of cognitive research on autobiographical
memory, suggests that individuals may
experience difficulty recounting their
whereabouts or behaviour on specific
days and times, and especially after
extended delays (see Conway, 2002, for
a review of the autobiographical memory
literature).
The manner in which memories are
stored and retrieved may therefore play a
role in the scepticism noted in alibi evalua-
tions. Should investigators trust an
individual who provides a general alibi
initially, but later recalls certain event
details upon reflection (and thereby alters
their alibi statement)? The nature of ‘‘gen-
erative retrieval’’ (Conway, 1996; Norman
& Brobow, 1979) suggests that, individuals
engage in a discrepancy-reducing process in
which specific retrieval cues are used to
search memory for relevant knowledge that
is then assessed for veracity prior to out-
put. Cognitive research suggests that gen-
eral event knowledge serves as the entry-
point for retrieval (Conway, 1996; Schac-
ter, Norman, & Koustaal, 1998), and it is
this level of access that is frequently
reported when individuals are prompted
to respond quickly with certainty. By
contrast, the direct retrieval of specific-
event knowledge occurs sporadically and
often results in response to a specific cue
that the individual later ties to the event.
As a result of these retrieval processes, it is
likely that an individual will recall only
general event details at the time of inter-
rogation, but may later, upon reflection,
derive a specific memory for their where-
abouts. This inconsistency may be a
natural form of autobiographical retrieval
that is unrelated to the veracity of the
memory.
Little published research has examined
the typology and precision of alibis that are
generated in a forensic context. Olson and
Wells (2004) were the first to suggest a
distinction in the taxonomy of alibi gen-
eration and resultant believability. The
authors combined the two alibi compo-
nents of alibi witnesses and physical alibi
evidence to create a grid for evaluating
alibi believability based upon its strength.
Culhane, Hosch, and Kehn (2008) have
examined differences in the generation of
alibi statements by distinct cultural groups
and examined the extent to which each
could provide alibi witnesses and physical
evidence to support alibis. Their research
asked participants to provide an honest
alibi for 9.00 p.m. two days earlier. Two
S. E. Culhane
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 621
timecard from work, etc.; Burke et al.,
2007). For example, Culhane and Hosch
(2004) and Olson and Wells (2004) each
examined issues of alibi witness’s motiva-
tion and its effect on evaluations of
defendant’s alibis. Each found that alibi
witnesses were only credible in the eyes of
the evaluator when the alibi witness had
no relationship to the defendant. This
work has been expanded by Hosch and
colleagues (2011), who found that within
subcategories of relationship types (biolo-
gical, affinal and social) differences in
evaluators’ scepticism exist, such that the
closer the alibi witness is to the suspect/
defendant, the greater the scepticism
displayed. This research provides a nice
start for understanding alibis, but there
are other aspects to an alibi. More
specifically, we know very little about
two steps in the alibi process: generation
and detection.
Experiment 1: Alibi Generation
According to Burke and colleagues
(2007), ‘‘there appears to be an implicit
assumption within the justice system that
people facing potential imprisonment . . .
should be able to recollect their actions
from an earlier date’’ (p. 12). Indeed,
researchers and scholars have often noted
the extent to which alibis are, on the one
hand, expected of suspects, but on the
other hand, sceptically dismissed as mis-
leading or false (Sanders, 1984). A history
of cognitive research on autobiographical
memory, suggests that individuals may
experience difficulty recounting their
whereabouts or behaviour on specific
days and times, and especially after
extended delays (see Conway, 2002, for
a review of the autobiographical memory
literature).
The manner in which memories are
stored and retrieved may therefore play a
role in the scepticism noted in alibi evalua-
tions. Should investigators trust an
individual who provides a general alibi
initially, but later recalls certain event
details upon reflection (and thereby alters
their alibi statement)? The nature of ‘‘gen-
erative retrieval’’ (Conway, 1996; Norman
& Brobow, 1979) suggests that, individuals
engage in a discrepancy-reducing process in
which specific retrieval cues are used to
search memory for relevant knowledge that
is then assessed for veracity prior to out-
put. Cognitive research suggests that gen-
eral event knowledge serves as the entry-
point for retrieval (Conway, 1996; Schac-
ter, Norman, & Koustaal, 1998), and it is
this level of access that is frequently
reported when individuals are prompted
to respond quickly with certainty. By
contrast, the direct retrieval of specific-
event knowledge occurs sporadically and
often results in response to a specific cue
that the individual later ties to the event.
As a result of these retrieval processes, it is
likely that an individual will recall only
general event details at the time of inter-
rogation, but may later, upon reflection,
derive a specific memory for their where-
abouts. This inconsistency may be a
natural form of autobiographical retrieval
that is unrelated to the veracity of the
memory.
Little published research has examined
the typology and precision of alibis that are
generated in a forensic context. Olson and
Wells (2004) were the first to suggest a
distinction in the taxonomy of alibi gen-
eration and resultant believability. The
authors combined the two alibi compo-
nents of alibi witnesses and physical alibi
evidence to create a grid for evaluating
alibi believability based upon its strength.
Culhane, Hosch, and Kehn (2008) have
examined differences in the generation of
alibi statements by distinct cultural groups
and examined the extent to which each
could provide alibi witnesses and physical
evidence to support alibis. Their research
asked participants to provide an honest
alibi for 9.00 p.m. two days earlier. Two
groups, non-Hispanic Whites and Hispa-
nics were compared in their abilities to
recall witnesses and physical evidence.
Findings indicate that participants relied
heavily on motivated alibi witnesses, or
individuals with some degree of motivation
to lie for the suspect/defendant, for alibis
(see Sullivan, 1971). Furthermore, the
purported ability to attain physical evi-
dence was substantially less than their
ability to recruit witnesses. This result is
disconcerting given the importance of
physical evidence in the eyes of those who
are asked to evaluate alibis (Culhane et al.,
2008). That is, most people find physical
evidence of great importance in evaluating
other’s alibis.
While this is an important start to the
understanding of generated alibis, further
research on the precision and typology of
alibis is warranted, particularly that which
might link alibi research with several
decades of research on autobiographical
memory. The study presents an attempt to
expand upon research conducted to date
and broaden the theoretical application
described.
Changes Made in Alibi Statements
Should investigators trust individuals who
provide a general alibi initially, but later
recall certain event details upon reflection
(and thereby alter their alibi statement)?
Kearns (1982) argued that alibis can
actually help a prosecutor in some
instances. Two examples are alibis offered
only after the trial has started and/or
those proven to be false. Thus, the
burden falls on the suspect to be accurate
in their generation and to present the
alibi within a reasonable amount of time.
Over half of the States and the Federal
Government in the USA require a defen-
dant to give the prosecution advanced
notice of an alibi (Gooderson, 1977).
Rule 12.1 of the Federal Rules of
Criminal Procedure directs that: (1) the
government may request written notice of
any intention to provide an alibi defence,
and (2) the defendant must serve such
notice within 10 days after the request.
Furthermore, the defendant has to iden-
tify specific places he or she claims to
have been at the time of the crime, and
information of alibi witnesses on whom
he or she intends to rely.
In the landmark case Williams v.
Florida (1970), the United States Supreme
Court discussed the validity of procedural
requirements in declaration of alibis. Spe-
cifically, the court found that the advanced
notice of intent to file an alibi on the part
of the defence is a legitimate request by the
government. The court saw this as an
opportunity to investigate credible alibis
and dismiss cases in which they are
proffered. Epstein’s (1964) research would
suggest that this is not the case, however.
His findings indicate that more than three-
quarters of prosecutors would proceed
with a criminal case, regardless of a
credible alibi determination. Instead, it
would be their intention to discredit the
defence’s alibi at trial.
Some defendants may have made alibi
statements early on during investigations
of the crime (e.g., an interrogation). An
important component of a generated alibi
statement is the potential for one to make a
change to his or her story. It would seem
that there are pros and cons to changing
one’s alibi statement. It may be beneficial
to a suspect if an alibi were to be
strengthened (e.g., an additional person
was included as a witness or a ‘‘new’’
witness was not as vested in the suspect) or
the individual can provide additional de-
tails about their whereabouts.
Two studies have touched on the
evaluation of changed stories. Culhane
et al.’s (2008) Alibi Belief Questionnaire
(ABQ) directly questioned the issue. Over
88% of participants agreed (responded
strongly agree, agree or somewhat agree)
with the question ‘‘If a defendant changes
Generation and
Detection of True
and False Alibi
Statements
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
622 S. E. Culhane et al.
his or her alibi after a police interview he or
she is probably lying.’’ More extensive
testing on the evaluation of changed alibi
statements was conducted by Culhane and
Hosch (2012). The participants in his
study included current law enforcement
officers, undergraduate students who in-
tended to enter law enforcement, and a
group of students who did not intend to
do so. The individuals evaluated alibi
statements that were changed (strength-
ened or weakened) or were maintained.
Their results suggested that all groups
were more favourable toward alibi state-
ments which were maintained than those
that were changed in either direction. It
should be noted that the change in the
suspect’s alibi was the alibi witness present
at the time. These findings suggest that
during the generation of an alibi the
suspect must be very careful and accurate
to maximize believability.
The question then becomes: ‘‘How
often does one make a change to a
generated autobiographical memory after
some time to reflect?’’ To date, only one
published article has directly examined the
degree to which generated alibis change
over time. Olson and Charman (2012)
questioned individuals about their where-
abouts for a particular time on two dates,
one recent (3 days) and one very distant (6–
14 weeks). An alibi was presented for those
dates at two different times during an
initial experimental session. The research-
ers conducted follow-up interviews after a
48-hr delay. Olson and Charman were able
to record changes in alibi statements and
their results showed that 11% of state-
ments changed in their content. That is, the
alibi was wrong with regards to their
original narrative. A further 25% of
individuals made changes to the evidence
elements provided in the initial alibi state-
ment. That is, they changed person evi-
dence or physical evidence associated with
their narrative. The higher change percen-
tages came from alibis that were more
distant in date for overall narrative, but the
changes in evidence were very much the
same for the two dates.
Hypotheses
The previous research on alibi generation
represents that same ‘‘small start to . . . a
potentially rich literature’’ that Olson and
Wells (2004, p. 159) identified for alibi
research as a whole. One important sub-
area of alibi research that has not been
investigated sufficiently is determining how
to distinguish between truthful and fabri-
cated alibi statements. The current research
was designed to begin to fill this void.
Specifically, the following questions were
focal: Are there differences in the types of
witnesses or the availability of corroborat-
ing evidence for true and false statements?
Furthermore, if individuals are providing
false alibis, what types of changes would
we see? Should investigators trust indivi-
duals who provide a general alibi initially,
but later recall a different event upon
reflection (and thereby altering their alibi
statement)?
It was predicted that the majority of
both true and false alibis provided would
include motivated witnesses (see Culhane
et al., 2008). In addition, false alibis
providers were predicted to make more
changes in alibi over time, and have less
supporting evidence than true alibi state-
ments. The time delay conditions should
reflect poorer performance with longer
delays, including more change in alibis,
and less evidence to validate the alibi
claims. We did not expect to see a high
number of complete changes to alibi
statements (a contradictory story at
Time 2).
Method
Participants
Two hundred and ninety-six under-
graduate students were recruited for this
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 623
his or her alibi after a police interview he or
she is probably lying.’’ More extensive
testing on the evaluation of changed alibi
statements was conducted by Culhane and
Hosch (2012). The participants in his
study included current law enforcement
officers, undergraduate students who in-
tended to enter law enforcement, and a
group of students who did not intend to
do so. The individuals evaluated alibi
statements that were changed (strength-
ened or weakened) or were maintained.
Their results suggested that all groups
were more favourable toward alibi state-
ments which were maintained than those
that were changed in either direction. It
should be noted that the change in the
suspect’s alibi was the alibi witness present
at the time. These findings suggest that
during the generation of an alibi the
suspect must be very careful and accurate
to maximize believability.
The question then becomes: ‘‘How
often does one make a change to a
generated autobiographical memory after
some time to reflect?’’ To date, only one
published article has directly examined the
degree to which generated alibis change
over time. Olson and Charman (2012)
questioned individuals about their where-
abouts for a particular time on two dates,
one recent (3 days) and one very distant (6–
14 weeks). An alibi was presented for those
dates at two different times during an
initial experimental session. The research-
ers conducted follow-up interviews after a
48-hr delay. Olson and Charman were able
to record changes in alibi statements and
their results showed that 11% of state-
ments changed in their content. That is, the
alibi was wrong with regards to their
original narrative. A further 25% of
individuals made changes to the evidence
elements provided in the initial alibi state-
ment. That is, they changed person evi-
dence or physical evidence associated with
their narrative. The higher change percen-
tages came from alibis that were more
distant in date for overall narrative, but the
changes in evidence were very much the
same for the two dates.
Hypotheses
The previous research on alibi generation
represents that same ‘‘small start to . . . a
potentially rich literature’’ that Olson and
Wells (2004, p. 159) identified for alibi
research as a whole. One important sub-
area of alibi research that has not been
investigated sufficiently is determining how
to distinguish between truthful and fabri-
cated alibi statements. The current research
was designed to begin to fill this void.
Specifically, the following questions were
focal: Are there differences in the types of
witnesses or the availability of corroborat-
ing evidence for true and false statements?
Furthermore, if individuals are providing
false alibis, what types of changes would
we see? Should investigators trust indivi-
duals who provide a general alibi initially,
but later recall a different event upon
reflection (and thereby altering their alibi
statement)?
It was predicted that the majority of
both true and false alibis provided would
include motivated witnesses (see Culhane
et al., 2008). In addition, false alibis
providers were predicted to make more
changes in alibi over time, and have less
supporting evidence than true alibi state-
ments. The time delay conditions should
reflect poorer performance with longer
delays, including more change in alibis,
and less evidence to validate the alibi
claims. We did not expect to see a high
number of complete changes to alibi
statements (a contradictory story at
Time 2).
Method
Participants
Two hundred and ninety-six under-
graduate students were recruited for this
project. Three students failed to return
for the second session and were subse-
quently dropped. The participants were
drawn from introductory courses in psy-
chology and criminal justice at two mid-
sized universities (168 from a Southwest
US university and 125 from a Rocky
Mountain US university). Two ethnic
groups comprised the majority of par-
ticipants: Hispanics and non-Hispanic
Whites. The students received either
course credit or extra credit for their
participation. Three participants were re-
moved because they had missing data on
their alibi statements or could not remem-
ber what they were doing at the time in
question.
Materials and Procedure
Experiment 1 used a 2 (true or false
statement) 62 (short or long delay) 62
(Time 1 or Time 2) mixed factorial
design. Participants were asked to provide
a true or false alibi statement regarding
their whereabouts at 9.30 p.m. either 5 or
12 days prior. False alibi providers were
told not to create an alibi based on what
they would normally be doing during that
time, but instead to generate a completely
novel alibi. After completing the written
statement, a subset of participants was
videotaped repeating their story which
formed the basis for materials in Experi-
ment 2. All participants were asked to
return two days later and provide evi-
dence that might support their alibi
statement, including statements from their
alibi witness, physical alibi evidence or
both. When the students returned, they
were again asked to provide an alibi for
the same day and time in question and
their supporting evidence was collected.
When providing the second alibi state-
ment, participants were instructed to
recall the actual alibi again, not simply
recall what they had written during the
previous session.
Results
Alibi statements were evaluated based
upon the type of alibi witness(es) provided,
the type of physical evidence provided, the
degree of change between statements pro-
vided and the level of detail provided as a
function of time delay and the true vs. false
alibi condition. Alibi statements were
coded by two independent raters at each
institution for typology, precision and
evidence information. For categorical cod-
ings (i.e., type of alibi, consistency of alibis,
witness type, etc.), kappa values ranged
from .83 to .91 at the Southwest institution
and .80 to .89 at the Rocky Mountain
institution. According to Landis and Koch
(1977), kappa values exceeding .79 are
representative of outstanding inter-rater
reliability. Coders met to discuss and
review coding results and then recoded
statements with discrepancies. Statements
were then rechecked in order to ensure all
differences were resolved.
Alibi Type
Alibi statements were coded as one of six
types: inconsistent (two different accounts),
home with friends/family, away from home
with friends/family, alone, out of town and
at work. There was no significant difference
for location, w
2
(5, N¼293) ¼10.12, p¼
ns,j¼.19. Therefore, data were col-
lapsed. As expected, most participants
reported motivated alibi witnesses that
could support their claim. The majority of
individuals (50.9%) reported being away
from home with friends/family, followed
by those who said they were home with
friends/family (25.6%), those who claimed
to be alone (8.9%), inconsistent (6.5%),
out of town (5.1%) or at work (3.1%). Chi-
square analyses found no significant effect
of delay, w
2
(5, N¼293) ¼1.18, p¼ns,j¼
.06. However, there was a significant effect
when truthful and fabricated alibi state-
ments were compared, w
2
(5, N¼293) ¼
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
624 S. E. Culhane et al.
31.24, p5.001, j¼.33. The greatest
difference showed that truthful participants
were more likely to report being home with
friends/family (35.9%) compared with par-
ticipants who provided a false alibi
(15.9%). A delay 6truth 6alibi type
chi-square test found that the home with
friends/family category was the only sig-
nificantly different category, w
2
(1,
N¼75) ¼4.24, p¼.039, j¼.24.
2
Witness Type
One hundred and ninety-five participants
were able to bring a witness statement as a
supporting piece of evidence. Thirty of
these participants were able to bring two
witness statements. The 225 alibi witnesses
purported in the statements were classified
as being one of the following types: family,
friends/familiar others, co-workers, signi-
ficant others or non-familiar others (Fig-
ure 1). There was a significant difference for
location, w
2
(4, N¼225) ¼36.32, p5.001,
j¼.38. The Southwestern students had a
much higher number of family member
alibi witnesses that did the Rocky Moun-
tain group (42 to 5), as well as significant
other alibi witnesses (25 to 11). The Rocky
Mountain students had a slightly higher
number of friends as alibi witnesses (59 to
74). This is similar to findings in previous
research (Culhane et al., 2008).
Because of the aforementioned signifi-
cant difference, individual chi-square ana-
lyses were conducted for each site. For the
Southwest participants, chi-square analyses
revealed a significant effect for truth con-
dition, w
2
(4, N¼133) ¼10.82, p¼.029,
j¼.29. However, there was not a sig-
nificant difference in the Rocky Mountain
sample, w
2
(4, N¼92) ¼4.94, p¼ns,
j¼.23. Follow-up analyses with all parti-
cipants revealed that participants who
generated false alibis mentioned at least
one friend as a witnesses (59.7%) more
often than those that generated truthful
alibis (40.3%), w
2
(1, N¼293) ¼5.71,
p¼.017, j¼.14. There was a similar
pattern of findings with respect to delay.
Specifically, the Southwest students had a
significant effect for delay condition, w
2
(4,
N¼133) ¼9.60, p¼.048, j¼.27, but,
there was no difference in the Rocky
Mountain sample, w
2
(4, N¼92) ¼4.32, p
ns,j¼.22.
Evidence
Forty-seven participants (16.0%) could not
provide any supporting alibi evidence for
the time in question. The remaining parti-
cipants brought witness statements
(62.8%) or physical evidence (16.4%). A
small portion of participants brought both
(4.8%). There was a significant difference
for location, as the Southwest sample was
more likely to procure physical evidence
than the Rocky Mountain participants.
Therefore, the two groups were analysed
separately. In this instance, however, the
focal variables were not significant in either
group’s analyses, largest w
2
(3, N¼121) ¼
3.08, p¼ns,j¼.16.
Analysis of covariance (ANCOVA),
controlling for data collection site, revealed
no significant main effects for the procure-
ment of supporting witness statements
(range 0 – 2). However, a truth 6delay
interaction approached traditional signifi-
cance, such that those participants who
gave false alibis for 12 days prior provided
significantly more witness statements than
any other group, F(1, 288) ¼3.57, p¼.06,
Z
2
¼.01. Figure 2 displays this interaction.
It appears that those falsifying their state-
ments following long delays worked harder
at generating a strong alibi claim.
There was a significant main effect for
the number of words provided in the
generated alibi at both Time 1 and Time
2. The falsely generated alibi statements
were longer than truthful ones, F(1,
288) ¼7.23, p¼.008, Z
2
¼.02 (Time 1)
and F(1, 288) ¼5.86, p¼.016, Z
2
¼.02
(Time 2). Therefore, being long winded in a
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 625
31.24, p5.001, j¼.33. The greatest
difference showed that truthful participants
were more likely to report being home with
friends/family (35.9%) compared with par-
ticipants who provided a false alibi
(15.9%). A delay 6truth 6alibi type
chi-square test found that the home with
friends/family category was the only sig-
nificantly different category, w
2
(1,
N¼75) ¼4.24, p¼.039, j¼.24.
2
Witness Type
One hundred and ninety-five participants
were able to bring a witness statement as a
supporting piece of evidence. Thirty of
these participants were able to bring two
witness statements. The 225 alibi witnesses
purported in the statements were classified
as being one of the following types: family,
friends/familiar others, co-workers, signi-
ficant others or non-familiar others (Fig-
ure 1). There was a significant difference for
location, w
2
(4, N¼225) ¼36.32, p5.001,
j¼.38. The Southwestern students had a
much higher number of family member
alibi witnesses that did the Rocky Moun-
tain group (42 to 5), as well as significant
other alibi witnesses (25 to 11). The Rocky
Mountain students had a slightly higher
number of friends as alibi witnesses (59 to
74). This is similar to findings in previous
research (Culhane et al., 2008).
Because of the aforementioned signifi-
cant difference, individual chi-square ana-
lyses were conducted for each site. For the
Southwest participants, chi-square analyses
revealed a significant effect for truth con-
dition, w
2
(4, N¼133) ¼10.82, p¼.029,
j¼.29. However, there was not a sig-
nificant difference in the Rocky Mountain
sample, w
2
(4, N¼92) ¼4.94, p¼ns,
j¼.23. Follow-up analyses with all parti-
cipants revealed that participants who
generated false alibis mentioned at least
one friend as a witnesses (59.7%) more
often than those that generated truthful
alibis (40.3%), w
2
(1, N¼293) ¼5.71,
p¼.017, j¼.14. There was a similar
pattern of findings with respect to delay.
Specifically, the Southwest students had a
significant effect for delay condition, w
2
(4,
N¼133) ¼9.60, p¼.048, j¼.27, but,
there was no difference in the Rocky
Mountain sample, w
2
(4, N¼92) ¼4.32, p
ns,j¼.22.
Evidence
Forty-seven participants (16.0%) could not
provide any supporting alibi evidence for
the time in question. The remaining parti-
cipants brought witness statements
(62.8%) or physical evidence (16.4%). A
small portion of participants brought both
(4.8%). There was a significant difference
for location, as the Southwest sample was
more likely to procure physical evidence
than the Rocky Mountain participants.
Therefore, the two groups were analysed
separately. In this instance, however, the
focal variables were not significant in either
group’s analyses, largest w
2
(3, N¼121) ¼
3.08, p¼ns,j¼.16.
Analysis of covariance (ANCOVA),
controlling for data collection site, revealed
no significant main effects for the procure-
ment of supporting witness statements
(range 0 – 2). However, a truth 6delay
interaction approached traditional signifi-
cance, such that those participants who
gave false alibis for 12 days prior provided
significantly more witness statements than
any other group, F(1, 288) ¼3.57, p¼.06,
Z
2
¼.01. Figure 2 displays this interaction.
It appears that those falsifying their state-
ments following long delays worked harder
at generating a strong alibi claim.
There was a significant main effect for
the number of words provided in the
generated alibi at both Time 1 and Time
2. The falsely generated alibi statements
were longer than truthful ones, F(1,
288) ¼7.23, p¼.008, Z
2
¼.02 (Time 1)
and F(1, 288) ¼5.86, p¼.016, Z
2
¼.02
(Time 2). Therefore, being long winded in a
Figure 1. Types of alibi witnesses reported for short and long delay separated by data collection site (N¼225).
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
626 S. E. Culhane et al.
generated alibi was indicative of deceit.
There was no significant effect of data site
collection or delay. Likewise, the interac-
tion of the two focal variables was not
significant.
Changes Made to Alibi Statements
Analyses show that complete changes to the
story provided were very rare (i.e., giving
the wrong alibi at Time 1 and subsequently
giving a different story at Time 2, such as
stating someone was home with friends at
Time 1 and then out of town at Time 2).
There was no difference between the data
collection sites for contradictory state-
ments, w
2
(1, N¼293) ¼0.05, p¼ns,j
7.01. Therefore, the data were collapsed.
Complete contradictions occurred in 54%
of the true cases and in only 11.9% of false
cases. While small, this difference was
statistically significant, w
2
(1, N¼293)
¼7.14, p¼.008, j¼.16. Thus, global
shifts in an alibi story are indicative of
deceit. There was no difference in delay,
however.
Discussion
The results of Experiment 1 suggest that it
may be difficult to distinguish between
honest and false alibi claims when con-
sidering the type of alibi provided. The
alibi ‘‘I was home with friends/family’’ was
indicative of a truth teller, which seems
contrary to anecdotal evidence. Further-
more, false alibi claims for longer delays
appeared to include a greater frequency of
alibi corroborating witnesses. However,
these witnesses were mostly friends and
therefore may be perceived as motivated to
lie for their friend. Nevertheless, it appears
that false alibi providers involving lengthier
delays may work harder to generate what
they perceive to be a more believable alibi
claim.
Previous research has shown that our
recall of autobiographical memories is
generally quite accurate, although fre-
quently lacking in event-specific details
(Howes & Katz, 1992). Our results suggest
that complete changes to the alibi story
may in fact be indicative of lying. For
Figure 2. Average number of witness statements provided by participants in Experiment 1.
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 627
generated alibi was indicative of deceit.
There was no significant effect of data site
collection or delay. Likewise, the interac-
tion of the two focal variables was not
significant.
Changes Made to Alibi Statements
Analyses show that complete changes to the
story provided were very rare (i.e., giving
the wrong alibi at Time 1 and subsequently
giving a different story at Time 2, such as
stating someone was home with friends at
Time 1 and then out of town at Time 2).
There was no difference between the data
collection sites for contradictory state-
ments, w
2
(1, N¼293) ¼0.05, p¼ns,j
7.01. Therefore, the data were collapsed.
Complete contradictions occurred in 54%
of the true cases and in only 11.9% of false
cases. While small, this difference was
statistically significant, w
2
(1, N¼293)
¼7.14, p¼.008, j¼.16. Thus, global
shifts in an alibi story are indicative of
deceit. There was no difference in delay,
however.
Discussion
The results of Experiment 1 suggest that it
may be difficult to distinguish between
honest and false alibi claims when con-
sidering the type of alibi provided. The
alibi ‘‘I was home with friends/family’’ was
indicative of a truth teller, which seems
contrary to anecdotal evidence. Further-
more, false alibi claims for longer delays
appeared to include a greater frequency of
alibi corroborating witnesses. However,
these witnesses were mostly friends and
therefore may be perceived as motivated to
lie for their friend. Nevertheless, it appears
that false alibi providers involving lengthier
delays may work harder to generate what
they perceive to be a more believable alibi
claim.
Previous research has shown that our
recall of autobiographical memories is
generally quite accurate, although fre-
quently lacking in event-specific details
(Howes & Katz, 1992). Our results suggest
that complete changes to the alibi story
may in fact be indicative of lying. For
Figure 2. Average number of witness statements provided by participants in Experiment 1.
honest alibi statements, errors in recall
could cluster around peripheral details that
were poorly encoded (Johnson & Sherman,
1990) or result from the blending of several
events in a schematic form (Schacter,
1999). Previous research has shown that
individuals can also produce erroneous
recall for events that were surprising,
important and emotional – so called
flashbulb memories (e.g. the Challenger
shuttle disaster, 9/11; Neisser & Harsch,
1992).
It is important to note some limitations
of Study 1. While many other studies do
not have issues using students instead of
community members (Bornstein, 1999),
alibi studies are different. The generated
alibis of older adults should differ from
college students. Anecdotes suggest that
the daily routine of a non-student seems to
be quite different from that of the average
university undergraduate. However, there
is more to the differences between the two
populations. Community members are
more likely to have additional family
members (spouse and children) and less
likely to spend the same amount of time
with friends.
Also, there was no sanction for perjury
by either the participant or the alibi
corroborator. In real-world situations
such penalties exist for both. However,
each would likely be affected differently. If
a guilty suspect is providing a false alibi,
the minor penalty of perjury would pale in
comparison with the more severe charge
the individual is accused of committing.
However, any penalty should scare poten-
tial false alibi corroborators from provid-
ing untrue statements.
Finally, a more comprehensive mea-
surement of time may influence alibi
results. For most adults the hours of 9.00
a.m. to 5.00 p.m. would result in a high
number of work-related alibis and even
students would have higher numbers of
school-related alibis. The 9.30 p.m. selec-
tion of time in our study allows
incorporates fewer crimes. According to
the National Crime Victimization Survey
(Bureau of Justice Statistics, 2007), only
29.7% of violent crimes and 12.9% of
property crimes occur between 6.00 p.m.
and midnight. More work with other
generated times will be useful, particularly
between the hours of 6.00 a.m. and 6.00
p.m.
Experiment 2: Alibi Discrimination
Ultimately, the alibi story may exclude the
person as a suspect, or it may cause more
suspicion to be placed on the person. This
depends on the believability and verifia-
bility of the alibi statement provided. The
question becomes, then, how good are
individuals at discriminating a person’s
guilt vs. innocence, or of his/her truth vs.
deception? To date, very little research has
examined whether or not naı
¨ve partici-
pants are capable of discriminating be-
tween truthful and deceptive alibi
statements (Stro
¨mwall, Granhag, & Jons-
son, 2003).
What we do know about individuals’
ability to discriminate between true and
false statements comes from the deception
literature (DePaulo, Lassiter, & Stone,
1982; Memon, Vrij, & Bull, 2003; Vrij,
2000). These studies, as well as others,
have consistently shown that naı
¨ve parti-
cipants, as well as trained professionals
are poor ‘‘lie detectors’’, most often
performing at the chance level of 50%
(Aamodt & Custer, 2006; Calcan
˜o, Keen,
Storey, Costello, & Aamodt, 2006). In-
deed, Bond and DePaulo’s (2008) meta-
analysis found that mean accuracy
was 556% after averaging across nearly
20,000 judges of veracity.
Most professionals display a guilt bias
(Meissner & Kassin, 2004), meaning that
they are more likely to state that any
particular individual is lying. While inves-
tigators appear to have a bias towards
mistrusting suspects’ statements, research
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
628 S. E. Culhane et al.
has generally indicated that untrained
individuals demonstrate a ‘‘truth’’ bias –
that is, they are often more trusting of an
individual’s statement (Bond & DePaulo,
2006; Vrij, 2000), including deception
judgements that are made across cultures
(Bond & Atoum, 2000). Furthermore,
training programmes have generally pro-
duced inconsistent improvements in per-
formance compared with control
conditions (Bull, 1989; Kassin & Fong,
1999; Porter, Woodworth, & Birt, 2000),
and police investigators and others with
relevant on-the-job experience have been
shown to perform only slightly better or no
better than naı
¨ve subjects (DePaulo &
Pfeifer, 1986; Ekman & O’Sullivan, 1991;
Ekman, O’Sullivan, & Frank, 1999; Kas-
sin, Meissner, & Norwick, 2005; Meissner
& Kassin, 2002). Thus, while many in the
law enforcement community assume, often
with great confidence, that investigators
can use verbal and non-verbal behavioural
cues to make accurate judgements of truth
and deception, there is little hard evidence
to support this assumption.
Autobiographical stories are quite dif-
ferent from statements used in much of the
deception-detection literature. The Un-
deutsch hypothesis (Stellar, 1989) suggests
that a statement extracted from an actual
experience should qualitatively differ from
a statement created by fantasy or imagina-
tion. For example, the stimulus materials in
a deception-detection study may involve
have participants doing or not doing some
task that is atypical to their daily lives (e.g.,
entering an office and stealing something or
not). The participants are then interviewed
and those interviews are recorded. These
recordings are then shown to or heard by a
new group of participants who make
judgements of their truthfulness/deception.
A genuine autobiographical memory
for a time and date is much more natural
to the participants when asked to recount
the events that took place. Furthermore,
they may prove more difficult to falsify, but
that remains to be seen. Therefore, it is
important to examine the participant’s
abilities to detect deception in autobiogra-
phical statements. Calcan
˜o et al. (2006)
videotaped students describing activities
over a one week period during a university
break. Half the students were lying and
half were truthful. Various groups (psy-
chologists, police officers, parole officers,
teachers, students and inmates) were then
presented the tapes in a deception-detec-
tion activity. Across groups mean accuracy
for the task was worse than chance
(42.76%).
Studies on alibi believability suggest
that naı
¨ve participants demonstrate a
proclivity towards scepticism when evalu-
ating an ‘‘alibi’’ claim (see Olson & Wells,
2004), such that the very context of the
criminal accusation may create a bias in
the evaluation of a suspect’s statement.
Researchers using forensically relevant
stimuli, however, have not generally found
such a scepticism bias when testing the
discrimination of true vs. false alibis. One
study is applicable to our questions
because of the use of alibi statements as
stimuli. Stro
¨mwall and colleagues (2003)
had pairs of participants generate true vs.
false alibis that would withstand interro-
gation, after which a separate pool of
participants evaluated the veracity of these
statements. A group of participants acted
in pairs that would either tell the truth or
be deceptive about a lunch they had. The
truthful pair actually went to lunch, where
the lying pair was given 30 minutes to
fabricate their story. All individuals were
then interrogated and their statements
were videotaped. A second group of
participants then watched the videotaped
interrogations and made veracity judge-
ments. Their results suggested a clear
‘‘truth bias’’ in the evaluation of alibi
statements.
Coffman, Meissner, and Marcon
(2005) similarly examined the extent to
which naı
¨ve participants could distinguish
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 629
has generally indicated that untrained
individuals demonstrate a ‘‘truth’’ bias –
that is, they are often more trusting of an
individual’s statement (Bond & DePaulo,
2006; Vrij, 2000), including deception
judgements that are made across cultures
(Bond & Atoum, 2000). Furthermore,
training programmes have generally pro-
duced inconsistent improvements in per-
formance compared with control
conditions (Bull, 1989; Kassin & Fong,
1999; Porter, Woodworth, & Birt, 2000),
and police investigators and others with
relevant on-the-job experience have been
shown to perform only slightly better or no
better than naı
¨ve subjects (DePaulo &
Pfeifer, 1986; Ekman & O’Sullivan, 1991;
Ekman, O’Sullivan, & Frank, 1999; Kas-
sin, Meissner, & Norwick, 2005; Meissner
& Kassin, 2002). Thus, while many in the
law enforcement community assume, often
with great confidence, that investigators
can use verbal and non-verbal behavioural
cues to make accurate judgements of truth
and deception, there is little hard evidence
to support this assumption.
Autobiographical stories are quite dif-
ferent from statements used in much of the
deception-detection literature. The Un-
deutsch hypothesis (Stellar, 1989) suggests
that a statement extracted from an actual
experience should qualitatively differ from
a statement created by fantasy or imagina-
tion. For example, the stimulus materials in
a deception-detection study may involve
have participants doing or not doing some
task that is atypical to their daily lives (e.g.,
entering an office and stealing something or
not). The participants are then interviewed
and those interviews are recorded. These
recordings are then shown to or heard by a
new group of participants who make
judgements of their truthfulness/deception.
A genuine autobiographical memory
for a time and date is much more natural
to the participants when asked to recount
the events that took place. Furthermore,
they may prove more difficult to falsify, but
that remains to be seen. Therefore, it is
important to examine the participant’s
abilities to detect deception in autobiogra-
phical statements. Calcan
˜o et al. (2006)
videotaped students describing activities
over a one week period during a university
break. Half the students were lying and
half were truthful. Various groups (psy-
chologists, police officers, parole officers,
teachers, students and inmates) were then
presented the tapes in a deception-detec-
tion activity. Across groups mean accuracy
for the task was worse than chance
(42.76%).
Studies on alibi believability suggest
that naı
¨ve participants demonstrate a
proclivity towards scepticism when evalu-
ating an ‘‘alibi’’ claim (see Olson & Wells,
2004), such that the very context of the
criminal accusation may create a bias in
the evaluation of a suspect’s statement.
Researchers using forensically relevant
stimuli, however, have not generally found
such a scepticism bias when testing the
discrimination of true vs. false alibis. One
study is applicable to our questions
because of the use of alibi statements as
stimuli. Stro
¨mwall and colleagues (2003)
had pairs of participants generate true vs.
false alibis that would withstand interro-
gation, after which a separate pool of
participants evaluated the veracity of these
statements. A group of participants acted
in pairs that would either tell the truth or
be deceptive about a lunch they had. The
truthful pair actually went to lunch, where
the lying pair was given 30 minutes to
fabricate their story. All individuals were
then interrogated and their statements
were videotaped. A second group of
participants then watched the videotaped
interrogations and made veracity judge-
ments. Their results suggested a clear
‘‘truth bias’’ in the evaluation of alibi
statements.
Coffman, Meissner, and Marcon
(2005) similarly examined the extent to
which naı
¨ve participants could distinguish
between the true and false alibis generated
in their naturalistic paradigm. In their
study, 20 male participants were asked to
provide alibis for their whereabouts during
a time in which a mock crime had been
committed either 2 or 5 days earlier. Only
half of the participants were, in fact, guilty
of the mock crime. All participants were
unexpectedly interviewed regarding their
whereabouts at the time of the event. That
is, participants were asked to return to the
lab several days later (for the guilty), or
they simply showed up for an experiment
(for the innocent). The 20 males were then
asked to describe their whereabouts for the
date/time in question. This ‘‘surprise inter-
view’’ prevented individuals from prepar-
ing an alibi ahead of time.
During the interview, participants were
given instructions to ‘‘provide a statement
of your whereabouts during the time of the
incident, being as specific as possible.’’
Guilty participants were instructed to
provide a convincing, yet false, alibi for
their whereabouts, while innocent partici-
pants were to provide an honest account of
their whereabouts. Each interview was
videotaped in its entirety.
The authors found that other partici-
pants’ discrimination performance was
significantly better for alibis generated
following a 2-day delay than for those
generated following a 5-day delay; how-
ever, no differences were observed in
measures of response criterion and partici-
pants generally demonstrated a ‘‘truth
bias’’ consistent with the previous decep-
tion literature. This may have been due to
the reduced the quality of the memories for
the 5-day truthful subjects, which would
make it more difficult to distinguish true vs.
false alibis. Taken together, it would
appear worthwhile to reconcile the rela-
tionship between judgements of believabil-
ity across alibi typology in previous studies
(Culhane & Hosch, 2004; Olson & Wells,
2004), and judgements of veracity in a
deception-detection paradigm.
One shortcoming of many deception-
detection studies, and the aforementioned
Stro
¨mwall et al. (2003) study, is the fact
that participants often times evaluate the
same story for various individuals. For
example, in the recent research by Klaver,
Lee, and Hart (2007), the generation of
false criminal statements were done so
identically. Participants were instructed to
view a videotaped crime and generate a
story from this incident for their false
statement. They were then taped giving a
false confession statement and those state-
ments were rated by undergraduate re-
search assistants. This means that the false
statements provided by the offenders were
generic in their description and raters heard
the same false story repeatedly. This is
problematic because participants could
make relative judgements by comparing
the various statements to each other and
not necessarily evaluate each individually.
The use of genuine alibi statements or self-
developed false alibis provides unique
information for detectors to consider.
Specifically, a person who makes veracity
judgements about such statements will hear
a different story during each discrimination
task.
It would appear that human judge-
ments of alibi veracity may prove no better
than chance detection, and further that an
important difference exists between the
‘‘scepticism’’ expressed by mock evaluators
in alibi evaluation studies (Olson & Wells,
2004) and the ‘‘truth bias’’ displayed in
studies on deception detection (Coffman
et al., 2005; Stro
¨mwall et al., 2003). More
work needs to be done to understand under
what conditions these divergent outcomes
are produced. However, no published
research has examined veracity judgements
of alibi statements. The purpose of the
current study was to evaluate naı
¨ve parti-
cipants’ ability to discriminate between
true and false alibi statements. We also
wanted to examine if the weak confidence-
accuracy relationship found in the
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
630 S. E. Culhane et al.
psycholegal literature was consistent for
the detection of deception of alibi state-
ments. Finally, we wanted to examine
whether gender (that of the speaker and
that of the evaluator) had an effect on
accuracy rates.
Hypotheses
As noted above, judgements of alibi state-
ments are qualitatively different from the
more traditional tests of deception detec-
tion. Because previous research has shown
that detection of honest and false autobio-
graphical statements are worse than chance
(Calcan
˜o et al., 2006), we expected that
participants would perform worse than
chance at detecting deceptive alibi state-
ments as a whole. We also explored the
relationship between participants’ confi-
dence ratings and their accuracy in dis-
criminating between true and false
statements. Again, previous research is
consistent in showing that no relationship
exists, and we do not expect one here.
However, a testing with alibi statements
has never been conducted.
Lastly, we tested four individual differ-
ences. Bond and DePaulo (2008) noted that
‘‘only a handful of investigators have
studied individual differences in lie detec-
tion ability’’ (p. 477). Further, their meta-
analysis suggested that while differences
among senders of deceptive (or not)
messages were relatively strongly related
to the detection of differences, perceiver
variables were not. Similarly, Aamodt and
Custer (2006) reported that there were not
enough studies to investigate the relation
between personality variables were related
to accuracy of detection of deception in
their meta-analysis. Therefore, we included
measures of individual’s intelligence, need
for cognition and need to belong in
addition to testing gender differences.
These measured individual difference di-
mensions were chosen because of their
widespread use and application in other
fields of psychology. We believe that
because they are among the strongest
predictors of other behaviours (e.g. general
intelligence as a predictor of the capacity to
process complex sets of information) they
might be among the potential explanatory
constructs in explaining alibi perception.
We admit that the work is purely explora-
tory, yet we can make predictions based on
previous research. To date, testing of
individual differences has yielded very little
explanation of variance in deception detec-
tion, and no research has tested any
individual difference measure when detect-
ing honest/false alibis.
Cacioppo and Petty (1982) argue that
need for cognition (NC) is ‘‘an individual’s
tendency to engage in and enjoy thinking’’
(p. 116). People with high levels of NC tend
to have a greater elaboration on informa-
tion presented, as well as more generated
thoughts about the information. A person
who is low in NC tends to avoid cognitive
effort. It is possible that high NC indivi-
duals are more accurate in their veracity
judgements because of this increased ela-
boration. However, this has yet to be
determined.
Recent research in the social psycho-
logical literature provides evidence that
individuals high in need to belong (NTB)
are particularly attentive and accurate in
decoding social cues (Pickett, Gardner, &
Knowles, 2004). In three studies, Pickett
et al. (2004) illustrated that individuals
high in NTB are more accurate in
detecting vocal tone and facial emotion,
which are both important in accurately
detecting deceitful statements (DePaulo
et al., 2003). Therefore, we included a
measure to determine if a correlation
between NTB and detecting accurate alibi
statements exists in this very specific task
of deception detection. The linkage be-
tween the aforementioned research should
predict that those high in NTB would be
more accurate and confident in their
judgements.
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 631
psycholegal literature was consistent for
the detection of deception of alibi state-
ments. Finally, we wanted to examine
whether gender (that of the speaker and
that of the evaluator) had an effect on
accuracy rates.
Hypotheses
As noted above, judgements of alibi state-
ments are qualitatively different from the
more traditional tests of deception detec-
tion. Because previous research has shown
that detection of honest and false autobio-
graphical statements are worse than chance
(Calcan
˜o et al., 2006), we expected that
participants would perform worse than
chance at detecting deceptive alibi state-
ments as a whole. We also explored the
relationship between participants’ confi-
dence ratings and their accuracy in dis-
criminating between true and false
statements. Again, previous research is
consistent in showing that no relationship
exists, and we do not expect one here.
However, a testing with alibi statements
has never been conducted.
Lastly, we tested four individual differ-
ences. Bond and DePaulo (2008) noted that
‘‘only a handful of investigators have
studied individual differences in lie detec-
tion ability’’ (p. 477). Further, their meta-
analysis suggested that while differences
among senders of deceptive (or not)
messages were relatively strongly related
to the detection of differences, perceiver
variables were not. Similarly, Aamodt and
Custer (2006) reported that there were not
enough studies to investigate the relation
between personality variables were related
to accuracy of detection of deception in
their meta-analysis. Therefore, we included
measures of individual’s intelligence, need
for cognition and need to belong in
addition to testing gender differences.
These measured individual difference di-
mensions were chosen because of their
widespread use and application in other
fields of psychology. We believe that
because they are among the strongest
predictors of other behaviours (e.g. general
intelligence as a predictor of the capacity to
process complex sets of information) they
might be among the potential explanatory
constructs in explaining alibi perception.
We admit that the work is purely explora-
tory, yet we can make predictions based on
previous research. To date, testing of
individual differences has yielded very little
explanation of variance in deception detec-
tion, and no research has tested any
individual difference measure when detect-
ing honest/false alibis.
Cacioppo and Petty (1982) argue that
need for cognition (NC) is ‘‘an individual’s
tendency to engage in and enjoy thinking’’
(p. 116). People with high levels of NC tend
to have a greater elaboration on informa-
tion presented, as well as more generated
thoughts about the information. A person
who is low in NC tends to avoid cognitive
effort. It is possible that high NC indivi-
duals are more accurate in their veracity
judgements because of this increased ela-
boration. However, this has yet to be
determined.
Recent research in the social psycho-
logical literature provides evidence that
individuals high in need to belong (NTB)
are particularly attentive and accurate in
decoding social cues (Pickett, Gardner, &
Knowles, 2004). In three studies, Pickett
et al. (2004) illustrated that individuals
high in NTB are more accurate in
detecting vocal tone and facial emotion,
which are both important in accurately
detecting deceitful statements (DePaulo
et al., 2003). Therefore, we included a
measure to determine if a correlation
between NTB and detecting accurate alibi
statements exists in this very specific task
of deception detection. The linkage be-
tween the aforementioned research should
predict that those high in NTB would be
more accurate and confident in their
judgements.
Participant gender should not be a
factor in the overall detection of state-
ments by members of the same/other
gender (Bond & DePaulo, 2008). How-
ever, we were particularly interested in
whether the genders differ in their truth/
lie biases for alibi statements. Deception-
detection interactions among gender and
the biases, as well as other variables, have
been found in previous research (de-
Turck, 1991; Tornqvist, 2002). Therefore,
we expected an interaction, such that
females are ‘‘better’’ at detecting truths,
while males are ‘‘better’’ at detecting lies.
This would suggest that males are more
prone to the lie bias and females to the
truth bias.
Method
Participants
Sixty-four undergraduate students (32 male
and 32 female) drawn from an introductory
criminal justice class at a large Rocky
Mountain university participated in Ex-
periment 2. The sample was mostly non-
Hispanic Whites (89%) and had a mean
age of 20.28 years (SD ¼4.36). One male
participant had missing data on a truth/lie
judgement and was excluded from subse-
quent analyses. Participants received extra
credit towards their class grade for
participation.
Materials
Shipley Institute of Living Scale (Zachery,
1991). The Shipley Institute of Living
Scale (SILS) is designed to assess general
intellectual functioning in adults and
adolescents. It is one of the oldest tests
to still be administered in its original
form. The SILS is a self-administered test
and consists of two subtests: vocabulary
and abstraction. The vocabulary subtest
has 40 multiple-choice questions in which
the respondent is asked to choose which
of four words is closest in meaning to a
target word. The vocabulary subtest relies
on verbal skills, which include reading
ability, verbal comprehension, acquired
knowledge, long-term memory and con-
cept formation. The abstraction subtest
has 20 questions in which sequences of
numbers, letters or words are presented
with the final element in each sequence
omitted. The respondent is required to
complete each of the sequences. The
abstraction subtest relies more heavily
on attentional abilities, letter, word and
number concept formation, abstract
thinking, cognitive flexibility, analysis
and synthesis, processing speed, long-
term memory, and specific vocabulary
and arithmetic skills.
Need for Cognition – Short Form (Caciop-
po, Petty, & Kao, 1984). Participants were
given the 18-item Need for Cognition –
Short Form (NFC-SF), which evaluates
an individual’s penchant for engaging in
cognitive processes (e.g. ‘‘The notion of
thinking abstractly is appealing to me’’).
Participants indicated how much they
agree with each statement on a 5-point
scale, ranging from extremely unlike me
(1) to extremely like me (5). The measure
has been shown to be reliable, valid and
appropriate for use in diverse populations
(Culhane, Morera, & Hosch, 2004; Cul-
hane, Morera, & Watson, 2006).
Need to Belong Scale (Leary, Kelly,
Cottrell, & Schreindorfer, 2007).
Participants were given a modified version
of the Need to Belong Scale (NTBS)
consisting of 10 items that assess the
degree to which participants desire to be
accepted by others, seek to become part of
social groups and react negatively to social
rejection. For example, sample items
included, ‘‘I do not like being alone’’ and
‘‘I seldom worry about whether other
people care about me’’ (reverse coded).
Participants responded on a five-point
scale ranging from 1 (strongly disagree) to
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
632 S. E. Culhane et al.
5 (strongly agree). Higher scores indicate a
greater need to belong.
Alibi statements. A random sample of 12
videotaped alibi statements (half true and
half false; half male and half female
speakers) was taken from a pool of 70
speakers collected in Experiment 1. The
individuals providing the alibis stated their
whereabouts for a specific time and date.
No two alibi statements were identical in
their description. Each video lasted
between 45 seconds and 1 minute.
Procedure
The project used a 2 (subject gender) 62
(speaker gender) 62 (true or false state-
ment) mixed design, with the latter two
variables representing within-subject con-
ditions. To control for specific speaker
effects, three versions of each condition
were randomly presented. Therefore, par-
ticipants heard 12 statements from 12
different speakers. No two stories were
identical, however, as alibis had different
activities and alibi witnesses proffered by
the speakers. The speakers selected were
chosen at random.
Students’ data were collected in groups.
Once informed consent was obtained,
participants were told that they would
view 12 statements that could either be
true or false. They were instructed to
determine for each speaker whether the
individual was lying or telling the truth.
They were told that all speakers may be
telling the truth, may be lying or that there
may be a combination of both. It was up to
them to make the determination. Partici-
pants were also asked to rate their con-
fidence in their decision after viewing each
video on a 1 (not at all confident) to 10
(extremely confident) scale. After all parti-
cipants made judgements for all 12 alibi
statements, they were then given the packet
of questionnaires. This packet contained a
demographics form, which was followed by
the four measures. The four individual
differences measures were randomized in
their presentation. After completion of the
packet, our participants were thanked, had
any questions answered and were
dismissed.
Results
Alibi Discrimination
In global judgement accuracy, the results of
this study echoed those of previous studies
from the deception literature. Contrary to
our expectations, the overall accuracy rate
for participants was 43.3% (SD ¼13.6%).
In terms of accuracy for true statements,
57.4% of participants accurately detected
when an individual was telling the truth.
Only 29.1% of participants were able to
detect a false alibi accurately. The overall
mean confidence was 6.96. The accuracy-
confidence relationship was non-significant,
r(61) ¼.09, p¼ns.
An analysis of variance (ANOVA) was
conducted on the 2 6262 mixed design.
The results found a significant main effect
for true and false alibi statements, F(1,
61) ¼77.73, p5.001, Z
2
¼.56. The par-
ticipants were ‘‘more accurate’’ for the
truthful statements and clearly demon-
strated the truth bias. Furthermore, there
was a significant interaction among parti-
cipant gender and accuracy of decisions in
true and false statements, F(1, 61) ¼5.10,
p¼.027, Z
2
¼.077. This effect can be seen
in Figure 3. As predicted, males were more
accurate at detecting lies, while females
were better at detecting truthful alibi
statements. There were no other significant
main effects or interactions, largest F(1,
61) ¼2.02, p¼.16, Z
2
¼.032.
Individual Difference Measures
The means, standard deviations, alpha
coefficients of the SILS, NFC-SF and
NTBS, as well as their correlations with
confidence and accuracy on true and false
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 633
5 (strongly agree). Higher scores indicate a
greater need to belong.
Alibi statements. A random sample of 12
videotaped alibi statements (half true and
half false; half male and half female
speakers) was taken from a pool of 70
speakers collected in Experiment 1. The
individuals providing the alibis stated their
whereabouts for a specific time and date.
No two alibi statements were identical in
their description. Each video lasted
between 45 seconds and 1 minute.
Procedure
The project used a 2 (subject gender) 62
(speaker gender) 62 (true or false state-
ment) mixed design, with the latter two
variables representing within-subject con-
ditions. To control for specific speaker
effects, three versions of each condition
were randomly presented. Therefore, par-
ticipants heard 12 statements from 12
different speakers. No two stories were
identical, however, as alibis had different
activities and alibi witnesses proffered by
the speakers. The speakers selected were
chosen at random.
Students’ data were collected in groups.
Once informed consent was obtained,
participants were told that they would
view 12 statements that could either be
true or false. They were instructed to
determine for each speaker whether the
individual was lying or telling the truth.
They were told that all speakers may be
telling the truth, may be lying or that there
may be a combination of both. It was up to
them to make the determination. Partici-
pants were also asked to rate their con-
fidence in their decision after viewing each
video on a 1 (not at all confident) to 10
(extremely confident) scale. After all parti-
cipants made judgements for all 12 alibi
statements, they were then given the packet
of questionnaires. This packet contained a
demographics form, which was followed by
the four measures. The four individual
differences measures were randomized in
their presentation. After completion of the
packet, our participants were thanked, had
any questions answered and were
dismissed.
Results
Alibi Discrimination
In global judgement accuracy, the results of
this study echoed those of previous studies
from the deception literature. Contrary to
our expectations, the overall accuracy rate
for participants was 43.3% (SD ¼13.6%).
In terms of accuracy for true statements,
57.4% of participants accurately detected
when an individual was telling the truth.
Only 29.1% of participants were able to
detect a false alibi accurately. The overall
mean confidence was 6.96. The accuracy-
confidence relationship was non-significant,
r(61) ¼.09, p¼ns.
An analysis of variance (ANOVA) was
conducted on the 2 6262 mixed design.
The results found a significant main effect
for true and false alibi statements, F(1,
61) ¼77.73, p5.001, Z
2
¼.56. The par-
ticipants were ‘‘more accurate’’ for the
truthful statements and clearly demon-
strated the truth bias. Furthermore, there
was a significant interaction among parti-
cipant gender and accuracy of decisions in
true and false statements, F(1, 61) ¼5.10,
p¼.027, Z
2
¼.077. This effect can be seen
in Figure 3. As predicted, males were more
accurate at detecting lies, while females
were better at detecting truthful alibi
statements. There were no other significant
main effects or interactions, largest F(1,
61) ¼2.02, p¼.16, Z
2
¼.032.
Individual Difference Measures
The means, standard deviations, alpha
coefficients of the SILS, NFC-SF and
NTBS, as well as their correlations with
confidence and accuracy on true and false
statements, are located in Table 1. Overall
intelligence was not related to either
accuracy or confidence, and neither subtest
produced significant results. The coefficient
alpha for the SILS overall was .80, with the
vocabulary subtest having an internal
consistency of .76 and .67 for the abstrac-
tion subtest. Finally, the NFC-SF (alpha ¼
.88) and the NTBS (alpha ¼.84) did not
correlate with either accuracy or
confidence.
Discussion
The primary aim of this Experiment 2 was
to examine whether individuals are capable
of discriminating between honest and false
alibi statements, and whether the confi-
dence in their decision is predictive of
accuracy. We expected that genuinely
developed autobiographical statements
would be easier to discriminate amongst
because of their true self-developed nature.
However, the obtained results parallel
Figure 3. Mean accuracy for males and females on veracity judgements in Experiment 2.
Table 1. Means, standard deviations, and correlations with accuracy and confidence for males and
females (n¼61).
Mean (SD) Correlations
Measure Males (n¼29) Females (n¼32) Accuracy Confidence
SILS 42.97 (5.38) 43.00 (6.26) 7.11 7.18
VOC 28.38 (3.78) 28.13 (5.02) 7.11 7.17
ABS 14.59 (2.44) 14.87 (2.32) 7.06 7.13
NFC-SF 60.52 (8.92) 63.55 (7.73) 7.11 .18
NTBS 28.69 (6.06) 30.53 (7.74) 7.04 7.02
Note. *p5.05. Correlations are collapsed across gender. Listwise deletion utilized. SILS, Shipley Institute of
Living Scale; VOC, vocabulary; ABS, abstraction; NFC-SF, Need for Cognition – Short Form; NTBS, Need to
Belong Scale.
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
634 S. E. Culhane et al.
those found in studies from the deception-
detection literature. Individuals are no
better than chance, and actually worse
than chance in the current study, in
detecting deceptive alibi statements.
Furthermore, naı
¨ve participants display
the commonly known truth bias, or ten-
dency to assume that speakers are being
honest, in their evaluation of alibi state-
ments. That is, they were more likely to
believe any given statement is true rather
than that it is a lie. The confidence
participants had in their decision was not
predictive of accuracy, a finding that is well
known from the eyewitness literature.
Furthermore, this is consistent with pre-
vious meta-analytic findings of a correla-
tion of .05 between confidence and
accuracy (Aamodt & Custer, 2006).
Gender also did not have a direct
impact upon whether or not an individual
was able to distinguish between a true
statement and a lie. However, the interac-
tion between statements and gender on
predicting accuracy was significant. In sum,
the findings of this study inform us that
individuals are not able to accurately
distinguish between a true alibi statement
and a false alibi statement. These results
mirror those found in the deception litera-
ture (Meissner & Kassin, 2004).
The individual difference measures were
underwhelming in their prediction of de-
tection accuracy and non-existent for con-
fidence. In general, researchers would be
discouraged by so many null findings in
their project. However, this is not an
uncommon finding in the deception-detec-
tion literature (Bond & DePaulo, 2008),
and such work assists future researchers
from wasting time and/or resources.
General Discussion
Wells and colleagues (1998) analysed the
first 40 DNA exoneration cases (see
www.innocenceproject.org) and found
that 28% of the cases (11 of 40) involved
some degree of scepticism regarding alibi
evidence that was provided. Given the
potential for alibi evidence to exonerate
those falsely accused, psychologists and
criminologists have begun to examine the
generation and assessment of alibis in
forensic contexts (for a review, see Burke
& Turtle, 2003; Burke et al., 2007). Never-
theless, a dearth of published research
exists on the psychological processes un-
derlying the generation and discrimination
of alibis.
The generation of alibis from Experi-
ment 1 replicated the findings of Culhane
et al. (2008). Specifically, participants rely
upon person evidence much more than
physical evidence to substantiate their
story. Also, the majority of participants
produce motivated individuals to serve as
alibi witnesses. This does not speak well for
an innocent individual, as most fact-finders
(e.g., jurors, police investigators) look for
evidence that most cannot produce. In-
deed, in an interrogation setting, many
individuals put themselves at risk, by
simply telling the truth about their alibi.
The deception ‘‘detection’’ in Experiment 2
research confirms the inability of under-
graduates to identify liars from their peer
group.
Limitations
Another limitation is the generation of
false alibis in Experiment 1 and their
subsequent use in Experiment 2. An
important addition to the generation of
false autobiographical memories would be
to gather data on the person’s true story. In
real-world contexts, the true story for a
false alibi provider would most likely be
that he or she was committing the crime,
but in the experimental laboratory we
could ascertain such information. This
could serve as an effective manipulation
check and support our expectation that the
false alibi information provided is genu-
inely 100% false. In addition, the stakes for
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 635
those found in studies from the deception-
detection literature. Individuals are no
better than chance, and actually worse
than chance in the current study, in
detecting deceptive alibi statements.
Furthermore, naı
¨ve participants display
the commonly known truth bias, or ten-
dency to assume that speakers are being
honest, in their evaluation of alibi state-
ments. That is, they were more likely to
believe any given statement is true rather
than that it is a lie. The confidence
participants had in their decision was not
predictive of accuracy, a finding that is well
known from the eyewitness literature.
Furthermore, this is consistent with pre-
vious meta-analytic findings of a correla-
tion of .05 between confidence and
accuracy (Aamodt & Custer, 2006).
Gender also did not have a direct
impact upon whether or not an individual
was able to distinguish between a true
statement and a lie. However, the interac-
tion between statements and gender on
predicting accuracy was significant. In sum,
the findings of this study inform us that
individuals are not able to accurately
distinguish between a true alibi statement
and a false alibi statement. These results
mirror those found in the deception litera-
ture (Meissner & Kassin, 2004).
The individual difference measures were
underwhelming in their prediction of de-
tection accuracy and non-existent for con-
fidence. In general, researchers would be
discouraged by so many null findings in
their project. However, this is not an
uncommon finding in the deception-detec-
tion literature (Bond & DePaulo, 2008),
and such work assists future researchers
from wasting time and/or resources.
General Discussion
Wells and colleagues (1998) analysed the
first 40 DNA exoneration cases (see
www.innocenceproject.org) and found
that 28% of the cases (11 of 40) involved
some degree of scepticism regarding alibi
evidence that was provided. Given the
potential for alibi evidence to exonerate
those falsely accused, psychologists and
criminologists have begun to examine the
generation and assessment of alibis in
forensic contexts (for a review, see Burke
& Turtle, 2003; Burke et al., 2007). Never-
theless, a dearth of published research
exists on the psychological processes un-
derlying the generation and discrimination
of alibis.
The generation of alibis from Experi-
ment 1 replicated the findings of Culhane
et al. (2008). Specifically, participants rely
upon person evidence much more than
physical evidence to substantiate their
story. Also, the majority of participants
produce motivated individuals to serve as
alibi witnesses. This does not speak well for
an innocent individual, as most fact-finders
(e.g., jurors, police investigators) look for
evidence that most cannot produce. In-
deed, in an interrogation setting, many
individuals put themselves at risk, by
simply telling the truth about their alibi.
The deception ‘‘detection’’ in Experiment 2
research confirms the inability of under-
graduates to identify liars from their peer
group.
Limitations
Another limitation is the generation of
false alibis in Experiment 1 and their
subsequent use in Experiment 2. An
important addition to the generation of
false autobiographical memories would be
to gather data on the person’s true story. In
real-world contexts, the true story for a
false alibi provider would most likely be
that he or she was committing the crime,
but in the experimental laboratory we
could ascertain such information. This
could serve as an effective manipulation
check and support our expectation that the
false alibi information provided is genu-
inely 100% false. In addition, the stakes for
providing alibis were low for our partici-
pants. Recent research has shown that
individuals who engaged in real mock
transgressions (i.e. eliciting true lies in a
laboratory setting through a cheating
paradigm; Russano, Meissner, Narchet, &
Kassin, 2005) in high stakes situations gave
off more diagnostic cues (i.e. non-verbal
cues) which enabled observers to make
more accurate judgements of veracity than
individuals who engaged in mock trans-
gressions (i.e. were told to tell a convincing
lie; Hatz, 2007).Therefore, researchers
should strive for higher stakes and real
transgressions in their alibi material
generation.
Future Research
There is still much we do not know about
all three phases of alibis. Recent research in
autobiographical memory has begun to
examine differences in the contents of recall
provided across cultures – particularly with
regard to independent vs. interdependent
cultures. For example, Conway, Wang,
Hanyu, and Haque (2005) examined the
autobiographical recall of participants
from several countries. While participants
demonstrated similar periods of recall
across the lifespan, content analyses in-
dicated significant differences in the typol-
ogy of memories produced such that
independent cultures (such as the majority
of the U.S.) produced memories that were
more self-focused while interdependent
cultures (such as China) generated mem-
ories that involved a more social
orientation.
There are many unresolved questions
about the detection of false alibi state-
ments. If the generation of autobiographi-
cal memories is different as the Undeutsch
hypothesis (Stellar, 1989) suggests, then
there should be viable ways to detect
honest and false statements. Recent re-
search (Sartori, Agosta, Zogmaister, Fer-
rara, & Castiello, 2008) has found that
autobiographical memories can be distin-
guished from others via reaction times.
Truly developed memories had faster
responses then false events not experienced.
The authors did not apply their work to
alibis, but it could easily be adapted.
Other research should evaluate whether
professionals with training in deception
detection perform better than naı
¨ve parti-
cipants when evaluating the veracity of
alibi statements. Bond’s (2008) research
found that experts in detection were highly
accurate in their veracity judgements. The
use of such experts in alibi statement
detection, and a subsequent comparison
with traditional methods of deception-
detection research, are both still lacking.
Only then can we determine whether
there are individual markers that aid
individuals when evaluating specific alibi
statements.
Acknowledgements
Thanks are extended to Amy Iberlin, Chelsie
Hess, Lisa Wempen, Rosie Aguilar and Zulema
Hermasillo for assistance with data collection.
Notes
1. Portions of Experiment 1 were presented as
a paper at the 2007 annual American
Psychology–Law Society conference in
Jacksonville, Florida and as a poster at
the 2007 Interrogations & Confessions: A
Conference Exploring Research, Practice,
and Policy in El Paso, TX. Portions of
Experiment 2 were presented as a paper at
the 2007 annual American Psychology-Law
Society conference in Jacksonville, Florida.
2. The alibi category ‘‘at work’’ could not be
compared, as only truthful alibis presented
such statements. That is, no false alibi
providers said ‘‘I was at work.’’
References
Aamodt, M.G., & Custer, H. (2006). Who can
best catch a liar? A meta-analysis of
individual differences in detecting deception.
Forensic Examiner,15(1), 6–11.
Bond, C.F. Jr, & Atoum, A.O. (2000). Interna-
tional deception. Personality & Social Psy-
chology Bulletin,26, 385–395.
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
636 S. E. Culhane et al.
Bond, C.F. Jr, & DePaulo, B.M. (2006). Accu-
racy of deception judgments. Personality &
Social Psychology Review,10, 214–234.
Bond, C.F. Jr, & DePaulo, B.M. (2008).
Individual differences in judging deception:
Accuracy and bias. Psychological Bulletin,
134, 477–492.
Bond, G.D. (2008). Deception detection ex-
pertise. Law and Human Behavior,32, 339–
351.
Bornstein, B.H. (1999). The ecological validity
of jury simulations: Is the jury still out?
Law and Human Behavior,23, 75–91.
Bull, R. (1989). Can training enhance the
detection of deception? In J.C. Yuille
(Ed.), Credibility assessment (pp. 83–99).
London: Kluwer Academic.
Bureau of Justice Statistics. (2007). National
Crime Victimization Survey. Available from
http://bjs.ojp.usdoj.gov/index.cfm?ty¼pb
detail&iid¼1743 (accessed 22 May 2010).
Burke, T.M., & Turtle, J.W. (2003). Alibi
evidence in criminal investigations and
trials: Psychological and legal factors. Ca-
nadian Journal of Police & Security Services,
1, 193–201.
Burke, T., Turtle, J.W., & Olson, E. (2007).
Alibi evidence in criminal investigations and
trials. In M. Toglia, J.D. Read, D.F. Ross,
& R.C.L. Lindsay (Eds.), Adult eyewitness
testimony: A handbook (pp. 157–174). Hills-
dale, NJ: Erlbaum.
Cacioppo, J.T., & Petty, R.E. (1982). The need
for cognition. Journal of Personality and
Social Psychology,42, 116–131.
Cacioppo, J.T., Petty, R.E., & Kao, C.F. (1984).
The efficient assessment of need for cogni-
tion. Journal of Personality Assessment,48,
306–307.
Calcan
˜o, M.M., Keen, A.M., Storey, A.L.,
Costello, A.R., & Aamodt, M.G. (2006,
October). Occupational differences in the
ability to detect deception. Paper presented
at the annual meeting of the Society for
Police and Criminal Psychology, Washing-
ton, DC.
Coffman, K.A.J., Meissner, C.A., & Marcon, J.
(2005, March). Detecting true and false alibis
in a mock crime paradigm. Paper presented
at the American Psychology–Law Society
Conference, La Jolla, CA.
Conway, M.A. (1996). Autobiographical memories
and autobiographical knowledge. In D.C.
Rubin (Ed.), Remembering our past: Studies
in autobiographical memory (pp. 67–93). Cam-
bridge, UK: Cambridge University Press.
Conway, M.A. (2002). Sensory perceptual
episodic memory and its context: Autobio-
graphical memory. In J.P. Aggleton & A.
Baddeley (Eds.), Episodic memory: New
directions in research (pp. 53–70). London:
Oxford University Press.
Conway, M.A., Wang, Q., Hanyu, K., &
Haque, S. (2005). A cross-cultural investiga-
tion of autobiographical memory: On the
universality and cultural variation of the
reminiscence bump. Journal of Cross-Cul-
tural Psychology,36, 739–749.
Culhane, S.E., & Hosch, H.M. (2004). An alibi
witness’s influence on juror’s decision mak-
ing. Journal of Applied Social Psychology,
34, 1604–1616.
Culhane, S.E., & Hosch, H.M. (2012). Chan-
ging your alibi: Current law enforcement,
future law enforcement, and layperson
reactions. Criminal Justice and Behavior,
39, 958–977.
Culhane, S.E., Hosch, H.M., & Kehn, A.
(2008). Alibi generation: Data from U.S.
Hispanics and U.S. Non-Hispanic Whites.
Journal of Ethnicity in Criminal Justice,6,
177–199.
Culhane, S.E., Morera, O.F., & Hosch, H.M.
(2004). The factor structure of the Need for
Cognition-short form in a Hispanic sample.
The Journal of Psychology,138, 77–88.
Culhane, S.E., Morera, O.F., & Watson, P.J.
(2006). The assessment of factorial invar-
iance in Need for Cognition using Hispanic
and Anglo samples. Journal of Psychology,
140, 53–67.
DePaulo, B.M., Lassiter, G.D., & Stone, J.I.
(1982). Attentional determinants of
success at detecting deception and truth.
Personality and Social Psychology Bulletin,
8, 273–279.
DePaulo, B.M., 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
DePaulo, B.M., & Pfeifer, R.L. (1986). On-the-
job experience and skill at detecting decep-
tion. Journal of Applied Social Psychology,
16, 249–267.
Dysart, J.E., & Strange, D. (2012). Beliefs about
alibis and alibi investigations: A survey of
law enforcement. Psychology, Crime & Law,
18, 11–25.
Ekman, P., & O’Sullivan, M. (1991). Who can
catch a liar? American Psychologist,46,
913–920.
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
Generation and Detection of True and False Alibi Statements 637
Bond, C.F. Jr, & DePaulo, B.M. (2006). Accu-
racy of deception judgments. Personality &
Social Psychology Review,10, 214–234.
Bond, C.F. Jr, & DePaulo, B.M. (2008).
Individual differences in judging deception:
Accuracy and bias. Psychological Bulletin,
134, 477–492.
Bond, G.D. (2008). Deception detection ex-
pertise. Law and Human Behavior,32, 339–
351.
Bornstein, B.H. (1999). The ecological validity
of jury simulations: Is the jury still out?
Law and Human Behavior,23, 75–91.
Bull, R. (1989). Can training enhance the
detection of deception? In J.C. Yuille
(Ed.), Credibility assessment (pp. 83–99).
London: Kluwer Academic.
Bureau of Justice Statistics. (2007). National
Crime Victimization Survey. Available from
http://bjs.ojp.usdoj.gov/index.cfm?ty¼pb
detail&iid¼1743 (accessed 22 May 2010).
Burke, T.M., & Turtle, J.W. (2003). Alibi
evidence in criminal investigations and
trials: Psychological and legal factors. Ca-
nadian Journal of Police & Security Services,
1, 193–201.
Burke, T., Turtle, J.W., & Olson, E. (2007).
Alibi evidence in criminal investigations and
trials. In M. Toglia, J.D. Read, D.F. Ross,
& R.C.L. Lindsay (Eds.), Adult eyewitness
testimony: A handbook (pp. 157–174). Hills-
dale, NJ: Erlbaum.
Cacioppo, J.T., & Petty, R.E. (1982). The need
for cognition. Journal of Personality and
Social Psychology,42, 116–131.
Cacioppo, J.T., Petty, R.E., & Kao, C.F. (1984).
The efficient assessment of need for cogni-
tion. Journal of Personality Assessment,48,
306–307.
Calcan
˜o, M.M., Keen, A.M., Storey, A.L.,
Costello, A.R., & Aamodt, M.G. (2006,
October). Occupational differences in the
ability to detect deception. Paper presented
at the annual meeting of the Society for
Police and Criminal Psychology, Washing-
ton, DC.
Coffman, K.A.J., Meissner, C.A., & Marcon, J.
(2005, March). Detecting true and false alibis
in a mock crime paradigm. Paper presented
at the American Psychology–Law Society
Conference, La Jolla, CA.
Conway, M.A. (1996). Autobiographical memories
and autobiographical knowledge. In D.C.
Rubin (Ed.), Remembering our past: Studies
in autobiographical memory (pp. 67–93). Cam-
bridge, UK: Cambridge University Press.
Conway, M.A. (2002). Sensory perceptual
episodic memory and its context: Autobio-
graphical memory. In J.P. Aggleton & A.
Baddeley (Eds.), Episodic memory: New
directions in research (pp. 53–70). London:
Oxford University Press.
Conway, M.A., Wang, Q., Hanyu, K., &
Haque, S. (2005). A cross-cultural investiga-
tion of autobiographical memory: On the
universality and cultural variation of the
reminiscence bump. Journal of Cross-Cul-
tural Psychology,36, 739–749.
Culhane, S.E., & Hosch, H.M. (2004). An alibi
witness’s influence on juror’s decision mak-
ing. Journal of Applied Social Psychology,
34, 1604–1616.
Culhane, S.E., & Hosch, H.M. (2012). Chan-
ging your alibi: Current law enforcement,
future law enforcement, and layperson
reactions. Criminal Justice and Behavior,
39, 958–977.
Culhane, S.E., Hosch, H.M., & Kehn, A.
(2008). Alibi generation: Data from U.S.
Hispanics and U.S. Non-Hispanic Whites.
Journal of Ethnicity in Criminal Justice,6,
177–199.
Culhane, S.E., Morera, O.F., & Hosch, H.M.
(2004). The factor structure of the Need for
Cognition-short form in a Hispanic sample.
The Journal of Psychology,138, 77–88.
Culhane, S.E., Morera, O.F., & Watson, P.J.
(2006). The assessment of factorial invar-
iance in Need for Cognition using Hispanic
and Anglo samples. Journal of Psychology,
140, 53–67.
DePaulo, B.M., Lassiter, G.D., & Stone, J.I.
(1982). Attentional determinants of
success at detecting deception and truth.
Personality and Social Psychology Bulletin,
8, 273–279.
DePaulo, B.M., 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
DePaulo, B.M., & Pfeifer, R.L. (1986). On-the-
job experience and skill at detecting decep-
tion. Journal of Applied Social Psychology,
16, 249–267.
Dysart, J.E., & Strange, D. (2012). Beliefs about
alibis and alibi investigations: A survey of
law enforcement. Psychology, Crime & Law,
18, 11–25.
Ekman, P., & O’Sullivan, M. (1991). Who can
catch a liar? American Psychologist,46,
913–920.
Ekman, P., O’Sullivan, M., & Frank, M.G.
(1999). A few can catch a liar. Psychological
Science,10, 263–266.
Epstein, D.M. (1964). Advance notice of alibi.
Journal of Criminal Law, Criminology, &
Police Science,55, 29–38.
Gooderson, R.N. (1977). Alibi. London: Heine-
mann Educational.
Hatz, J. (2007). Do deceptive behaviors and lie
detection abilities vary as a function of the
method used for eliciting lies? Unpublished
dissertation, University of Wyoming.
Hosch, H.M., Culhane, S.E., Jolly, K.W.,
Chavez, R.M., & Shaw, L.H. (2011). Effects
of an alibi witness’ relationship to the
defendant on mock jurors’ judgments. Law
and Human Behavior,35, 127–142.
Howes, J.L., & Katz, A.N. (1992). Remote
memory: Recalling autobiographical
and public events from across the
lifespan. Canadian Journal of Psychology,
46, 92–116.
Johnson, M.K., & Sherman, S.J. (1990). Con-
structing and reconstructing the past and
the future in the present. In E.T. Higgins &
R.M. Sorrentino (Eds.), Handbook of moti-
vation and cognition: Foundations of social
behavior (Vol. 2) (pp. 482–526). New York:
Guilford Press.
Kassin, S.M., & Fong, C.T. (1999). ‘‘I’m
innocent!’’ Effects of training on
judgments of truth and deception in the
interrogation room. Law & Human Beha-
vior,23, 499–516.
Kassin, S.M., Meissner, C.A., & Norwick, R.J.
(2005). ‘‘I’d know a false confession if I saw
one’’: A comparative study of college
students and police investigators. Law &
Human Behavior,29, 211–228.
Kearns, R.P. (1982). Recent developments in the
law concerning alibi and identification evi-
dence. The Criminal Law Quarterly,25, 47–
65.
Klaver, J.R., Lee, Z., & Hart, S.D. (2007).
Psychopathy and nonverbal indicators of
deception in offenders. Law and Human
Behavior,31, 337–351.
Landis, J.R., & Koch, G.G. (1977).
The measurement of observer agreement
for categorical data. Biometrics,33, 159–
174.
Leary, M.R., Kelly, K.M., Cottrell, C.A., &
Schreindorfer, L.S. (2007). Individual differ-
ences in the need to belong: Mapping the
nomological network. Unpublished manu-
script, Duke University.
Meissner, C.A., & Kassin, S.M. (2002). ‘‘He’s
guilty!’’: Investigator bias in judgments of
truth and deception. Law & Human Beha-
vior,26, 469–480.
Meissner, C.A., & Kassin, S.M. (2004). ‘‘You’re
guilty, so just confess!’’ Cognitive
and behavioral confirmation biases in the
interrogation room. In D. Lassiter (Ed.),
Interrogations, confessions, and entrapment.
New York: Kluwer Academic/Plenum
Press.
Memon, A., Vrij, A., & Bull, R. (2003).
Psychology and law: Truthfulness, accuracy
and credibility. London: Jossey-Bass.
Neisser, U., & Harsch, N. (1992). Phantom
flashbulbs: False recollections of hearing the
news about Challenger. In E. Winograd &
U. Neisser (Eds.), Affect and accuracy in
recall: Studies of ‘‘flashbulb memories’’ (pp.
9–31). Cambridge, UK: Cambridge Univer-
sity Press.
Norman, D.A., & Brobow, D.G. (1979). De-
scriptions and intermediate stage in
memory retrieval. Cognitive Psychology,
11, 107–123.
Olson, E.A., & Charman, S.D. (2012). ‘But can
you prove it?’ – Examining the quality of
innocent suspects’ alibis. Psychology, Crime
& Law,18, 453–471.
Olson, E.A., & Wells, G.L. (2004). What makes
a good alibi: A proposed taxonomy. Law
and Human Behavior,28, 157–176.
Olson, E.A., & Wells, G.L. (2012). The alibi-
generation effect: Alibi-generation experi-
ence influences alibi evaluation. Legal and
Criminological Psychology,17, 151–164.
Pickett, C.L., Gardner, W.L., & Knowles, M.
(2004). Getting a cue: The need to belong
and enhanced sensitivity to social cues.
Personality and Social Psychological Bulle-
tin,30, 1095–1107. doi: 10.1177/0146167203
262085
Porter, S., Woodworth, M., & Birt, A.R. (2000).
Truth, lies, and videotape: An investigation
ofthe ability of federal parole officers to
detect deception. Law and Human Behavior,
24, 643–658.
Russano, M.B., Meissner, C.A., Narchet, F.M.,
& Kassin, S. (2005). Investigating true and
false confessions within a novel experimen-
tal paradigm. Psychological Science,16,
481–486.
Sanders, R. (1984). Helping the jury evaluate
eyewitness testimony: The need for addi-
tional safeguards. American Journal of
Criminal Law,12, 189–220.
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014
638 S. E. Culhane et al.
Sartori, G., Agosta, S., Zogmaister, C., Ferrara,
S.D., & Castiello, U. (2008). How to
accurately detect autobiographical events.
Psychological Science,19, 772–780.
Schacter, D.L. (1999). The seven sins of
memory: Insights from psychology and
cognitive neuroscience. American Psycholo-
gist,54, 182–203.
Schacter, D.L., Norman, K.A., & Koustaal, W.
(1998). The cognitive neuroscience of con-
structive memory. Annual Review of Psy-
chology,49, 289–318.
Sommers, S.R., & Douglas, A.B. (2007). Con-
text matters: Alibi strength varies according
to evaluator perspective. Legal and Crimin-
ological Psychology,12, 41–54.
Stellar, M. (1989). Recent developments in
statement analysis. In J.C. Yuille (Ed.),
Credibility assessment (pp. 135–154). Dor-
drecht, the Netherlands: Kluwer.
Stro
¨mwall, L.A., Granhag, P.A., & Jonsson, A-
C. (2003). Deception among pairs: ‘‘Let’s
say we had lunch and hope they will
swallow it!’’ Psychology, Crime, & Law,9,
109–124.
Sullivan, T.P. (1971). Presentation of the
defense. The Journal of Criminal Law,
Criminology, and Police Science,62, 153–
172.
Tornqvist, J.S. (2002). The detectability and
believability of lies and truths in the e-mail
medium. Unpublished dissertation, Univer-
sity of Virginia.
Vrij, A. (2000). Detecting lies and deceit:
Psychology of lying and the implications for
professional practice. Chichester, UK:
Wiley.
Wells, G.L., Small, M., Penrod, S., Malpass,
R.S., Fulero, S.M., & Brimacombe, C.A.E.
(1998). Eyewitness identification proce-
dures: Recommendations for lineups and
photospreads. Law & Human Behavior,22,
603–647.
Williams v. Florida, 399 U.S.78 (1970).
Zachery, R.A. (1991). Shipley Institute of Living
Scale, Revised manual. Los Angeles: Wes-
tern Psychological Services.
Downloaded by [The University of Texas at El Paso] at 12:21 06 January 2014