Content uploaded by Julia Shaw
Author content
All content in this area was uploaded by Julia Shaw on Jan 29, 2015
Content may be subject to copyright.
Psychological Science
1 –11
© The Author(s) 2015
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0956797614562862
pss.sagepub.com
Research Article
If one wants to recall the details of an important life
event, subjectively it may seem easy to dig into the
secure memory vaults of one’s mind and retrieve the
relevant information. This assumption of memory as a
largely reliable process traditionally forms a major part
of the foundation of the legal system, in which memory
accounts of witnesses—and when they confess, defen-
dants—can play a key role in judicial decision making.
Such individuals are often expected to reliably recall
details of a crime, and their memory statements are
generally assumed to be valid. Indeed, a confession is
one of the most potent forms of legal evidence (e.g.,
Cutler, 2012; Gudjonsson & Pearse, 2011; Kassin,
Bogart, & Kerner, 2012). However, although the assump-
tion of memory being generally reliable may be intui-
tively appealing, many studies have yielded evidence
of reconstructive processes and distortions in memory
in many legally relevant situations (e.g., Frenda,
Nichols, & Loftus, 2011; Nash & Wade, 2008), and only
in a few situations does it seem that memory is particu-
larly resistant to distortion (e.g., Oeberst & Blank,
2012).
Researchers have been able to induce participants to
generate various types of false autobiographical accounts,
including accounts of getting lost in a shopping mall
(Loftus, 1997), being involved in an accident at a family
wedding (Hyman, Husband, & Billings, 1995), having tea
with Prince Charles (Strange, Sutherland, & Garry, 2006),
being attacked by a vicious animal (Porter, Yuille, &
Lehman, 1999), and cheating on a recent test (Russano,
Meissner, Narchet, & Kassin, 2005). These memories may
feel “real” because rememberers can generate event
details that were never mentioned by the interviewer.
The mind seems to be able to construct information from
internal and external sources to generate a coherent but
false picture of what occurred (e.g., Frenda etal., 2011).
These plausible confabulations are likely constructed
from real autobiographical memory fragments but are
configured in ways that depict events that did not occur
562862PSSXXX10.1177/0956797614562862Shaw, PorterRich False Memories of Crime
research-article2014
Corresponding Author:
Julia Shaw, University of Bedfordshire–Psychology, A209, University
Square, Luton LU1 3JU, United Kingdom
E-mail: julishaw@gmail.com
Constructing Rich False Memories of
Committing Crime
Julia Shaw1 and Stephen Porter2
1University of Bedfordshire and 2University of British Columbia
Abstract
Memory researchers long have speculated that certain tactics may lead people to recall crimes that never occurred, and
thus could potentially lead to false confessions. This is the first study to provide evidence suggesting that full episodic
false memories of committing crime can be generated in a controlled experimental setting. With suggestive memory-
retrieval techniques, participants were induced to generate criminal and noncriminal emotional false memories, and
we compared these false memories with true memories of emotional events. After three interviews, 70% of participants
were classified as having false memories of committing a crime (theft, assault, or assault with a weapon) that led to
police contact in early adolescence and volunteered a detailed false account. These reported false memories of crime
were similar to false memories of noncriminal events and to true memory accounts, having the same kinds of complex
descriptive and multisensory components. It appears that in the context of a highly suggestive interview, people can
quite readily generate rich false memories of committing crime.
Keywords
false memory, episodic memory, legal confession, adolescent delinquency
Received 2/5/14; Revision accepted 11/14/14
Psychological Science OnlineFirst, published on January 14, 2015 as doi:10.1177/0956797614562862
by guest on January 15, 2015pss.sagepub.comDownloaded from
2 Shaw, Porter
(Conway, 2002). They are known as honest lies
(Moscovitch, 1989), pseudomemories (Lindsay, Hagen,
Read, Wade, & Garry, 2004), phantom recollective experi-
ences (Brainerd & Reyna, 2002), or autobiographical
false memories (Loftus, 1997).
In such a situation, the recollective experience may
signal to a rememberer that what is in his or her mind is
a memory of an actual autobiographical experience.
Finding false memories even in superior-memory indi-
viduals suggests that these reconstructive mechanisms
underlying false memory may be fundamental to epi-
sodic remembering (Patihis etal., 2013). Even memories
for stressful and emotional events seem highly vulnerable
to modification by exposure to misinformation (Morgan,
Southwick, Steffian, Hazlett, & Loftus, 2013). Studies also
suggest that false memories can be largely indistinguish-
able from true memories in both emotional content
(Laney & Loftus, 2008) and brain activation (Stark, Okado,
& Loftus, 2010).
During interviewing, asking leading questions, intro-
ducing new and inaccurate information, and pressuring
or expecting the interviewee to report memory details
may facilitate such an inaccurate account (Loftus, 2005).
In legal contexts, interviewing techniques such as guilt-
presumptive and confrontational approaches are thought
to facilitate false confessions and promote inaccurate wit-
ness accounts (e.g., Kassin etal., 2010), which can ulti-
mately lead to procedural injustice and wrongful
imprisonment (Leo & Davis, 2010). Although evidence
supports the existence of other kinds of false confessions
as well (e.g., voluntary false confessions and compliant
false confessions; see Kassin etal., 2012), it has been sug-
gested that so-called internalized false confessions
involve the individuals actually coming to believe that
they have committed a crime (e.g., Gudjonsson &
Lebegue, 1989; Kassin & Kiechel, 1996).
A post hoc analysis of wrongful-conviction cases sug-
gests that a predictable sequence of events typically
occurs prior to and during the internalization of a false
accusation. As detailed by Kassin et al. (2012), part of this
process can involve the presentation of allegedly incon-
trovertible evidence, such as false eyewitness evidence,
by the investigator. The suspect may then be led to pre-
sume that he or she must have repressed or otherwise
forgotten the event. At this point, the individual may
make an admission of possible guilt, using inferential lan-
guage. This admission of possible guilt may be aggres-
sively pursued, and the suspect may begin to incorrectly
create specific details of his or her involvement in the
crime in memory. Although case studies and legal anec-
dotes substantiate this process, no research has examined
the extent to which such false memories for crime are
possible or whether they can be distinguished from real
memories (Kassin & Kiechel, 1996; Laney & Takarangi,
2013). An empirical demonstration of such false memo-
ries for crime and corresponding false confessions would
carry major legal implications.
The Present Study
In the research reported here, we explored whether com-
plete false memories of committing crimes involving
police contact could be generated in a controlled experi-
mental setting. If so, we wanted to explore how prevalent
they would be and how their features would compare
with those of both false memories of other emotional
events and true memories. If supposed corroboration by
caregivers informs young adults that they committed a
crime during adolescence, can they generate such false
memories, or do they reject the notion?
Method
Sample
One hundred twenty-six undergraduate students at a
Canadian university were included in the screening phase
of this study. Of this sample, 70 students met the partici-
pation criteria, and the first 60 eligible students partici-
pated in the interview stage (Phase 2) in exchange for
$50. Participants in Phase 2 were on average 20 years old
(range: 18–31), were predominantly Caucasian (5 were
non-Caucasian), were in their second year of their bach-
elor’s program, were English native speakers (5 were
nonnative speakers), and were predominantly female (43
females).
Procedure
This study used a modified familial-informant false-
narrative paradigm to attempt to convince young adult
participants that they had committed a crime when they
were between the ages of 11 and 14. We followed the
same basic procedure as previous studies (e.g., Lindsay
et al., 2004; Porter et al., 1999; Wade, Garry, Read, &
Lindsay, 2002) and used the same basic interview script
as Porter et al. (1999). The only modification that was
made to the script was that instead of participants being
asked only to recall the true memory during the first
interview session, they were asked about both the true
and the false memory in each of the three interviews.
Participants were told that this was done because the
researchers wanted to get as much information as possi-
ble for both memories. We made this modification to try
to minimize participants’ potential suspicion as to why
the experimenter scheduled them for three sessions.
In the screening phase, 126 undergraduates provided
consent for the researchers to send an extensive memory
by guest on January 15, 2015pss.sagepub.comDownloaded from
Rich False Memories of Crime 3
questionnaire to their primary caregivers. The question-
naires were returned by the caregivers of 91 participants,
of whom 70 were deemed eligible to participate. Eligibility
was based on the caregiver reporting that the participant
had experienced at least one highly emotional event in
the specified time frame, had not experienced any of the
target criminal events, and had never had police contact.
Additionally, the caregiver had to report in some detail at
least one highly emotional event (of any kind). Individuals
were ineligible if their caregivers mentioned any kind of
police contact or reported events that resembled the tar-
get events at any point during adolescence. On the ques-
tionnaire, caregivers were asked whether their child had
experienced any of six negative emotional events, three
of which were criminal (assault, assault with a weapon,
and theft) and three of which were noncriminal (an acci-
dent, an animal attack, and losing a large amount of
money). For each recalled event, caregivers were asked
to write a description of what they could remember,
including the location, people present, time of year, age
of the participant, and how confident they were that the
event had occurred. The questionnaire consent form and
cover letter instructed caregivers to not discuss any of the
events with the participants under any circumstances
until the end of the study. (For more information on the
screening questionnaire, see the Supplemental Material
available online.)
After the questionnaires were returned, eligible par-
ticipants were identified and contacted to schedule the
interview component of the experiment. Data collection
was halted earlier than anticipated because the higher-
than-expected rate of success in inducing false memories
allowed the hypotheses of interest to be tested with a
sample size of 60.
In Phase 2 of the study, participants completed three
interviews, at approximately 1-week intervals. The inter-
views were on average around 40 min long. The same
researcher, who used a scripted interview for all sessions,
conducted all interviews. In the first interview, two of the
events from the questionnaire, one that the participant
had experienced (true event) and one that the participant
had not experienced (false event), were verbally pre-
sented to the participant. The true event was always pre-
sented first in an effort to maximize the researcher’s
credibility.
Participants were randomly assigned to one of two
false-memory conditions. Participants in the criminal
condition were told that they had committed a crime
resulting in police contact; one third of them were told
that they had committed assault, another third that they
had committed assault with a weapon, and the remainder
that they had committed theft. Participants in the non-
criminal condition were told that they had experienced
an emotional event; one third of them were told that they
had had a powerful emotional experience during which
they injured themselves, another third that they had been
attacked by a dog, and the remainder that they had lost a
large sum of money and gotten in a lot of trouble with
their parents. Thirty participants were assigned to each
condition, and 10 were assigned to each specific event.
We used three events of each type in the interest of
increasing generalizability, not with the aim of comparing
the events within a condition with one another.
Participants were asked to explain what happened
during each of the events in turn, after the interviewer
provided some accurate cues from the caregiver ques-
tionnaire, including the city that the participant lived in
and the name of a friend the participant had at the time
of the alleged event (a friend who was supposedly pres-
ent during the event). The interviewer also provided a
number of cues, including the participant’s age at the
time of the event, the season when it took place, and an
indication that the caregiver was involved after the event
occurred; for the true event, these were accurate cues,
and for the false event, they were randomly assigned
inaccurate cues. As expected, participants successfully
provided an account of the true event but were unable to
provide an account of the false event in the first inter-
view. The fact that no participants immediately recalled
the false event helped rule out the possibility that partici-
pants had actually experienced such an event (see Porter
et al., 1999). When participants had difficulty recalling
the false event, the interviewer encouraged them to try to
remember it, and (falsely) told them that most people can
remember these kinds of memories if they try hard
enough. Then, participants were told that the study was
an examination of memory-retrieval methods, and they
were asked to use context reinstatement and guided
imagery to retrieve the memory. They also were told to
practice visualization of the false event each night at
home. These methods have been shown to effectively
generate details that form the foundations of false memo-
ries (e.g., Henkel & Carbuto, 2009).
In all three interviews, as in the study by Porter etal.
(1999), the interviewer kept as close to the script as pos-
sible and tried to behave in a consistent manner with all
participants. To this end, and to try to maximize the
chances of inducing false memories, the interviewer used
a number of verbal and behavioral tactics consistently
and systematically in all interviews and both conditions.
For a basic measure of interview consistency, we con-
ducted a word-count analysis (using Linguistic Inquiry
and Word Count; Pennebaker, Booth, & Francis, 2007)
examining whether the total number of words spoken by
the interviewer in each interview differed between the
two memory conditions (criminal and noncriminal). In a
two-way between-subjects t test, the effect of condition
was not significant, t(58) = 0.72, p = .476, d = 0.19, which
by guest on January 15, 2015pss.sagepub.comDownloaded from
4 Shaw, Porter
indicated that the interviewer used a consistent number
of words in the two conditions.
The strategies that were employed throughout all
interviews in this study were based on literature regard-
ing factors that facilitate the generation of false confes-
sions (e.g., Kassin et al., 2012). The tactics that were
scripted into all three interviews included incontrovert-
ible false evidence (“In the questionnaire, your parents/
caregivers said. . .”), social pressure (“Most people are
able to retrieve lost memories if they try hard enough”),
and suggestive retrieval techniques (including the scripted
guided imagery). Other tactics that were consistently
applied included building rapport with participants (e.g.,
asking “How has your semester been?” when they entered
the lab), using facilitators (e.g., “Good,” nodding, smil-
ing), using pauses and silence to allow participants to
respond (longer pauses seemed to often result in partici-
pants providing additional details to cut the silence), and
using the open-ended prompt “what else?” when probing
for additional memory details. We also used the tactic of
presumed additional knowledge if participants asked
about the accuracy of details. In other words, participants
were told that the interviewer had very detailed informa-
tion about the event from their caregiver but was able
only to vaguely confirm details (e.g., “this sounds like
what your parents described,” “I can’t give you more
details because they have to come from you”). Further,
when participants reported that they could not recall the
false memory, the interviewer seemed disappointed but
sympathetic (while saying the scripted line “That’s ok.
Many people can’t recall certain events at first because
they haven’t thought about them for such a long time.”)
and scribbled down a note on her clipboard. Finally, the
interview office had a bookshelf intentionally filled with
very visible books on memory and memory retrieval to
help increase the interviewer’s credibility as a memory
researcher.
In the second and third interviews, participants again
were asked to provide as many details as possible for
both the true and false events. The nature of participants’
memory for the true and false events was probed each
time an event was recalled by asking follow-up questions
regarding their perspective in the memory (i.e., whether
they recalled it from their own perspective or could see
themselves in the memory), the vividness of the memory,
sensory details included in the memory, and their confi-
dence in the memory. Participants were also asked to rate
the anxiety they experienced at the time of the event. At
the conclusion of the third interview, participants were
paid $50 for their participation and informed that their
second memory was false. Next, before a more extensive
debriefing that explained the false-memory process, par-
ticipants were asked how often they had visualized the
memory at home, how surprised they were that one of
the memories was false (scale from 1 to 7), and how sus-
picious of the interviewer they had been (scale from 1 to
7). They were also asked whether they had believed that
the false event had actually happened. (For more infor-
mation on the interviews, including the full interview
scripts, sample transcript excerpts, and the debriefing
script, see the Supplemental Material. Also, the primary
author, and sole interviewer, is available to provide inter-
view training for researchers who hope to replicate this
study.)
Analysis
One hundred eighty videotaped interviews (60 partici-
pants interviewed three times each) were transcribed,
and the transcriptions of the memories were coded for
details by two independent researchers who had an
excellent interrater Krippendorf’s alpha of .89, as calcu-
lated with ReCal (Freelon, 2010). Details were coded as
“general” if they related to single units of information
pertaining directly or indirectly to the event in question,
as “police-specific” if they were single units of informa-
tion pertaining directly to the police contact for the event
in question, and as “cognitive operations” if they per-
tained to intrinsic perceptions of the event (e.g., emo-
tions, thoughts, and tactile feelings). The number of each
type of detail was tallied across the three interviews for
each participant.
Memory taxonomy. The partial-complete memory
dichotomy used in previous research is a useful catego-
rization for examining the extent to which participants
confabulate details and accept an account as their own
memory. Indeed, we intended to use this dichotomy.
However, the methodology in this study unexpectedly
facilitated the confabulation of very extensive false-
memory details in a high proportion of participants,
such that very few participants would have qualified as
having experienced partial rather than complete false
memories. Therefore, we adopted a different approach
in an attempt to meaningfully differentiate among par-
ticipants’ responses.
Our categories were loosely adopted from the inter-
nalized/compliant dichotomy Kassin et al. (2012) used
for false confessions and the partial/complete dichotomy
from the false-memory literature. Participants who were
classified as having false memories (as defined in the fol-
lowing section) could be said to be most akin to what
have been referred to as participants with “internalized”
false memories in the previous literature. Participants
who provided 10 or more details of the false event but
did not claim at the debriefing that they had believed the
event actually happened were classified as compliant;
they could be seen as having simply acquiesced to the
by guest on January 15, 2015pss.sagepub.comDownloaded from
Rich False Memories of Crime 5
situational demands. On the flip side, participants who
provided fewer than 10 details but claimed at debriefing
that they had believed the event actually happened were
classified as being accepting of the false memory event.
They seemed immune to significant memory generation
despite appearing to believe that the event had happened
to them. The acceptant and compliant groups most
closely correspond to the partial-false-memory category
in previous studies, as participants in that category were
presumed to be merely accepting that the false memory
occurred or to be speculating about it (Lindsay et al.,
2004). Finally, participants who provided fewer than 10
details and asserted at debriefing that they had not
believed the event happened to them were classified as
having no memory of the false event. (See Fig. 1 for a
schematic of the memory taxonomy we used.)
Defining false memories. We used a conservative def-
inition of false memory that was modeled after the defini-
tions used in previous studies. For example, according to
Hyman and Billings’s (1998) definition, participants
exhibited false memories only if their reports included
critical misinformation (of spilling punch) and if the elab-
orations and details in their reports were consistent.
Wade etal. (2002) used a similar definition, according to
which participants had to report details of the critical
event and provide elaborations. The most comprehensive
operational definition of false memory to date was elabo-
rated by Porter etal. (1999). In their study, participants
were considered to have false memories if they reported
remembering the event, agreed with or incorporated
information cues that had been provided, reported details
beyond the four cues provided, did not remember the
incident immediately, and reported during debriefing that
they had not discussed the event outside the lab.
Guided by these definitions, we used the following
objective and participant-subjective criteria to identify
which participants had generated a false memory. First,
the individual had to indicate that he or she remembered
the suggested event during the final interview by report-
ing details about it. Second, the participant’s report by
the third interview had to include the critical pieces of
false information presented by the interviewer (including
at least the location and the name of the friend who was
supposedly there when asked, “Where exactly did the
event occur?” and “Who was present during the event”).
Third, the individual had to provide a basic account of
the false event in response to the instruction “tell me
everything you remember from start to finish,” and this
account had to include more details than those provided
by the experimenter (at least 10 unique details in total).
Fourth, the participant could not have recalled the false
event immediately upon its initial presentation. Fifth, the
participant had to indicate that he or she had not talked
to his or her primary caregivers about any part of the
parental memory questionnaire (i.e., during debriefing,
answered “no” to the question “Did you talk to your par-
ents?”). Sixth, after being informed that the false event
had not actually happened (during debriefing), the par-
ticipant had to answer “yes” to the question “Did you
believe that you had forgotten the event and that it actu-
ally happened?” Overall, this definition of false memory
is a very conservative one.
Results
This section presents the results of the memory inter-
views. For the false events, all results presented are for
only those subjects who were classified as having false
memories. For the true events, results are reported for all
subjects, as well as separately for only those subjects
who were classified as having false memories. The for-
mer statistics provide a general overview of the true
memories, and the latter allow the reader to see within-
subjects differences between true and false memories.
Results for participants categorized as compliant, as
acceptant, or as having no false memories are not dis-
cussed here because the samples were very small.
< 10 Details Reported
Believed Event
Happened Acceptance (n = 6)
Did Not Believe
Event Happened
≥ 10 Details Reported
False Memory (n = 44)
Compliance (n = 6) No False Memory (n = 4)
Fig. 1. Classification taxonomy. Participants were classified into one of four categories accord-
ing to the number of unique details of the false event they reported over the course of the three
memory interviews and according to whether they responded “yes” or “no” at the debriefing to
the question “Did you believe that you had forgotten the event and that it actually happened?”
The diagram shows the number of participants classified in each category after completion of
the last interview.
by guest on January 15, 2015pss.sagepub.comDownloaded from
6 Shaw, Porter
However, for compliant and acceptant participants, the
values for all variables are presented in the Supplemental
Material.
Criminal false memories
Of the participants assigned to the criminal condition, 21
(70%) were classified as having false memories of being
involved in the criminal event resulting in police contact.
Of those 21, 8 provided an account involving assaulting
another person, 6 provided an account involving a theft,
and 7 provided an account involving assaulting another
person with a weapon. Eleven (73.33%) of the partici-
pants who were classified as having false memories of
assault or assault with a weapon reported information
describing the nature of their police contact (e.g., physi-
cal descriptions of the police officers), recalling a mean
of 12.18 police-specific details, 95% confidence interval
(CI) = [8.14, 16.22]. Only 2 participants who were classi-
fied as having false memories of a theft reported details
of the police contact, providing 4.00 police-specific
details on average, 95% CI = [2.05, 5.95].
Participants who were classified as having false mem-
ories of criminal events provided a mean of 71.76 (95%
CI = [55.72, 87.80]) details for these events. More specifi-
cally, on average they provided 59.52 (95% CI = [46.49,
72.56]) general details, 5.48 (95% CI = [4.09, 6.87])
cognitive- operations details, and 6.76 (95% CI = [2.94,
10.59]) police-specific details. Broken down by type of
criminal event, the means for the total number of details
were 75.63 (95% CI = [61.32, 89.93]) for assault, 71.29
(95% CI= [30.79, 111.78]) for assault with a weapon, and
67.17 (95% CI = [37.29, 97.04]) for theft.
Table 1 summarizes the ratings of anxiety at the time
of the criminal event, vividness of memory for the event,
and confidence in memory for the event among partici-
pants who were classified as having false memories of
crime. Table 2 summarizes these participants’ reports of
whether their memories of the criminal event had visual,
auditory, olfactory, and tactile components. During the
debriefing, participants who were classified as having
false memories of committing a crime indicated that they
had tried to recall and visualize the false event at home a
mean of 5.05 times (95% CI = [4.29, 5.82]), reported hav-
ing low suspicion that the interviewer was trying to
manipulate them somehow (M = 2.43, 95% CI = [1.54,
3.32]), and reported having been surprised by the true
nature of the study (M = 4.95, 95% CI = [3.94, 5.97]).
Noncriminal false memories
We included the noncriminal condition so that we could
examine whether there were qualitative or quantitative
differences between false memories of criminal events
and false memories of noncriminal events. Of the partici-
pants assigned to the noncriminal condition, 23 (76.67%)
were classified as having false memories. Of those 23, 7
provided an account involving an animal attack, 8 pro-
vided an account involving an accident resulting in an
injury, and 8 provided an account involving losing a large
sum of money. A chi-square analysis was conducted to
examine whether the criminal and noncriminal condi-
tions differed in the proportion of false memories gener-
ated, and no statistically significant difference was found,
χ2(1, N = 60) = 0.635, p = .425, r = .103. Participants who
were classified as having false memories of noncriminal
emotional events reported a mean of 53.30 (95% CI =
[43.11, 63.49]) details for these events. A two-tailed
independent- samples t test revealed no statistically sig-
nificant difference between the criminal and noncriminal
conditions in the total number of details reported by par-
ticipants who were classified as having false memories,
t(42)= 1.94, p = .06, d = 0.59. On average, participants
who were classified as having noncriminal false memo-
ries provided 49.17 (95% CI = [39.94, 58.4]) general details
and 4.09 (95% CI = [2.43, 5.74]) cognitive-operations
details. Broken down by type of noncriminal event, the
means for the total number of details were 52.00 (95%
Table 1. Participants’ Ratings of Their Anxiety During the Remembered Events, Their Confidence in Their
Memories, and the Vividness of Their Memories
Condition and memory type
Anxiety Confidence Vividness
Mean 95% CI Mean 95% CI Mean 95% CI
Criminal condition (n = 21)
False memory 5.48 [5.09, 5.86] 2.86 [2.37, 3.36] 2.68 [2.19, 3.17]
True memory 4.76 [4.26, 5.27] 5.30 [4.92, 5.67] 4.73 [4.29, 5.17]
Noncriminal condition (n = 23)
False memory 4.52 [3.96, 5.08] 2.76 [2.27, 3.24] 2.59 [2.11, 3.07]
True memory 5.30 [4.92, 5.68] 5.24 [4.80, 5.67] 4.66 [4.18, 5.13]
Note: This table reports data only for those participants who were classified as having false memories. Ratings of anxiety,
confidence, and vividness were made on Likert scales from 1 to 7. CI = confidence interval.
by guest on January 15, 2015pss.sagepub.comDownloaded from
Rich False Memories of Crime 7
CI= [34.27, 69.73]) for the animal attack, 46.61 (95% CI =
[32.90, 60.35]) for the accident resulting in injury, and
61.13 (95% CI = [39.78, 82.47]) for losing a large sum of
money.
Tables 1 and 2 report the results for ratings of anxiety,
vividness, and confidence and for the presence of sen-
sory components for participants who were classified as
having false memories of the noncriminal event. During
the debriefing, these participants indicated that they had
tried to recall and visualize the false event at home a
mean of 4.58 times (95% CI = [3.84, 5.32]), reported hav-
ing low suspicion that the interviewer was trying to
manipulate them somehow (M = 2.70, 95% CI = [1.79,
3.6]), and indicated that they had been surprised by the
true nature of the study (M = 4.65, 95% CI = [3.84, 5.47]).
Repeated t tests with Bonferroni correction were con-
ducted on the number and type of details reported; on
the ratings of confidence, vividness, and anxiety; and on
the presence of sensory components (visual, auditory,
olfactory, tactile, gustatory) in the memory reports. These
tests revealed no statistically significant differences
between criminal and noncriminal false memories.
Additionally, no significant gender differences were found.
True memories
We asked participants to describe true memories so that
we could examine whether a given individual’s recall of
a true memory differed qualitatively or quantitatively
from his or her recall of a false memory. The 60 partici-
pants reported a mean of 91.98 (95% CI = [82.04, 101.92])
details for their true memories. On average, they pro-
vided 85.75 (95% CI = [76.48, 95.02]) general details and
6.23 (95% CI = [4.95, 7.51]) cognitive-operations details.
Note that the veracity of the details for the true events
was confirmed only broadly by the written accounts pro-
vided by the participants’ caregivers. Thus, the specific
details of the true memories remain largely unverified.
On average, participants rated their anxiety at the time
of the true event as 5.0 (95% CI = [4.68, 5.32]), the vivid-
ness of their memories of the true event as 4.67 (95%
CI= [4.35, 4.99]), and their confidence in these memories
as 5.20 (95% CI = [4.91, 5.49]). There was a strong posi-
tive correlation between the total number of details
reported and confidence rating for both false memories
(r = .57, n = 44, p < .001) and true memories (r = .54, n=
60, p < .001), indicating that confidence may be generally
related to the number of details generated in interviews
for memories.
Because within-participant comparisons are consid-
ered the most meaningful way to explore differences
between true and false memories, we focus on those
comparisons here. Table 1 summarizes ratings of anxiety,
vividness, and confidence for the true memories among
participants who formed false memories, and Table 2
summarizes these participants’ ratings of the presence of
sensory components in their memories of the true event.
For participants who were categorized as having false
memories, we conducted a series of two-tailed dependent-
samples t tests with Bonferroni correction (adjusting p to
< .003) to compare true and false memories. Participants
reported significantly more event details for true than for
false memories, t(43) = 5.49, p < .0001, d = 1.66; had
more confidence in true than in false memories, t(43) =
9.87, p < .001, d = 3.01; and reported that their true mem-
ories were more vivid than their false memories, t(43) =
7.99, p < .001, d = 2.44. The t tests also revealed that for
participants classified as having false memories, there
were no significant differences between the true and
false memories in the number of cognitive-operations
details, reported anxiety during the event, or the pres-
ence of any of the sensory components. Finally, a two-
tailed Fisher’s exact test was conducted, and participants
were found to be significantly more likely to report
adopting multiple perspectives (i.e., being able to see
themselves in the memory as well as to see things from
Table 2. Sensory Components of the True and False Memories
Condition and memory type
Visual Auditory Olfactory Tactile
Mean 95% CI Mean 95% CI Mean 95% CI Mean 95% CI
Criminal condition (n = 21)
False memory 0.86 [0.74, 0.99] 0.39 [0.21, 0.56] 0.14 [0.01, 0.26] 0.30 [0.13, 0.46]
True memory 0.95 [0.88, 1.03] 0.48 [0.30, 0.66] 0.23 [0.08, 0.38] 0.48 [0.30, 0.66]
Noncriminal condition (n = 23)
False memory 0.85 [0.72, 0.97] 0.41 [0.23, 0.58] 0.15 [0.02, 0.27] 0.29 [0.13, 0.45]
True memory 0.94 [0.85, 1.02] 0.50 [0.32, 0.68] 0.24 [0.09, 0.39] 0.48 [0.30, 0.65]
Note: This table reports data only for those participants who were classified as having false memories. For a given memory, participants
indicated whether each sensory component was present (1) or absent (0). CI = confidence interval. Gustatory details are omitted from this
table because of very low prevalence rates.
by guest on January 15, 2015pss.sagepub.comDownloaded from
8 Shaw, Porter
their own perspective) in the true than in the false mem-
ory (p = .0079).
Discussion
This study provides evidence that people can come to
visualize and recall detailed false memories of engaging
in criminal behavior. Not only could the young adults in
our sample be led to generate such memories, but their
rate of false recollection was high, and the memories
themselves were richly detailed. Additionally, false mem-
ories for perpetrating crime showed signs that they may
have been generated in a way that is similar to the way
in which false memories for noncriminal emotional
memories are generated. False memories for committing
crime also shared many characteristics with true memo-
ries. Finally, we have proposed a novel taxonomy for
classifying false memories that is more in line with the
current standards for false confessions than previous tax-
onomies are.
Our results align with the literature suggesting that
exposure to misinformation provided by interviewers
can lead to major distortions in memory (Morgan etal.,
2013), and that malleable reconstructive mechanisms
may be fundamental to episodic remembering (Patihis
etal., 2013). A number of current theories, such as fuzzy-
trace theory (Brainerd & Reyna, 2002), propose that a
memory may be retrieved not by accessing a fixed rep-
resentation of a past event, but rather by reactivating
incomplete fragments that can be either distorted or
accurate, and that may have arisen from other real events
(Stark etal., 2010). This implies that false memories may
actually be recalled in a way that is surprisingly similar
to how memories for real events are retrieved.
Consequently, as the results here indicate, true and false
memories have many similar features—including being
highly detailed and multisensory. These results are also
in line with neuroimaging research showing that true
and false memories evoke similar brain activation pat-
terns (Stark etal., 2010), and that even highly emotional
content may not reliably indicate memory accuracy
(Laney & Loftus, 2008). Therefore, it may prove difficult
in the real world to reliably tell the difference between
true and false memories without independent corrobo-
ration (Bernstein & Loftus, 2009).
Our use of a context-reinstatement exercise, in which
participants were to picture what it would have been like
to engage in the false events, may also help explain our
findings. Imagination exercises such as this one have
been repeatedly associated with the generation of false
memories (Pezdek, Blandon-Gitlin, & Gabbay, 2006). The
relevance of imagination for false memories may be par-
tially explained by the source-monitoring framework
(e.g., Johnson, Hashtroudi, & Lindsay, 1993), which refers
to people’s tendency to confuse imagination with reality.
Individuals who are recalling details from a visualization
exercise or experimenter misinformation can forget the
source of their ideas and may think they are recalling
details from a genuine experience. Additionally, explana-
tory coherence has been demonstrated to play a role in
the memory errors that result from suggestive forensic
interviews (Chrobak & Zaragoza, 2013). In particular, it
has been shown that forced-fabrication paradigms, such
as the one used here, lead participants to incorporate
causally relevant misinformation into memory over time
so as to help make sense of events that participants accept
or believe happened but cannot remember (e.g., Chrobak
& Zaragoza, 2008). In other words, imagined memory ele-
ments regarding what something could have been like
can turn into elements of what it would have been like,
which can become elements of what it was like. Although
the interviewer in this study provided only a small, prede-
termined set of misinformation and did not add novel
misinformation across the interviews, it is possible that
participants increasingly tried to make sense of the intro-
duced false events by spinning explanatory frameworks
around what they thought could have happened.
To help make sense of why participants were willing to
accept the kind of erroneous memory cues provided in
this study, we turn to the literature on the effects of vari-
ous kinds of misinformation on memory. Research by
Desjardins and Scoboria (2007) has demonstrated that
rehearsal of self-relevant details like the participant-
specific misinformation provided to participants in this
study can significantly increase false-memory rates. This
effect could be due to superior encoding and retrieval of
information relevant to the self, along with a shift in
beliefs about the plausibility of an event having hap-
pened. This plausibility shift has been supported by the
work of Mazzoni, Loftus, and Kirsch (2001), who sug-
gested that perceived plausibility needs to pass only a
relatively low threshold in order for a personalized manip-
ulation to produce changes in belief that may then be
incorporated into memory. This may help explain why,
despite possible concerns regarding event plausibility
(Pezdek etal., 2006), our participants were as willing to
accept false criminal accounts as they were to accept false
noncriminal accounts (cf. supporting findings by Bays,
2011, and Wade etal., 2002). Incorporating true details
into the false-memory account—especially the caregiver-
provided details regarding the city the participant lived in
and the name of a friend the participant had at the time
of the alleged event—likely constituted a personalized
manipulation in our study. Including these details may
have contributed to increased fluency (Kelley & Jacoby,
1998) and familiarity (e.g., Koriat & Levy-Sadot, 2001) of
event details, giving participants pieces of real memories
that they could use as the foundation upon which to build
by guest on January 15, 2015pss.sagepub.comDownloaded from
Rich False Memories of Crime 9
false memories. Combined with general participant com-
pliance, these processes likely all contributed to partici-
pants reporting detailed false memories in this study.
Some methodological aspects of our study point to
questions for further research that would carry major
implications for understanding the malleability of mem-
ory. As in Lindsay etal. (2004), only one interviewer con-
ducted all of the interviews, to satisfy a requirement of
the research ethics board. This modification may partially
account for the high success rate in implanting false
memories, as the sole interviewer was a senior Ph.D. stu-
dent who was well trained in police interview tactics and
is extraverted—a personality characteristic that has been
demonstrated to be related to high success rates for gen-
erating false memories (Porter, Birt, Yuille, & Lehman,
2000). Future research needs to examine the role of
interviewer- specific characteristics and whether these
can be modified to minimize the risk of inducing false
memories in interviewees.
Future studies also need to examine the importance of
each of the interview tactics used in the present study to
see which are most relevant for understanding the social
processes involved in the formation of false memories.
Such fine-tuned examination was not the goal of the cur-
rent study, but would make a significant contribution to
understanding the effects of each of these tactics and how
well they would map onto actual police behavior. Also,
unlike in a regular police interrogation, there were prob-
ably no perceived negative consequences of confessing to
the criminal or noncriminal event in the present study.
This leads to questions regarding the applicability of this
study to real-world policing situations.
Another important question raised by this study is the
extent to which participants succumbed to lingering
demand characteristics when asked after the interviews
were complete whether they had believed the event actu-
ally happened. Although participants seemed surprised to
learn that the study concerned false memory, and it seems
unlikely that they would have perceived that telling the
truth would lead to adverse consequences, it is very hard
to say with certainty that participants were not deceptive
in answering this question. Anecdotally, the primary
investigator had contact with a number of the participants
through university classes months after the study had fin-
ished, and they routinely brought up their study experi-
ences and proclaimed their astonishment that they could
have been so easily fooled to accept a false memory.
Finally, in our analysis, we did not distinguish between
false memories and false beliefs, and it will be critical for
future research to require participants to rate whether
they “remember” or “know/believe” their reported false
memories (Zaragoza, Belli, & Payment, 2007). It has been
argued that false beliefs are qualitatively different from
false memories, and Tulving’s (1985) remember/know
paradigm has been effectively applied to address this
concern.
Legal systems around the world rely heavily on memory-
related evidence, and the present study can help address
issues of concern related to the accuracy of such accounts.
Our finding that young adults generated rich false memo-
ries of committing criminal acts during adolescence sup-
ports the notion that false confessions and gross
confabulations can take place within interview settings.
The Innocence Project (2012) has shown that about 25%
of false convictions are attributable to faulty confession
evidence, which is often obtained via questionable Reid-
model interrogation tactics (e.g., Kassin et al., 2010),
some of which mirror the false-memory-inducing strate-
gies used in the present study. The kind of research pre-
sented here is essential in the quest to help prevent
memory-related miscarriages of justice.
Author Contributions
J. Shaw and S. Porter developed the study concept and design.
Data collection, analysis, and interpretation were performed by
J. Shaw under the supervision of S. Porter. J. Shaw wrote the
manuscript and revised it in response to peer-review sugges-
tions, and S. Porter provided feedback. Both authors approved
the final version of the manuscript for submission.
Acknowledgments
We would like to acknowledge the extensive transcription and
coding done by Kimberly Crosby, Megan Udala, Skyler Rabbit,
and Lauren Currie. We would also like to thank our colleagues
who have provided tremendous support in helping us revise this
manuscript after pointing out areas in need of improvement.
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with
respect to their authorship or the publication of this article.
Funding
We acknowledge the financial support of the University of
British Columbia through the Lashley and Mary Haggman
Memory Research Award and the financial support of the Social
Sciences and Humanities Research Council of Canada.
Supplemental Material
Additional supporting information can be found at http://pss
.sagepub.com/content/by/supplemental-data
References
Bays, R. (2011). Investigating a model of false memory construc-
tion: Is seeing believing? (Doctoral dissertation). Retrieved
from http://scholarworks.gsu.edu/epse_diss/72
Bernstein, D. M., & Loftus, E. F. (2009). How to tell if a particu-
lar memory is true or false. Perspectives on Psychological
Science, 4, 370–374. doi:10.1111/j.1745-6924.2009.01140.x
by guest on January 15, 2015pss.sagepub.comDownloaded from
10 Shaw, Porter
Brainerd, C. J., & Reyna, V. F. (2002). Fuzzy-trace theory
and false memories. Current Directions in Psychological
Science, 11, 164–169. doi:1.1111/1467-8721.00192
Chrobak, Q. M., & Zaragoza, M. S. (2008). Inventing stories:
Forcing witnesses to fabricate entire fictitious events leads
to freely reported false memories. Psychonomic Bulletin &
Review, 15, 1190–1195. doi:10.3758/PBR.15.6.1190
Chrobak, Q. M., & Zaragoza, M. S. (2013). When forced fabrica-
tions become truth: Causal explanations and false memory
development. Journal of Experimental Psychology: General,
142, 827–844. doi:10.1037/a0030093
Conway, M. A. (2002). Sensory-perceptual episodic memory
and its context: Autobiographical memory. In A. Baddeley,
J. P. Aggleton, & M. A. Conway (Eds.), Episodic memory:
New directions in research (pp. 53–70). New York, NY:
Oxford University Press.
Cutler, B. L. (2012). Conviction of the innocent: Lessons
from psychological research. Washington, DC: American
Psychological Association. doi:1.1037/13085-001
Desjardins, T., & Scoboria, A. (2007). “You and your best friend
Suzy put Slime in Ms. Smollett’s desk”: Producing false
memories with self-relevant details. Psychonomic Bulletin
& Review, 14, 1090–1095.
Freelon, D. G. (2010). ReCal: Intercoder reliability calculation
as a Web service. International Journal of Internet Science,
5, 20–33. Retrieved from http://www.ijis.net/ijis5_1/ijis5_1_
freelon.pdf
Frenda, S. J., Nichols, R. M., & Loftus, E. F. (2011). Current
issues and advances in misinformation research. Current
Directions in Psychological Science, 20, 20–23. doi:1.1177/
0963721410396620
Gudjonsson, G. H., & Lebegue, B. (1989). Psychological and
psychiatric aspects of a coerced-internalized false confes-
sion. Journal of the Forensic Science Society, 29, 261–269.
doi:1.1016/S0015-7368(89)73261-8
Gudjonsson, G. H., & Pearse, J. (2011). Suspect interviews
and false confessions. Current Directions in Psychological
Science, 20, 33–37. doi:1.1177/0963721410396824
Henkel, L. A., & Carbuto, M. (2009). Remembering what we
did: How source misattributions arise from verbalization,
mental imagery, and pictures. In M. R. Kelley (Ed.), Applied
memory (pp. 213–234). Hauppauge, NY: Nova Science.
Hyman, I. E., & Billings, F. J., Jr. (1998). Individual differences
and the creation of false childhood memories. Memory, 6,
1–20. doi:10.1080/741941598
Hyman, I. E., Jr., Husband, T. H., & Billings, F. J. (1995). False
memories of childhood experiences. Applied Cognitive
Psychology, 9, 181–197. doi:1.1002/acp.2350090302
Innocence Project. (2012). False confessions. Retrieved from
www.innocenceproject.org/understand/False-Confessions
.php
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993).
Source monitoring. Psychological Bulletin, 114, 3–28.
doi:10.1037/0033-2909.114.1.3
Kassin, S. M., Bogart, D., & Kerner, J. (2012). Confessions that
corrupt: Evidence from the DNA exoneration case files.
Psychological Science, 23, 41–45.
Kassin, S. M., Drizin, S. A., Grisso, T., Gudjonsson, G. H., Leo,
R. A., & Redlich, A. D. (2010). Police-induced confessions:
Risk factors and recommendations. Law and Human
Behavior, 34, 3–38. doi:1.1007/s10979-009-9188-6
Kassin, S. M., & Kiechel, K. L. (1996). The social psychology
of false confessions: Compliance, internalization, and con-
fabulation. Psychological Science, 7, 125–128. doi:1.1111/j
.1467-928.1996.tb00344.x
Kelley, C. M., & Jacoby, L. L. (1998). Subjective reports and
process dissociation: Fluency, knowing, and feeling.
Acta Psychologica, 98, 127–140. doi:10.1016/S0001-6918
(97)00039-5
Koriat, A., & Levy-Sadot, R. (2001). The combined contri-
butions of the cue-familiarity and accessibility heuris-
tics to feelings of knowing. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 27, 34–53.
doi:10.1037/0278-7393.27.1.34
Laney, C., & Loftus, E. F. (2008). Emotional content of true
and false memories. Memory, 16, 500–516. doi:10.1080/
09658210802065939
Laney, C., & Takarangi, M. K. T. (2013). False memories for
aggressive acts. Acta Psychologica, 143, 227–234. doi:10
.1016/j.actpsy.2013.04.001
Leo, R. A., & Davis, D. (2010). From false confession to wrong-
ful conviction: Seven psychological processes. Journal of
Psychiatry & Law, 38(1–2), 9–56.
Lindsay, D., Hagen, L., Read, J., Wade, K. A., & Garry,
M. (2004). True photographs and false memories.
Psychological Science, 15, 149–154. doi:1.1111/j.0956-7976
.2004.01503002.x
Loftus, E. F. (1997). Creating childhood memories. Applied
Cognitive Psychology, 11, S75–S86. doi:1.1002/(SICI)1099-
0720(199712)11:7<S75::AID-ACP514>3.CO;2-F
Loftus, E. F. (2005). Planting misinformation in the human
mind: A 30-year investigation of the malleability of mem-
ory. Learning & Memory, 12, 361–366. doi:1.1101/lm.94705
Mazzoni, G. A. L., Loftus, E. F., & Kirsch, I. (2001). Changing
beliefs about implausible autobiographical events: A
little plausibility goes a long way. Journal of Experi-
mental Psychology: Applied, 7, 51–59. doi:10.1037/1076-
898X.7.1.51
Morgan, C., Southwick, S., Steffian, G., Hazlett, G. A., & Loftus,
E. F. (2013). Misinformation can influence memory for
recently experienced, highly stressful events. International
Journal of Law and Psychiatry, 36, 11–17. doi:10.1016/
j.ijlp.2012.11.002
Moscovitch, M. (1989). Confabulation and the frontal systems:
Strategic versus associative retrieval in neuropsychological
theories of memory. In H. L. Roediger, III, & F. I. M. Craik
(Eds.), Varieties of memory and consciousness: Essays in hon-
our of Endel Tulving (pp. 133–160). Hillsdale, NJ: Erlbaum.
Nash, R. A., & Wade, K. A. (2008). Innocent but proven guilty:
Eliciting internalized false confessions using doctored-
video evidence. Applied Cognitive Psychology, 23, 624–637.
doi:10.1002/acp.1500
Oeberst, A., & Blank, H. (2012). Undoing suggestive influ-
ence on memory: The reversibility of the eyewitness mis-
information effect. Cognition, 125, 141–159. doi:10.1016/j
.cognition.2012.07.009
Patihis, L., Frenda, S. J., LePort, A. R., Petersen, N., Nichols,
R.M., Stark, C. L., . . . Loftus, E. F. (2013). False memories
by guest on January 15, 2015pss.sagepub.comDownloaded from
Rich False Memories of Crime 11
in highly superior autobiographical memory individuals.
Proceedings of the National Academy of Sciences, USA, 110,
20947–20952. doi:10.1073/pnas.1314373110
Pennebaker, J. W., Booth, R. J., & Francis, M. E. (2007).
Linguistic Inquiry and Word Count: LIWC [Computer soft-
ware]. Austin, TX: LIWC.net.
Pezdek, K., Blandon-Gitlin, I., & Gabbay, P. (2006). Imagination
and memory: Does imagining implausible events lead to
false autobiographical memories? Psychonomic Bulletin &
Review, 13, 764–769. doi:10.3758/BF03193994
Porter, S., Birt, A. R., Yuille, J. C., & Lehman, D. R. (2000).
Negotiating false memories: Interviewer and rememberer
characteristics relate to memory distortion. Psychological
Science, 11, 507–510. doi:10.1111/1467-9280.00297
Porter, S., Yuille, J. C., & Lehman, D. R. (1999). The nature
of real, implanted, and fabricated memories for emo-
tional childhood events: Implications for the recovered
memory debate. Law and Human Behavior, 23, 517–537.
doi:1.1023/A:1022344128649
Russano, M. B., Meissner, C. A., Narchet, F. M., & Kassin, S.M.
(2005). Investigating true and false confessions within a
novel experimental paradigm. Psychological Science, 16,
481–486. doi:10.1111/j.0956-7976.2005.01560.x
Stark, C. L., Okado, Y., & Loftus, E. F. (2010). Imaging the
reconstruction of true and false memories using sensory
reactivation and the misinformation paradigms. Learning &
Memory, 17, 485–488. doi:10.1101/lm.1845710
Strange, D., Sutherland, R., & Garry, M. (2006). Event plausibil-
ity does not determine children’s false memories. Memory,
14, 937–951. doi:1.1080/09658210600896105
Tulving, E. (1985). Memory and consciousness. Canadian
Psychology/Psychologie canadienne, 26, 1–12. doi:10.1037/
h0080017
Wade, K. A., Garry, M., Read, J., & Lindsay, S. (2002). A picture
is worth a thousand lies: Using false photographs to create
false childhood memories. Psychonomic Bulletin & Review,
9, 597–603. doi:10.3758/BF03196318
Zaragoza, M. S., Belli, R. S., & Payment, K. E. (2007). Misinformation
effects and the suggestibility of eyewitness memory. In
M. Garry & H. Hayne (Eds.), Do justice and let the sky fall:
Elizabeth F. Loftus and her contributions to science, law, and
academic freedom (pp. 35–63). Hillsdale, NJ: Erlbaum.
by guest on January 15, 2015pss.sagepub.comDownloaded from
