Content uploaded by Kim van Oorsouw
Author content
All content in this area was uploaded by Kim van Oorsouw on Sep 13, 2021
Content may be subject to copyright.
Alcohol Intoxication Impairs Memory and Increases Suggestibility for a Mock Crime:
A Field Study
K. VAN OORSOUW*, H. MERCKELBACH and T. SMEETS
Forensic Psychology Section, Maastricht University, Maastricht, The Netherlands
Summary: Defendants and witnesses are often intoxicated by alcohol. We investigated whether memory and resistance to sugges-
tive cues are undermined at blood alcohol concentrations (BACs) that were (close to) zero (M
BAC
= 0.01%), moderate
(M
BAC
= 0.06%), or high (M
BAC
= 0.16%). Participants (N = 67) were approached in bars and instructed to commit a mock crime.
Immediately after this, their memory and susceptibility to suggestive questions were tested, and these were re-tested during a sober
follow-up 3–5 days later. Compared with sober participants, moderate and severe intoxication was associated with lower levels of
correctly recalled crime details during both test sessions (i.e. intoxicated and sober). Also, during both sessions, severely intoxi-
cated participants displayed a greater tendency to go along with suggestive cues compared with sober participants. Thus, intox-
ication impaired memory and increased suggestibility during an immediate interview, and both effects persisted when sober again.
Copyright © 2015 John Wiley & Sons, Ltd.
INTRODUCTION
Alcohol interferes with the transference of information from
short- to long-term memory because it disturbs the encoding
and consolidation of new information (Ray & Bates, 2006;
Söderlund, Parker, Schwartz, & Tulving, 2005; Verster,
Van Duin, Volkerts, Schrueder, & Verbaten, 2003; White,
2003). During intoxication, encoding is more superficial
due to a lack of rehearsal and other mnemonic strategies
(Saults, Cowan, Sher, & Moreno, 2007). Even a relatively
modest but sudden rise in blood alcohol concentrations
(BACs) may produce fragmentary blackouts or grayouts
(Perry et al., 2006; Ray & Bates, 2006; Wetherill & Fromme,
2011; White, 2003). This means that parts of the event(s)
that took place during intoxication are not consolidated and
not remembered afterwards. Higher BACs may even lead
to a complete (i.e. en bloc) blackout, that is, a total inability
to recall events at a later point in time (Goodwin, Othmer,
Halikas, & Freeman, 1970; Kalant, 1996; Mintzner &
Griffiths, 2002; Van Oorsouw, Merckelbach, Ravelli,
Nijman, & Mekking-Pompen, 2004). Number of drinks
and also the rate at which they are consumed are important
predictors of grayouts and blackouts. The faster the con-
sumption rate, the more rapid intoxication levels rise (see
also Goodwin, Crane, & Guze, 1969; White, 2003). In the
study by Van Oorsouw et al. (2004), participants with black-
outs reported having consumed about 15 drinks within
4 hours. This corresponds with BACs of up to 0.26%. To
put this further into perspective, in the Netherlands, the max-
imum concentration of blood alcohol at which it is legally
permitted to drive a motor vehicle is 0.02% (approximately
2 drinks within 1 hour); in the United States, this level is
0.08% (approximately 5.5 drinks within 1 hour).
1
Suspects, victims, and witnesses of crimes are often under
the influence of alcohol (Evans, Schreiber Compo, &
Russano, 2009; Haggard-Grann, Hallqvist, Langstrom, &
Moller, 2006). For example, 35% of the offenders who have
been convicted for violent crimes claim alcohol-related am-
nesia for their offence (Cima, Nijman, Merckelbach, Kremer,
& Hollnack, 2004; Kopelman, 1995). Although some of them
may have been simulating their amnesia (e.g. van Oorsouw &
Merckelbach, 2009), many of them are likely to have been
intoxicated during the crime.
Given that alcohol may impair memory, one would expect
police officers to refrain from interrogating an intoxicated
suspect or witness. Nevertheless, Evans et al. (2009) noted
in their survey conducted among police investigators that
22% of them said that they were willing to interrogate an in-
toxicated suspect. Only 7% of the police investigators said
that they would allow the suspect to become sober before
they would start the interrogation (Evans et al., 2009; see
also Sigurdsson & Gudjonsson, 1994; Santtila, Ekholm, &
Niemi, 1998). Likewise, an analysis of police files by
Palmer, Flowe, Takarangi, and Humphries (2013) demon-
strated that police investigators asked intoxicated witnesses
to describe the culprit and to take an identification test just
as often as they asked sober witnesses. For the subgroup of
intoxicated suspects, it was not clear whether they were
actually questioned about the crime while still being intoxi-
cated. However, the majority of the arrested offenders were
taken into custody while under the influence of alcohol,
rendering it likely that they were still intoxicated when
questioned.
A recent police survey in the Netherlands revealed that
50% of the professionals believe that perpetrators, including
those who are intoxicated, are more likely to disclose crucial
information in an immediate interrogation (Van Oorsouw,
Merckelbach, & Willems, 2013). This assumption is not sur-
prising because police investigators are familiar with the
phenomenon that memory fades over time.
Given that police investigators often deal with intoxicated
suspects and witnesses and question them in both intoxicated
and sober states, it is important to study the dose–response
effect of alcohol on memory. Previous studies in this domain
have yielded mixed results. For example, regarding intoxi-
cated eyewitnesses (BAC range: 0.06–0.12%), Hagsand,
Roos af Hjelmsäter, Granhag, Fahlke, and Söderpalm-Gordh
1
For a male person weighing 80 kg.
*Correspondence to: Kim van Oorsouw, Department of Clinical Psychological
Science, Forensic Psychology Section, Maastricht University, PO Box 616,
6200 MD, Maastricht, The Netherlands.
E-mail: K.vanoorsouw@maastrichtuniversity.nl
Copyright © 2015 John Wiley & Sons, Ltd.
Applied Cognitive Psychology,Appl. Cognit. Psychol. 29: 493–501 (2015)
Published online 17 April 2015 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/acp.3129
(2013a) demonstrated that at moderate dosages (mean BAC
0.06%), alcohol impaired the number of details that partici-
pants recalled of a crime video. However, intoxication level
did not affect memory accuracy. Similar findings were re-
ported by Hildebrand Karlén, Roos af Hjelmsäter, Fahlke,
Granhag, and Söderpalm-Gordh (2014), but only for women
(mean BAC 0.08%) and not for men (mean BAC 0.07%). On
the other hand, Schreiber Compo et al. (2012) found that in-
toxicated individuals (with BACs ranging between 0.06%
and 0.08%) who watched a staged theft did not differ from
sober participants in either memory completeness or accu-
racy. In another study, Hagsand and colleagues investigated
the effect of similar dosages on participants’identification
accuracy (i.e. identifying a culprit) in a simultaneous
target-absent or target-present lineup, after they had watched
a crime video (Hagsand, Roos af Hjelmsäter, Granhag,
Fahlke, & Söderpalm-Gordh, 2013b). Participants who had
been intoxicated while watching the movie later performed
at a similar level—which was indicative of overall poor
lineup performance (i.e. 25% correct identifications)—as
their sober counterparts. Similar findings were reported by
Harvey, Kneller, and Campbell (2013) at slightly higher dos-
ages. These researchers employed a slide sequence depicting
a theft. Again, no effect of alcohol was found on subsequent
identification accuracy in a simultaneous target-absent or
target-present lineup. Importantly, in both studies, partici-
pants were sober when tested. Thus, it appears that recogni-
tion memory is relatively unaffected by alcohol intoxication.
An exception to this rule is the study by Dysart, Lindsay,
MacDonald, and Wicke (2002). These authors did find a det-
rimental effect of alcohol on identification accuracy, but only
in target-absent show-ups. However, their study differed in
important respects from the studies cited earlier. First, partic-
ipants were involved in a real-life interaction rather than pas-
sively watching a movie or slide sequence. Second, partici-
pants were still intoxicated (BACs ranging between 0.00%
and 0.20%) when asked to identify the person they had
interacted with. Third, the authors did not use a lineup but
a single picture (i.e. a show-up). A meta-analysis by Steblay,
Dysart, Fulero, and Lindsay (2003) demonstrated that line-
ups and show-ups do not generate different rates of false
identifications. Still, the Dysart et al. finding of a heightened
rate of false identifications in intoxicated witnesses might be
the result of a subtle interaction between suggestive show-
ups and intoxication.
In the older study of Yuille and Tollestrup (1990), sober
and intoxicated individuals were exposed to a mock crime.
Although free recall of the intoxicated participants (BACs
ranging between 0.06% and 0.12%) was impaired, they
did not differ from sober participants in the accuracy with
which they identified the presence or absence of the culprit
in a lineup 1 week later, when sober again. Taken together,
these studies suggest that identification accuracy is only
undermined when participants are intoxicated at the time
of testing (Dysart et al., 2002), and not when sober again
(Yuille & Tollestrup, 1990).
Only two studies have looked at intoxicated offenders’
memories for crime details. Using a staged mock crime,
Read, Yuille, and Tollestrup (1992) investigated intoxicated
participants’memory for what they did (as perpetrators) and
their memory for what they saw (as witnesses) in an imme-
diate and/or delayed (1 week later) memory test. Alcohol
intoxication (with BACs up to 0.10%) undermined perpe-
trators’memory for the mock crime, but not witnesses’
identification accuracy. The memory-undermining effect of
alcohol was not affected by the time of testing (immediate
and/or delayed). Yet, repeated testing did improve memory,
irrespective of intoxication level. In a field study, Van
Oorsouw and Merckelbach (2012) tested memory for a staged
robbery filmed from the perpetrator’s perspective when
participants were sober again. The authors noted that intoxi-
cation (with BACs up to 0.24%) during exposure to the
film fragment impaired completeness and accuracy of
delayed (3–5 days later) recall.
Most studies on alcohol and memory focussed on
witnessed events rather than enacted events. This is an im-
portant limitation: Crimes mostly involve action rather than
passive states, and memory for enacted events is better than
memory for witnessed events (Engelkamp, 1995). The rea-
son for this is that arousal is likely to be higher during an
enacted event as compared with passively watching a crime
video. Raised arousal levels might counteract the memory-
undermining effect of alcohol (Engelkamp, 1995; Read
et al., 1992).
With these considerations in mind, it is ecologically rele-
vant to employ mock crimes when testing the effects of alco-
hol on crime-related memories. In doing so, it is informative
to look not only at memory performance but also at suggest-
ibility. Note that only a few studies have addressed the link
between alcohol intoxication and suggestibility. Germane
to this is the well-documented discrepancy-detection princi-
ple that dictates that people are more prone to misleading
cues when their memory of the original event is poorer
because it becomes more difficult for them to detect discrepan-
cies between truly encoded details and merely suggested
details (Peterson, Rothfleisch, Zelazo, & Pihl, 1990; Schooler
& Loftus, 1986). Because alcohol undermines the encoding of
detailed information, one expects intoxication to lead to prob-
lems in detecting discrepancies between stored information
and information merely suggested during, for example, a po-
lice investigation. This, in turn, would lead to increased sug-
gestibility by rendering the discrimination between falsely
suggested and actual details exceedingly difficult. In line with
this, Nash and Takarangi (2011) found in their survey among
individuals with blackout experiences that they often tend to
rely on less credible resources to fill in the gaps. Consequently,
being exposed to inaccurate information (e.g. in a suggestive
interview, or by co-witnesses or other external sources) could
more easily make people with blackouts believe inand remem-
ber experiences that never occurred (Loftus, 2005; Mazzoni &
Kirsch, 2002; Nash & Takarangi, 2011).
So far, only Santtila, Ekholm, and Niemi (1999) directly
investigated the link between alcohol and suggestibility.
These authors found that intoxicated individuals exhibited
areduced susceptibility to go along with leading questions,
as measured with the Gudjonsson suggestibility scale
(Gudjonsson, 1997). They concluded that intoxicated indi-
viduals may become less vulnerable to suggestive influences
because of the anxiolytic properties of alcohol (Santtila et al.,
1999). The rationale behind this idea is that anxiety—for
494 K. van Oorsouw et al.
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
example, fear of giving the wrong answer during an
interview—is an important antecedent of suggestibility.
However, Santtila et al. relied on an experimental procedure
that deviated strongly from the order of events in real-life
settings. Their participants first experienced an event; next,
they were given alcohol, only to be questioned in a sugges-
tive way about the target event in the final phase of the pro-
cedure. Thus, Santtila et al. did not address the short- and
long-term suggestibility of offenders or witnesses who were
intoxicated while undergoing the critical event. This is an
important omission because, in reality, the perpetrator is of-
ten questioned about events that took place while he or she
was intoxicated. In addition, when alcohol is consumed after
encoding of an event, it may paradoxically enhance memory
for the event through its inhibition of retrograde interference
(i.e. the retrograde enhancement effect; Bruce & Pihl, 1997;
Knowles & Duka, 2005). This effect could explain why
Santilla et al. did not find an increase but rather a decrease
in suggestibility after intoxication. Thus, the Santtila et al.
study does not provide an optimal test of the idea that alco-
hol induced discrepancy-detection problems may contribute
to increased suggestibility in former intoxicated offenders.
The aims of the present study were threefold. First, we
tested whether alcohol would undermine memory for an
event (i.e. mock crime) enacted in a perpetrator role. Second,
we investigated whether intoxication would increase sug-
gestibility in an interrogation situation. And third, we ex-
plored whether participants who were intoxicated during
the critical event remember more correct details when inter-
rogated immediately (while still intoxicated) or when sober
again in a repeated interview about the event.
We hypothesized that compared with low and moderate
levels of intoxication, higher levels (i.e. BACs above
0.11%) at the time of encoding (i.e. during the mock crime)
would be detrimental to the accuracy and completeness of
memory for crime details when participants are tested imme-
diately following the mock crime. Based on the idea that in-
toxication interferes with encoding, this effect was expected
to persist during a follow-up test when participants were so-
ber again. With respect to suggestibility, we hypothesized
that alcohol intoxication would increase suggestibility owing
to difficulties in discrepancy detection that arise when people
have memory deficits due to alcohol.
METHOD
Participants
Potential participants (mixed sample) were approached in
three local bars in Maastricht, the Netherlands. These were
regular bars with a mixed audience (e.g. students and work-
ing class people). Each participant was approached and
tested individually. Participants were approached when it
was evident that they were adults (>18 years), and when
they made a healthy impression. When they appeared to be
heavily intoxicated, they were not approached for participa-
tion. Participants were not further screened in any way. In
total, 100 volunteers agreed to participate, and 67 of them (54
men) completed both the session in the bar and the follow-up
session. Their mean age was 22.2 years (range, 18–36 years;
SD = 2.53). The study was approved by the local standing
Ethical Research Committee. Participants were asked not to
talk to others about the experiment until after the second
session.
Mock crime
The mock crime consisted of an introduction story and a
criminal act that participants had to perform. The 2-minute
introductory crime story was presented to them via head-
phones. This was carried out to eliminate distraction noise
and to ensure that the story was read to each participant in
the same pitch and volume. The introduction was as follows:
You have been working at this bar for a while now.
Usually you get paid on the last Friday of the month.
However, your boss has not paid you for the last two
months. Tonight you have a day off and you are in this
bar with your friends. You know your boss keeps the
daily cash turnover in a briefcase that is in his office,
where the coats of the staff are also stored. You suspect
that the key to the briefcase is inside his coat pocket.
You decide to sneak into your boss’office and steal some
of his money you know he keeps in a briefcase.
2
Next, participants were asked to enter a room that was deco-
rated as an office (e.g. a coat stand with coats, posters on the
wall, a table with flowers, and a briefcase). They were
instructed to search for a key that was in one of the coats
and to use it to open the briefcase. They were then asked
to steal the money from the briefcase. There were also other
objects (e.g. an orange and a picture) in the briefcase. How-
ever, the instructions did not mention any details they were
going to encounter (e.g. ringing of a phone and objects inside
the briefcase). After the instruction had been given, partici-
pants were asked whether they understood the procedure.
As a manipulation check, participants were asked to rate
their emotional involvement and ability to empathize with
the main character using two 100 mm (1 = not emotionally
involved/extremely difficult to empathize, 100 = very emo-
tionally involved/very easy to empathize) visual analogue
scales.
Procedure
Participants were approached in bars between 20.00 h and
04.00 h and were invited to participate in a study on alcohol
and cognition. They signed an informed consent form after
which their BAC was measured, using the Lion Alcometer
SD400. This breath analyser converts the breath alcohol ratio
into blood alcohol ratio. Next, participants were asked sev-
eral questions about their drinking habits and drug use
(Van Oorsouw & Merckelbach, 2012). We also asked them
to provide us with contact information so that we could send
them a debriefing form and their BAC level afterwards. At
that moment, nothing was mentioned about follow-up testing
because we did not want to influence participants in any
way. Next, participants were taken into a quiet room where
they received the instructions for the mock crime. They were
2
A detailed description can be obtained from the first author.
Intoxication, memory, and suggestibility 495
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
asked to identify with the main character of the story that
they were about to hear. The story was administered to them
through headphones. Following this, they were given the in-
structions to commit the actual mock crime as described ear-
lier. After having carried out the mock crime, participants
were taken into a different, adjacent room where their mem-
ory and level of suggestibility were tested (i.e. immediate
test). The interview was conducted by a different person than
the one who approached participants and explained the pro-
cedure. Participants’responses were audiotaped using voice
recorders. There was an inte rval of approximately 10 min be-
tween the invitation to participate in the study and the start of
the interview. When all tests were completed, participants
were thanked and informed that they would be contacted
within the next 3–5 days for a follow-up test. Participants re-
ceived a candy bar for their participation.
Three to five days later, participants were contacted by
phone and asked to complete several tests about the mock
crime. When they agreed, the free and cued recall instruc-
tions and suggestibility items employed during test 1 were
once more administered. Thus, they were instructed to report
as completely as possible about the mock crime and its intro-
ductory story. The response rate was 67%. Prior to the
follow-up test, participants were asked whether they had
been drinking alcohol. To reduce the chance of participants
being intoxicated during the follow-up test, testing took
place at weekdays between 9.00 h and 16.00 h. None of the
participants said they had been drinking alcohol prior to
the second test. Again, responses were audiorecorded. After
completion of the follow-up test, participants were thanked
and provided with information about their level of intoxica-
tion during the first test session. After the study was finished,
all participants were fully debriefed by e-mail.
Memory testing and scoring
Participants’memory for the mock crime was tested both im-
mediately after their return from the office space and at the
follow-up session.
3
Memory testing consisted of a free recall
test and a cued recall test. For the free recall test, participants
were asked to give a detailed description of their motive, the
surroundings, and the objects they had seen, as well as their
own actions. Thus, they were instructed to recall everything
they remembered of both the introductory story and the theft.
The cued recall test consisted of 15 memory questions of the
following type: ‘When did you usually get paid?’and ‘What
objects were located in the briefcase?’The questions
pertained to the details mentioned in the introductory story
as well as details of the mock crime.
An a priori scoring protocol was employed to evaluate
participants’free and cued recall (see for a similar method
Van Oorsouw & Merckelbach, 2012). In total, 27 informa-
tional units were identified in the introductory story and 29
units in the mock crime. Thus, the maximum number of de-
tails was 56. Examples of informational units that were
scored in the story are: ‘my boss did not pay me’(1 point);
‘for two months’(1 point). As for the mock crime, points
were accredited when participants correctly described ac-
tions, for example, that they took the key out of the coat
pocket (1 point) and that they opened the briefcase (1 point).
Extra points were given when participants correctly specified
the amount of stolen money (e.g. 50 euros), mentioned the
colour of the wallet, and so on. One point was accredited
to each unit that was accurately recalled. The number of cor-
rect units represented the total free recall score.
The 15 cued recall questions covered a total of 21 critical
information units. For each correct answer to a question, par-
ticipants received 1 point, except for the question about the
objects in the briefcase. Here, participants could earn 7 points
if they mentioned all objects correctly, adding to a total free
recall score of 21. Most questions pertained to the enacted
event and the objects involved [e.g. the briefcase (17 points)],
and only four questions were about the introductory story and
referred to the motive for stealing the money (4 points).
Apart from the numbers of correctly free- and cued-
recalled details, we evaluated two types of errors. Commis-
sion errors were defined as the introduction of an entirely
new and false piece of information (e.g. ‘the briefcase
contained a cell phone’). Distortion errors were defined as
pieces of information that were essentially correct, but
misrepresented during the cued recall test. An example of a
distortion would be ‘I took the key from the blue coat’when
in fact the coat was grey.
Free and cued recall tests were scored by the first author and
an independent second rater. Both were blind as to partici-
pants’intoxication levels. Pearson correlations between the
two raters for the number of correctly recalled units during free
(range: 0–56) and cued (range: 0 –21) recall were 0.98 and
0.91, respectively. For commission and distortion errors, these
correlations were 0.85 and 0.76, respectively (all ps<0.01).
Suggestibility measure
Suggestibility was assessed using 15 misleading questions
that were intermixed with the 15 cued recall memory ques-
tions. The misleading questions pertained to details that
had not been present in the story or during the mock crime.
For instance, we asked participants ‘Did the wallet contain
50 or 100 euros?’when in fact it contained 70 euros. We also
gave participants questions containing two false alternatives
such as ‘Was there an apple or a banana in the briefcase?’
when in fact the briefcase contained an orange. When partic-
ipants yielded to a misleading question, this was scored with
1 point. In line with the procedure described by Gudjonsson
(1997), participants received negative feedback after com-
pletion of the 30 cued recall and misleading questions. That
is, they were told that they made quite a few mistakes and
were asked to answer the questions for a second time. We
were interested in whether negative feedback would lead to
changes (shifting) in participants’answers to the second
series of (misleading) questions. This way, we were able to
calculate three suggestibility parameters: (i) the tendency to
go along with misleading questions immediately (yield 1);
(ii) the tendency to accept misleading cues after negative
feedback (yield 2); and (iii) the tendency to change an an-
swer after negative feedback (shift).
3
Participants were first contacted after 3 days but were not always available
to undergo the follow-up test. In those cases, the follow-up test was sched-
uled for the next day or 2 days later. The 3- to 5-day interval was also used in
previous studies (e.g., Van Oorsouw & Merckelbach, 2012).
496 K. van Oorsouw et al.
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
RESULTS
Statistical analyses
To examine how different levels of intoxication affected
memory and suggestibility, both recall data and suggestibil-
ity scores were subjected to 3 (groups: nearly sober vs. mod-
erately intoxicated vs. severely intoxicated) × 2 (sessions:
immediate vs. follow-up) analyses of variance (ANOVAs),
with the last factor being a repeated measure. Follow-up
pairwise comparisons between groups were carried out. Re-
gression analyses were conducted to explore whether BAC
and/or memory affected suggestibility.
Blood alcohol concentrations and memory performance
Blood alcohol concentrations ranged from 0.00% to 0.26%,
with an average BAC of 0.09% (SD = 0.07).
Following the approach of Van Oorsouw and
Merckelbach (2012), participants were subdivided in three
groups: sober (BACs <0.02%),
4
moderately intoxicated
(BACs between 0.02% and 0.11%), and severely intoxicated
(BACs >0.11%). The sober group consisted of 14 partici-
pants (M
BAC
= 0.01%, SD = 0.01), the moderately intoxicated
group had 27 participants (M
BAC
= 0.06%, SD = 0.02), and
the severely intoxicated group had 26 participants
(M
BAC
= 0.16%, SD =0.04). As was to be expected, the
groups differed in the number of alcoholic beverages partic-
ipants reported to have drank on the night of testing [F(2,
64) = 36.75, p<0.01, η2
p= 0.53]. Sober, moderately intoxi-
cated, and severely intoxicated participants reported to have
consumed respectively 4 (SD = 2.7), 9 (SD = 3.7), and 17
(SD = 6.3) alcoholic beverages.
5
Groups did not differ in
self-reported number of drinking nights per week [F(2, 64) =
1.00, p= 0.37], or the number of drinks on a typical drinking
occasion [F(2, 64) = 1.28, p= 0.28].
On the whole, participants reported they were able to iden-
tify with the main character of the mock crime (M= 71.2,
SD = 18.6) and felt emotionally involved in the mock crime
(M= 66.6, SD = 20.3). There were no group differences in
the ability to empathize or in emotional involvement in role
playing the thief [both Fs<1.0].
Table 1 shows memory performance data (correct free re-
call, cued recall, and errors for the story and actions, respec-
tively) of the three groups during the two test sessions for
free recall and cued recall.
A repeated-measure ANOVA performed on free recall of
the introductory story, with group (sober, moderately, and se-
verely intoxicated) as a between-subject factor and session as
a repeated measure, revealed no interaction between groups
and session [F(2, 64) <1.0, η2
p= 0.04], or an effect of session
[F(1, 64) = 2.43, p= 0.12, η2
p= 0.04]. Yet, a significant main
effect of group [F(2, 64) = 24.00, p<0.01, η2
p= 0.45] did
emerge. Bonferroni corrected post hoc t-tests revealed that
during both sessions, the three groups differed significantly
in the number of correctly recalled story details, with intoxi-
cation resulting in significant lower levels of correct recall
(Table 1). No significant main effects or an interaction effect
was found for number of errors made in recalling the intro-
ductory story (all Fs<1.40, all ps>0.05).
For free recall of the enacted event, no interaction [F(2,
58) = 1.11, p= 0.33, η2
p= 0.03]
6
was found. But main effects
emerged for session [F(1, 58) = 20.93, p<0.01, η2
p= 0.26]
and group [F(2, 58) = 7.21, p<0.01, η2
p= 0.20]. Sober partic-
ipants recalled significantly more correct details as compared
with both intoxication groups during both sessions (both
ts>2.23, both ps<0.05), and all groups recalled
4
Although participants in this group had consumed alcohol, their BAC was
below the Dutch legal driving limit (0.02%) according to which they would
be legally perceived as sober. Therefore, this group will be referred to as the
sober group.
5
An alcoholic beverage refers to a standard glass of beer, wine, spirits, or a
mixed drink.
Table 1. Number of correctly recalled details and errors for the three groups on the free and cued recall test, during session 1 (T1) and the
follow-up test (T2)
Sober (n= 14) Moderately intoxicated (n= 27) Severely intoxicated (n= 26)
T1 T2 T1 T2 T1 T2
Free recall story
Correct total 14.15 (4.93)
a,b
13.30 (4.31)
a,b
11.18 (3.51)
c
10.07 (3.51)
c
6.23 (3.59) 5.90 (3.20)
Errors (com/dis) 1.76 (1.16) 1.92 (1.38) 1.88 (1.36) 1.70 (1.26) 2.04 (2.26) 1.23 (1.04)
Free recall actions
Correct total 8.38 (5.31)
a,b
12.93 (5.57)
a,b,
* 5.70 (3.97) 7.88 (4.21)* 5.04 (2.72) 7.42 (4.35)*
Errors (com/dis) 0.31 (0.48) 0.76 (0.83) 0.18 (0.48) 0.92 (1.46) 0.47 (0.81) 0.61 (0.92)
Cued recall story
Correct total 2.71 (0.91) 2.50 (1.16) 2.44 (1.25) 2.29 (1.20) 2.03 (1.39) 1.80 (1.47)
Errors (com/dis) 1.14 (0.94) 1.35 (1.00) 1.18 (1.37) 1.14 (1.09) 1.69 (1.37) 1.65 (1.44)
Cued recall actions
Correct total 11.71 (1.47)
b
12.35 (1.90)
b
10.59 (2.37)
c
10.51 (1.98) 8.80 (2.36) 10.03 (3.05)*
Errors (com/dis) 0.42 (0.85) 0.21 (0.80) 0.62 (1.00) 0.05 (1.06) 1.23 (1.72) 1.11 (1.27)
Totals and errors (commission and distortion errors; com/dis) are displayed separately for the introductory story (recall story) and enacted event (recall actions).
a
p<0.05 between sober and moderately intoxicated groups.
b
p<0.05 between sober and severely intoxicated groups.
c
p<0.05 between moderately and severely intoxicated groups.
*p<0.05 within this group between sessions.
6
For six participants, data were partly missing on either the first or the sec-
ond session.
Intoxication, memory, and suggestibility 497
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
significantly more correct details of the enacted event during
the second test session (all ts>2,22, all ps<0.05; Table 1).
Regarding errors made in recalling the enacted event, only a
significant effect of session was found [F(2, 58) = 7.64,
p<0.01, η2
p= 0.12]. Follow-up tests indicated that this effect
was carried entirely by the moderately intoxicated partici-
pants, who made more errors in recalling the enacted event
when they were sober as compared with when they were in-
toxicated [t(26) = 2.92, p<0.01].
As for cued recall of the introductory story, only a significant
main effect of session was found [F(1, 64) = 5.04,
p<0.05, η2
p= 0.07]. All groups recalled fewer correct
details of the introductory story during the second (sober)
test. An ANOVA on the number of errors that were made
during cued recall of the introductory story yielded no
effects (all Fs<1.30).
Regarding cued recall of the enacted event, significant
main effects for session [F(1, 64) = 7.56, p<0.01, η2
p
= 0.11] and group [F(2, 64) = 6.74, p<0.01, η2
p= 0.17] and
a significant interaction between group and session was
found [F(2, 64) = 5.04, p<0.01, η2
p= 0.11]. To break down
this interaction, we carried out univariate follow-up
ANOVAs for both sessions separately, with group as a
between-subject factor. Significant between group differ-
ences were found in correct cued recall scores obtained dur-
ing session 1 [F(2, 64) = 9.10, p<0.01, η2
p= 0.22] and ses-
sion 2 [F(2, 64) = 4.23, p<0.05, η2
p= 0.12]. During the first
test session, severely intoxicated participants recalled signif-
icantly fewer correct details of the enacted event relative to
the other two groups [both ts>2.96, both ps<0.05]. This
effect persisted during the second session [t(64) = 2.85,
p<0.05], but only so when severely intoxicated participants
were compared with sober participants. Also, participants in
the severe intoxication group recalled significantly more cor-
rect details during the second (sober) than during the first test
[t(25) = 3.12, p<0.01]. The group effect in the number of
errors made in recalling the enacted event was close to
reaching significance [F(2, 64) = 2.98, p= 0.058, η2
p= 0.08].
That is, compared with sober participants, participants in
the severe intoxication group had a tendency to make more
errors during both sessions.
Suggestibility
Yield and shift scores of the three groups are shown in
Table 2. For yield scores, no interaction effect [F(2, 64) <
1.0], or an effect of session emerged [F(1, 64) <1.0]. Yet,
there was a significant main effect of group [F(2, 64) = 4.49,
p<0.01, η2
p= 0.12]. During both sessions, severely intoxi-
cated participants displayed a stronger tendency to go along
with misleading questions compared with sober participants
[t(39) = 2.94, p= 0.01]. No other group differences in yield
scores were apparent [ts<1.78, ps>0.24].
Shift scores revealed no interaction between intoxication
level and session [F(2, 64) <1.0]. Neither was there a signif-
icant group effect for the tendency to shift answers after
negative feedback [F(2, 64) <1.0]. However, a significant
main effect was found for sessions [F(1, 64) = 19.50,
p<0.01, η2
p= 0.24]. Exploratory analyses showed that shift
scores decreased significantly over sessions for both intoxi-
cation groups [both ts>2.95, both ps<0.01], but not for so-
ber participants [t= 1.53, p= 0.15].
Regression analyses
For both free and cued recall tests, we calculated Pearson
correlations between BACs, free and cued recall for the in-
troductory story and enacted event, and suggestibility (yield-
ing and shifting). The correlational patterns for both sessions
are given in the Supporting information.
There was a negative correlation between BAC and mem-
ory performance ranging from r=0.32, p<0.05 to
r=0.60, p<0.01 during both sessions. Similarly, signifi-
cant negative correlations existed between suggestibility
and several, but not all, memory output measures (Table 1
Supporting information). Yet, BAC only correlated nega-
tively with yielding to suggestive questions (r= 0.32,
p<0.01 for sessions 1 and 2).
Using regression analyses, we investigated whether mem-
ory performance mediated the relationship between BAC
level and yielding. Such pattern would be in line with the no-
tion that failure to detect discrepancies between misleading
information and the (poorly encoded) actual event underlies
increased suggestibility in intoxicated individuals. A linear
regression analysis was conducted with yield scores as de-
pendent variable and BAC and proportions correct free recall
and cued recall elements for story and actions as independent
variables. Backwards elimination with a criterion of p<0.05
resulted in a model in which free recall of story details was
predictive of yield scores during the first test session.
This model accounted for 22% of the variance. A similar
analysis was carried out for the follow-up test session,
now including BAC and all memory performance vari-
ables for the first and second test sessions. Yield scores
during the second test were predicted by free recall of
story and action details at session 1, accounting for 28%
of the variance (Table 3).
Table 2. Yield and shift scores for the three groups during session 1 (T1) and the follow-up test (T2)
Sober (n= 14) Moderately intoxicated (n= 27) Severely intoxicated (n= 26)
T1 T2 T1 T2 T1 T2
Yield 3.00 (3.01)
a
2.80 (2.96)
a
3.93 (2.67) 4.56 (2.72) 5.42 (2.85) 5.73 (3.31)
Shift 3.86 (4.52) 2.64 (3.00) 3.41 (2.81) 1.89 (2.19)* 4.54 (4.31) 2.29 (3.26)*
a
p<0.05 between sober and severely intoxicated groups.
*p<0.05 within this group between sessions.
498 K. van Oorsouw et al.
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
DISCUSSION
A quick glance at the extant literature on alcohol intoxication
and memory in the domain of legal psychology would lead
one to believe that the memory-undermining effects of alco-
hol on, for example, identifications are limited (e.g. Harvey
et al., 2013). We examined how alcohol affects free and cued
recall at immediate and follow-up tests. Based on the well-
established observation that alcohol intoxication impairs
consolidation (Ray & Bates, 2006; Söderlund et al., 2005),
we expected memory of the mock crime details to be im-
paired, even at the follow-up test when participants were so-
ber again, and particularly so in participants who had
previously been highly intoxicated. By and large, this was
borne out by our data. To begin with, and in line with previ-
ous findings (e.g. Read et al., 1992; van Oorsouw &
Merckelbach, 2012), alcohol intoxication was associated
with lower levels of correct recall, although this effect was
more straightforward for free than for cued recall. An excep-
tion to this pattern was the error rate for action details in the
cued recall that was higher for intoxicated than for sober
people.
Second, the memory-undermining effect increased with
higher levels of intoxication. Thus, in accordance with previous
research (e.g. Perry et al., 2006), we found a dose–response
relationship. Third, the memory-undermining effect emerged
when participants were tested immediately (i.e. when in the
same state as while encoding) and persisted when sober
again. Last, higher levels of intoxication were associated
with a stronger tendency to go along with misleading ques-
tions (i.e. yielding) during both test sessions. This effect was
largely mediated by less complete recall, a pattern that fits
nicely with the idea of discrepancy detection (e.g. Peterson
et al., 1990; Schooler & Loftus, 1986) and the work of
Nash and Takarangi (2011).
As was the case in previous studies (e.g. Van Oorsouw &
Merckelbach, 2012), intoxicated participants reported fewer
correct details about the mock crime as compared with sober
participants at both test sessions. Importantly, the
undermining effect of intoxication was also evident when
we looked specifically at memory for the enacted event
(i.e. stealing the money). Apparently, and in contrast to the
idea that memory traces for enacted events are more robust
against intoxication because they are better consolidated
(Engelkamp, 1995), alcohol impairs consolidation of mem-
ory traces regardless of whether or not the event is enacted.
Furthermore, the lost material cannot be readily retrieved
either when intoxicated or when sober again. We found free
recall to be already impaired at moderate levels of intoxica-
tion, but the effects were more straightforward when
intoxication exceeded a BAC of 0.10%. That intoxication
impaired free recall more than cued recall is in line with pre-
vious studies that found recognition memory—for example,
identification in line-ups—to be relative robust against at
least moderate levels of intoxication (Hagsand et al.,
2013b; Harvey et al., 2013).
One could argue that our findings are of little practical
relevance because they relied on an artificial mock crime
procedure. For example, one could question whether our
participants performed to the best of their abilities. However,
we found that participants were able to identify with the
main character.
7
Furthermore, they had no motive to
underperform, nor do we have reasons to assume that their
memory performance was shaped by demand characteristics.
After all, for a naïve participant, it is difficult to anticipate
what a discrepancy-detection-like pattern of performance
looks like. Thus, we believe that our findings bear relevance
to the practical context of intoxicated defendants and wit-
nesses. Is it better to interrogate such a defendant or witness
immediately or wait until he or she is sober again? Police of-
ficers regularly interview a suspect or witness while he or she
is still intoxicated, and also hold the belief that more critical
information is disclosed during an immediate interrogation
(Evans et al., 2009; van Oorsouw et al., 2013). The present
findings demonstrate that during the second (sober) test, all
participants were better at a free recall test of the enacted part
of the mock crime as compared with the first test. In addition,
only participants who had initially been severely intoxicated
demonstrated an improvement in cued recall memory during
the sober test. Given this significant increase in correct re-
call, police officers are well advised to have the suspect or
witness sober up before questioning.
One explanation for the increase in recall of action details
over sessions might be reminiscence. Reminiscence refers to
the phenomenon that repeated testing elicits memory details
that were inaccessible during a previous test (Kelley &
Nairne, 2003; Odinot, Wolters, & Van Giezen, 2013; Otani,
Von Glahn, Goenert, Migita, & Widner, 2009). One reason
why we obtained a reminiscence pattern for free recall of
action elements in all participants might be that during the
second session, testing conditions were more favourable
(i.e. quiet environment and testing during the day) as com-
pared with the first session (noisy bar and testing late at
night). However, the reminiscence pattern was only apparent
for free recall and action elements. Most importantly, ini-
tially intoxicated groups recalled overall fewer items on both
sessions compared with sober participants.
The present findings lend support to the idea that poor
memory makes it difficult to detect discrepancies between
misleading information and what really happened. This ex-
plains why high levels of intoxication (i.e. BAC above
0.11%) increased the tendency to go along with misleading
information (i.e. yielding). In comparison with sober partici-
pants, severely intoxicated participants were more likely to
7
Although one could argue, of course, that these self-reports were less cred-
ible owing to intoxication.
Table 3. Summary of backwards linear regressions, with yield for
sessions 1 and 2 as dependent variable and BAC, FRStory, FRAct,
cued recall story, and cued recall actions for sessions 1 and 2 as in-
dependent variables
BSEβtp
Yield 1
FRStory1 0.29 0.06 0.48 4.39 0.000
Yield 2
FRStory1 0.25 0.07 0.39 3.36 0.001
FRAct1 0.22 0.09 0.28 2.46 0.017
BAC, blood alcohol concentration; FRStory, free recall story; FRAct, free
recall actions.
Intoxication, memory, and suggestibility 499
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
opt for one of two false alternatives, or to endorse a sug-
gested event. Severely intoxicated participants displayed this
tendency both when questioned immediately as well as when
sober again. According to the principle of discrepancy detec-
tion, poor memory for an event makes it difficult for individ-
uals to detect discrepancies between what actually happened
and what is only suggested (Peterson et al., 1990; Schooler &
Loftus, 1986). We believe that this principle helps to under-
stand the pattern found in the current study: Lacking specific
and detailed memories of an event due to impaired encoding
makes intoxicated participants more prone to go along with
misleading questions. Because raised suggestibility levels
were only evident for severe and not moderate intoxication,
one could speculate that suggestibility increases merely when
correct recall is compromised to a considerable extent.
We found that it was the tendency to go along with
leading questions rather than the tendency to change answers
after negative feedback (‘shift’) that was affected by higher
levels of intoxication during encoding. This is in accordance
with the Santtila et al. (1999) study, which also noted that
intoxication was not related to an increased propensity to
change answers after receiving negative feedback. Appar-
ently, yielding and shifting are two distinct types of suggest-
ibility that are not affected by intoxication in a similar way.
We would argue that the memory-undermining effect of
intoxication and its inference with discrepancy detection pro-
motes yield suggestibility (i.e. acceptance of misinforma-
tion), while the anxiolytic properties of alcohol work against
shift suggestibility (i.e. acceptance of misinformation under
the pressure of negative feedback).
One limitation of our study was that the second test
session was conducted over the telephone. This may have
reduced participants’anxiety in comparison with the first
session. The decrease in shift scores over sessions seems to
support this interpretation. Unfortunately, we did not ask
participants how anxious they felt. The links between intox-
ication, anxiety, and suggestibility clearly warrants further
research. What can be said with some confidence on the
basis of the current findings, however, is that poor memory
mediates the increased tendency to go along with leading
questions, especially when participants are immediately
questioned in a misleading way.
Another limitation of our study is the repeated testing de-
sign that we employed. One could speculate that repetition of
the same memory test within a timeframe of a few days in
sober–sober versus intoxicated–sober participants contrib-
uted to artificial group differences (e.g. sober participants
who feel familiar with the test vs. previously intoxicated par-
ticipants who feel less familiar with it). As mentioned earlier,
reminiscence in free recall of action details was demon-
strated for all groups, indicating some beneficial effects of
repeated testing in all participants. Yet, only previously se-
verely intoxicated participants exhibited an increase over
sessions in cued recall, perhaps reflecting a regression to
the mean. Nevertheless, overall, previously intoxicated par-
ticipants performed significantly worse than control partici-
pants. Our point is that these findings cannot be explained
solely in terms of practice effects. More generally, some
authors have argued that we should not exaggerate the effects
of repeated testing (e.g. Greiffenstein, 2009). Still, a design
including delayed-test only groups would, of course, be more
ideal and future studies may want to include such groups in
order to exclude practice effects.
Another limitation that is typical for this sort of field study
is that it lacks control over potential confounders such as
health factors and the amount of alcohol participants con-
sume after the experimenters leave the scene. However, such
confounders will introduce noise, and it is likely that they lead
to an underestimation of intoxication-memory effects. Another
factor that could have obscured the results is fatigue. That is,
some participants were tested early in the evening and others
in the middle of the night. But again, time of testing will have
introduced error noise and therefore will have led to underesti-
mation of effects.
8
The same is true for the relatively short
interval between mock crime and first interview. Of course, a
police interrogation is unlikely to be conducted some 10 minutes
after the crime took place. But with a longer time interval—say
30–60 minutes—and keeping the slow breakdown of alcohol in
mind (0.02% per hour), it is plausible that stronger memory-
undermining effects of intoxication will occur.
Despite these limitations, our results indicate that it could
be beneficial to wait with interrogating an intoxicated sus-
pect or witness until he or she has sobered up. Evans and col-
leagues (2009) found that intoxicated suspects are more
likely to waive their Miranda rights and incriminate them-
selves. These are alarming observations considering that
we found intoxication to be associated with poorer memory
and heightened suggestibility.
Criminal courts frequently face complications when defen-
dants or witnesses have been intoxicated. In such cases, expert
witnesses are often appointed to examine, for example, claims
of amnesia. Given the limited number of studies in this do-
main, we need more research on the links between alcohol,
memory, and suggestibility so as to be better able to inform tri-
ers of fact in intoxication cases. The present study is one stepin
that direction. It clearly shows that intoxication interferes with
memory completeness in a dose-dependent manner.
REFERENCES
Bruce, K. R., & Pihl, R. O. (1997). Forget ‘drinking to forget’: Enhanced
consolidation of emotionally charged memory by alcohol. Experimental
Clinical Psychopharmacology,5. doi: 10.1037//1064-1297.5.3.242
Cima, M., Nijman, H., Merckelbach, H., Kremer, K., & Hollnack, S.
(2004). Claims of crime-related amnesia in forensic patients. Interna-
tional Journal of Law and Psychiatry,27, 215–221. doi: 10.1016/j.
ijlp.2004.03.007
Dysart, J. E., Lindsay, R. C. L., MacDonald, T. K., & Wicke, C. (2002). The
intoxicated witness: Effects of alcohol on identification accuracy from
showups. Journal of Applied Psychology,87, 170–175. doi: 10.1037//
0021-9010.87.1.170
Engelkamp, J. (1995). Visual imagery and enactment of actions in memory.
British Journal of Psychology,86, 227–240. doi: 10.1111/j.2044-
8295.1995.tb02558.x
Evans, J. R., Schreiber Compo, N., & Russano, M. B. (2009). Intoxicated
witnesses and suspects: Procedures and prevalence according to law en-
forcement. Psychology, Public Policy, and Law,15, 194–221. doi:
10.1037/a0016837
8
There was no interaction between time of testing (before or after midnight)
and group, or a main effect of time of testing on any of the dependent vari-
ables (all Fs<3.4, all ps>0.05).
500 K. van Oorsouw et al.
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
Goodwin, D. W., Crane, J. B., & Guze, S. B. (1969). Phenomenological
aspects of the alcoholic blackout. British Journal of Psychiatry,115,
1033–1038. doi: 10.1192/bjp.115.526.1033
Goodwin, D. W., Othmer, E., Halikas, J. A., & Freeman, F. (1970). Loss of
short-term memory as a predictor of the alcoholic black-out. Nature,227,
201–202. doi: 10.1038/227201a0
Greiffenstein, M. (2009). Clinical myths of forensic neuropsychology. The Clin-
ical Neuropsychologist,23,286–296. doi: 10.1080/13854040802104873
Gudjonsson, G. H. (1997). The Gudjonsson suggestibility scales manual.
Hove: Psychology Press. doi: 10.4135/9781412959537.n131
Haggard-Grann, U., Hallqvist, J., Langstrom, N., & Moller, J. (2006). The role
of alcohol and drugs in triggering criminal violence: A case cross-over
study. Addiction,101,100–108. doi: 10.1111/j.1360-0443.2005.01293.x
Hagsand, A., Roos af Hjelmsäter, E., Granhag, P. A., Fahlke, C., & Söderpalm-
Gordh, A. (2013a). Bottled memories: On how alcohol affects eyewitness recall.
Scandinavian Journal of Psychology,54,188–195. doi: 10.1111/sjop.12035
Hagsand, A., Roos af Hjelmsäter, E., Granhag, P. A., Fahlke, C., &
Söderpalm-Gordh, A. (2013b). Do sober witnesses outperform alcohol
intoxicated eyewitnesses in a lineup? The European Journal of Psychol-
ogy Applied to Legal Context,5,23–47. Retrieved from http://scielo.
isciii.es/pdf/ejpalc/v5n1/original2.pdf
Harvey, A. J., Kneller, W., & Campbell, A. C. (2013). The elusive effects of
alcohol intoxication on visual attention and eyewitness memory. Applied
Cognitive Psychology. doi: 10.1111/j.1360-0443.2005.01293.x
Hildebrand Karlén, M., Roos af Hjelmsäter, E., Fahlke, C., Granhag, P. A.,
& Söderpalm-Gordh, A. (2014). Alcohol-intoxicated eyewitnesses’
memory of intimate partner violence. Psychology Crime and Law,21,
156–171. doi: 10.1080/1068316x.2014.951644
Kalant, H. (1996). Intoxicated automatism: Legal concept vs. scientific evi-
dence. Contemporary Drug Problems,23, 631–648. Retrieved from
http://heinonline.org/HOL/Page?handle=hein.journals/
condp23&div=41&g_sent=1&collection=journals#665
Kelley, M. R., & Nairne, J. (2003). Remembering the forgotten? Reminis-
cence, hypermnesia and memory for order. The Quarterly Journal of Ex-
perimental Psychology,56, 577–599. doi: 10.1080/02724980244000530
Knowles, S. K., & Duka, T. (2005). Does alcohol affect memory for emo-
tional and non-emotional experiences in different ways? Behavioral
Pharmacology,15, 111–121. doi: 10.1097/00008877-200403000-00003
Kopelman, M. D. (1995). The assessment of psychogenic amnesia. In A. D.
Baddeley, B. A. Wilson, & F. N. Watts (Eds.), Handbook of memory dis-
orders (pp. 427–448). New York: Wiley.
Loftus, E. F. (2005). Planting misinformation in the human mind: A 30-year
investigation into the malleability of memory. Learning and Memory,12,
361–366. doi: 10.1101/lm.94705
Mazzoni, G., & Kirsch, I. (2002). Autobiographical memories and beliefs: A
preliminary metacognitive model. In T. J. Perfect & B. L. Schwarz (Eds.),
Applied metacognition (pp. 121–146). Cambridge: Cambridge University
Press. doi: 10.1037/e537102012-123
Mintzner, M. Z., & Griffiths, R. R. (2002). Alcohol and triazolam: Differen-
tial effects on memory, psycho-motor performance and subjective ratings
of effects. Behavioral Pharmacology,13, 653–658. doi: 10.1097/
00008877-200212000-00007
Nash, R. A., & Takarangi, M.(2011). Reconstructing alcohol-induced memory
blackouts. Memory,19,566–573. doi: 10.1080/09658211.2011.590508
Odinot, G., Wolters, G., & Van Giezen, A. (2013). Accuracy, confidence
and consistency in repeated recall of events. Psychology, Crime & Law,
19, 629–642. doi: 10.1080/1068316x.2012.660152
Otani, H., Von Glahn, N. R., Goenert, P. N., Migita, M., & Widner, R. L.
(2009). Hypermnesia, reminiscence and repeated testing. In M. R. Kelley
(Ed.), Applied memory (pp. 89–108). New York: Nova Science Publishers.
Palmer, F. T., Flowe, H. D., Takarangi, M. K. T., & Humphries, J. E.
(2013). Intoxicated witnesses and suspects: An archival analysis of their
involvement in criminal case processing. Law and Human Behavior,
37,54–59. doi: 10.1037/lhb0000010
Perry, P. J., Argo, T. R., Barnett, M. J., Liesveld, J. L., Liskow, B., Hernan, J.
M., Trnka, M. G., & Brabson, M. A. (2006). The association of alcohol-
induced blackouts and grayouts to blood alcohol concentrations. Journal
of Forensic Sciences,51,896–899. doi: 10.1111/j.1556-4029.2006.00161.x
Peterson, J. B., Rothfleisch, J., Zelazo, P. D., & Pihl, R. O. (1990). Acute
alcohol intoxication and cognitive functioning. Journal of Studies on
Alcohol,51, 114–122.
Ray, S., & Bates, M. E. (2006). Acute alcohol effects on repetition priming
and word recognition memory with equivalent memory cues. Brain and
Cognition,60, 118–127. doi: 10.1016/j.bandc.2005.07.009
Read, J. D., Yuille, J. C., & Tollestrup, P. (1992). Recollections of a rob-
bery: Effects of arousal and alcohol upon recall and person identification.
Law and Human Behavior,16, 425–446. doi: 10.1007/bf02352268
Santtila, P., Ekholm, M., & Niemi, P. (1998). Factors moderating the effects
of alcohol on interrogative suggestibility. Psychology Crime and Law,4,
139–152. doi: 10.1080/10683169808401754
Santtila, P., Ekholm, M., & Niemi, P. (1999). The effects of alcohol on in-
terrogative suggestibility: The role of state-anxiety and mood states as
mediating factors. Legal and Criminological Psychology,4,1–13. doi:
10.1348/135532599167707
Saults, J. S., Cowan, N., Sher, J. J., & Moreno, M. V. (2007). Differential
effects of alcohol on working memory: Distinguishing multiple pro-
cesses. Experimental and Clinical Psychopharmacology,15, 576–587.
doi: 10.1037/1064-1297.15.6.576
Schooler, J. W., & Loftus, E. F. (1986). Individual differences and experi-
mentation: Complementary approaches to interrogative suggestibility.
Social Behaviour,1, 105–112.
Schreiber Compo, N., Evans, J. R., Carol, R. N., Villalba, D., Ham, L. S.,
Garcia, T., & Rose, S. (2012). Intoxicated eyewitnesses: Better than their
reputation? Law and Hu man Behavior,36,77–86. doi: 10.1 037/h0093951
Sigurdsson, J. F., & Gudjonsson, G. H. (1994). Alcohol and drug
intoxication during police interrogation and the reasons why suspects
confess to the police. Addition,89, 985–997. doi: 10.1111/j.1360-
0443.1994.tb03358.x
Söderlund, H., Parker, E. S., Schwartz, B. B., & Tulving, E. (2005). Mem-
ory encoding and retrieval on the ascending and descending limbs of the
blood alcohol concentration curve. Psychopharmacology,182, 305–317.
doi: 10.1007/s00213-005-0096-2
Steblay, N., Dysart, J., Fulero, S., & Lindsay, R. C. L. (2003). Eyewitness
accuracy rates in police showup and lineup presentations: A meta-
anlaytic comparison. Law and Human Behavior,27, 523–540. doi:
10.1023/a:1025438223608
Van Oorsouw, K. I. M., & Merckelbach, H. L. G. J. (2009). Detecting
malingered memory problems in the civil and criminal arena. Legal and
Criminological Psychology,15,97–115. doi: 10.1348/135532509x451304
Van Oorsouw, K., & Merckelbach, H. (2012). The effects of alcohol on
crime-related memories: A field study. Applied Cognitive Psychology,
26,82
–90. doi: 10.1002/acp.1799
Van Oorsouw, K., Merckelbach, H., Ravelli, D., Nijman, H., & Mekking-
Pompen, I. (2004). Alcohol blackouts for criminally relevant behavior. Jour-
nal of the American Academy of Psychiatry and the Law,32, 364–370.
Retrieved from http://www.jaapl.org/content/32/4/364.full.pdf+html
Van Oorsouw, K., Merckelbach, H., & Willems, N. (2013). Als verdachten
of getuigen een slokje op hebben. Expertise en Recht,5, 204–209. [When
suspects or witnesses had a drink too many]
Verster, J. C., Van Duin, D., Volkerts, E. R., Schrueder, A. H. C. M. L., &
Verbaten, M. N. (2003). Alcohol hangover effects on memory functioning
and vigilance performance after an evening of binge drinking.
Neuropsychopharmacology,28,740–746. doi: 10.1038/sj.npp.1300090
Wetherill, R. R., & Fromme, K. (2011). Acute alcohol effects on narrative
recall and contextual memory: An examination of fragmentary blackouts.
Addictive Behaviors,36,886–889. doi: 10.1016/j.addbeh.2011.03.012
White, A. M. (2003). What happened? Alcohol, memory blackouts, and the
brain. Alcohol Research & Health,27, 186–196.
Yuille, J. C., & Tollestrup, P. A. (1990). Some effects of alcohol on eyewit-
ness memory. Journal of Applied Psychology,75, 268–273. doi:
10.1037//0021-9010.75.3.268
SUPPORTING INFORMATION
Supporting information may be found in the online version
of this article.
Intoxication, memory, and suggestibility 501
Copyright © 2015 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 29: 493–501 (2015)
