J. Behav. Ther. & Exp. Psychiat.

44 (2013) 179e185

Contents lists available at SciVerse ScienceDirect

Journal of Behavior Therapy and

Experimental Psychiatry
journal homepage: www.elsevier.com/locate/jbtep

Unconscious thought reduces intrusion development: A replication and extension

Julie Krans a, *, Drte Janecko b, Maarten W. Bos c, d
a
University of New South Wales,1 School of Psychology, Sydney NSW 2052, Australia
b
Radboud University Nijmegen, Nijmegen, The Netherlands
c
Harvard Business School,1 Negotiation, Organizations & Markets, Boston, USA
d
Harvard Medical School,1 Department of Psychiatry, Center for Sleep and Cognition, Boston, USA

a r t i c l e i n f o a b s t r a c t

Article history: Background and Objectives: Intrusive images after a traumatic event, a hallmark feature of post-traumatic
Received 16 January 2012 stress disorder, are suggested to develop because the trauma memory is disorganized and not integrated
Received in revised form into autobiographical memory. Unconscious Thought Theory predicts that information can be concep-
24 July 2012
tually organized after a period of unconscious thought (UT), more so than after conscious thought (CT).
Accepted 31 August 2012
We aimed to test the hypothesis that UT decreases intrusions and increases conceptual organization in
memory.
Keywords:
Methods: Participants were shown a stressful lm and were required to perform an UT task, a CT task, or
Intrusions
Trauma lm paradigm
a distraction task. Intrusions of the lm, intrusion qualities, and sequence memory were measured
Unconscious Thought Theory afterwards.
PTSD Results: We conrmed our hypothesis that UT (versus CT or mere distraction) leads to fewer intrusions,
thereby replicating earlier research. Contrary to prediction, we found no difference between the
conditions on sequence memory. In addition, conscious thought appeared to increase intrusion nowness
and arousal.
Limitations: The analogue design and healthy participant sample prevent from generalizing results to
other populations. Intrusion frequency and qualities were assessed immediately after the lm thereby
prohibiting us from drawing conclusions about any long-term effects.
Conclusions: Engaging in unconscious thought after a stressful lm can reduce intrusion frequency. This
has potential implications for clinical interventions to prevent initial stress symptoms. The underlying
mechanism remains unclear for now and provides an avenue for future research.
2012 Elsevier Ltd. All rights reserved.

1. Introduction manipulation immediately after an analogue traumatic event

Intrusive images are a hallmark feature of posttraumatic stress
disorder (PTSD; American Psychiatric Association, 2000) and are
increasingly recognized as being a signicant symptom of other
disorders, such as depression (Newby & Moulds, 2011). Effective
treatment options are available for chronic PTSD. However, there is
a lack of successful interventions that can effectively prevent the
initial development of symptoms by acting immediately after the
trauma (National Institute for Clinical Excellence, 2005; Sijbrandij
et al., 2007). The present study tested a theory-driven
* Corresponding author. Tel.: 61 (2) 93858249; fax: 61 (2) 93853641.
E-mail address: j.krans@unsw.edu.au (J. Krans).
1 a visuospatial task after viewing a trauma lm compared to
Data collection was conducted at the Behavioural Science Institute, Radboud
University Nijmegen, The Netherlands. Authors JK and MWB have moved during
the process of writing this article.
specically targeted at intrusions.
Intrusive images are suggested to develop because the trauma
memory is disorganized and not readily integrated into autobio-
graphical memory (Brewin, Dalgleish, & Joseph, 1996; Ehlers &
Clark, 2000). A memory that is disorganized may be fragmented
and not adhering to chronological order. This can make it difcult
for the trauma survivor to make sense of what happened and can
lead to recurring intrusive images that have a sense of nowness
(Ehlers & Clark, 2000). It can thus be hypothesized that a better
organized memory would not be intrusive as often, and would not
be as vivid, experienced as being here-and-now, and arousing.
There is previous research aimed at developing evidence-based
interventions to prevent initial post-traumatic stress symptoms.
Several studies found an ameliorating effect of performing
a verbal task or a no task control condition (Deeprose, Zhang,
DeJong, Dalgleish, & Holmes, 2012; Holmes, James, Coode-Bate, &

0005-7916/$ e see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jbtep.2012.08.004
180 J. Krans et al. / J. Behav. Ther. & Exp. Psychiat. 44 (2013) 179e185
Deeprose, 2009; Holmes, James, Kilford, & Deeprose, 2010). A
cognitive training towards more positive self-efcacy and meta-
cognitive beliefs about stress symptoms after a trauma lm was
shown to reduce intrusion development compared to a training
towards more negative self-efcacy beliefs (Woud, Holmes,
Postma, Dalgleish, & Mackintosh, 2011). Other studies have
shown that worrying or ruminating after a distressing lm
increased intrusions (Butler, Wells, & Dewick, 1995; Wells &
Papageorgiou, 1995) although others have found no signicant
effects (Ehring, Szeimies, & Schaffrick, 2009; Williams & Moulds,
2007).
Few studies have directly targeted the hypothesis that intru-
sions develop from a lack of memory organization although this is
an important claim made by models of PTSD and intrusive memory
(Brewin et al., 1996; Ehlers & Clark, 2000). Zetsche, Ehring, and
Ehlers (2009) presented participants with a distressing lm after
which they were assigned to a rumination, memory organization or
distraction condition. Intrusions were measured during a rst
session and reported in a one-week diary following the lm.
Overall, the experimental conditions did not have a signicant
effect on intrusion development, although, when split by gender,
men in the rumination condition reported more intrusions in the
rst session than men in the memory organization condition. Men
(but not women) in the memory organization condition, in turn,
reported fewer intrusions than men in the distraction condition.
Despite the unexpected gender effect, this gives some credence to
the idea that memory organization has an inuence on intrusion
frequency.
In another study using an analogue trauma design, participants
were shown a lm with trafc accidents, and then received a cued-
recognition memory test about the lm. The items were presented
in chronological order (Krans, Nring, Becker, & Holmes, 2009).
Participants then reported the recurrence of intrusive images of the
lm in a diary during the following week. After one week they
returned and completed a cued-recall test for the lm. The results
showed that intrusion frequency was lower for the part of the lm
for which they had received the recognition memory test.
Furthermore, distress associated with the intrusive images showed
a trend towards being lower for that part of the lm. This suggests
that imposing a chronological structure to a traumatic event could
reduce the occurrence of intrusions and possibly distress associated
with the intrusions.
The studies described above are part of a recent surge in efforts
to discover effective experimental interventions to prevent the
initial development of intrusions. There is, however, still a lack of
studies targeting memory organization as a possible benecial
factor. Interestingly, organization of new information in memory
has been described by Unconscious Thought Theory (UTT;
Dijksterhuis & Nordgren, 2006). UTT distinguishes between
conscious and unconscious thought, where conscious thought
processes are dened as schematic and top-down thinking about
an object or task using conscious attention allocated to that task or
object. It is thought to be of limited capacity, meaning that only few
chunks of information can be handled simultaneously (Dijksterhuis
& Nordgren, 2006). Unconscious thought processes, on the other
hand, are dened as thought processes that operate while
conscious attention is directed elsewhere. For example, if people
are faced with a problem that requires a large amount of infor-
mation to be processed, they are often advised to sleep on it (Bos,
Dijksterhuis, & Van Baaren, 2011; Dijksterhuis, 2004; Dijksterhuis,
Bos, Nordgren, & van Baaren, 2006a, 2006b, 2006c). This mode of
thought is suggested to be of high capacity and using a bottom-up
approach (Bos & Dijksterhuis, 2011). Importantly, unconscious
thought is assumed to be a goal-dependent process, which distin-
guishes it from a mere distraction effect. This assumption has been
recently supported with experimental evidence (e.g., Bos,
Dijksterhuis, & Van Baaren, 2008, 2011).
Most UTT research has focused on decision-making. The
common nding is that unconscious thought leads to better deci-
sions (Bos et al., 2008; Dijksterhuis, 2004; Dijksterhuis, Bos,
Nordgren, & Van Baaren, 2006; Dijksterhuis & Meurs, 2006; Ham,
Van den Bos, & Van Doorn, 2009; Lerouge, 2009), although some
research reports failures to replicate these effects (Acker, 2008;
Calvillo & Penaloza, 2009; Gonzalez-Vallejo, Lassiter, Bellezza, &
Lindberg, 2008; Newell, Wong, Cheung, & Rakow, 2009; Payne,
Samper, Bettman, & Luce, 2008; Thorsteinson & Withrow, 2009;
Waroquier, Marchiori, Klein, & Cleeremans, 2009). However,
a recent meta-analysis including more than 80 studies indicated
that, overall, there is a benet for unconscious thought in decision-
making although the overall effect size is small (for more infor-
mation, please see Strick et al., 2011).
Importantly for the research at hand, several experiments have
shown that unconscious thought can aid in organizing, clustering
and restructuring previously unorganized information in a mean-
ingful way (Bos et al., 2008; Dijksterhuis, 2004). For example, in the
experiment of Dijksterhuis (2004, Experiment 5; see also Bos et al.,
2008; Experiment 2), participants read 18 sentences describing
behaviors (for instance: Jeroen usually gets high grades in school)
that related to three overarching traits (intelligence, idealism and
extraversion) of a hypothetical person, presented in random order.
They were then instructed to recall these behaviors. Sequence
memory scores showed that participants in the unconscious
thought condition, more so than in the conscious thought condi-
tion, recalled the behaviors in the order according to the over-
arching traits rather than in the order in which they were
presented. That is, memory appeared to be more conceptually
organized after unconscious thought than after conscious thought.
Unconscious thought theoretically has the potential to organize
traumatic information in a more conceptual way than dealing with
many overwhelming details of the event by thinking it over
consciously. Having a more structured memory of an event may
reduce the negative effects of trauma (Ehlers & Clark, 2000).
Although there have been efforts at nding conscious strategies to
aid the integration of information in memory (e.g. Michael & Ehlers,
2007), this requires one to consciously revisit the negative events.
Unconscious thought potentially offers a way to aid the integration
of information in memory without consciously focusing on the
negative event. Krans and Bos (2012) tested whether unconscious
thought after a trauma lm would be successful at reducing
intrusion frequency. Participants were shown a negative lm clip
known for inducing intrusive images and were then given the
instruction to think about the lm before a memory test (conscious
thought condition), or performed an unrelated cognitively
demanding task, which prevented conscious processing. Of the
participants who received this distraction task, half were told that
the experiment was over (mere distraction condition), whereas the
other half were told that a memory test of the lm would be pre-
sented afterwards (unconscious thought condition). Comparing the
latter two conditions shows that unconscious thought is a goal-
dependent process.2 After the memory test (presented in all
conditions), participants reported their intrusive images of the lm
during an intrusion provocation task. This involved presenting the
participants with still frames from the lm that did not depict
anything negative. After this presentation, participants monitored
2
When participants are told the task is over, they no longer have a goal to
process information, whereas telling the participants that the experiment will
continue after the distraction task introduces the goal to unconsciously process the
information (Bos et al., 2008).
 J. Krans et al. / J. Behav. Ther. & Exp. Psychiat. 44 (2013) 179e185
their thoughts for 2 min and pressed a button every time they
experienced an intrusive image of the lm. Intrusion frequency was
signicantly lower in the unconscious thought condition than in
both other conditions. However, the free recall memory task that
was designed to measure memory organization for the lm did not
show any differences between the groups.
To summarize, unconscious thought decreased intrusion
frequency as expected, but the proposed mechanism by which this
was achieved remained unclear. The particular memory task may
not have been sensitive enough to pick up differences in memory
organization between the three conditions. Another limitation in
the Krans and Bos (2012) study was that only intrusion frequency
was assessed. In the present study, we aimed to (a) replicate the
effect of unconscious thought, (b) measure qualities of the intrusive
images to compare between conditions, and (c) measure memory
organization with a different measure.
In order to address our research goals, participants were shown
a negative lm clip and were then randomly assigned to one of
three conditions as in Krans and Bos (2012): conscious thought
(thinking about the lm before a memory test), unconscious
thought (performing a cognitively demanding task before the
memory test), or mere distraction (performing the task whilst
being unaware of the upcoming memory test). Participants
completed a structured memory test and intrusion frequency,
vividness, nowness and arousal were measured.
Based on models of PTSD (e.g., Ehlers & Clark, 2000), UTT
(Dijksterhuis & Nordgren, 2006), we predicted that (a) intrusion
frequency would be lowest in the unconscious thought condition,
with no differences between the two other conditions, and (b) that
memory organization would be higher in the unconscious thought
condition than in the other two conditions. In relation to the
intrusion qualities (vividness, nowness, and arousal), we predicted
that (c) all would be lower in the unconscious thought condition
than in the other two conditions. This prediction was based on
models from the PTSD (e.g., Ehlers & Clark, 2000) and memory
literature (Conway & Pleydell-Pearce, 2000) proposing that- by
denition- memory representations that are processed on a more
conceptual level will be less vivid, arousing and here-and-now than
memory representations at the lowest sensory levels (i.e.,
intrusions).
2. Method
Approval for this study was obtained from the Ethical
Committee of the Behavioural Science Institute (BSI) of the Rad-
boud University Nijmegen (number ECG 06042009), where data
was collected.
2.1. Participants
Participants were undergraduate students from the Radboud
University Nijmegen. They were recruited through an online
participation system and were rewarded with course credit. For
ethical reasons, participants were informed beforehand of the
exclusion criteria, that they would be watching a negative lm clip,
and that they could terminate the study at any time without
consequence for their credit. Exclusion criteria were as follows:
PTSD (current and lifetime), major depression (current and life-
time), psychotic episodes (current and lifetime), panic attacks
(current), panic disorder (current), intoxication during or shortly
prior to testing, history of fainting, and having to take exams in the
3
Data from 18 additional participants was lost due to a technical error.
181
week after the experiment. Complete data were collected from 78
participants.3
2.2. Negative lm clip
Participants viewed an 8.42 min scene from the lm Irrversible
by Gaspar No (2002). The lm clip shows a violent murder in
a nightclub. Participants viewed the lm on a PC with a 17 inch
monitor and listened to the sound through headphones.
2.3. Experimental conditions
In the conscious thought condition (CT), participants were
instructed to think about the lm for 4 min and were informed that
they would receive a memory test afterwards. In the unconscious
thought condition (UT), participants were also informed about the
memory test but they were instructed to rst complete a 2-back
task for 4 min (Jonides et al., 1997; Kirchner, 1958; Krans & Bos,
2012). In the 2-back task, participants see a series of numbers
presented on the screen in rapid succession (1 s presentation time,
0.5 s interval). Whenever the number they saw on the screen was
the same as two numbers before, participants were required to
press the spacebar on the computer. This task was chosen as this is
a highly demanding task leaving very little opportunity for
conscious thinking about the lm. Participants in the mere
distraction condition (MD) were informed that the 2-back task was
the end of the experiment and that they would not be questioned
about the lm.
2.4. Randomization, manipulation, compliance, and demand
Individual differences in trait anxiety were measured with the
Dutch trait version of the State-Trait Anxiety Inventory (STAI-T; Van
der Ploeg, 1980). This questionnaire contains 20 items to which
participants respond on a scale from 1 (almost never) to 4 (almost
always). Testeretest reliability (0.75) and internal consistency
(a 0.85) are good (Van der Ploeg, 1980).
To measure the emotional impact of the lm, a mood ques-
tionnaire (MoodQ; Holmes, Brewin, & Hennessy, 2004; Krans,
Nring, Speckens, & Becker, 2011) and the Dutch state version of
the State Trait Anxiety Inventory (STAI-S; Spielberger, Gorsuch,
Lushene, Vagg, & Jacobs, 1983; Van der Ploeg, 1980) were used.
The MoodQ consists of ve items about current happiness, fear,
horror, depression and anger to which participants respond on
a scale from 0 (not at all) to 10 (extremely). The STAI-S contains 20
items and participants respond on a scale from 1 (not at all) to 4
(very much so). Testeretest reliability for the STAI-S has been re-
ported at a 0.25, internal consistency at a 0.85 (Van der Ploeg,
1980).
Participants rated their attention for the lm on an 11-point
scale from 0 (not at all) to 10 (completely). In addition, partici-
pants in all conditions were asked to indicate how much they had
thought about the lm during the experimental task (continuous
scale from not at all to very much). Participants in the UT and MD
conditions were additionally asked whether their thinking of the
lm was deliberate or spontaneous. Lastly, participants indicated
whether they expected their task to increase, decrease or have no
effect on their memory and intrusions of the lm.
2.5. Experimental measures
2.5.1. Intrusion frequency
The intrusion provocation procedure (Lang, Moulds, & Holmes,
2009; Michael, Ehlers, Halligan, & Clark, 2005) was used to elicit
intrusions from the lm clip. Participants were presented with ten
182 J. Krans et al. / J. Behav. Ther. & Exp. Psychiat. 44 (2013) 179e185
still pictures from the lm representing neutral scenes that
preceded the negative events in the lm. The pictures were pre-
sented as a slideshow with a presentation time of 4 s per picture.
Intrusion frequency was measured with an adapted version of the
widely used Intrusion subscale of the Impact of Events scale (IES;
Horowitz, Wilner, & Alvarez, 1980). This subscale has a good
internal consistency (a 0.86) and good testeretest reliability
(a 0.89) (Sundin & Horowitz, 2002). The 8 items were reworded
to refer to the lm and participants responded on a scale from
0 (not at all) to 4 (very much). Participants were instructed to rate
the items in terms of their appropriateness during the course of the
present study (rather than during the last seven days as in the
original version).
2.5.2. Intrusion characteristics
Intrusion characteristics were measured with an adapted
version of the Intrusion Questionnaire (IQ; Hackmann, Ehlers,
Clark, & Speckens, 2004; Michael et al., 2005). Four items
assessed the mean vividness of the intrusions that participants had
experienced during the provocation task. Participants responded
on a 1 (vague and dim) to 4 (perfectly clear and real) scale, origi-
nally derived from the Vividness of Visual Imagery Questionnaire
(VVIQ; Marks, 1973). Four items assessed the experience of
nowness of the intrusions, i.e., a sense of nowness accompanied by
the original emotions as during encoding of the lm. Five items
measured physical and emotional arousal accompanying the
intrusions (based on the Hyperarousal subscale of the IES-Revised;
Weiss & Marmar, 1996). These items were answered using a 1
(completely disagree) to 4 (completely agree) scale.
2.5.3. Memory organization
In the Krans and Bos (2012) study, participants were asked to
say out loud everything they remembered about the lm. This was
audio-recorded and later scored for sequence memory. However,
no differences were found between the experimental conditions.
Upon further inspection of the data it appeared that only the rst,
last, and most salient events from the lm were mentioned by
almost all participants with big gaps in between. In the present
study therefore a structured recall task was used in order to direct
participants towards recalling all events they could. Participants
were instructed to list all the events in the lm within given rows
(one event per row). An example of an event was given to provide
an illustration of the amount of detail that was expected: If you had
watched a lm clip about a trafc accident, the events could, for
example, be: a car is driving through a red light e a child is waiting to
cross at the trafc light e the medical staff carry a woman into the
ambulance. First, the order of events as it was recorded in the
participants list was coded in comparison to the order of events in
a previously constructed master list (Krans & Bos, 2012). Next,
a sequence memory score was calculated by computing a Spearman
correlation between the participants list and the master list (see
Jelinek, Randjbar, Seifert, Kellner, & Moritz, 2009; Kindt & van den
Hout, 2003; Kindt, van den Hout, & Buck, 2005; Krans & Bos, 2012;
Wegner, Quillian, & Houston, 1996).
2.6. Procedure
The total experiment took approximately 40 min depending on
individual speed. All measures were presented on a PC using
Macromedia Authorware software (version 7.01). Potential partic-
ipants were rst screened for exclusion criteria. If they could
continue in the study they received written information and signed
an informed consent form. The experiment started with instruc-
tions on the 2-back task and practice trials. To avoid an experi-
mental confound, these instructions were provided in all
conditions. Demographic information was obtained and partici-
pants continued on the STAI-T, STAI-S, and MoodQ, after which they
were shown the lm. They were instructed to completely attend to
the lm as if they were a bystander and not to look away or
disengage in any way. After the lm, the experimental instructions
(CT, UT, MD) were given randomly with the restriction that an
identical number of participants were allocated to each condition.
The recall task was then administered, followed by the STAI-S and
MoodQ. Participants then completed the intrusion provocation
task, the IQ, compliance and demand questions, and IES Intrusion
subscale. The interval between the lm and the IES administration
was approximately 12 min, depending on individual speed. Finally,
they were thanked for their participation and debriefed.
3. Results
3.1. Outliers and statistical approach
Outliers were detected according to the procedure described by
Tabachnick and Fidell (1996). Across and within groups, Mahala-
nobis distances were calculated to detect multivariate outliers, but
none were found. Univariate outliers were checked across and
within groups and three were detected. These were changed into
the rst unit that was not a statistical outlier based on three stan-
dard deviations from the mean.
Directional hypotheses were tested one-tailed using Helmert
contrasts. Wherever one-tailed tests were used, this is mentioned
explicitly. In case of non-signicant or opposite effects two-tailed
values are reported. The level of signicance was an a of 0.05 for
all tests; multiple comparisons were Bonferroni corrected.
3.2. Compliance and demand
In order to be included in the analysis, participants in the CT
condition were required to think about the lm at least half of the
time during the conscious thought task (Krans & Bos, 2012). Eight
participants were excluded from the dataset because they did not
meet this criterion. To be included in the analysis, participants in
the UT and MD conditions were not allowed to have thought about
the lm deliberately during the 2-back task (Krans & Bos, 2012).
Based on this criterion, 11 participants from the UT condition and
eight participants from the MD condition were excluded from
analysis. The nal dataset included 51 participants; 17 in each
condition.
Overall, attention for the lm was rated as high, M 8.82,
SD 1.35; with no signicant difference between conditions
(p 0.76). Performance on the 2-back task was comparable
between the UT and MD condition (p 0.93), M 109.15, SD 5.94.
Participants intrusion/sequence memory scores did not depend on
whether they expected their experimental task to increase,
decrease or not affect their intrusions/memory of the lm; for
intrusions: F(2, 48) 0.52, p 0.60, or for sequence memory: F(2,
48) 0.73, p 0.49.
3.3. Randomization check
There was no signicant difference in participant age between
conditions (p 0.23), M 19.76, SD 1.69. Distribution of men and
women was comparable across conditions, p 0.21; with 47
women and 4 men in total. There was no signicant difference in
trait anxiety (STAI-T) between conditions (p 0.98), overall
M 35.71, SD 7.96. The conditions were comparable on pre-lm
state anxiety (STAI-S; p 0.71), overall M 40.29, SD 5.30; and
pre-lm negative mood (MoodQ; p 0.35), overall M 12.92,
SD 4.87.
 J. Krans et al. / J. Behav. Ther. & Exp. Psychiat. 44 (2013) 179e185 183

3.4. Manipulation check 4. Discussion

Change in state anxiety from pre- to post lm was tested with
a mixed model repeated measures ANOVA, with time (pre-lm,
post-lm) as the within-subjects factor, condition (CT, UT, MD) as
the between-subjects factor, and state anxiety (STAI-S scores) as
the dependent variable. There was a predicted signicant effect of
time, F(1, 48) 145.60, p < 0.001, f 1.74, indicating an overall
signicant increase in state anxiety from pre- to post-lm. The
main effect of condition was not signicant, F(2, 48) 1.62, p 0.21.
There was a signicant time  condition interaction, F(2,
48) 5.48, p 0.01, f 0.48. Pairwise comparisons of the anxiety
change scores (post minus pre) showed that the increase in state
anxiety was signicantly larger in the CT than in the UT condition,
p 0.01, and MD, p 0.03; whereas the latter two conditions did
not differ signicantly from each other, p > 0.99.
The same pattern emerged for negative mood (MoodQ): there
was a signicant main effect of time (pre- to post-lm), F(1,
48) 247.25, p < 0.001, f 2.27; but no signicant main effect of
condition (CT, UT, MD), F(2, 48) 0.93, p 0.40. The
time  condition interaction effect was signicant, F(2, 48) 3.51,
p 0.04, f 0.38, though pairwise comparisons on the change
scores of negative mood (post-lm minus pre-lm) did not reach
signicance. The descriptive statistics are reported in Table 1.
3.5. Experimental effects
3.5.1. Intrusion frequency and quality
As predicted, Helmert contrasts showed that scores on the
intrusion subscale of the IES were signicantly lower in the UT
condition compared to the other two conditions, p < 0.05 (one-
tailed). There was no signicant difference between the CT and MD
conditions, p 0.45 (two-tailed).
Contrary to prediction, Helmert contrasts showed no signicant
difference between the UT condition and the CT and MD condition
on intrusion vividness, nowness or arousal, (all ps > 0.12, two-
tailed). However, there was a signicant difference between the
CT and MD condition on nowness and arousal (all ps  0.02, two-
tailed). Follow-up contrasts (two-tailed) showed a signicantly
higher feeling of nowness in the CT than in the UT and MD, p 0.02,
and a higher arousal level in the CT condition compared to the MD
condition, p 0.05, but not compared to the UT condition, p .18.
3.5.2. Sequence memory
Contrary to prediction, Helmert contrasts showed no signicant
difference on sequence memory between the UT condition and the
CT and MD conditions (p 0.82, two-tailed). There was no signif-
icant difference between the CT and MD conditions either (p 0.88,
two-tailed).
In the present study, we aimed to (a) replicate the effect of
unconscious thought on intrusion frequency, (b) extend earlier
research by including intrusion qualities, and (c) test the organi-
zation of the trauma lm in memory with a structured recall test.
The main nding was that participants who were given an
unconscious thought instruction after viewing the trauma lm
reported a reduction in intrusive experiences compared to partic-
ipants who were instructed to think about the lm deliberately, or
participants in the distraction control condition. This nding
replicates our earlier work (Krans & Bos, 2012) and supports the
goal-dependent asset of unconscious thought as opposed to it
being a mere distraction effect (Bos et al., 2008, 2011). This time, we
included a widely used and well validated measure of intrusive
experience, namely the Impact of Event Scale (Horowitz et al.,
1980). Thus, the effect of unconscious thought reducing intrusion
frequency has now been found using two difference intrusion
measures, which increases the validity of the effect.
Contrary to expectations, we did not nd an effect of uncon-
scious thought on intrusion qualities. If anything, it appeared that
consciously thinking about the lm increased intrusion nowness
and arousal. This is in line with the nding that state anxiety and
negative mood increased to a larger extend in the conscious
thought condition than in the other two conditions. Although this
was not in line with our expectations, this opens up an interesting
possibility that can be addressed in future studies. Conscious
thought is better at encoding details than at processing the
conceptual meaning of new information (Dijksterhuis & Nordgren,
2006). Intrusive images, by denition, consist mostly of sensory
impressions rather than more abstract representations. Thus, more
conscious thought may have resulted in more arousing and here-
and-now intrusions because of enhanced sensory encoding.
As in Krans and Bos (2012), we were unable to nd differences in
sequence memory between the three conditions despite using
a structured recall test to measure memory organization for the
lm. According to PTSD models (Brewin et al., 1996; Ehlers & Clark,
2000), intrusions occur when the trauma information is disorga-
nized and not well integrated into memory. As unconscious
thought is suggested to be superior in organizing information in
a conceptual way (Dijksterhuis & Nordgren, 2006) this should have
been detected in both the intrusion and recall measure. One
possibility is that the recall measure that was used here was still
insensitive to differences in sequence memory. Another possibility
is that sequence memory is not a good measure of the conceptu-
alization of the lm information. For example, asking about the
meaning or general gist of the lm may have resulted in differences
between groups. According to UTT, unconscious thought would
lead to a more global assessment of meaning (e.g., vengeance)

Table 1
Means and standard deviations for control and experimental variables within and across conditions.

Conscious thought Unconscious thought Mere Distraction Total

M SD M SD M SD M SD
STAI-T 36.06 9.58 35.53 6.72 35.53 7.80 35.71 7.96
STAI-S Pre-lm 39.59 3.78 40.18 6.26 41.12 5.75 40.29 5.30
Post-lm 55.76 5.54 49.24 7.64 50.88 5.98 51.96 6.91
MoodQ Pre-lm 11.82 3.88 12.71 4.83 14.24 5.70 12.92 4.87
Post-lm 34.88 5.84 29.59 9.36 30.24 7.01 31.57 7.68
Attention 9.00 1.37 8.82 1.29 8.65 1.46 8.82 1.35
Sequence memory 0.81 0.37 0.77 0.37 0.79 0.36 0.79 0.36
Intrusion vividness (IQ) 5.06 3.31 3.94 2.28 3.29 2.59 4.10 2.80
Intrusion nowness (IQ) 9.29 1.57 7.47 1.88 7.41 2.24 8.06 2.07
Intrusion arousal (IQ) 12.12 2.26 10.59 2.81 10.12 1.73 10.94 2.42
Post-lm intrusions (IES) 9.29 6.47 5.87 3.63 7.88 5.73 7.68 5.49
Post-lm avoidance (IES) 10.33 7.89 9.59 5.50 11.59 4.49 10.50 6.06
184 J. Krans et al. / J. Behav. Ther. & Exp. Psychiat. 44 (2013) 179e185
whereas conscious thought should lead to more detailed informa-
tion (e.g., a man assaults another man in a bar). This is another
testable hypothesis that can instigate future research.
Another necessary avenue for future research is to test the effect of
unconscious thought in comparison to different forms of conscious
thought. In the rumination literature, for example, a distinction is
made between abstract and concrete thought (Watkins, 2008).
Whereas abstract thought is hypothesized to increase intrusions
through ruminative processes, concrete thought may work to coun-
teract cognitive avoidance and should enhance emotional processing
(Borkovec, Alcaine, & Behar, 2004). Several studies have compared
these forms of conscious thought but none included an unconscious
thought condition (e.g., Williams & Moulds, 2007; Zetsche et al.,
2009). The present study only included a general conscious
thought condition thus this comparison still open for investigation.
Our study has several limitations. First, the sample consisted of
healthy students and a lm was used as an analogue traumatic event.
Therefore, the results may not generalize to other populations or
real-life events such as a trauma. The implications of this study are
limited to the phenomenon of intrusions and do not inform about
other PTSD symptoms or, for example, biological markers of trauma.
Second, we did not include a natural group that was not instructed
to think in a certain way about the trauma lm or was immediately
distracted. Theoretically, it is possible that all three conditions had
an effect on intrusion development compared to a more natural state
of processing. It can also not be ascertained how other natural
behaviors, such as avoidance, rumination, or spontaneous conscious
or unconscious thought processes would interact with our manip-
ulations. However, it is difcult to come up with a natural condition
that is matched on all methodological aspects, e.g., the 4 min delay to
recall that is present in the three experimental conditions of this
study. Third, intrusive experiences were measured immediately
after the lm. Future studies are needed that test the longer-term
effect of the different modes of thought. Additionally, the measure
used (IES Intrusion scale) is retrospective in nature and is therefore
sensitive to memory biases. Next, as in Krans and Bos (2012) a fair
number of participants had to be excluded because they did not
fulll the task criteria. However, there is no easy solution for this
problem as instructing participants not to think about the lm could
result in unwanted suppression effects (Krans & Bos, 2012; Wegner,
1994). Most trauma victims likely try not to think about their trauma,
and engage in distracting activities. A strength of the unconscious
thought approach is that, in fact, participants are not required to
think about the trauma. Instead, it is made explicit that the event
will be revisited at some time in the future. This is clearly different
from any existing situation so far.
4.1. Conclusions
To conclude, the present study replicated the effect that uncon-
scious thought after a trauma lm reduces intrusion frequency. We
extend previous ndings by having a widely used measure for intru-
sions (the IES) and by incorporating a measure of intrusion quality. As
UTT is a very recent theory, several questions about the underlying
mechanisms of the effects remain unanswered here. We suggested
several areas of future research that can help extend our knowledge in
this eld. Additionally, if future research can conrm the effect of
unconscious thought on intrusion frequency when more time has
passed between the trauma and the intervention, then results of the
presented study may proof to be clinically promising as well.
Funding
Julie Krans is currently supported by a Rubicon fellowship from
the Netherlands Organisation for Scientic Research.
Acknowledgements
We would like to thank Gosia Fietz, Imke Veenstra, and Karin
van Roosmalen for their help with data collection. The data was
collected at the Radboud University Nijmegen in the Netherlands
with support of the Behavioural Science Institute.
References
Acker, F. (2008). New ndings on unconscious versus conscious thought in decision
making: additional empirical data and meta-analysis. Judgment and Decision
Making, 3, 292e303.
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental
disorders (4th ed. Text Revision). Washington, D.C.: APA.
Borkovec, T. D., Alcaine, O. M., & Behar, E. (2004). Avoidance theory of worry and
generalized anxiety disorder. In R. G. Heimberg, C. L. Turk, & D. S. Mennin (Eds.),
Generalized anxiety disorder: Advances in research and practice (pp. 77e108).
New York, NY: Guilford.
Bos, M. W., Dijksterhuis, A., & Van Baaren, R. B. (2008). On the goal-dependency
of unconscious thought. Journal of Experimental Social Psychology, 44, 1114e
1120.
Bos, M. W., & Dijksterhuis, A. (2011). Unconscious thought works bottom-up and
conscious thought works top-down when forming an impression. Social
Cognition, 29, 727e737.
Bos, M. W., Dijksterhuis, A., & Van Baaren, R. B. (2011). The benets of sleeping on
things: unconscious thought leads to automatic weighting. Journal of Consumer
Psychology, 21(1), 4e8.
Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of post-
traumatic stress disorder. Psychological Review, 103, 670e686.
Butler, G., Wells, A., & Dewick, H. (1995). Differential effects of worry and imagery
after exposure to a stressful stimulus: a pilot study. Behavioural and Cognitive
Psychotherapy, 23, 45e56.
Calvillo, D. P., & Penaloza, A. (2009). Are complex decisions better left to the
unconscious? Further failed replications of the deliberation-without-attention
effect. Judgment and Decision Making, 4, 509e517.
Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobio-
graphical memories in the self-memory system. Psychological Review, 107,
261e288.
Deeprose, C., Zhang, S., DeJong, H., Dalgleish, T., & Holmes, E. A. (2012). Imagery in
the aftermath of viewing a traumatic lm: using cognitive tasks to modulate
the development of involuntary memory. Journal of Behavior Therapy Experi-
mental Psychiatry, 43, 758e764.
Dijksterhuis, A. (2004). Think different: the merits of unconscious thought in
preference development and decision making. Journal of Personality and Social
Psychology, 87, 586e598.
Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & van Baaren, R. B. (2006a). On making
the right choice: the deliberation-without-attention effect. Science, 311, 1005e
1007.
Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & van Baaren, R. B. (2006b). Complex
choices better made unconsciously? Science, 313, 760e761.
Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & van Baaren, R. B. (2006c). Making
choices without deliberation. Science, 312, 1472.
Dijksterhuis, A., & Meurs, T. (2006). Where creativity resides: the generative power
of unconscious thought. Consciousness and Cognition, 15, 135e146.
Dijksterhuis, A., & Nordgren, L. F. (2006). A theory of unconscious thought.
Perspectives on Psychological Science, 1, 95e109.
Ehlers, A., & Clark, D. M. (2000). A cognitive model of posttraumatic stress disorder.
Behaviour Research and Therapy, 38, 319e345.
Ehring, T., Szeimies, A. K., & Schaffrick, C. (2009). An experimental analogue study
intro the role of abstract thinking in trauma-related rumination. Behaviour
Research and Therapy, 47, 285e293.
Gonzalez-Vallejo, C., Lassiter, D. G., Bellezza, L. S., & Lindberg, M. J. (2008). Save
angels perhaps: a critical examination of unconscious thought theory and the
deliberation-without-attention effect. Review of General Psychology, 12, 282e
296.
Hackmann, A., Ehlers, A., Clark, D. M., & Speckens, A. (2004). Characteristics of
intrusive memories in PTSD and their response to treatment. Journal of Trau-
matic Stress, 17, 231e240.
Ham, J., Van den Bos, K., & Van Doorn, E. A. (2009). Lady justice thinks uncon-
sciously: unconscious thought can lead to more accurate justice judgments.
Social Cognition, 27, 509e521.
Holmes, E. A., Brewin, C. R., & Hennessy, R. G. (2004). Trauma lms, information
processing, and intrusive memory development. Journal of Experimental
Psychology: General, 133, 3e22.
Holmes, E. A., James, E. L., Coode-Bate, T., & Deeprose, C. (2009). Can playing the
computer game Tetris reduce the build-up of ashbacks for trauma? A
proposal from cognitive science. PLoS ONE, 4(1), e4153.
Holmes, E. A., James, E. L., Kilford, E. J., & Deeprose, C. (2010). Key steps in devel-
oping a cognitive vaccine against traumatic ashbacks: visuospatial Tetris
versus pub quiz. PLoS ONE, 5(11), e13706.
Horowitz, M. J., Wilner, N., & Alvarez, W. (1980). Impact of event scale: a measure of
subjective stress. Psychosomatic Medicine, 41, 209e218.
 J. Krans et al. / J. Behav. Ther. & Exp. Psychiat. 44 (2013) 179e185
Jelinek, L., Randjbar, S., Seifert, D., Kellner, M., & Moritz, S. (2009). The organization
of autobiographical memory and nonautobiographical memory in post-
traumatic stress disorder (PTSD). Journal of Abnormal Psychology, 118, 288e298.
Jonides, J., Schumacher, E. H., Smith, E. E., Lauber, E. J., Awh, E., Minoshima, S., et al.
(1997). Verbal working memory load affects regional brain activation as
measured by PET. Journal of Cognitive Neuroscience, 9, 462e475.
Kindt, M., & van den Hout, M. (2003). Dissociation and memory fragmentation:
experimental effects on meta-memory but not on actual memory performance.
Behaviour Research and Therapy, 41, 167e178.
Kindt, M., van den Hout, M., & Buck, N. (2005). Dissociation related to subjective
memory fragmentation and intrusions but not to objective memory distur-
bances. Journal of Behavior Therapy and Experimental Psychiatry, 36, 43e59.
Kirchner, W. K. (1958). Age differences in short-term retention of rapidly changing
information. Journal of Experimental Psychology, 55, 352e358.
Krans, J., & Bos, M. W. (2012). To think or not to think about trauma? An experi-
mental investigation into unconscious thought and intrusion development.
Journal of Experimental Psychopathology, .
Krans, J., Nring, G., Holmes, E. A., & Becker, E. S. (2009). Tell me more: can
a memory test reduce analogue traumatic intrusions? Behaviour Research and
Therapy, 47, 426e430.
Krans, J., Nring, G., Speckens, A., & Becker, E. S. (2011). Eyewitness or earwitness:
the role of mental imagery in intrusion development. International Journal of
Cognitive Therapy, 4, 154e164.
Lang, T. J., Moulds, M. M., & Holmes, E. A. (2009). Reducing depressive intrusions via
a computerized cognitive bias modication of appraisals task: developing
a cognitive vaccine. Behaviour Research and Therapy, 47, 139e145.
Lerouge, D. (2009). Evaluating the benets of distraction on product evaluations:
the mindset effect. Journal of Consumer Research, 36, 367e379.
Marks, D. F. (1973). Visual imagery differences in the recall of pictures. British
Journal of Psychology, 64, 17e24.
Michael, T., & Ehlers, A. (2007). Enhanced perceptual priming for neutral stimuli
occurring in a traumatic context: two experimental investigations. Behaviour
Research and Therapy, 45, 341e358.
Michael, T., Ehlers, A., Halligan, S. L., & Clark, D. M. (2005). Unwanted memories of
assault: what intrusion characteristics are associated with PTSD? Behaviour
Research and Therapy, 43, 613e628.
National Institute for Health and Clinical Excellence. (2005). Post-traumatic stress disorder
(PTSD): The management of PTSD in adults and children in primary and secondary care
(No. CG026). London National Institute for Health and Clinical Excellence.
Newby, J. M., & Moulds, M. L. (2011). Characteristics of intrusive memories in
a community sample of depressed, recovered depressed and never-depressed
individuals. Behaviour Research and Therapy, 49, 234e243.
Newell, B. R., Wong, K. Y., Cheung, J. C. H., & Rakow, T. (2009). Think, blink or sleep
on it? The impact of modes of thought on complex decision making. Quarterly
Journal of Experimental Psychology, 62, 707e732.
(Director) No, G. (2002). Irreversible [motion picture]. France: Mars Distribution.
185
Payne, J. W., Samper, A., Bettman, J. R., & Luce, M. F. (2008). Boundary on uncon-
scious thought in complex decision making. Psychological Science, 19, 1118e
1123.
Sijbrandij, M., Olff, M., Reitsma, J. B., Carlier, I. V. E., de Vries, M. H., & Gersons, B. P. R.
(2007). Treatment of acute stress disorder with brief cognitive behavioural
therapy: a randomized controlled trial. American Journal of Psychiatry, 164, 82e
90.
Spielberger, C. D., Gorsuch, R. L., Lushene, P. R., Vagg, P. R., & Jacobs, A. G. (1983).
Manual for the state-trait anxiety inventory. Palo Alto: Consulting Psychologists
Press, Inc.
Strick, M., Dijksterhuis, A., Bos, M. W., Sjoerdsma, A., van Baaren, R. B., &
Nordgren, L. F. (2011). A meta-analysis on unconscious thought effects. Social
Cognition, 29, 738e762.
Sundin, E. C., & Horowitz, M. J. (2002). Impact of event scale: psychometric prop-
erties. British Journal of Psychiatry, 180, 205e209.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New
York: Harper Collins.
Thorsteinson, T. J., & Withrow, S. (2009). Does unconscious thought outperform
conscious thought on complex decisions? A further examination. Judgment and
Decision Making, 4, 235e247.
Van der Ploeg, H. M. (1980). Validity of the Zelf-Beoordelings-Vragenlijst. A Dutch
version of the Spielberger state-trait anxiety inventory. Nederlands Tijdschrift
Voor De Psychologie En Haar Grensgebieden, 35, 243e249.
Waroquier, L., Marchiori, D., Klein, O., & Cleeremans, A. (2009). Methodological
pitfalls of the unconscious thought paradigm. Judgment and Decision Making, 4,
610.
Watkins, E. (2008). Constructive and unconstructive repetitive thought. Psycho-
logical Bulletin, 134, 163e206.
Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101,
34e52.
Wegner, D. M., Quillian, F., & Houston, C. E. (1996). Memories out of order: thought
suppression and the disturbance of sequence memory. Journal of Personality and
Social Psychology, 71, 680e691.
Weiss, D. S., & Marmar, C. R. (1996). The impact of event scale-revised. In
J. P. Wilson, & T. M. Keane (Eds.), Assessing psychological trauma and PTSD (pp.
399e411). New York, NY: Guilford Press.
Wells, A., & Papageorgiou, C. (1995). Worry and the incubation of intrusive images
following stress. Behaviour Research and Therapy, 33, 579e583.
Williams, A. D., & Moulds, M. L. (2007). The impact of ruminative processing on the
development of intrusive memories. Behaviour Change, 24, 55e69.
Woud, M. L., Holmes, E. A., Postma, P., Dalgleish, T., & Mackintosh, B. (2011).
Ameliorating intrusive memories of distressing experiences using computer-
ized reappraisal training. Emotion, . http://dx.doi.org/10.1037/a0024992.
Zetsche, U., Ehring, T., & Ehlers, A. (2009). The effects of rumination on mood and
intrusive memories after exposure to traumatic material: an experimental
study. Journal of Behavior Therapy and Experimental Psychiatry, 40, 499e514.