J. Behav. Ther. & Exp. Psychiat.

40 (2009) 374383

Contents lists available at ScienceDirect

Journal of Behavior Therapy and Experimental Psychiatry

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

Visual false memories in post-traumatic stress disorder (PTSD)

Lena Jelinek a, *, Birgit Hottenrott a, Sarah Randjbar a, Maarten J. Peters b, Steffen Moritz a
Department of Psychiatry and Psychotherapy, University Medical Center Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg; Germany Department of Clinical Psychological Science, Faculty of Psychology and Neuroscience, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands
b a

a r t i c l e i n f o
Article history: Received 23 July 2008 Received in revised form 22 January 2009 Accepted 10 February 2009 Keywords: False memory Post-traumatic stress disorder Trauma Depression Dissociation Source monitoring

a b s t r a c t
There is an ongoing debate whether or not patients with posttraumatic stress disorder (PTSD) are more prone to produce false memories. The present study investigated this question using a visual variant of the DeeseRoedigerMcDermott (DRM) paradigm, additionally addressing underlying mechanisms of false memory production (e.g., depression, dissociation, emotional valence, arousal). The visual paradigm was administered to 48 traumatized individuals with (n 20) and without PTSD (n 28) and 28 non-traumatized controls. Groups did not differ with regard to memory performance and memory condence. False memories were correlated with depression. We recommend that future studies employ trauma-related material to further explore memory aberrations in PTSD. 2009 Elsevier Ltd. All rights reserved.

1. Theoretical background Victims of traumatic events are often haunted by stressful memories of the experience in form of nightmares or ashbacks. Such intrusive symptoms are part of the diagnosis post-traumatic stress disorder (PTSD; American Psychiatric Association, 1994), but it remains unclear how accurate and valid these involuntary as well as deliberately retrieved traumatic memories are. Several reports about fabricated and invalid (recovered) traumatic memories, in particular those related to childhood sexual

* Corresponding author. Tel.: 49 40 7410 57539; fax: 49 40 7410 57566. E-mail address: l.jelinek@uke.uni-hamburg.de (L. Jelinek). 0005-7916/$ see front matter 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbtep.2009.02.003

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

375

abuse, have led to an enormous debate and much controversy in the past years (Loftus & Davis, 2006; McNally, 2003). Research has shown that memories in general are susceptible to distortions and false memories can be induced rather easily (Loftus & Davis, 2006). One experimental paradigm to provoke false memories is the DeeseRoedigerMcDermott (DRM; Deese, 1959; Roediger & McDermott, 1995) paradigm. Typically word lists are presented to participants: Each word list is highly associated with one word that is not part of this learning list, the socalled critical lure. For example, when door, glass, pane, shade, ledge, sill, house, open, curtain, frame, view, breeze, sash, screen, and shutter are presented to the participants, the critical lure window is often falsely recalled or recognized in later trials. In healthy participants this mechanism has been demonstrated with a prevalence rate of up to 80% for false recall or recognition (Stadler, Roediger, & McDermott, 1999). The DRM paradigm has already been administered to PTSD patients to investigate whether this population is more susceptible to memory distortions (Bremner, Shobe, & Kihlstrom, 2000; Brennen, Dybdahl, & Kapidzic, 2007; Zoellner, Foa, Brigidi, & Przeworski, 2000). It has been assumed that PTSD patients are more prone to produce false memories arguing that PTSD is linked to dissociation (as a personality trait see, e.g., Bremner, Southwick, Brett, Fontana, Rosenheck, & Charney, 1992; Dancu, Riggs, Hearst-Ikeda, Shoyer, & Foa, 1996; for dissociative experiences at the time of trauma, see Ozer, Best, Lipsey, & Weiss, 2003) and dissociative symptoms have been associated with false memories (e.g., Clancy, Schacter, McNally, & Pitman, 2000; Hyman & Billings, 1998; Winograd, Peluso, & Glover, 1998). Results are yet inconclusive. On the one hand, Bremner et al. (2000) showed that sexually abused women with PTSD were more prone to falsely recall critical lures than abused women without PTSD and women without a history of sexual abuse. On the other hand, Zoellner et al. (2000) found that trauma rather than PTSD was associated with false memories as in their study assault victims with and without PTSD falsely recalled more critical lures than non-traumatized individuals. While the two traumatized groups did not differ statistically with regard to false memories, the false recall of critical lures was correlated with PTSD severity in the PTSD group (n 14). While the two aforementioned studies implemented neutral material to create false memories, Brennen et al. (2007) additionally administered trauma-related word lists to war-exposed participants with and without PTSD. Their ndings showed that groups were equally prone to critical lures of non-trauma lists, whereas the PTSD group incorrectly recalled more critical lures of the trauma-related lists. One explanation for these mixed ndings could well be the inuence of one or several moderating variables obscuring the relationship between PTSD and false memories. For example, in a recent study by Corson and Verrier (2007) it was found that false memories were signicantly more frequent under conditions of high arousal, independent of valence. In addition to false memories, meta-memory was also investigated in the aforementioned studies. By implementing the rememberknow paradigm (Tulving, 1985), it was proposed that PTSD patients should display more remember responses (i.e. vivid recall of episode) for false memories (Brennen et al., 2007; Zoellner et al., 2000). While the results of Brennen et al. (2007) provided some evidence in this direction (i.e. more remember responses of trauma-related false memories), results were statistically insignicant. Another way to investigate meta-memory is to assess memory condence. This was done in the study by Bremner et al. (2000) by asking participants whether an item of the recognition list was probably or denitely new versus old. Unfortunately, results on these indices were not reported. While the investigation of memory condence is highly relevant for the clinical and forensic eld, where trauma victims are asked to disclose their trauma-related memories, clear insights on the accuracy-condence relationship are thus lacking. Also important for the evaluation of testimonies of trauma victims is that until now only verbal material (presented orally or visually) has been used in studies employing the DRM paradigm in PTSD research (for comparisons of participants with and without histories of trauma on other paradigms, such as the directed forgetting, see, e.g., DePrince & Freyd, 2004; McNally, Ristuccia, & Perlman, 2005). To the best of our knowledge, visual picture cues were never used in PTSD, despite high ecologic validity (cf. Miller & Gazzaniga, 1998; Moritz, Woodward, & Rodriguez-Raecke, 2006), as virtually all traumatic incidents involve the visual sense. Moreover, differences in memory impairment in visual versus verbal material have been reported in PTSD (for meta-analysis see Brewin, Kleiner, Vasterling, & Field, 2007), emphasizing the need to investigate visual false memories in PTSD.

376

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

For the rst time, we investigated false memories in PTSD with a visual DRM paradigm using varying emotional content (Moritz et al., 2006). Moreover, condence ratings for recognition were included as well as measures of depression and dissociation as studies suggested that false memories are related to dissociation (e.g., Clancy et al., 2000; Hyman & Billings, 1998; Winograd et al., 1998) and depression (Brennen et al., 2007). Besides memory for critical lures, the DRM paradigm allows one to investigate the capacity of memory for presented items, which has been found to be impaired in PTSD (for meta-analysis see Brewin et al., 2007). We hypothesized that traumatized participants with PTSD (1) produce more visual false memories and (2) show impaired memory accuracy in comparison with traumatized participants without PTSD and non-traumatized participants. Third, we hypothesized that false memories are related to depression and dissociation. Finally, the study aimed to explore condence in false memories. 2. Methods 2.1. Participants For the present study 48 traumatized participants were recruited via the University Medical Center Hamburg-Eppendorf. Thirty-one were victims of assaults and 17 victims of (trafc) accidents. All of the events fullled the trauma criteria according to the DSM-IV and none of the participants had a history of psychotic symptoms (i.e. hallucinations, delusions, or mania), substance or alcohol dependence, or any neurological disorder including traumatic brain injury (exclusion criteria). Seventeen participants were diagnosed with full PTSD according to DSM-IV and three with subsyndromal PTSD, that is, DSM-IV criteria A, B, E, and F were met and either criterion C or criterion D (arousal or avoidance) was fullled, but not both (cf. Blanchard et al., 1996). For statistical purposes traumatized participants with full and partial PTSD were combined in one group (PTSD group). The remaining 28 traumatized participants did not meet criteria for full or partial PTSD (Non-PTSD group). Additionally, a non-traumatized healthy control group was recruited (Non-Trauma group, n 28) via an established subject pool with none of the participants having ever experienced a traumatic event according to DSM-IV criteria. After complete description of the study, written informed consent was obtained from all participants before assessment. The study was approved by the local ethics committee. 2.2. Measures 2.2.1. Psychopathology All traumatized participants were interviewed with the PTSD section of the Structured Clinical Interview for DSM-IV (SCID; First, Spitzer, Gibbon, & Williams, 1996) and completed the Post-traumatic Diagnostic Scale (PDS; Foa, Cashman, Jaycox, & Perry, 1997). The PDS is a self-report questionnaire that allows diagnosing PTSD according to DSM-IV criteria. It also assesses PTSD symptom severity and has demonstrated high reliability and validity (Foa et al., 1997). The MINI neuropsychiatric interview (Sheehan et al., 1998) was used to check for comorbid disorders in traumatized participants and to verify healthy psychiatric status in the Non-Trauma group. The Beck Depression Inventory (BDI; German version, Hautzinger, Bailer, Worall, & Keller, 1995) as well as the Hamilton Depression Rating Scale (HDRS; Hamilton, 1960) were used to quantify depression in all participants. The Dissociative Experiences Scale (DES; Bernstein & Putnam, 1986) was used to measure dissociation. The DES is a self-report questionnaire that comprises 28 items which tap into dissociative lifetime experiences (e.g., feelings of derealization, depersonalization). Participants rated the frequency of dissociative experiences on a scale from 1 (never) to 10 (very often) (cf. Tyler, Cauce, & Whitbeck, 2004). The total score was computed by the sum of the 28 items divided by the number of items. Moreover, a multiple choice vocabulary test (MWT-B; Lehrl, 1995) was employed to estimate verbal intelligence in participants. 2.2.2. False memory experiment For the assessment of false memories, a visual false memory paradigm was administered that was developed by Moritz et al. (2006) following the principles of the DRM paradigm (Miller & Gazzaniga,

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

377

1998; Roediger & McDermott, 1995). In this test, four pictures that each depict a prototypical scene (classroom, beach, funeral, room surveillance) are shown to participants on a computer screen (for material see Moritz et al., 2006 and Appendix A). While constructing the test, several details that were central to each scene were left out (e.g., beach towels in the beach scene) that later served as critical lures. The pictures were equal in terms of perceptual complexity but differed with regard to emotional valence (classroom scene neutral, beach positive, funeral and room surveillance negative). To check for idiosyncratic differences, participants rated the pictures with regard to valence (1 very positive up to 9 very negative) and arousal (1 very aroused up to 9 no arousal at all) by means of the Self Assessment Manikin (SAM; Lang, 1985). For recognition, 12 previously presented items (old items) were selected from each scene. Moreover, 12 distractor items that were not presented (new items) were compiled for each scene: four items unrelated to the scene (e.g., television in beach scene) and eight items representing critical lures (i.e. prototypical but not presented aspects of the scene, e.g., beach towel in beach scene). In the recognition trial, these items (in total 48 old and 48 new items) were presented in written form to participants. For false memory assessment, participants were seated approximately 20 inches from a 15-inch monitor and were instructed to carefully attend to four pictures that would be presented on the screen. They were told that they would later have to recognize details from each of the pictures. Then, the four scenes were presented each for 40 s in randomized order. While being presented, each picture lled out the entire computer screen. Before the recognition trial, participants were requested to complete a vocabulary test (see above) which took approximately 10 min. In the recognition trial, old and new items were presented in written form on the computer screen, one item at a time in randomized order but clustered for each scene. Additionally a contextual cue was given (classroom, beach, funeral, and room surveillance). To assess recognition and memory condence, a four-point scale was presented to participants: old and condent (1); old and not condent (2); new and condent (3); new and not condent (4) (cf. Moritz et al., 2006). 2.3. Statistical analyses An alpha level of .05 (two-tailed) was used for all statistical tests. Effect sizes (h2) were calculated for ANOVAs following Cohens (1988) conventions for small (.01 < h2  .09), medium (.09 < h2 < .25), and large (h2  .25) effects. 3. Results 3.1. Participants Groups did not differ on any major socio-demographic background variable. However, group differences occurred with regard to depression and PTSD symptoms (see Table 1). 3.2. Valence and arousal To check for differences regarding valence ratings, a mixed betweenwithin-subjects ANOVA with Scenario Type as within-subject factor and Group as between-subject factor was calculated. Valence ratings served as dependent variables (1 very positive up to 9 very negative). As expected, scenarios differed with regard to valence, F(3, 207) 74.08, p < .001, with the beach scene being rated as rather positive (M 2.61, SD 1.62), whereas the classroom scene was judged as neutral (M 3.50, SD 1.58) and the funeral (M 5.63, SD 1.54) and room surveillance scenarios (M 5.71, SD 1.86) were rated as more negative. Contrast analyses revealed that except for the difference between the funeral and the room surveillance scene, all valence ratings differed signicantly from each other (at least p < .01). Moreover, a signicant group effect emerged, F(2, 69) 4.41, p < .05, indicating that traumatized participants without PTSD generally rated scenarios more positive than non-traumatized

378

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

Table 1 Sociodemographic and psychopathological characteristics of the samples. Variables PTSD (P) (n 20) Age Gender (Female/Male) Education (years) Vocabulary test (verbal intelligence) BDI HDRS PDS DES
a

Non-PTSD (N) (n 28) 37.18 (13.85) 15/13 11.75 (1.48) 58.21 (8.59) 2.75 2.86 4.36 2.11 (3.33) (3.04) (4.76) (0.72)

Non-Trauma (T) (n 28) 40.82 (10.82) 17/11 11.21 (1.64) 59.36 (7.96) 4.32 (3.83) 2.75 (2.32)

Statisticsa

39.45 (13.80) 14/6 10.95 (1.73) 55.20 (11.45) 15.10 13.85 25.25 2.30 (6.43) (7.25) (11.07) (0.99)

F(2,73) 0.58; ns c2(2) 1.32; ns F(2,73) 1.58; ns F(2,73) 1.22; ns F(2,73) 49.42* (P > N, T) F(2,73) 47.12* (P > N, T) t(24.06) 7.93* t(46) 0.79, ns

Post-hoc test: Bonferroni; ns non-signicant; *p < .001.

healthy controls (Bonferroni corrected post-hoc tests, p < .05). The interaction between Group and Scenario Type, however, did not yield signicance, p > .3. For arousal ratings (1 very aroused up to 9 no arousal at all), the effect of Scenario type, F(3, 207) 10.47, p .001, was again signicant but neither Group (p > .2) nor the Group Scenario type interaction (p > .7) reached signicance. Numerically, the funeral scene (M 7.15, SD 1.79) was rated the most arousing followed by the room surveillance scene (M 7.19, SD 1.84), the beach scene (M 7.88, SD 1.48), and the classroom scene (M 8.17, SD 1.40). The only signicant difference emerged between the funeral and classroom scene (p < .05). 3.3. False memories and memory accuracy A mixed betweenwithin-subjects ANOVA was conducted with Group (PTSD, Non-PTSD, NonTrauma) as the between-subject factor and Scenario Type (classroom, beach, funeral, room surveillance) and Item Type (old, new-related, new-unrelated) as the within-subject factors. Percentage of old responses served as the dependent variable. While the main effect of Group did not achieve signicance, F(2, 73) 1.16, p .32; hpartial .03, the main effect of Item Type, F (2, 146) 823.32, p < .001, hpartial .92, was signicant, indicating that, as expected, more old responses were given for presented material than for non-presented material. Supporting the validity of our experiment, critical lures (new-related items) were more likely classied as old than unrelated distractor items (see Fig. 1). Moreover, a signicant main effect of Scenario Type was observed, F(3, 219) 8.64, p < .001, hpartial .11, indicating that the classroom and the funeral scene were remembered with more superiority. However, in these two scenes the percentage of false memories was also the highest. None of the interactions reached signicance (all ps > .18). For an in-depth investigation, hits and false memories (related and unrelated false memories) were combined into one score: the corrected recognition index. This score was calculated by subtracting the false memories (related and unrelated false memories) from the hits. Groups differed on trend level regarding this score, F(2, 73) 2.75, p .07; hpartial .07 (see Fig. 2). Exploratory single contrasts revealed that participants with PTSD performed worse on the corrected recognition index than participants without PTSD and non-traumatized participants (ps  .05), while participants without PTSD and non-traumatized participants did not differ (p > .8). Results regarding false memories and memory accuracy remained essentially the same when participants with sub-syndromal PTSD (n 3) were removed from the analyses. 3.4. Condence A mixed between-within-subjects ANOVA was conducted with Item Type (old, new-related, new unrelated) and Response Type (old, new) as within-subject factors, and Group as the between-subject

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

379

"Old"-responses in %

100 75 50 25 0

beach

surveillance

classroom

funeral

old
PTSD Non-PTSD Non-Trauma

"Old"-responses in %

100 75 50 25 0

beach

surveillance

classroom

funeral

new-related
PTSD Non-PTSD Non-Trauma

c
"Old"-responses in %

100 75 50 25 0

beach

surveillance

classroom

funeral

new-unrelated
PTSD Non-PTSD Non-Trauma

Fig. 1. (ac) Percentage of old responses in the PTSD, Non-PTSD, and Non-Trauma group with regard to Scene and Item Type: (a) old, (b) new-related, (c) new-unrelated.

factor. The percentage of high condent responses (i.e. 1 or 3) served as dependent variable. While groups were comparable with regard to memory condence, F(2, 73) 0.51, p .60, hpartial .01, signicant differences appeared with regard to Response Type, F(1, 73) 234.98, p < .001, hpartial .76, and Item Type, F(2, 146) 1059.73, p < .001, hpartial .94. Moreover, the Item Type Response Type interaction was signicant, F(2;146) 894.05, p < .001, hpartial .93, indicating that correct responses (i.e. old items that were rated as old and new items that were rated as new) were more likely to be made with high condence than incorrect responses. However, for new items this was only valid for unrelated items. Condence ratings for new related items (critical lures) were rather similar, no matter

380

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

60

Corrected recall ratio

40 PTSD Non-PTSD Non-Trauma 20

0
Fig. 2. Corrected recognition index (hits false positives) in the PTSD, Non-PTSD, and Non-Trauma group. For results see text.

whether they were correctly classied as new or incorrectly classied as old (see Fig. 3). This result was not further moderated by Group (ps > .19). 3.5. Relationships with psychopathology PTSD severity in the traumatized group correlated negatively at trend level with the number of correctly identied old-items (i.e. hits; r .25, p .09), but not with the extent of false memories (r .12, p > .4). Moreover, none of the experimental parameters correlated with the DES (rrsr < .17, ps > .14). Concerning self (BDI) and clinician (HDRS) assessed depression, signicant positive correlations were found between depression and false memories calculated over all participants (BDI: r .26, p < .05; HDRS: r .23, p < .05). When only calculated for traumatized participants this relation was non-signicant (rrsr  .20, ps > .16). 4. Discussion For the rst time, visual false memories were investigated with the DeeseRoedigerMcDermott (DRM) paradigm in trauma victims with and without PTSD and non-traumatized participants. Whereas

High confident responses in %

40 30 20 10 0

old

newrelated old

newunrelated

old

newrelated new

newunrelated

Resonse and Item Type

PTSD Non-PTSD Non-Trauma

Fig. 3. Percentage of high-condence ratings in the PTSD, Non-PTSD, and Non-Trauma group. For results see text.

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

381

PTSD participants displayed numerically more false memories than traumatized participants without PTSD and non-traumatized participants, this difference did not achieve statistical signicance and the effect size was rather small. Moreover, PTSD severity did not correlate with false memories and groups did not differ with regard to old studied items. At trend level, differences emerged when two memory scores (hits and false memories) were combined into one index. Moreover, greater forgetting correlated at trend level with PTSD severity. This shows that the current PTSD sample was prone to forgetting but not prone to false remembering. Until now, few studies have systematically investigated the prevalence of false memories in trauma victims. Moreover, the few studies that were carried out differ highly with regard to the trauma samples assessed. Bremner et al. (2000) and Brennen et al. (2007) both investigated victims of longlasting traumatic events, such as war and childhood sexual abuse. Our sample can be best compared with that of the study by Zoellner et al. (2000), who also investigated victims of single traumatic events (sexual or physical assault; our study also included victims of (trafc) accidents). The results of our study partly correspond with their study in that they found similar levels of false memories for neutral critical lures in traumatized subjects with and without PTSD. However, in their study differences concerning the production of false memories were revealed between traumatized and non-traumatized participants, which could not be found in our study and may be explained by differences in material (visual vs. verbal). A further aim of the current study was to investigate memory condence for false memories in PTSD. In line with previous ndings implementing the rememberknow paradigm (Brennen et al., 2007; Zoellner et al., 2000) the PTSD group did not display higher condence in falsely remembered critical lures. However, in the study by Brennen et al. (2007), although statistically non-signicant, tentative evidence suggested that PTSD patients only display more remember answers for traumarelated critical lures. Additionally, we aimed to investigate the relationship between false memories and depression (e.g., Brennen et al., 2007) as well as dissociation (e.g., Clancy et al., 2000). In line with our hypothesis, we found a positive relation between false memories and depression when calculated across all groups. However, when only traumatized participants were considered, this relation was reduced to nonsignicance. While Brennen and colleagues argue that depression and not PTSD symptomatology may be the driving force behind the production of false memories (Brennen et al., 2007, p. 883), our results do not fully conrm this result. However, Brennen et al. (2007) used a larger traumatized sample. Regarding dissociation, no relation was revealed for false memories in the trauma sample, which may be due to sample selection as a relation between dissociation (as a personality trait) and false memories has been reported before with correlations ranging from r .28 to .48 (e.g., Clancy et al., 2000; Hyman & Billings, 1998; Winograd et al., 1998). While the relation between dissociation and false memories has not been conrmed in all studies (e.g., Bremner et al., 2000; Geraerts, Smeets, Jelicic, van Heerden, & Merckelbach, 2005), dissociation is proposed to serve as a mediator between PTSD and increased false memories (see introduction). Thus, the lack of group differences in dissociation scores (see Table 1) may be a potential explanation for our negative ndings. While there is evidence for a more general impairment of declarative memory functioning in PTSD (Brewin et al., 2007), our groups did not differ with regard to studied items. This result might be due to a more pronounced impairment of verbal than visual memory in PTSD which has been claimed in a meta-analysis by Brewin et al. (2007). The present study holds several strengths as well as limitations. We would like to emphasize that for the rst time a visual measure for the production of false memories was employed in PTSD. As outlined above, the existing evidence is exclusively based on word stimulus material. However, differences for the impairment of verbal versus visual memory are evident in PTSD (see meta-analyses, Brewin et al., 2007). Therefore, it is important to further expand false memory research in PTSD to visual material. Additional advantages of a visual paradigm include strong contextual cues in the recognition phase and the economic administration (i.e. several critical lures can be included in one picture). At the same time, some limitations are also noteworthy. Firstly, our PTSD sample only comprised 17 participants suffering a full-blown PTSD and three participants diagnosed with subsyndromal PTSD (according to Blanchard et al., 1996). However, when these three participants were removed from the analysis, our results did not change and the sample size is still competitive with the

382

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

study by Bremner et al. (2000) (sample sizes between 11 and 23) and Zoellner et al. (2000) (sample size n 14 for each group). Secondly, our ndings are limited by the fact that we did not integrate trauma-related material in our design. Brennen et al. (2007) were able to show with a verbal version of the DRM paradigm that participants with PTSD falsely recalled more trauma-related lures than traumatized participants without PTSD, while both groups (similar to the results of our study) were equally error-prone to critical lures that were not trauma related. In the current study, apart from neutral material emotional material (e.g., funeral scene) was implemented and its emotional valence was conrmed by participants ratings. However, no interaction between Group and Scenario Type occurred. A potential explanation could be that our material in contrast to the material used by Brennen et al. (2007) was not specically more arousing for PTSD patients as indicated by their ratings. As stated in the introduction, arousal may represent a moderating variable for the production of false memories (Corson & Verrier, 2007; Moritz, Voigt, Arzola, & Otte, 2008) and might also explain the results of more falsely recalled trauma-related critical lures in PTSD (cf. Brennen et al., 2007). Future studies implementing traumarelated material and using more objective measures of arousal (i.e. heart rate) could shed further light on this issue. Finally, to improve our understanding of memory functioning in PTSD, false memory research should choose material that more closely resembles autobiographical experiences of trauma victims. Ecologically valid false memory paradigms may prove useful to provide indices for the validity of eyewitness testimonies, which are often requested from trauma survivors. Acknowledgement We would like to thank Marit Hauschildt for her support with the manuscript. Appendix A Example of memory material (beach scene; for other scenes see Moritz, Woodward, & RodriguezRaecke, 2006).

References
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. Bernstein, E. M., & Putnam, F. W. (1986). Development, reliability, and validity of a dissociation scale. Journal of Nervous and Mental Disease, 174, 727735.

L. Jelinek et al. / J. Behav. Ther. & Exp. Psychiat. 40 (2009) 374383

383

Blanchard, E. B., Hickling, E. J., Buckley, T. C., Taylor, A. E., Vollmer, A., & Loos, W. R. (1996). Psychophysiology of posttraumatic stress disorder related to motor vehicle accidents: replication and extension. Journal of Consulting and Clinical Psychology, 64, 742751. Bremner, J. D., Shobe, K. K., & Kihlstrom, J. F. (2000). False memories in women with self-reported childhood sexual abuse: an empirical study. Psycholgical Science, 11, 333337. Bremner, J. D., Southwick, S., Brett, E., Fontana, A., Rosenheck, R., & Charney, D. S. (1992). Dissociation and posttraumatic stress disorder in Vietnam combat veterans. American Journal of Psychiatry, 149, 328332. Brennen, T., Dybdahl, R., & Kapidzic, A. (2007). Trauma-related and neutral false memories in war-induced Posttraumatic Stress Disorder. Consciousness and Cognition, 16, 877885. Brewin, C. R., Kleiner, J. S., Vasterling, J. J., & Field, A. P. (2007). Memory for emotionally neutral information in posttraumatic stress disorder: a meta-analytic investigation. Journal of Abnormal Psychology, 116, 448463. Clancy, S. A., Schacter, D. L., McNally, R. J., & Pitman, R. K. (2000). False recognition in women reporting recovered memories of sexual abuse. Psycholgical Science, 11, 2631. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Corson, Y., & Verrier, N. (2007). Emotions and false memories: valence or arousal? Psychological Science, 18, 208211. Dancu, C. V., Riggs, D. S., Hearst-Ikeda, D., Shoyer, B. G., & Foa, E. B. (1996). Dissociative experiences and posttraumatic stress disorder among female victims of criminal assault and rape. Journal of Traumatic Stress, 9, 253267. Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology: Learning, Memory, and Cognition, 58, 1722. DePrince, A. P., & Freyd, J. J. (2004). Forgetting trauma stimuli. Psychological Science, 15, 488492. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1996). Structured clinical interview for DSM-IV axis I disorders. Washington, DC: American Psychiatric Press, Inc. Foa, E. B., Cashman, L., Jaycox, L., & Perry, K. (1997). The validation of a self-report measure of posttraumatic stress disorder: the Posttraumatic Diagnostic Scale. Psychological Assessment, 9, 445451. Geraerts, E., Smeets, E., Jelicic, M., van Heerden, J., & Merckelbach, H. (2005). Fantasy proneness, but not self-reported trauma is related to DRM performance of women reporting recovered memories of childhood sexual abuse. Consciousness and Cognition, 14, 602612. Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry, 23, 5662. Hautzinger, M., Bailer, M., Worall, H., & Keller, F. (1995). Beck-Depressions-Inventar (BDI) [Beck Depression Inventory]. Bern: Huber. Hyman, I. E., & Billings, F. J. (1998). Individual differences and the creation of false childhood memories. Memory, 6, 120. Lang, P. J. (1985). The cognitive psychophysiology of emotion: anxiety and the anxiety disorders. Hillsdale, NJ: Lawrence Erlbaum. Lehrl, S. (1995). Mehrfachwahl-Wortschatz-Intelligenztest: MWT-B [Multiple choice vocabulary test: MWT-B] (3rd ed.). Balingen: PERIMED-spitta. Loftus, E., & Davis, D. (2006). Recovered memories. Annual Review of Clinical Psychology, 2, 469498. McNally, R. J. (2003). Remembering trauma. Cambridge, MA: Harvard University Press. McNally, R. J., Ristuccia, C. S., & Perlman, C. A. (2005). Forgetting of trauma cues in adults reporting continuous or recovered memories of childhood sexual abuse. Psychological Science, 16, 336340. Miller, M. B., & Gazzaniga, M. S. (1998). Creating false memories for visual scenes. Neuropsychologia, 36, 513520. Moritz, S., Voigt, K., Arzola, M., & Otte, C. (2008). When the half full glass is appraised as half empty and memorized as completely empty: mood-congruent true and false recognition in depression is modulated by salience. Memory, 16, 810820. Moritz, S., Woodward, T. S., & Rodriguez-Raecke, R. (2006). Patients with schizophrenia do not produce more false memories than controls but are more condent in them. Psychological Medicine, 36, 659667. Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2003). Predictors of posttraumatic stress disorder and symptoms in adults: a meta-analysis. Psychological Bulletin, 129, 5273. Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803814. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., et al. (1998). The MINI International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview. Journal of Clinical Psychiatry, 59(suppl. 20), 2223. Stadler, M. A., Roediger, H. L., & McDermott, K. B. (1999). Norms for word lists that create false memories. Memory & Cognition, 27, 494500. Tulving, E. (1985). Memory and consciousness. Canadian Psychologist, 26, 112. Tyler, K. A., Cauce, A. M., & Whitbeck, L. (2004). Family risk factors and prevalence of dissociative symptoms among homeless and runaway youth. Child Abuse & Neglect, 28, 355366. Winograd, E., Peluso, J. P., & Glover, T. A. (1998). Individual differences in susceptibility to memory illusions. Applied Cognitive Psychology, 12, 527. Zoellner, L. A., Foa, E. B., Brigidi, B. D., & Przeworski, A. (2000). Are trauma victims susceptible to false memories? Journal of Abnormal Psychology, 109, 517524.