Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Predictors of analogue intrusive thoughts : a structural equation model Regambal, Marci Joan 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-ubc_2006-0634.pdf [ 1.95MB ]
JSON: 831-1.0092786.json
JSON-LD: 831-1.0092786-ld.json
RDF/XML (Pretty): 831-1.0092786-rdf.xml
RDF/JSON: 831-1.0092786-rdf.json
Turtle: 831-1.0092786-turtle.txt
N-Triples: 831-1.0092786-rdf-ntriples.txt
Original Record: 831-1.0092786-source.json
Full Text

Full Text

PREDICTORS OF A N A L O G U E INTRUSIVE THOUGHTS: A S T R U C T U R A L EQUATION M O D E L by M A R C I JOAN R E G A M B A L B.Sc , University of Calgary, 2001 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF ARTS in THE F A C U L T Y OF G R A D U A T E STUDIES (Psychology) THE UNIVERSITY OF BRITISH C O L U M B I A August 2006 © Marci Joan Regambal, 2006 Abstract Structural equation modeling was used to examine the prediction of intrusive thoughts from the theoretically proposed pathways in the Ehlers and Clark model (Behav. Res. Ther., 38 (2000) 319), and to examine the interface of emotion and preexisting vulnerability factors with the cognitive processes considered central to the model. Using a prospective design, 148 undergraduate students viewed a distressing film and recorded their film-related intrusive thoughts over the following week. The results indicated that a preexisting emotional vulnerability factor (i.e., trait anxiety and depression), predicted dysfunctional processing (i.e., dissociation and data-driven processing). Furthermore, emotional reactivity and dysfunctional coping strategies (i.e., rumination, thought suppression, and the use of safety behaviours) directly predicted intrusive thoughts. The observed model was consistent with Ehlers and Clark's proposed model, placing it in a broader context. Implications for the interface of these factors are discussed. i i Table of Contents Abstract i i Table of Contents i i i List of Tables v List of Figures vi Acknowledgements • vii Dedication viii Introduction. : - I Method 8 Overview 8 Participants 8 Materials 9 Measures • 9 Procedure 12 Data Analytic Approach 13 Results 13 Assumptions • 13 Descriptive Statistics • 14 Measurement Model 14 Structural Model 15 Direct and Indirect Effects 16 Discussion • 17 Conclusions 23 Tables 25 Figures 30 Footnotes 33 References 34 Appendices 42 Appendix A : Emotional Reactivity Questionnaire 42 Appendix B: Data-driven Processing Scale. 43 Appendix C: Response to Intrusions Questionnaire 44 Appendix D: Safety Behaviours Questionnaire 45 Appendix E: Intrusion Diary 46 Appendix F: Experimental Checks < 48 Appendix G: U B C Research Ethics Board Certificate of Approval 49 iv List of Tables Table 1 Descriptive statistics and bivariate correlations among model variables. 25 Table 2 Final structural model factor loadings for latent variables .26 Table 3 Model modification statistics .27 Table 4 Factor correlations 28 Table 5 Total, direct and indirect effects on intrusive thoughts 29 v List of Figures Figure 1 Predicted model : 30 Figure 2 Model 1: Initial structural equation model 31 Figure 3 Model 2: Final structural equation model 32 v i Acknowledgements I owe particular thanks to my supervisor Dr. Lynn Alden for with every aspect of this thesis, including her time, thoughtful feedback and patient questioning to help me think more deeply. I thank my committee, Dr. Peter Graf and Dr. Wolfgang Linden, for their constructive feedback. I also thank Dr. Jeremy Biesanz for his valuable feedback and direction with the statistical analysis. I would also like to thank Julie Kim and Lacey Baker for their dedication and hard work with data collection. Finally, I thank Ben Olivier for his endless support and help with this thesis. Dedication My Husband and Parents Introduction Posttraumatic stress disorder (PTSD) is a clinical reaction to a traumatic event. According to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) (American Psychiatric Association, 1994), PTSD involves three sets of symptoms: re-experiencing the event in the form of intrusive memories, nightmares or flashbacks; behavioral avoidance of trauma stimuli; and emotional numbing, and ongoing arousal. A key feature of PTSD is intrusive memories of the trauma. Early theories conceptualized such intrusive thoughts as an emotional memory phenomenon (for review see Witvliet, 1997). Drawing heavily from those theories, contemporary writers underscore the role of cognitive processes in the onset and maintenance of cognitive intrusions and other PTSD symptoms (Brewin, Dalgleish, & Joseph, 1996; Ehlers & Clark, 2000). Ehlers and Clark (2000), for example, proposed a model in which cognitive processing during the traumatic event, and the behavioural and cognitive responses adopted to reduce distress after the event, are crucial contributors to intrusions. The purpose of the current study was to examine the theoretically proposed pathways in the Ehlers and Clark model and to examine the interface of emotion and preexisting vulnerability factors with the cognitive processes considered central to the model. Ehlers and Clark postulated that under certain conditions (e.g., high arousal), events are more likely to be processed in terms of sensory information rather than meaning (i.e., dysfunctional processing), resulting in memories that are poorly elaborated and contextually integrated in terms of time, place, and related information (see also Brewin et al., 1996). According to Ehlers and Clark, sensory encoded memories are more likely to be unintentionally triggered by sensory cues reminiscent of the event. In addition, the persistence of PTSD symptoms, including intrusive thoughts, is made more likely by the use of dysfunctional post-1 trauma coping strategies, such as suppression, rumination, and avoidance, which are adopted to reduce distress. Such strategies are hypothesized to maintain symptoms by preventing changes in the trauma memory that would allow it to be integrated into autobiographical memory. Thus, Ehlers and Clark have proposed a clear sequence of events that lead to ongoing intrusive thoughts. The Ehlers and Clark model's emphasis on dysfunctional processing and encoding is in line with other contemporary theories of PTSD (e.g., Brewin et al., 1996; Foa & Rothbaum, 1998). A growing body of studies has examined the relationship between dysfunctional processing and PTSD. In particular, many theorists consider dissociation to be a marker of dysfunctional processing. Dissociation is defined by The American Psychiatric Association (APA) as a "disruption of the usually integrated functions of consciousness, memory, identity or perception of the environment" (APA, 1994, p. 822), and is measured by self-reported alterations in cognitive and perceptual functioning. Individuals frequently report experiencing dissociation during traumatic events, with one of the more common sensations being time distortion (Shalev, Peri, Canetti, & Schreiber, 1996; Ursano et al. 1999). A large body of literature has linked peritraumatic dissociation to the development and maintenance of PTSD and intrusive thoughts (Birmes et al., 2003; Clohessy & Ehlers, 1999), even when controlling for trait dissociation (Ehlers, Mayou, & Bryant, 1998; Holmes, Brewin, & Hennessy, 2004; Murray, Ehlers, & Mayou, 2002) and prior PTSD diagnosis (Ursano et al., 1999; see, however, Marshall & Schell, 2002; Shalev et al., 1998). Recently, Engelhard and her colleagues supported the assertion that dissociation may be responsible for the processing of sensory and emotional impressions of the event, which later characterizes the trauma memory (Engelhard, van den Hout, Kindt, Arntz, & Schouten, 2003). 2 To better operationalize dysfunctional cognitive processing, Ehlers and Clark drew on research in cognitive psychology to develop a measure they called data-driven processing (Halligan, Michael, Clark, & Ehlers, 2003). Data-driven processing is said to be characterized by sensory driven processing, as opposed to meaning-based processing. However, research examining the link between data-driven processing and intrusive thoughts has been scarce and has yielded mixed results. Halligan, Clark and Ehlers (2002) attempted to manipulate data-driven and conceptual processing (i.e., processing the meaning of the event) in an analogue study, but found no between-group differences in PTSD "symptoms", including intrusive thoughts. Those authors then repeated the experiment, selecting participants with high trait tendencies to engage in conceptual or data-driven processing. Compared to the trait conceptual processing group, the data-driven group demonstrated greater avoidance, arousal, and intrusive recollections, which was consistent with the Ehlers and Clark model. Further evidence linking data-driven processing and PTSD symptoms was found in a sample of motor vehicle victims (Murray et al., 2002). Overall, more research relating dysfunctional processing to intrusive cognitions is needed to firmly establish that link. The second tenet of the Ehlers and Clark model is the assertion that the link between dysfunctional cognitive processing and ongoing PTSD symptoms is dependent, in part, upon the use of dysfunctional coping strategies, such as suppression, rumination and safety behaviours. Although those strategies are adopted to control symptoms, research indicates that they actually perpetuate them. For example, a number of studies have demonstrated a paradoxical increase in intrusive thoughts following attempts to suppress them (e.g., Wegner, 1994), and support exists for this effect in the context of PTSD (Davies & Clark, 1998a; Davies & Clark, 1998b; Ehlers et al., 1998; Harvey & Bryant, 1998). Rumination about the trauma has also been associated with 3 P T S D severity (Clohessy & Ehlers, 1999; Laposa & Alden, 2003; Mayou, Ehlers, Bryant, 2002). Ehlers and Clark proposed that the "what i f nature of rumination (e.g., what i f I would have left the house five minutes earlier?) may prevent the trauma memory from being integrated into autobiographical memory. Finally, Ehlers and Clark postulated that safety behaviours maintain P T S D symptoms. These behaviours (e.g., vigilance to threat cues) are meant to prevent or minimize further "catastrophes" associated with the trauma. Little research has explored the role of safety behaviours in the context of P T S D , however, steps taken to feel safe by assault victims (e.g., sleeping with lights on, carrying a weapon) were associated with greater P T S D severity both in the short (i.e., 6 months post-assault) and long term (i.e., 9 months post-assault) (Dunmore, Clark, & Ehlers, 2001). Safety behaviours can also include complete avoidance of trauma reminders, and the role of avoiding trauma reminders in maintaining P T S D is well documented (Bryant & Harvey, 1995; Charlton & Thompson, 1996; Dunmore et al., 2001). In summary, there is consistent support for the hypothesized role o f dysfunctional post-trauma coping strategies in maintaining P T S D . In addition to the cognitive processes described above, contemporary P T S D models acknowledge the contribution of preexisting psychological and emotional vulnerabilities to the onset of P T S D , although those variables receive less emphasis. Preexisting emotional vulnerabilities have been shown to be associated with P T S D development and severity in clinical samples. Two well established vulnerability factors are depression (e.g., Breslau, Davis, Peterson, & Schultz, 1997; Ozer, Best, Lipsey, & Weiss, 2003; Ursano et al., 1999) and trait anxiety (e.g., Frommberger et al., 1998; Tampke & Irwin, 1999). However, little research has examined how preexisting emotional vulnerabilities may exert their influence on the development of P T S D . 4 Ozer and her colleagues concluded in a recent meta-analysis of PTSD predictors that there is a need determine the proximal predictors (e.g., dissociation and peritraumatic emotionality) that explain the statistical relationships between the distal predictors (e.g., depression and trait anxiety) and PTSD (Ozer, et al., 2003). Dysfunctional processing and dysfunctional coping strategies are both proximal predictors from the Ehlers and Clark model that could explain the relationship between preexisting emotional vulnerabilities and PTSD severity, and more specifically the frequency of intrusive thoughts. In an analogue trauma study, trait anxiety was associated with higher levels of trait data-driven processing during a distressing movie (Halligan et al., 2002). Furthermore, in a sample of assault victims, depression measured shortly after an assault correlated with peritraumatic data-driven processing and persistent dissociation (Halligan et al., 2003). In contrast, however, one prospective analogue study failed to find a direct relationship between depression and intrusive thoughts (Davies & Clark, 1998a). Preexisting vulnerability factors may also affect post-event coping. Suppression is a common form of mental control in adults with depression and in those at risk of developing depression (Wegner & Zanakos, 1994; Wenzlaff & Bates, 1998). Furthermore, trait anxiety is associated with a general tendency to suppress thoughts (Muris, Merckelbach, Horselenberg, 1996). Rumination is also a characteristic of both depressive (Kuehner & Weber, 1999; Nolen-Hoeksema, 2000) and anxious thinking (Muris, Roelofs, Rassin, Franken, & Mayer, 2005). Overall, the process through which preexisting factors increase vulnerability to PTSD, and the effect of vulnerability factors on the other elements of the cognitive model, requires further study. Most theorists, as well as the DSM-IV, emphasize the importance of individuals' emotional reactions to the traumatic situation as a key factor in PTSD. Responding to the event 5 with fear, helplessness and horror are required for a PTSD diagnosis. In addition, there is now evidence to suggest that other emotions, such as sadness, anger and shame, are experienced and associated with the trauma features that are later reexperienced (Holmes, Grey, & Young, 2005). According to earlier emotional theories, heightened levels of positive or negative emotions increases sympathetic nervous system responding and thus influences what is remembered (Witvliet, 1997). Furthermore, there is a large body of evidence supporting the role of emotional arousal in the enhanced memory consolidation of emotional events (Cahill & McGaugh, 1998). A pioneering trauma analogue study conducted by Davies and Clark (1998a) observed that mood change (i.e., change in anxiety, depression, happiness and anger) proved to be the strongest and most consistent predictor of movie-related intrusive thoughts. Cognitive models acknowledge the potential for higher emotional reactivity (i.e., state emotion) to influence both the quality of processing and the content of intrusive memories. In one study, it was found that individuals higher on trait data-driven processing were higher on state anxiety, but that they did not differ from individuals lower on trait data-driven processing on increases in anxiety during the video (Halligan et al., 2003). However, the influence of peritraumatic emotional reactivity has not been consistently explored in the context of contemporary cognitive models. The goal of the current study was to use structural equation modeling (SEM) to simultaneously examine the associations between the various elements of contemporary . cognitive models in the context of a laboratory trauma analogue task. The complexity of the relationships among those components, and the incorporation of emotional reactivity, make S E M an ideal statistical procedure to address the current research questions. A few authors have applied S E M to causal models of preexisting vulnerabilities and the development of PTSD in war veterans (King, King, Foy, & Gudanowski, 1996; King, King, Fairbank, Keane, & Adams, 6 1998; Vuksic-Mihaljevic, Mandic, Bensic, & Mihaljevic, 2000). However, those studies did not include the cognitive processes implicated in contemporary PTSD models, and given the methodological challenges of prospective measurement, have relied on retrospective reports. The laboratory analogue methodology allows for the truly prospective measurement of preexisting factors and the direct and immediate assessment of peritraumatic variables. This methodology allows stronger causal conclusions to be drawn from S E M methods. Another advantage of analogue studies is that the nature of the distressing event is precisely controlled, resulting in each person being exposed to the exact same stimulus. The hypothesized model, which is based on Ehlers and Clark's cognitive theory, is found in Figure 1. This model addressed hypotheses one two and three: 1) A preexisting emotional vulnerabilitiy (i.e., depression and trait anxiety) will directly increase a) emotional reactivity b) levels of dysfunctional processing (i.e., data-driven processing and dissociation) and c) dysfunctional post-trauma coping strategies (i.e., thought suppression, rumination and safety behaviours). 2) Greater emotional reactivity will result in more dysfunctional processing. 3) Higher levels of dysfunctional processing and dysfunctional coping strategies will be associated with a greater frequency of intrusive thoughts. 4) We 1 were uncertain i f dysfunctional processing would completely mediate the affect of emotional reactivity on intrusion frequency. Therefore, following the recommendation by MacCallum, Roznowski, & Necowitz (1992), we proposed a second a priori model rather than rely solely on modification statistics, which can capitalize on chance and not replicate. In this model, hypotheses 1-3 remained the same with the addition of a direct pathway from emotional reactivity to intrusion frequency. 7 Method Overview Participants completed measures of depression, trait anxiety and emotional state, and then watched a short movie of a medical event. After the movie, participants completed measures of their cognitive processing during the movie and reassessed their emotional state. Over the next week, participants recorded any spontaneous thoughts they had about the movie. When participants returned to the lab, they completed questionnaires measuring their reactions to their spontaneous thoughts. Participants Participants consisted of undergraduate students who were enrolled in at least one psychology course at the University of British Columbia. Participants received class credit for their participation. There were three exclusion criteria: 1) current mental health treatment, 2) recent experience (i.e., within the last 6 months) with a car accident 3) recent death of a relative or friend due to a car accident. The exclusion criteria were meant to exclude any participants who may be adversely affected by the video, due to potential car-accident related trauma symptoms, and to ensure participants' responses during the study were not a result of a past car accident. 159 participants completed session one. Five participants failed to complete session two and one participant was excluded because he failed to watch the entire movie, resulting in 153 participants who completed both sessions. From this sample, five participants were classified as multivariate outliers and were subsequently removed from the analysis. The resulting sample size for the analysis was 148 participants. 8 Participants were primarily single (73%), female (70%) and had completed an average of 1.86 years of university (SD=1.19). The average age was 20.23 years (SD=2.90). The most common cultural backgrounds consisted of European Canadian (33%), Indo Canadian (23%) and Asian Canadian (19%). In the past, 7% of participants had been in a serious car accident and 7% had a friend or relative die in a car accident. Materials Participants viewed a nine-minute film of graphic real-life footage depicting a car accident patient undergoing medical treatment and dying from his injuries. This video was used in a previous analogue study and was shown to produce intrusive thoughts (Laposa & Alden, 2006). Participants viewed the film on a 27-inch television screen from a distance of approximately 5 feet from the screen. Measures Preexisting emotional vulnerabilities. Vulnerability was assessed with measures of depression and trait anxiety. Depression was assessed with the Beck Depression Inventory-II (BDI; Beck, Steer, & Brown, 1996). The BDI consists of 21 items, each rated on a scale ranging from 0 to 3, with higher scores indicating greater severity. The BDI has high internal consistency (e.g., a = .92) and test-retest reliability (r =. 93) in student and outpatient populations (Beck, et al., 1996). Cronbach's alpha in this sample was 0.86. The State Trait Anxiety Inventory- Trait subscale (STAIT) was used to assess trait anxiety (Speilberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983). The STAIT consists of 20 items rated on a 4-point scale in relation to various anxiety symptoms. The STAIT has good internal consistency (a = .86 - .95), test-retest reliability (r =. 71-.86) and convergent validity (Antony, Orsillo, & Roemer, 2001). Chronbach's alpha for this sample was .92. 9 Emotional reactivity. Four measures were used to assess emotional reactivity during the film. Current levels of depression, happiness and anger were assessed using visual analogue scales (see Appendix A). Congruent with Davies and Clark (1998a), participants were asked to rate each emotion "at this moment" on scales ranging from 0 (not at all) to 100 (extremely). State anxiety was measured using the State Trait Anxiety Inventory- State subscale (STAIS; Spielberger et al., 1983). The STAIS has been shown to have acceptable reliability and validity (Speilberg et al., 1983). Cronbach's alpha in this sample was .93. Dysfunctional processing. We included two measures of dysfunctional processing. Data -driven processing was assessed using the data-driven subscale of the Cognitive Processing Questionnaire (DDP; Ehlers, 1998). The DDP was developed through a series of studies with various types of trauma victims (e.g., Ehlers et al., 1998; Halligan et al., 2002; Murray et al., 2002). The score on the data-driven scale (eight items) reflects the degree to which an individual focused on sensory or surface information during a target event (see Appendix B). Items are scored on a 4-point Likert-type scale with a higher score reflecting more data-driven processing. The DDP has been shown to have good internal consistency (a =. 70) and predictive ability of PTSD in M V A victims (Halligan et al. 2002). In this sample, Chronbach's alpha was 0.78. Dissociation was assessed using the Peritraumatic Dissociation Experiences Questionnaire (PDEQ; Marmar, Weiss, & Metzler, 1997). The PDEQ assesses experiences such as depersonalization, derealization and altered sense of time during the critical incident. There are 10 items rated on a 5-point Likert-type scale. The PDEQ has shown internal consistency (a =. 81) and is strongly associated with other measures of general dissociation (Marmar et al., 1997). In this sample, the internal consistency was also adequate (a =0.84). 10 Dysfunctional coping strategies. Maladaptive coping strategies were assessed using three measures. Rumination and thought suppression was measured by the respective subscales from The Response to Intrusions Questionnaire (RIQ; Clohessy & Ehlers, 1999) (see Appendix C). Each subscale is rated on a seven-point Likert-type Scale. The rumination subscale consists of three items (a = .31), and the suppression subscale has four items (a =.73) (Laposa & Alden, 2006). Chronbach's alphas for the rumination and suppression subscales in this study were .69 and .64, respectively. Safety behaviours were measured using a three-item measure developed for this study (see Appendix D). This measure built on a previous measure of avoidance used by Laposa & Alden (2006). The revisions targeted more subtle forms of behavioural changes after viewing the film, as past research suggests participants are unlikely to completely avoid road-related situations as a result of viewing the film. The measure consisted of three questions which asked participants "since viewing the film last week did you notice an increase in": 1) discomfort associated with road-related situations 2) hesitancy surrounding road related situations and 3) vigilance while in a car. Participants rated their responses on 9-point Likert scales ranging from 0 (not at all) to 9 (very much). Chronbach's alpha for this sample was .93. Intrusive thoughts. The number of movie-related intrusive thoughts was measured using an intrusion diary (see Appendix E). The diary is a commonly used measure in analogue studies, and has been shown to correlate with well established questionnaire measures of intrusive recollections (e.g., Holmes et al., 2004; Laposa & Alden, 2006). Participants recorded the date and time for each spontaneous intrusive thought of the film, gave a description of the thought and indicated how much distress it caused them. These instructions were included with the diary. 11 The number of distressing intrusions (i.e., those rated as causing any distress) was summed to create an intrusion frequency total score. Experimental Checks Attention. Participants rated the amount of attention they paid to the film on a scale of 0 (not at all) to 10 (full attention) (cf. Laposa & Alden, 2006) (see Appendix F). Diary compliance. Participants were asked to rate the extent to which they forgot to record their intrusions on a scale of 0 (not at all true of me) to 10 (extremely true of me) (cf. Holmes et al., 2004) (see Appendix F). Prior frequency of medical footage exposure. Participants rated their tendency to watch medical footage on a scale of 0 (never) to 10 (almost every day) (cf. Laposa & Alden, 2006) (see Appendix F). Procedure Participants individually attended two sessions, one week apart. In the first session, they completed the preexisting emotional vulnerability measures of depression and anxiety, the STAIS and emotional reactivity questionnaire. Participants were then left alone to watch the film. Immediately after the film, participants then completed the dysfunctional processing questionnaires, STAIS and emotional reactivity questionnaire. Participants received verbal and written instructions on how to complete the diary. Specifically, they were instructed to only record thoughts that occurred when they did not intend to think about the film. During session two, participants completed the subsections of the RIQ and safety behaviours questionnaire. Participants were fully debriefed. 12 Data Analytic Approach Change scores for the individual emotional reactivity measures and STAIS were derived by taking the unstandardized residuals after regressing participants' post movie scores on their pre-movie scores. LISREL 8.72 (Joreskog & Sorbom, 2005) was used for all modeling procedures. Because of a small proportion of missing data, the raw data file was analyzed using full information maximum likelihood (FIML). FIML has been demonstrated to be superior to other missing data techniques (i.e., list-wise deletion and pairwise deletion) in determining parameter estimates, parameter estimate efficiency, model goodness of fit and reducing model convergence failures (Enders & Bandalos, 2001). Factors were scaled and the model identified by setting one indicator for each latent variable to a unit weight of 1.0. FIML provides two indices of model fit: the root mean square error (RMSEA) and chi-square (^2) statistic. Good model fit was defined by a R M S E A value less than .06 (Hu & Bentler, 1999), and a nonsignificant chi-square value. Results Assumptions Two main assumptions of FIML are multivariate normality and linearity (Enders, 2001). Multivariate normality was indirectly assessed by examining the univarite distributions. Test statistics revealed that one distribution was significantly skewed and kurtotic (p < .001). Five other distributions were significantly, but mildly, skewed. The most common way to address the impacts of nonnormality with a full data set (i.e., no missing data) is through a statistical correction. However, there is no readily available correction for data sets with missing data and multivariate nonnormality, and those available are still in their infancy and require further 13 assessment (e.g., Enders, 2001). Given the mildness of skew, no corrections were made to the univariate distributions . To address the assumption of linearity scatter plots were examined and there was no evidence of nonlinearity. Using Malahanobis distance (p < .001), five multivariate outliers were detected and deleted. Descriptive Statistics Table 1 displays the descriptive information and bivariate correlations among the observed variables. Participants experienced an average of 2.76 (SD = 2.69) intrusive thoughts over a one week period, with a range of 0-11. There were no significant correlations between participants' prior exposure to medical footage and any of the outcome measures. There was a small relationship between the residualized change score for depression and attention to the film, r (146) = .27, p < .05. Finally, there were small correlations with the degree to which participants forgot to record their intrusive thoughts in their diary and the following measures: thought suppression, r (147) =.17,p < .05; rumination, r (146) = .19, p < .05; safety behaviours, r (142) =.20, p <.05; and the BDI, r (147) = . .20, p < .05. Given the small magnitude of the associations, the analyses were conducted without any corrections. Measurement Model The development of the final structural model followed the two-step method outlined by Kline (2005). The two-step method is useful as it eliminates the factor structure as a source of model misfit when modifying the structural model. Following this method, the structural model is first specified as a confirmatory factor analysis in which the factors are allowed to freely correlate. If the measurement model provides an acceptable fit to the data, the structural model is 14 tested. There should be only small discrepancies between the factor loadings of an adequately fit measurement model and the final structural model. The measurement model contained four latent variables. The latent variable of preexisting emotional vulnerability was composed of the total scores on the BDI and STAIT. The latent variable of emotional reactivity consisted of the residualized change scores for happiness, depression and anger, as well as the residualized change of the total STAIS score. The latent variable of dysfunctional processing was composed of the total scores for the DDP and PDEQ. The latent variable of dysfunctional coping strategies consisted of the composite scores on the measures of rumination, thought suppression and safety behaviours. Finally, intrusions were represented by total number of distressing intrusive thoughts as recorded in the diary. A l l latent variables were considered causal for observed scores. To test the measurement model all variables were allowed to freely correlate with each other, and the covariance among the residuals was set to zero. The results for this model.suggested good fit, x2 (45) = 55.34,/) > .05, and RMSEA= .039, indicating that it was appropriate to test the more restrictive structural model. Congruent with an adequate fitting measurement model, the factor loadings were only slightly different from the values in the measurement model therefore, Table 2 contains the composition of the latent variables and factor loading information from the final structural model. Structural Model To examine the more restrictive structural model the variables were no longer allowed to freely correlate and were restrained to the hypothesized relationships (see Figure 1). Only marginal support was found for the hypothesized model, (49) = 100.45, p < .001, RMSEA=.084 (see Figure 2). To improve model fit we tested the second a priori hypothesis and 15 allowed emotional reactivity to have a direct effect on the frequency of intrusive thoughts, rather r than its effect to be completely mediated by dysfunctional processing. This modification resulted in a significant improvement in model fit, x2 (48) = 88.50, p < .001, R M S E A = .076 (see Table 3), however the model fit was not yet acceptable. Post hoc model modifications were performed in an attempt to develop a better fitting model. However, to reduce data-driven modifications a few restrictions were set: 1) due to the adequate fit of the measurement model, modification indices were not considered for factor loadings; 2) to maintain the causal nature of the model, pathways that did not follow the temporal order of data collection were not considered; 3) the model did not meet specifications to be empirically identified for bi-directional pathways. With these restrictions, a significant model misfit was identified by not allowing a direct pathway between dysfunctional processing and dysfunctional coping strategies. When this was accomplished, the final model fit was good, X2 (47) = 58.25,/) >.05, RMSEA= .04 (see Figure 3) and produced a significant increase in model fit (see Table 3). The correlations among the latent variables are presented in Table 4. Direct and Indirect Effects The direct and indirect effects on intrusive thoughts are summarized in Table 5. A preexisting emotional vulnerability was not significantly related to emotional reactivity. Consistent with the hypothesis, a preexisting emotional vulnerability was associated with greater levels of dysfunctional processing during the movie (P = .28, p < .05). Although a preexisting emotional vulnerability predicted dysfunctional coping strategies in Model 1, this was no longer significant when dysfunctional processing directly predicted dysfunctional coping strategies (i.e., Model 2). However, a preexisting emotional vulnerability exerted a significant indirect effect on dysfunctional coping strategies through dysfunctional processing (P = .20, p < .05). 16 Emotional reactivity ((3 = .32,/? < .05) and dysfunctional coping strategies ( P = .48, p < .05) were both significant predictors of the frequency of intrusive thoughts. Thus, individuals who displayed greater increases in negative emotions during the film experienced more frequent intrusive thoughts, as did people who engaged in dysfunctional coping strategies. Interestingly, dysfunctional processing during the movie was not a significant predictor of intrusive thoughts once emotional reactivity was allowed a direct pathway to intrusive thoughts. However, dysfunctional processing exerted an indirect effect on intrusive thoughts through dysfunctional coping strategies (P = .27, p < .05). Discussion We used structural equation modeling (SEM) to predict analogue distressing intrusive thoughts, by incorporating preexisting emotional vulnerabilities and peritraumatic emotional reactivity with cognitive processes central to the Ehlers and Clark cognitive model of PTSD. Of particular interest from the Ehlers and Clark model was peritraumatic dysfunctional cognitive processing and dysfunctional post-trauma coping strategies. To our knowledge, this is the first study to incorporate these four sets of variables in the prediction of intrusive thoughts. This incorporation established that cognitive variables have meaningful links to both preexisting vulnerabilities and emotional reactivity variables. Furthermore, all variables contributed, either directly or indirectly, to the development of intrusive thoughts. Specifically, a preexisting emotional vulnerability was associated with increased dysfunctional processing and, indirectly with the use of dysfunctional coping strategies. Increased emotional reactivity was related to increases in dysfunctional processing. In turn, dysfunctional processing was indirectly related to an increase in intrusive thoughts through dysfunctional coping strategies. The findings will be discussed below following the temporal sequence that underlies the proposed model. 17 Preexisting Emotional Vulnerabilities A body of literature suggests that negative affect (i.e., depression and anxiety) increases an individual's risk of developing PTSD, and more specifically trauma related intrusive thoughts. This study not only confirmed these results, but also elucidated the mechanism through which a preexisting emotional vulnerability contributes to the development of intrusive thoughts, namely dysfunctional processing. Furthermore, the truly prospective measure of preexisting vulnerabilities afforded by analogue research strengthened our confidence in this result. According to our model, general negative affect, as captured by the preexisting emotional vulnerability factor, results in certain types of dysfunctional cognitive processing. Traumatic and distressing events may exacerbate such dysfunctional cognitive processing, which, in turn, furthers the severity of the trauma related symptoms, such as intrusive thoughts. This result has important implications for the potential screening of individuals who may be at particular risk for developing PTSD. However, this result needs to be confirmed in clinical samples. The standard practice in the PTSD literature has been to explore the role of anxiety and depression separately in the development of PTSD. One limitation to this approach is that any common etiology or mechanism between the two constructs is lost. In this study, S E M allowed for the advantage of combining these two constructs and our results indicated that both constructs were readily viewed as markers of a latent emotional vulnerability variable. That finding is consistent with earlier research on tripartite models of negative affect, which indicated that depression and anxiety share a nonspecific core feature of general distress, or negative affect (e.g., Mineka, Watson, & Clark, 1998), which this study suggests increases vulnerability to intrusive recollections of distressing information. 18 The hypothesis that a preexisting emotional vulnerability would affect emotional reactivity was not supported. A few explanations may exist for this finding. First, is it possible that individuals higher on the factor of preexisting emotional vulnerability have a higher baseline of emotional arousal and therefore, the magnitude of change in negative affect is not as large as individuals lower on this variable. Second, individuals without a higher baseline of general negative affect may rate their emotional change differently. These individuals may rate the change more accurately than individuals with higher negative affect, or exaggerate the change since it is different from their typical emotional state. In summary, these results helped to clear up a point of confusion in the literature, which is how preexisting emotional vulnerabilities are related to the development of distressing intrusive thoughts. In line with Ozer et al.'s conceptualization of PTSD predictors, it is likely a preexisting emotional vulnerability is a distal predictor that interacts with the proximal predictor of dysfunctional processing. Emotional Reactivity Most theorists conclude that levels of emotion are an important determinant of PTSD symptoms. Our results were consistent with such assertions and past research (Davies & Clark 1998a), as emotional reactivity was strongly associated with dysfunctional processing and directly related to the frequency of intrusive thoughts. One explanation for the direct pathway from emotional reactivity to intrusive thoughts is the enhanced retrieval of emotionally relevant material. Ehlers and Clark proposed two mechanisms of retrieval that may influence the frequency of intrusive thoughts. The first mechanism of retrieval involves the implicit memory process of priming, "the facilitated identification of perceptual objects as a result of prior exposure to those objects" (Ehlers, 19 Michael, Chen, Payne, & Shan, 2006, p. 317). Clinical populations have demonstrated enhanced priming for trauma related material (Michael, Ehlers, & Halligan, 2005), and analogue research has supported enhanced priming for degraded analogue stimuli (Ehlers et al., 2006). In both populations, priming predicted the frequency of intrusive thoughts. Emotion is thought to enhance priming through its activation of the amygdala. The amygdala can discern which stimuli are most threatening, and anticipate fearful stimuli based on only fragmentary cues or sensory stimuli (Joseph, 1997). The second process of intrusive memory retrieval proposed by Ehlers and Clark is associated learning- the process through which cues perceptually similar to trauma stimuli trigger both the trauma memory and the affective response during the trauma. Analogue research has supported the role of associated learning in the development of intrusive thoughts (Ehlers et al. 2006). Once again, the amygdala is postulated to play a vital role in this process. Elzinga & Bremner (2002) suggested that while witnessing a traumatic event there is a rapid reaction and release of adrenaline, which acts on the amygdala producing enhanced consolidation of the event. The result of this reaction is a strengthened emotional memory trace and fear conditioning that automatically triggers the autonomic, emotional and behavioural responses associated with the trauma. Ehlers and Clark proposed that it is the reexperiencing of those strong affective responses that contribute to the distressing nature of the intrusive thoughts. This is in line with the finding that distressing intrusions are more persistent than non-distressing intrusions (Michael, Ehlers, Halligan, & Clark, 2005). Thus, emotional reactivity may play a vital role in the retrieval of traumatic memories and our results indicate this is plausible explanation. However, our results suggested that 20 emotional reactivity also indirectly affects the frequency of intrusions through dysfunctional processing. Cognitive Variables: Dysfunctional Processing Ehlers and Clark emphasized the influence of dysfunctional processing (i.e., quality of encoding) in the development of the trauma memory and, more specifically, intrusive thoughts. They also hypothesized that dysfunctional processing is influenced by state factors (i.e., degree of arousal and fear). This study underscored the relationship between general emotional reactivity and dysfunctional processing. Furthermore, the importance of dysfunctional processing in the development of "long-term" intrusions was elucidated by the indirect effect of dysfunctional processing on intrusions through dysfunctional coping strategies. Emotional reactivity may assist with certain memory processes, but the subjective experience of emotional arousal may cause some individuals to engage in dysfunctional processing. For example, some writers postulate that dissociation is a "hardwired response" for dealing with extreme anxiety, which combines increased alertness and the inhibition of emotion. During this time, an individual may intensely focus on external stimuli while attention is taken away from internal experiences. This has been termed "detachment" by some theorists (e.g., Holmes et al., 2005). This sense of being separated from the experience and environment, and narrowed attention, may suspend higher cognitive structures and deter the encoding of peripheral environmental information. Conceptual processing involves the encoding of information that helps to place the event into context, which decreases the likelihood the memory will be unintentionally retrieved by sensory cues (Ehlers & Clark, 2000). Thus, narrowed attention would detract from conceptual encoding, and contribute to data-driven processing. There is 21 speculation that the experience of dissociation, and associated reduction in emotional experience, results from the partial inhibition of the amygdala (Elzinga & Bremner, 2002). Since emotion has memory enhancing effects implicated in PTSD symptoms, it is unlikely that the entire emotional experience is suspended during dissociation. Furthermore, some writers assert that dissociation may be a transient experience during a traumatic event (Allen, Console, & Lewis, 1999) allowing for periods of time when emotion is felt more strongly. Sierra and Berrios (1998) put forth a neurobiological model of depersonalization to explain the concurrent experiences of emotional numbing/ reduction and vivid sensory processing. According to those authors, once a threshold of anxiety is reached, the medial prefrontal cortex inhibits the emotional processing of material by the amygdala. There is a reduced emotional experience and dampening of sympathetic output (e.g., reduced heart, skin conductance). The hypothesized result is qualitative changes in perception, cognition and feelings of reality. However, the authors suggest that the central nucleus of the amygdala is spared this inhibitory control, allowing for the hyperalertness and sensory processing that typically accompanies dissociation. This sensory processing can be conceptualized as a component of the dysfunctional processing variable proposed by this study. Although these results are promising, causal statements must be tempered because only the pre-movie emotion ratings were measured before the processing questionnaires, the post-movie emotion ratings were measured at the same time as the dysfunctional processing questionnaires. Cognitive Variables: Dysfunctional Coping Strategies Our results supported Ehlers and Clark's emphasis on post-event dysfunctional coping strategies in the development of intrusive thoughts. Furthermore, the prediction of dysfunctional 22 coping strategies from dysfunctional processing elucidated a series of events leading to long-term intrusive thoughts. Intrusive thoughts are a common reaction to witnessing distressing and traumatic events (Michael et al., 2005). The experience of these initial distressing sensory laden memories causes some people to engage in dysfunctional coping strategies (i.e., suppression and rumination), resulting in ongoing intrusions. Conversely, it is plausible that people who cope with initial distressing intrusions in a more functional way would experience fewer intrusions, even i f they engaged in dysfunctional processing during the trauma. For these people, there may still be a significant direct relationship between dysfunctional processing and intrusions, especially at symptom onset. For example, Dunmore et al. (2001) found that after controlling for initial PTSD severity, peritraumatic cognitive processing variables were no longer correlated with PTSD severity at 6 and 9 months post-assault. However, avoidance/safety behaviours were still associated with symptom severity. Thus, peritraumatic variables were more important at predicting only initial PTSD severity. However, this hypothesis must be tested directly with further investigation into dysfunctional processing and the use of dysfunctional, as well as functional, coping strategies. Conclusions The results of this study are tempered by the following limitations. First, it is unclear to what extent individuals can comment on their internal cognitive processes when reporting dissociation and data-driven processing. Second, the generalization of these findings beyond university undergraduate students who are not receiving mental health treatment, or recently involved in a car accident, remains to be determined. Finally, the severity of intrusive imagery exists on a continuum, on which the analogue paradigm fits (Holmes, 2003-2004). However, it 23 necessarily produces intrusive imagery to a lesser degree and severity than either directly witnessing or experiencing a real-life trauma. However, consistent with the assertion of existing on a continuum, the intrusive thoughts experienced by the participants were mainly visual in nature and described as distressing, qualities of intrusive thoughts reported by clinical populations (Ehlers & Steil, 1995; Ehlers et al., 2002). A number of strengths also characterized this study. The distressing film paradigm allowed for the prospective measure of preexisting depression and trait anxiety, strengthening causal statements about the association between theses variables and other variables implicated in the development of intrusive thoughts. Secondly, the use of S E M allowed for the exploration of recursive pathways amongst key variables in the Ehlers and Clark model of PTSD and intrusive thought development. This recursive nature highlighted both unique variance and error associated with the prediction of each component in the model. Finally, the factor approach to variables commonly associated with similar concepts (i.e., preexisting vulnerabilities, dysfunctional processing) allowed for a more parsimonious exploration of highly associated predictors in the development of intrusive thoughts. The current results support the role of preexisting vulnerabilities, emotional reactivity, dysfunctional processing and dysfunctional coping strategies in the development of intrusive thoughts. Moreover, these four sets of variables were dynamically inter-related, underscoring the importance of using techniques such as SEM. Future research would add additional components to the model, for example priming, and continue to incorporate preexisting vulnerabilities and emotional reactivity with contemporary cognitive constructs. 24 Table 1 Descriptive Statistics and Bivariate Correlations Among Model Variables Variables 1 2 3 4 5 6 7 8 9 10 11 12 1. Happiness 2. Depression -.40** ) 3. Anger -.29** .39** 4. STAIS -.66** ** .61 .39 5. DDP -.28** .30 .23 .53** 6. PDEQ -.29** .23 .24 .39** .72** 7. Intrusions ** -.29 _ .** .24 .11 .40** .35** _ . *» .24 8. Suppression -.07 .08 .11 .13 .25** .25** .33 9. Rumination -.18* .23 .08 .25** .43** .38** .33 .27** 10. Safety -.15 .11 .06 .16 .40** .40** .37 .39** .56** .19* l l . B D I -.03 .05 .14 .07 .18* .30** _ _ * * .23 .11 .25** 12. STAIT -.05 .04 .08 .12 .23 .37 .23 .13 .27** .21* .70** Mean 0.00 0.00 0.00 0.00 7.30 16.79 2.76 11.83 8.85 6.68 8.48 39.9^ SD 18.51 17.26 12.20 9.49 4.63 6.30 2.69 4.37 4.16 6.98 6.34 9.92 Note: Happiness = residualized change score for happiness; Depression = residualized change score for depression; Anger = residualized change score for anger; STAIS = residualized change score for State Trait Anxiety Inventory- State Version; DDP = Data-driven Processing; PDEQ = Peritraumatic Dissociation Experiences Questionnaire; Safety = Safety Behaviours Questionnaire; BDI = Beck Depression Inventory; STAIT = State Trait Anxiety Inventory- Trait Version. * p < .05 **p<.001 Table 2 Final Structural Model Factor Loadings for Latent Variables Parameter Unstandardized SE Standardized Factor Loadings Preexisting: BDI (PI) 1.00a - .78 Preexisting: STAIT (P2) 1.81 .38 .90 Reactivity: Happiness (RI) -1.34 .16 -.67 Reactivity: Depression(R2) 1.16 .15 .62 Reactivity: Anger (R3) .54 .11 .40 Reactivity: STAIS (R4) 1.00a - .97 Processing: DDP (PR1) 1.00a - .89 Processing: PDEQ (PR2) 1.23 .13 .81 Response: Suppression (RP1) 1.00a - .47 Response: Rumination (RP2) 1.44 .29 .72 Response: Safety (RP3) 2.60 .52 .77 Measurement Error Variances E p i 15.92 5.06 .40 Ep2 18.96ns 15.61 .19 ERI 187.86 24.81 .55 ER2 181.41 23.03 .61 ER3 123.48 14.68 .84 ER4 4.45n s 5.99 .05 EPRI 4.27 1.45 .20 EpR2 13.55 2.63 .34 ERPI 14.72 1.87 .77 ERP2 8.40 1.44 .48 EpR3 19.77 4.05 .41 Note: BDI = Beck Depression Inventory; STAIT = State Trait Anxiety Inventory- Trait Version; Happiness = residualized change score for happiness; Depression = residualized change score for depression; Anger = residualized change score for anger; STAIS = residualized change score for State Trait Anxiety Inventory- State Version; DDP=Data-driven Processing; PDEQ = Peritraumatic Dissociation Experiences Questionnaire; Safety = Safety Behaviours Questionnaire Standardized estimates for measurement error are proportions of unexplained variance a marker variables set to 1.00 p < .05 for all unstandardized estimates unless otherwise noted nsp>.05 26 Table 3 Model Modification Statistics A from base A from model previous Model x2 df P X2 df P x2 df p Base (measurement model) 55.34 45 <.00 Initial model 100.45 49 <.oo 45.11 4 <.05 Add path from reactivity 88.50 48 <.oo 33.16 3 <.001 -11.95 -1 <.001 to intrusions Add path from processing 58.25 47 >.05 2.91 2 >.05 -30.25 -1 <.001 To responses 27 Table 4 Factor Correlations 1 2 3 4 1. Preexisting Vulnerability 2. Emotional Reactivity .13 3. Dysfunctional Processing .35 .56 4. Dysfunctional Strategies .38 .34 .63 5. Intrusions .18 .41 .35 .51 Note: Significance values are not available for these parameters. Table 5 Total, Direct and Indirect Effects on Intrusive Thoughts Latent Variable Total effect Direct Effect Indirect Effect Preexisting Vulnerability .18* .18* Emotional Reactivity .39* .32* .07 Dysfunctional Processing .14 -.13 .27* Dysfunctional Strategies .48* .48* * p < .05 29 30 Figure 2 Model 1: Initial Structural Equation Model *p<.05 Figure 3 Model 2: Final Structural Equation Model 32 Footnotes 1 This study was supervised by Dr. Lynn Alden. 2 Three consistent results have emerged from Monte Carlo studies on the impacts of nonnormality on maximum likelihood estimations in structural equation modeling (Enders, 2001): parameter estimates are generally unbiased; parameter standard errors are negatively biased (i.e., smaller); and model fit statistics yields excessive model rejection. 3 3 References Allen, J. G., Console, D. A. , & Lewis, L. (1999). Dissociative detachment and memory impairment: Reversible amnesia or encoding failure? Comprehensive Psychiatry, 40, 160-171. American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Antony, M . M . , Orsillo, S. M . , & Roemer, L. (Eds.). (2001). Practitioner's guide to empirically based measures of anxiety. New York: Kluwer Academic/Plenum Publishers. Beck, A . T., Steer, R. A. , & Brown, G. K. (1996). Beck depression inventory- second edition manual. San Antonio, T X : The Psychological Corporation. Birmes, P., Brunei, A. , Carreras, D., Ducasse, J., Charier, J., Lauque, D., et al. (2003). The predictive power or peritraumatic dissociation and acute stress symptoms for posttraumatic stress symptoms: A three-month prospective study. American Journal of Psychiatry, 160, 1337-1339. Breslau, N . , Davis, G. C , Peterson, E. L. , & Schultz, L. R. (2000). A second look at comorbidity in victims of trauma: The posttraumatic stress disorder-major depression connection. Biological Psychiatry. 48, 902-909. Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of posttraumatic stress disorder. Psychological Review, 103, 670-686. Bryant, R. A. , & Harvey, A . G. (1995). Avoidant coping style and post-traumatic stress following motor vehicle accidents. Behavior Research and Therapy, 33, 631-635. Cahill, L., & McGaugh, J. L. (1998). Mechanisms of emotional arousal and lasting declarative memory. Trends in Neurosciences, 21, 294-299. 34 Charlton, P. F. C , & Thompson, J. A . (1996). Ways of coping with psychological distress after trauma. British Journal of Clinical Psychology, 35, 517-530. Clohessy, S., & Ehlers, A . (1999). PTSD symptoms, response to intrusive memories and coping in ambulance service workers. British Journal of Clinical Psychology, 38, 251-265. Davies, M . I., & Clark, D. M . (1998a). Predictors of analogue post-traumatic intrusive cognitions. Behavioural and Cognitive Psychotherapy, 26, 303-314. Davies, M . I., & Clark, D. M . (1998b). Thought suppression produces a rebound effect with analogue post-traumatic intrusions. Behaviour Research and Therapy, 36, 571-582. Dunmore, E., Clark, D. M . , & Ehlers, A . (2001). A prospective investigation of the role of cognitive factors in persistent posttraumatic stress disorder (PTSD) after physical or sexual assault. Behaviour Research and Therapy, 39, 1063-1084. Ehlers, A . (1998). Data-driven versus conceptual processing questionnaire. Unpublished manuscript. Ehlers, A. , & Clark, D. M . (2000). A cognitive model of posttraumatic stress disorder. Behaviour Research and Therapy, 38, 319-345. Ehlers, A. , Hackman, A. , Steil, R., Clohessy, S., Wenninger, K. , & Winter, H . (2002). The nature of intrusive memories after trauma: The warning signal hypothesis. Behaviour Research and Therapy, 40, 995-1002. Ehlers, A. , Mayou, R. A. , & Bryant, B. (1998). Psychological predictors of chronic posttraumatic stress disorder after motor vehicle accidents. Journal of Abnormal Psychology, 107, 508-519. Ehlers, A. , Michael, T., Chen, Y . P., Payne, E., & Shan, S. (2006). Enhanced perceptual priming 35 for neutral stimuli in a traumatic context: A pathway to intrusive memories? Memory, 14, 316-328. Ehlers, A. , & Steil, R. (1995). Maintenance of intrusive memories in PTSD: A cognitive approach. Behavioral and Cognitive Psychotherapy, 23, 217-249. Elzinga, B. M . , & Bremner, J. D. (2002); Are the neural substrates of memory the final common pathway in posttraumatic stress disorder (PTSD)? Journal of Affective Disorders, 70, 1-17. ' Enders, C. K . (2001). The impact of nonnormality on full information maximum-likelihood estimations for structural equation models with missing data. Psychological Methods, 6, 352-370. Enders, C. K. , & Bandalos, D. L. (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling, 8, 430-457. Engelhard, I. M . , van den Hout, M . A. , Kindt, M . , Arntz, A . , & Schouten, E. (2003). Peritraumatic dissociation and posttraumatic stress after pregnancy loss: A prospective study. Behavior Research and Therapy, 41, 67-78. Foa, E. B., & Rothbaum, B. O. (1998). Treating the trauma of rape: Cognitve and behavioral therapy for PTSD. New York: Guilford Press. Frommberger, U . H. , Stieglitz, R. D., Nyberg, E., Schlickewei, W., Kuner, E., & Berger, M . (1998). Prediction of posttraumatic stress disorder by immediate reactions to trauma: A prospective study in road traffic accident victims. European Archives of Psychiatry and Clinical Neuroscience, 248, 316-321. Halligan, S. L. , Clark, D. M . , & Ehlers, A . (2002). Cognitive processing, memory and the 36 development of PTSD symptoms: Two experimental analogue studies. Journal of Behavior Therapy and Experimental Psychiatry, 33, 73-89. Halligan, S. L., Michael, T., Clark, D. M., & Ehlers, A. (2003). Posttraumatic stress disorder following assault: The role of cognitive processing, memory and appraisals. Journal of Consulting and Clinical Psychology, 71,419-431. Harvey, A. G., & Bryant, R. A. (1998). The effect of attempted thought suppression in acute stress disorder. Behaviour Research and Therapy, 36, 583-590. Holmes, E. A. (2003-2004). Intrusive, emotional mental imagery and trauma: Experimental and clinical clues. Imagination, Cognition and Personality, 23, 147-154. Holmes, E. A., Brewin, C. R., & Hennessy, R. G. (2004). Trauma films, information processing, and intrusive memory development. Journal of Experimental Psychology: General, 133, 3-22. Holmes, E. A., Brown, R. J., Mansell, W., Fearon, R. P., Hunter, E. C. M., Frasquilho, F., et al. (2005). Are there two qualitatively distinct forms of dissociation? A review and some clinical implications. Clinical Psychology Review, 25, 1-23. Holmes, E. A., Grey, N., & Young, K. A. D. (2005). Intrusive images and "hotspots" of trauma memories in posttraumatic stress disorder: An exploratory investigation of emotions and cognitive themes. Journal of Behavior Therapy and Experimental Psychiatry, 36, 3-17. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Coventional criteria versus new alternatives. Structural Equation Modeling, 6,1-55. Joreskog, K., & Sorbom, D. (2005). Lisrel 8.72 Student Edition [Computer software]. Lincolnwood, IL: Scientific Software International, Inc. Joseph, R. (1996). Neuropsychiatry, neuropsychology, and clinical neuroscience (2nd ed.). 37 Baltimore: Williams & Wilkins. King, D. W., King, L. A . , Foy, D. W., & Gudanowski, D. M . (1996). Prewar factors in combat-related posttraumatic stress disorder: Structural equation modeling with a national sample of female and male Vietnam veterans. Journal of Consulting and Clinical Psychology, 64,520-531. King, L. A. , King, D. W., Fairbank, J. A. , Keane, T. M . , & Adams, G. A . (1998). Resilience-recovery factors in post-traumatic stress disorder among female and male Vietnam veterans: Hardiness, postwar social support and additional stressful life events. Journal of Personality and Social Psychology, 74, 420-434. Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed). New York: Guilford Press. Kuehner, C , & Weber, I. (1999). Responses to depression in unipolar depressed patients: An investigation of Nolen-Hoeksema's response styles theory. Psychological Medicine, 29, 1323-1333. Laposa, J. M . , & Alden, L. E. (2003). Posttraumatic stress in the emergency room: Exploration of a cognitive model. Behaviour Research and Therapy, 41, 49-65. Laposa, J. M . , & Alden, L. E. (2006). An analogue study of intrusions. Behaviour Research and Therapy, 44, 925-946. Marmar, C. R., Weiss, D. S., & Metzler, T. J. (1997). The Peritraumatic Dissociative Experiences Questionnaire. In J. P. Wilson & T. M . Keane (Eds.), Assessing psychological trauma and PTSD (pp. 412-428), New York: Guilford Press. Marshall, G. N . , & Schell, T. L. (2002). Reappraising the link between peritraumatic dissociation 38 and PTSD symptom severity: Evidence from a longitudinal study of community violence survivors. Journal of Abnormal Psychology, 2002, 626-636. MacCallum, R. C , Roznowski, M . , & Necowitz, L. B. (1992). Model modifications on covariance structure analysis: The problem of capitalization on chance. Psychological Bulletin, 111,490-504. Mayou, R. A. , Ehlers, A. , & Bryant, B. (2002). Posttraumatic stress disorder after motor vehicle accidents: 3-year follow-up of a prospective longitudinal study. Behaviour Research and Therapy, 40, 665-675. Michael, T., Ehlers, A. , & Halligan, S. L. , (2005). Enhanced priming for trauma-related material in posttraumatic stress disorder. Emotion, 5, 103-112. 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, 613-628. Mineka, S., Watson, D., & Clark, L. (1998). Comorbidity of anxiety and unipolar mood disorders. Annual Review of Psychology, 49, 377-412. Muris, P., Merckelbach, H. , & Horselenberg, R. (1996). Individual differences in thought suppression. The white bear suppression inventory: Factor structure, reliability, validity and correlates. Behavior Research Therapy, 34, 501-513. Muris, P., Roelofs, J., Rassin, E., Franken, I., & Mayer, B. (2005). Mediating effects of rumination and worry on the links between neuroticism, anxiety and depression. Personality and Individual Differences, 39,1105-1111. Murray, J., Ehlers, E., & Mayou, R. A. , (2002). Dissociation and post-traumatic stress disorder: 39 Two prospective studies of road traffic accident survivors. British Journal of Psychiatry, 180,363-368. Nolen-Hoeksema, S. (2000). The role of rumination in depressive disorders and mixed anxiety/depressive symptoms. Journal of Abnormal Psychology, 109, 504-511. 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, 52-73. Phelps, E. A . (2006). Emotion and cognition: Insights from studies of the human amygdala. Annual Review of Psychology, 57, 27-53. Shalev, A. Y . , Peri, T., Canetti, L. , & Schreiber, S. (1996). Predictors of PTSD in injured trauma survivors: A prospective study. The American Journal of Psychiatry, 153, 219-225. Shalev, A . Y . , Sahar, T., Freedman, S., Tuvia, P., Glick, N . , Brandes, D., et al. (1998). A prospective study of heart rate response following trauma and the subsequent development of posttraumatic stress disorder. Archives of General Psychiatry, 55, 553-559. Sierra, M . , & Berrios, G. E. (1998). Depersonalization: Neurobiological perspectives. Biological Psychiatry, 44, 898-908. Speilberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, G. A . (1983). Manual for the State-Trait Anxiety Inventory. Palo Alto, C A : Consulting Psychologist Press. Tampke, A. K. , & Irwin, H . J. (1999). Dissociative processes and symptoms of posttraumatic stress in Vietnam veterans. Journal of Traumatic Stress, 12, 725-738. Ursano, R. J., Fullerton, C. S., Epstein, R. S., Crowley, B. , Vance, K . , Kao, T. et al. (1999). Peritraumatic dissociation and posttraumatic stress disorder following motor vehicle accidents. American Journal of Psychiatry, 156, 1808-1810. 40 Vuksic-Mihaljevic, Z., Mandic, N . , Bensic, M . , & Mihaljevic, S. (2000). Posttraumatic stress disorder among Croatian veterans: A causal model. Psychiatry and Clinical Neurosciences, 54, 625-636. Wegner, D. M . (1994). Ironic processes of mental control. Psychological Review, 101, 34-54. Wegner, D. M . , & Zanakos, S. (1994). Chronic thought suppression. Journal of Personality, 62, 615-640. Wenzlaff, R. M . , & Bates, D. E. (1998). Unmasking a cognitive vulnerability to depression: How lapses in mental control reveal depressive thinking. Journal of Personality and.Social Psychology, 75, 1559-1571. Witvliet, C. V . (1997). Traumatic intrusive imagery as an emotional memory phenomenon: A review of research and explanatory information processing theories. Clinical Psychology Review, 17, 509-536. 41 Appendix A Emotional Reactivity Questionnaire Please rate how you feel at this moment: 1. Happy 0 10 20 30 40 50 60 2. Depressed 0 10 20 30 40 50 60 3. Angry 0 10 20 30 40 50 60 Appendix B Data-driven Processing Scale During the event this applied to me: Not at all 1. I couldn't really take it all in. 0 2. I did not fully understand what was going on. 0 3. It was just like a stream of unconnected impressions following each other. 4. I could not think clearly. 5. I was overwhelmed by sensations and couldn't put everything together. 6. I was confused and could not fully make sense of what was happening. 7. My mind was fully occupied with what I saw, heard, smelled, and felt. 8. My mind was full of impressions and my reactions to them. 0 0 0 A little 1 1 1. Moderately Strongly Very strongly 2 2 2 2 2 4 4 4 4 4 43 Appendix C Response to Intrusions Questionnaire Suppression and Rumination Subscales Think back to the intrusive memories you had about the film. What did you do when the recollections occur? Never 3.1 try to push them out of my mind. 1 2 4.1 dwell on them. 1 2 5.1 think about something else. 1 2 6.1 worry that something like that could happen to me 1 2 or my family. 7.1 watch TV, listen to music, or read. 1 2 8.1 drink alcohol or smoke. 1 2 9.1 think about what could have been done differently. 1 2 Appendix D Safety Behaviours Questionnaire Since viewing the film last week did you notice an increase in: Not at all Very Much 1. Discomfort associated with road-related situations 0 1 2 3 4 5 6 7 8 9 2. Hesitancy surrounding road-related situations 0 1 2 3 4 5 6 7 8 9 3. Vigilance while in a car 0 1 2 3 4 5 6 7 8 9 4 5 Appendix E Intrusion Diary Over the course of the next week, starting as soon as you leave this room, please fill out the appropriate columns of the chart each time you spontaneously have a memory/image/thought about the film that occurs when you do not intend to think about the film. Each time this happens, please enter the date and time, rate how distressing it was, give a description of it's content, and indicate whether it took the form of a thought, image, or combination of the two. It is very important that you try to fill out the columns as soon as possible after it occurs. Please return this form to the experimenter at your appointment next week. s ' Date and Time (please enter each one) How distressing was it? 0 10 not extremely at all (please rate each one) Description of its contents (please describe each one) Was it a thought(T), image(I), or Combination (C)? (Please circle the appropriate letter for each one) 1 T . I C 2 T I C 3 T I C 4 T . I C 5 T I C 6 T I C 46 7 T I C 8 T I C 9 T I C 47 Appendix F Experimental Checks Attention How much attention did you pay to the film? 0 1 2 3 4 5 6 7 8 9 10 not at all full attention Prior frequency of medical footage exposure To what extent is this statement true of you? I watch surgical T V programs, for example like those on The Learning Channel never almost everyday 0 1 2 3 4 5 6 7 Diary Compliance To what extent is this statement true of you: I was often unable (or forgot) to record my spontaneous memories/images/thoughts in the diary chart. Not at all Extremely true of me true of me 0 1 2 3 4 5 6 7 8 9 10 48 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items