| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the University of Plymouth, School of Psychology (M.C.C., I.D.), Plymouth, UK; the Department of Clinical Psychology, University of Manchester (Y.E.), Manchester, UK; the University of Birmingham, School of Health Sciences (J.W.), Birmingham, UK; and the Department of Psychology, University of Wolverhampton (S.F.), Wolverhampton, UK.
Address correspondence and reprint requests to Man Cheung Chung, PhD, School of Psychology, University of Plymouth, Drake Circus, Plymouth, Devon, PL4 8AA, UK. E-mail: m.chung{at}plymouth.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Materials and Methods: One hundred forty-eight community residents, after exposure to the aircraft or train crash, were assessed using the Impact of Event Scale, the General Health Questionnaire-28, the Death Anxiety Scale, the Eysenck Personality Questionnaire, and the Ways of Coping Checklist. The control group (n = 90) comprised members of the general public, who had not been exposed to the disasters, from another city.
Results: The model showed significant associations between the impact of the disaster and general health problems, which varied depending on where community residents lived in relation to the disaster site, whether they were present when the disaster occurred, and the type of disaster. The model also suggested that death anxiety was associated with type of disaster and neuroticism. The model supported the interactive model in that personality factors interacted with coping strategies in maintaining or generating PTSD and general health problems.
Conclusions: After exposure to technological disasters, community residents could develop PTSD and general health problems; however, increased death anxiety was a separate psychological reaction. The interaction between certain personality traits and coping strategies was one reason for PTSD and general health problems.
Key Words: posttraumatic stress • death anxiety • personality • coping • technological disaster
Abbreviations: MIMIC = multiple-indicator multiple-cause model; PTSD = posttraumatic stress disorder; POWS = prisoners of war; IES = The Impact of Event Scale; GHQ-28 = The General Health Questionnaire; DAS = The Death Anxiety Scale; EPQ-R = The Eysenck Personality Questionnaire-R Short Scale; WOC = The Ways of Coping Checklist; ANOVA = analysis of variance; RMSEA = root mean square error of approximation.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
The mediational model suggests that personality factors determine particular coping strategies, which might result in maladjustment. However, much of PTSD research has not focused on this model, and the way in which personality traits may affect people's ability to cope with recurring trauma memories is still very much underresearched (9). Focusing on people exposed to war, limited studies have shown that prisoners of war (POWs) with high sensation-seeking traits tend to use active coping strategies that are associated with better adjustment to PTSD from captivity. On the other hand, POWs with low sensation-seeking traits tend to cope by detachment and denial, which are associated with poorer adjustment (10). Veterans with the dysfunctional passive–aggressive personality style or avoidant style tend to use more confrontive coping and escape–avoidance, which are associated with greater PTSD severity (9). Also, civilians who have been exposed to missile attack and who have a secure attachment style tend to seek social support more than those with an ambivalent attachment style. The latter use more emotion-focused coping and report more distress (11).
The additive model suggests that personality and coping make independent and unique contributions to maladjustment. PTSD research has placed more emphasis on this model than on the mediational model. With regard to personality, neuroticism has on the whole been shown to be a significant predictor of PTSD symptoms and general health problems among victims of different disasters (12,13). Veterans with combat-related PTSD tend to be more neurotic than those without PTSD and to have personality disorders, antisocial disorders, and psychopathic tendencies associated with neuroticism (14–16). One study also showed that among peacekeepers, neuroticism measured before their deployment was a significant predictor of PTSD symptoms after peacekeeping duties (17). High neuroticism was also found to be associated with PTSD symptoms among victims of natural disasters (18) and emergency service personnel after their duties (19–22). The association between neuroticism and PTSD symptoms and diagnoses has been shown to last as long as 3 years after the rescue for some firefighters (23), 6 months after accidents for motor vehicle victims (24), and 12 months after discharge from the hospital for burns (25). Indeed, this is consistent with the finding that neuroticism predicted lifetime diagnosis of PTSD among young adults in an urban area (26). In addition to neuroticism, introversion was another significant predictor of chronic PTSD among firefighters (19,20) and Vietnam veterans (27). Introversion was also associated with PTSD diagnoses 4 and 12 months after discharge from the hospital for burns (25) and with PTSD symptoms 3 months after an aircraft crash (28).
Turning to the contribution that coping can make independently to PTSD, after disasters, research suggests that emotion-focused coping is a predictor of greater psychiatric, physical, and somatic symptoms among, for example, veterans involved in war (29,30). Also, veterans who seek mental health treatments for their PTSD tend to use emotion-focused coping (31), whereas problem-focused coping moderates the detrimental effects of emotion-focused coping on mental health (32). Emotion-focused coping measured at 2 to 3 weeks and 4 months after some military training accidents predicted PTSD symptoms at 12 months (33). Specifically, the escape–avoidance coping strategy, one type of emotion-focused coping, is often associated with PTSD. For example, to cope with war memories, veterans predominantly use escape–avoidance strategies, which are associated with increased PTSD severity (9). Similarly, to cope with the traumatic effects of their involvement in the rescue operation, disaster personnel have been known to use escape–avoidance or emotional distancing. These coping strategies are associated with high levels of PTSD (34,35). In addition to disaster personnel, escape–avoidance has also been shown to predict severe PTSD for victims of accidents, terrorism, and crime (12), as well as acute stress reactions (ie, 2–3 weeks) for victims after fatal training accidents (33) and shipwrecks (36).
With regard to the interactive model, the claim is that personality factors interact with coping strategies in maintaining or generating maladjustment. Like with the mediational model, PTSD studies that investigate personality, coping, and PTSD in the light of this interactive model are rare. For example, one study followed up Israeli soldiers who served during the 1982 Lebanon war. When studied 2 and 3 years after engagement in combat, they showed a decline in PTSD symptoms, which was associated with soldiers using more internal locus of control, less emotion-focused coping, and perceiving more social support (37).
Death Anxiety and Posttraumatic Stress Disorder
Death anxiety research has highlighted the link among death anxiety, overall psychological health, and exposure to life-threatening events. This link bears potentially important implications for PTSD research. According to the two-factor model of death anxiety (38,39), the degree of death anxiety is determined by two factors; one is general psychological health and the other is life experiences related to death. That is, the degree of death anxiety is associated with our pervasive psychopathological condition (people who suffer from, for example, depression or anxiety disorders may suffer from a high degree of death anxiety) or with, for example, life-threatening experiences (people who have experienced life-threatening events may suffer from increased death anxiety). It seems plausible that these two factors are interrelated in that, after exposure to life-threatening events, the degree of death anxiety may increase and PTSD symptoms may develop. Once this psychopathological condition (PTSD) has developed, it may heighten the degree of death anxiety even further. If death anxiety were shown to be associated with PTSD symptoms, it would shed new light on our understanding of PTSD responses and may have important implications for diagnosing PTSD.
This exposition of literature has revealed two things. First, the complex relationship among personality, coping, and PTSD in the light of the mediational, the additive, and the interactive models has been relatively unexplored in PTSD research, although there is a substantial amount of evidence recorded in terms of the additive model. Second, little is known about the relationship between death anxiety and PTSD responses after life-threatening events. To address these issues, in this study, we primarily aimed to build a multiple-indicator multiple-cause (MIMIC) model, which would describe the relationship among personality factors, coping strategies and PTSD responses, general health problems, and death anxiety among community residents exposed to the technological disasters of aircraft and train crashes. By definition, technological disasters result from failure of manmade products and include aircraft crashes and train derailments or collisions (40). On the basis of this literature, we hypothesized 1) that the model would reveal associations among PTSD responses, general health problems, and death anxiety. In particular, we expected that intrusion and avoidance, anxiety, and depression would be associated with death anxiety. We also hypothesized 2) that the model would support the interactive model, ie, personality factors would interact with coping strategies in maintaining or generating PTSD reactions and general health problems.
Disasters
Disaster 1
The aircraft disaster in the present study occurred in Coventry in the United Kingdom in 1994. The aircraft was returning to Coventry from Amsterdam after having unloaded a cargo of livestock. At around 9 am, the aircraft lost control, struck a pylon, started to descend, clipped the gable ends of two houses, and eventually crashed into a woodland area close to the edge of a housing estate. Although all five crewmembers on board died, hundreds of residents escaped death.
Disaster 2
In 1996, at 11:00 pm, in Stafford, United Kingdom, a train collision occurred between a freight train carrying liquid carbon dioxide tank wagons and a post office train. Although the driver of the freight train was not injured (because it was not a head-on crash), all of the employees on the post office train were and one post office employee was killed. There were residential houses on both sides of the collision site. Residents were evacuated in an operation to remove the carbon dioxide.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Procedure
The study began approximately 6 to 7 months after the disasters. The researchers designated an area of each of the two communities within 10 to 300 meters of the crash site. Approximately 150 households near the aircraft crash and 95 households near the train collision (on both sides of the track) were situated in the designated areas. We distributed letters, explaining the purpose of the research, to all of these households. The letters explained that the purpose was to investigate to what extent the residents had been affected by the disasters, how they coped subsequently, and whether personality factors had a role to play in their coping styles. The letter informed community residents that the research team would be visiting them for an interview. Twenty-four community residents did not want to participate in the study, two were not relevant because they had moved there after the crash, and two had left the area since the crash. Sixty-nine community residents were not at home when visited and revisited (two or three times) and the remaining 148 were interviewed. This yielded a 60% response rate.
A control group consisting members of the general public from another city in the United Kingdom was recruited by means of advertisements in leisure clubs, shops, churches, and a university campus, and by approaching people on high streets. Members of the control group confirmed that they had not been exposed to either of the two disasters in this study. To establish whether there was a greater incidence of general health problems in people exposed to the disasters, the two disaster groups and the controls were asked to complete the General Health Questionnaire-28. Other measures were administered to the disaster groups to investigate the factors predicting posttraumatic stress responses and general health problems in members of the disaster groups. Respondents from these groups were asked to indicate the methods they used for dealing with the aftermath of the disasters by completing the Ways of Coping Checklist and were also asked to complete the Impact of Event Scale, the Death Anxiety Scale, and the Eysenck Personality Questionnaire-R.
Measures
The Impact of Event Scale (IES) (41) is a 15-item, four-point scale (0 = not at all, 1 = rarely, 3 = sometimes, 5 = often) self-report instrument, which measures experiences of intrusion and avoidance. The scale has been shown to have a high internal consistency of the two subscales (intrusion = 0.78, avoidance = 0.82), and a high test–retest reliability of 0.89 for intrusion and 0.79 for avoidance.
The General Health Questionnaire (GHQ-28) (42) aims to estimate the likelihood of participants being assessed as psychiatric cases at interview. There are four subscales to this questionnaire: Somatic, Anxiety, Social Dysfunction, and Depression. As the total GHQ score exceeds the recommended cutoff point of four, the probability of becoming a psychiatric case increases. On the basis of this cutoff, the GHQ-28 has shown a sensitivity value of 88% at a specificity of 84.2% and an overall misclassification rate of 14.5%.
The Death Anxiety Scale (DAS) (43) is a self-report instrument for measuring death anxiety, which consists of 15 true or false items. The questionnaire has a test–retest reliability of 0.83 and an internal consistency coefficient of 0.73.
The Eysenck Personality Questionnaire-R Short Scale (EPQ-R) (44) is a 48-item questionnaire measuring character traits of psychoticism, extroversion, and neuroticism. It also includes a lie scale. The manual reports test–retest reliabilities for males and females. For males, it has reliabilities of 0.77, 0.83, 0.76, and 0.76 for psychoticism, extroversion, neuroticism, and the lie scale, respectively. The corresponding reliabilities for females are 0.81, 0.89, 0.81, and 0.80.
The Ways of Coping Checklist (WOC) (45) consists of 67 items that require responses on a four-point scale (0 = not used, 1 = used somewhat, 2 = used quite a bit, 3 = used a great deal). It aims to explore the role of coping in the relationship between stress and adaptational outcomes. The items on the original WOC were classified on the basis of "problem-focused" or "emotion-focused" functions of coping. Problem-focused coping refers to efforts undertaken to manage or alter the troubled person–environment relationship that is the source of stress. Emotion-focused coping refers to efforts undertaken to regulate stressful emotions. This checklist measures eight coping strategies, namely, confrontive coping (
= 0.70), distancing ( = 0.61), self-controlling ( = 0.70), seeking social support ( = 0.76), accepting responsibility ( = 0.66), escape–avoidance ( = 0.72), planful problem-solving ( = 0.68), and positive reappraisal ( = 0.79). Problem-focused coping is composed of the strategies of planful problem-solving and confrontive coping. Emotion-focused coping is composed of the strategies of seeking social support, distancing, escape–avoidance, self-controlling, accepting responsibility, and positive reappraisal.
The data analysis plan was organized in the following way. We compared the mean scores of the IES, GHQ-28, DAS, EPQ-R, and WOC for the two disaster groups and the control using t tests and analysis of variance (ANOVA). What followed was the building of the MIMIC model. The fit of the measurement component of the MIMIC model, which relates the observed variables to the latent variables (general health problems, impact of event, coping strategies, and death anxiety), was assessed first. Once a satisfactory measurement model had been achieved, personality and other predictor variables were added to the model. All possible paths from predictors to latent variables were considered but only significant paths were retained in the final model.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
|
In terms of personality, the results showed no differences in psychoticism (t = 0.89, NS), extraversion (t = –0.54, NS), and neuroticism (t = 0.02, NS) between the two disaster groups. With regard to WOC, the results showed that the community residents near the aircraft disaster scored higher than those exposed to the train disaster in coping strategies. In particular, the community residents near the aircraft disaster used significantly more confrontive coping (t = 3.51, p <.001), self-controlling (t = 3.49, p <.001), escape–avoidance (t = 4.32, p <.001), and positive reappraisal coping strategies (t = 4.75, p <.001) than those near the train disaster.
To build a MIMIC model, which explored the relationship between the foregoing measures and other variables, scale scores were calculated for each participant on each of the measures listed in Table 2. In view of the limited sample size, and to limit the number of parameters in the fitted model, the eight coping strategies were collapsed to emotion-focused coping and problem-focused coping. To identify the model, multiple indicators of death anxiety were needed. This was achieved by arbitrarily assigning the 15 items of the death anxiety questionnaire to three item parcels (Death A, B, and C).
|
Some participants did not complete all the scales. There were none missing in GHQ-28 and IES; 1, 6, and 9 people did not fill in the WOC, DAS, and EPR-R, respectively. The missing scale scores were dealt with using the full-information maximum-likelihood procedures in MPlus. There were a number of cases in which participants completed a scale but omitted some of the items. These were dealt with by using regression imputation to impute the missing responses from the responses to other items on the same scale. Overall, 1.06% of responses were imputed in this way.
The maximum likelihood-fitting criterion used here assumes that the distribution of the indicator variables is multivariate normal. As can be seen from Table 2, the raw scale scores showed some substantial departures from this assumption. Accordingly, monotonic transformations were selected to bring the scale scores closer to univariate normality. The transformation used and the coefficients of skew and kurtosis after transformation are shown in Table 3. It should be noted that the procedures used by MPlus for MIMIC models do not make any distributional assumptions concerning the background variables (46). Table 4 shows the correlation matrix for transformed variables used in the modelling.
|
|
The model shown in Figure 1 was first fitted with only the indicator variables but without the background variables to assess the measurement aspect of the model. The model was initially fitted without any correlations between the residuals. This gave a significant chi square value (chi square = 69.1 with 38 df, p = .0015), an RMSEA of 0.077, and a CFI of 0.969. The significant chi square and the RMSEA value indicated some significant departure between the observed covariance matrix and that predicted by the model. Inspection of modification indices suggested two post hoc modifications to the measurement part of the model to improve its fit. The first of these was allowing the residuals on GHQC (social dysfunction) and GHQD (depression) to correlate. Thus, there was an additional correlation between social dysfunction and depression over and above that arising from the relationship that each had to general health. The second adjustment allowed the residuals on intrusions and problem focused coping to correlate, indicating that there was a specific relationship between experiencing intrusions and adopting problem-focused coping. These two residual correlations are shown dotted in the figure. Both of these adjustments were considered to be theoretically plausible, and both residual correlations were found to be significant. However, as post hoc modifications to the model, they needed to be viewed with caution. Whether they were included has little impact on the conclusions that we reach later concerning the impact of background variables on the four main constructs in the model. With these two amendments, the model gave a nonsignificant chi square (chi square = 50.14 with 36 df, p = .059) and an improved RMSEA of 0.053 and CFI of 0.986. The 90% confidence interval for the RMSEA extended from 0 to 0.086.
|
The estimated correlations between the four latent variables in the measurement model are shown above the diagonal in Table 5. All four constructs were positively related. Based on the estimated standard errors, all of the correlations shown were significant at 1% or better. It is tempting to assume that these correlations demonstrated, for example, that experiencing PTSD after the technological disaster (as manifested in impact-of-event scores) leads to heightened death anxiety. However, an alternative explanation would be that there is some third factor such as a particular personality type, which is associated both with a greater likelihood of PTSD and with higher death anxiety. The MIMIC model tests whether any of the background variables that it included are acting in this way.
|
The background factors incorporated in the MIMIC model were the three personality dimensions (neuroticism, extraversion, and psychoticism), age, gender, and three binary variables coding: 1) which of the two disasters the community resident was exposed to (type), 2) whether the resident was at home at the time of the disaster (present), and 3) whether the home was very near or slightly less near the disaster (distance). Estimated correlations of these variables with the indicators are shown in Table 4.
Initially, all of the background factors were considered as potential predictors of the four latent variables in the model. However, initial fitting of the MIMIC model produced no evidence that psychoticism, age, or gender were predictive of any of the four factors. Hence, these three variables were not included in the final model. Paths from the five remaining background variables to each of the four factors were included in the model. For the MIMIC model in this form, with no direct paths from the background variables to the measured variables, fit appeared somewhat worse than for the measurement model (chi square = 114.34 with 71 df, p = .0009, RMSEA = 0.066, CFI = 0.961).
Inspection of the residuals suggested that the model would be improved by the addition of some direct links between the personality variables and specific indicators, as shown in Figure 1. The paths added were from extraversion to GHQC (social dysfunction) and from both extraversion and neuroticism to GHQD (depression). These paths implied that general health problems might manifest somewhat differently for different personality types with introverts showing enhanced susceptibility to social dysfunction and depression, and those higher on neuroticism also showing an increased likelihood of depression. These amendments were felt to be theoretically plausible, all three added paths were significant, and between them they produced a clear improvement in fit. The revised model, including the three paths, gave an improved fit (chi square = 89.34 with 68 df, p = .048, RMSEA = 0.048, CFI = 0.981, 90% confidence interval on RMSEA extending from 0.011 to 0.073). The chi square difference test comparing the model with the three direct paths with the original model demonstrated that the latter was clearly a significantly worse fit (chi square difference = 25.00 with 3 df, p = .00002). Given the marked improvement in fit as a result of introducing the three direct paths, the parameter estimates reported here are based on the model, including these paths. However, given that they are post hoc modifications, they should be treated with caution. Estimates of standardized path coefficients from the amended model are shown in Figure 1. Only significant paths are included in the figure. The estimated correlations between factors, once the background factors have been controlled, are also shown below the diagonal in Table 5.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
The MIMIC model also suggested that PTSD reactions varied depending on which disaster was experienced. Seemingly, the aircraft disaster gave rise to greater traumatic impact and general health problems than the train disaster. One might speculate that this was the result of the greater impression of horror, near-death or near-injury, surrounding the aircraft disaster. The fact that the aircraft had descended, clipping the roofs of two houses on the way, showed the closeness of the danger, horror, and indeed the near-death experience. On impact, there was a massive explosion and the aircraft was on fire for some time. More people died in this impact than in the train disaster.
However, the MIMIC model did not support the first hypothesis in that once personality had been controlled, it showed no association among PTSD responses, general health, and death anxiety. In other words, death anxiety seemed to be a separate psychological reaction from PTSD reactions and general health problems. This finding basically contradicted the claim in factor one of the two-factor model of death anxiety that the development of some psychopathological conditions, in this case PTSD or general health problems, was not necessarily associated with the level of death anxiety.
Rather, the MIMIC model suggested that death anxiety was associated with type of disaster. The community residents near the aircraft disaster were more anxious about death than those exposed to the train disaster. This was not too surprising taking into account the greater impression of horror, near-death and near-injury, mentioned earlier, connected with the aircraft disaster. This was consistent with the claim in factor two of the two-factor model of death anxiety that the degree of death anxiety is associated with life experiences concerning the topic of death (in this case, the life-threatening experience of an aircraft or train crash). Also, death anxiety was found not to be affected by whether the community residents were present at the time of the disaster or how close they lived to the disaster site. In other words, although the community residents reacted to the disaster with death anxiety, the degree of death anxiety that they experienced was not associated with proximity to the disaster. This in fact strengthened further the foregoing association between death anxiety and type of disaster, which seemed to be over and above any other possible association (eg, associations between death anxiety and PTSD reactions, and between death anxiety and proximity to the disaster).
What are the implications of the findings for the three models of the relationship among coping, personality, and maladjustment (ie, PTSD and general health problems) discussed previously? The MIMIC model did not support the mediational model. The mediational model assumes the absence of a direct link between personality factors and PTSD or general health problems. However, the MIMIC model clearly showed a direct link between neuroticism and a general health problem of depression and the impact of the disasters. A direct link was also found between extraversion and the impact of the disasters and the general health problems of social dysfunction and depression.
In terms of the additive model, this model implies 1) effects of personality on PTSD and general health problems, 2) effects of coping on PTSD and general health problems, and 3) effects of (1) and (2) being independent and additive (ie, the additive model assumes that use of coping strategies is not directly affected by personality). The MIMIC model showed evidence for (1) in that neuroticism was associated with general health problems and the impact of the disasters. The model also showed that extraversion was in fact a protective factor against general health problems. The more extraverted the community residents were, the less they experienced social dysfunction and depression. This was not consistent with other findings (21) in which extraversion was positively associated with general health problems, in particular, social dysfunction and depression. The MIMIC model also showed evidence for (2) in that the community residents used both problem-focused and emotion-focused coping strategies, which were associated with intrusive thoughts of the disaster, avoidance, and subsequent general health problems. These results contradicted other research, which suggested that emotion-focused coping was mainly associated with psychological distress (29–31). However, the MIMIC model offered no support to (3) because the MIMIC model clearly revealed a direct path between neuroticism and coping. In other words, they were not independently associated with PTSD and general health problems.
The MIMIC model, however, did offer support to the interactive model, as we hypothesized. The MIMIC model showed that an element of the effect of neuroticism on PTSD and general health problems was direct. However, there was also a second indirect element whereby there was a relationship between neuroticism and coping, which, in turn, was associated with PTSD reactions and general health problems. One could say that this finding reflects a version of the dispositional approach to coping. That is, in coping with stressful situations (technological disasters), there are relatively stable personality factors that underpin or interact with the choice of coping strategy. To put it another way, one could say that this finding reflects criticisms of the contextual approach to coping such as the appraisal-based model proposed by Lazarus and his colleagues (2,49). The basic assumption from this appraisal-based model is that coping is a response to specific stressful situations as opposed to a stable feature of our personality. According to this appraisal-based model, coping is a dynamic process and is changeable over time, which makes it possible to respond to changing demands of a situation by reappraising it.
However, we feel that the MIMIC model could be interpreted as reflecting an integrative approach combining both the dispositional and the contextual aspects. That is, when the community residents experienced the technological disasters, they could have reacted to the disasters in terms of their disposition, ie, the interaction between stable personality factors and coping strategies. However, this dispositional approach would not prevent these residents from going through a cognitive appraisal process. That is, they could still cognitively appraise the disasters and perhaps adjust their coping responses. The degree to which these residents could adjust their coping strategies would probably depend partly on the strength of their dispositional factors. Seemingly, going through this process would lead to the development of PTSD symptoms and general health problems for some of the residents. One could argue that this interpretation bears similarities to the integrative conceptual framework proposed by Moos and Schaefer (4).
Some limitations of the study need to be noted. First, the sample was drawn from two populations exposed to technological disasters. Consequently, the nature of the study might limit the generalizability of the results to the victims of disasters of a different nature. To an extent, this limitation is inevitable because research has suggested that psychological responses to traumatic events do vary depending on the type of disaster (40). Thus, the generalizability of research findings will always be an issue in PTSD research because most PTSD studies focus on specific incidents. Future studies may, nevertheless, be carried out to improve the generalizability of these results by encompassing disasters of different natures. This, of course, would have important implications on time and cost.
Second, this study has not escaped an inherent methodologic difficulty in PTSD research, which is as follows. The existing community residents could represent a biased sample in that those who had participated in the study could have been those who were experiencing severe PTSD and viewed this participation as an opportunity for help. On the other hand, one could argue that it is equally possible that those who refused to participate were exhibiting a classic PTSD reaction, namely, avoidance. This would mean that those who participated in the study were the least affected.
To conclude, in this study, we developed a model that suggested that after technological disasters, community residents could develop PTSD symptoms and general health problems, and that these residents seemed to have developed death anxiety associated with the type of technological disaster that they experienced. However, the model showed that death anxiety was a psychological reaction separate from PTSD reactions and general health problems. The model also revealed that the most appropriate way of conceptualizing the complex roles that personality and coping played in the development of PTSD and general health problems among these community residents was through the interactive model. In other words, one could not treat personality and coping as mediating or independent factors. Instead, one should treat them as interactive factors, which combined to bring about the PTSD and general health problems of this group of residents.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Received for publication November 14, 2003; revision received September 13, 2004.
DOI:10.1097/01.psy.0000155675.56550.5f
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. P. Ginsberg, E. D. Ayres, L. B. Burriss, and D. A. Powell A Study of the Association between Retrospective Appraisal of Childhood Reactivity and Post-Discharge Traumatic Stress in Combat Veterans Traumatology, March 1, 2006; 12(1): 61 - 82. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
