Coping of school-age children in the sealed room during scud missile bombardment and postwar stress reactions.

Acknowledgement:

For the 6 weeks of the Persian Gulf war, the Israeli civilian population was attacked by 18 waves of Iraqi scud missiles. The possibility of chemical or biological warfare dictated a unique form of civil defense: the sealed room. Families hermetically closed a room or area in their home, and at the sound of the air-raid alarm, entered it and put on a gas mask. The sound of the alarm gave 2–5 min warning before missile impact. Detonation marked a second period of waiting: Until it was certain that no chemical or biological agents were present, gas masks remained on and the room stayed sealed. This second waiting period lasted a number of hours at the beginning of the war and was reduced to 1 hr as test procedures were expedited. In short, the multiple missile attacks posed a repeated demand for coping with the threat of loss of property and life.

Coping is conceptualized as “the person's cognitive and behavioral efforts to manage (reduce, minimize, master or tolerate) the internal and external demands of the person-environment transaction that is appraised as taxing or exceeding the person's resources” (Folkman, Lazarus, Gruen, & DeLongis, 1986, p. 572). Two major forms of coping have been suggested by Lazarus and Folkman (1984): problem-focused and emotion-focused coping. In problem-focused coping, the individual attempts to alter the sources of distress (e.g., rational efforts to solve problems or aggressive interpersonal efforts to change the situation). In emotion-focused coping, the person tries to regulate the emotional reactions to the threat by activities such as distraction, self-controlling escape, or avoidance.

The efficacy of these coping types is thought to be situationally contingent (Folkman et al., 1986). Emotion-focused coping is believed to be particularly useful in situations in which the source of threat cannot be influenced, whereas problem-focused coping is thought to be most effective in situations in which threat can be altered. For example, in the case of an incurable disease, once all that should be done has been done, reliance mainly on emotion-focused coping is considered effective and appropriate. By contrast, resorting to relaxation exercises to reduce tension caused by not studying for an examination would not be considered appropriate in place of the direct action of studying for the test (Roskies & Lazarus, 1980).

In the sealed room, little can be done to change the source of the threat by direct action. Thus, we hypothesized that children who persisted in activities aimed at changing the unchangeable situation would cope less effectively and have more severe postwar stress reactions. By contrast, children who relied more heavily on emotion-focused coping in the sealed room were expected to report less severe stress responses.

The theory of coping in stress situations presented here is based mainly on work with adults; there is no recognized taxonomy of coping behaviors and specifically for trauma of a repetitive nature (Hoffman, Levy-Shiff, Solberg, & Zarizki, 1992). Nonetheless, it is evident that the use of coping strategies differs with age. As Pynoos and Eth (1985) pointed out, adolescents have a greater range of cognitive skills, particularly those involving reappraisal and planning. These cognitive skills allow for coping in a more flexible and situationally adaptive mode. Indeed, Monaco and Gaier (1987) have shown that children's coping is related to their level of cognitive development. Therefore, we predicted that adolescents would be more likely to use situationally effective forms of coping than would younger children.

Subjects were drawn from nine schools from the central coastal region of Israel and totaled 492 children (227 boys and 265 girls) from Grades 5, 7, and 10. The ethnic composition of the sample comprised 74% of children of Middle-Eastern origin and 26% of Western origin.

Emotional reactions and coping behaviors

A questionnaire was developed to assess emotional responses and coping behaviors in the sealed room during missile attacks. Items were chosen on the basis of the coping literature (Folkman & Lazarus, 1980; Lazarus & Folkman, 1984; Vitaliano, Russo, Carr, Maiuro, & Becker, 1985) as well as pilot interviews with children regarding their experiences in the sealed room. The questionnaire included 41 items, 16 items dealing with emotional reactions and 25 with coping behaviors (see Table 1). Subjects were asked to indicate the degree to which each item described their characteristic reactions and behaviors in the sealed room on a 3-point scale: not at all (0), a little (2), and a lot (4).

Stress Reaction Questionnaire

This questionnaire consisted of 26 items. The first 20 questions were drawn from Frederick and Pynoos's (1988) Child Post-Traumatic Stress Reaction Index interview, converted to questionnaire format. In order to allow for a structured assessment of posttraumatic stress disorder (PTSD) according to the revised third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM–III–R; American Psychiatric Association, 1987), an additional six items were added so as to cover the DSM criteria not assessed by the Frederick and Pynoos scale.

Global symptom score (GSS)

The Frederick and Pynoos (1988) index is a continuous measure of reaction to extreme stress that is based on a structured interview. The interview was transformed into a questionnaire to permit data collection from a large sample within a short period of time as close to the war as possible. Examples of items are “I get scared when I think about the war,” “Since the war, I do things that I did when I was younger but have stopped doing (such as sucking my thumb, or sleeping with a teddy bear),” and “Since the war, I have more headaches, stomach aches or other aches.” Respondents were asked to rate each item on a 3-point scale: not at all (0), a little (2), and a lot (4), following Frederick and Pynoos (1988). On the basis of the responses to the 20 items, a global symptom score (GSS) was calculated as the sum of the item scores. This measure served as an independent variable to assess the efficacy of coping in the sealed room. For this purpose, subjects were divided into three groups on the basis of their percentile standing on the GSS: low, medium, and high.

PTSD criteria

According to the DSM–III–R (American Psychiatric Association, 1987), the diagnosis of PTSD requires the fulfillment of four criteria: (a) The person must be exposed to a traumatic event. This criterion was met by all of the subjects. (b) The person must reexperience the trauma as seen by at least one of the following: recurrent and intrusive distressing recollection of the event; recurrent, distressing dreams of the event; acting or feeling as if the traumatic event were recurring; and distress at exposure to events that symbolize the traumatic event. (c) The person avoids stimuli associated with the trauma or has a numbing of responsiveness. It requires at least three out of a list of seven, such as efforts to avoid thoughts or feelings associated with the trauma. (d) The person demonstrates increased arousal following the trauma, as seen by at least two of six symptoms, such as sleep difficulties.

Assessment of PTSD was based on responses to 17 Stress Reaction Questionnaire items equivalent to the 17 DSM–III–R symptoms of PTSD. For this purpose, 11 of the GSS items were used, along with an additional 6 items added to cover the criteria of the DSM–III–R for PTSD. These items were written using the same format as the other questions. Examples of items are “I am constantly tense and waiting to hear a siren,” “Since the war, I have more fears,” and “Today it is more difficult for me to get along with friends and family.” PTSD was scored dichotomously following the criteria of the DSM-III-R. The use of a questionnaire to assess PTSD has previously been demonstrated to be a reliable and valid approach (Solomon, Schwarzwald, Weisenberg, & Mikulincer, 1987).

Cronbach alpha reliability for the GSS measure was high (.86), indicating an internally consistent questionnaire. The validity of the PTSD and the GSS measures was assessed by comparing these scores with homeroom teachers' ratings of changes in academic performance since the war. Although the students completed the questionnaires, the teachers were asked to indicate for each of their students whether his or her academic functioning had improved, remained the same, or deteriorated since the war. Among those rated by teachers as showing a deterioration, 32% were categorized as suffering from PTSD, whereas for those rated as showing no change or improvement, only 18% were so categorized, χ2(1, N = 488) = 7.52, p < .01. Regarding GSS scores, students whose performance deteriorated scored higher, F(1, 485) = 13.18, p < .001, on GSS (M = 20.91, SD = 14.27) than their counterparts who were rated as showing no change or improvement (M = 15.52, SD = 11.30).

Recruitment of subjects and administration of questionnaires was subject to the review and supervision of the Israel Ministry of Education. Subjects were assured of confidentiality. Staff used identification numbers rather than names on all files, a procedure that permitted matching teacher and subject evaluations. It was also made clear to the students that participation was voluntary.

Data were collected in the schools during regular class time starting 3 weeks following the end of the war and lasting for 1 week. Subjects were told that the aim of the study was to document their experiences and feelings during the missile attacks and at present. It took approximately 25–45 min to complete the questionnaires.

Findings are presented in two sections. The first describes children's emotional responses and coping behaviors in the sealed room. The second section examines the relation between types of coping behaviors in the sealed room and postwar stress reactions.

Table 1 shows the means, standard deviations, and percentages of positive endorsement (reported at least “a little”) for emotional responses and coping behaviors in the sealed room. Although most children reported feeling tense (75.8%), all of the other highly endorsed items had a positive bent. Examples are “I felt sure that everything will be OK” (80.0%), “I felt relaxed” (78.2%), and “I acted as if everything was as usual” (74.2%). By contrast, infrequently endorsed items had a negative emotional bent. Examples are “I cried” (17.4%) and “I thought I was almost going crazy” (17.8%). Moderately endorsed items included mainly somatic responses: perspiring (31.0%), stomachaches (32.0%), and headaches (44.5%).

Examination of the coping behaviors indicated some activities reported by almost everyone. These common activities included information seeking by listening to the radio or watching TV (96.6%), talking to others in the sealed room (90.6%), and listening to noises coming from the outside (89.0%). Another set of commonly shared activities could be categorized as monitoring and assisting others. These included checking to see if everyone was okay (89.8%), helping others (89.8%), calming others (85.4%), and checking gas masks (82.6%). A third group of shared responses involved wishful thinking: wishing for a miracle (88.5%), wishing it was a false alarm (82.2%), and wishing the missiles would fall elsewhere (81.2%). All other activities were reported at a moderate level.

For data reduction purposes, each of the coping behavior items was initially analyzed by a one-way analysis of variance (ANOVA) to determine whether GSS scores or PTSD diagnoses would distinguish between respondents. Items that did not differentiate significantly were not analyzed further.

A principal-components factor analysis with varimax rotation was then computed for the remaining items. The factor analysis yielded five factors (see Table 2) accounting for 56.1% of the variance before rotation. The first factor, Checking, described activities primarily related to the continuous checking of the precautionary measures available in the sealed room. The second factor, Verbal Distraction, represented interpersonal, verbal distraction activities in which others were used as a distraction from the immediate threat of the missiles. The third factor, Reassurance Request, consisted of attempts to find solace in parents or from seeing the situation as usual rather than as a threat. The fourth factor, Distraction–Avoidance, was similar to the second factor and dealt with distraction behaviors, but here the activities were at the individual, intrapersonal level. The fifth factor—Wish Fulfillment—contained a wishful thinking motif. On the basis of factor analysis outcome, five coping factor scores were computed.

In order to determine whether the type of coping would vary across PTSD status, GSS score, grade level, and sex, a set of two-way multivariate ANOVAs (MANOVAs) was performed. Each of these variables was treated, in turn, as a single between-subjects variable and type of coping served as the within-subjects variable. In cases in which the interactions were significant, the MANOVAs were followed by one-way ANOVAs conducted separately for each coping factor. The relevant means are shown in Table 3. Because no significant variation arose for the fifth factor score, we do not discuss it further.

PTSD status

The MANOVA for PTSD status yielded significant effects for PTSD, F(1, 455) = 18.74, p < .001; for coping, F(4, 452) = 6.22, p < .001; and for the PTSD × Coping interaction, F(4, 452) = 16.92, p < .001. To determine the source of the interaction, we conducted one-way ANOVAs separately for each of the coping factors.

As can be seen in Table 3, PTSD children obtained higher Checking scores, F(1, 455) = 25.64, p < .001, higher Reassurance Request scores, F(1, 455) = 48.87, p < .01, and lower Verbal Distraction scores, F(1, 455) = 7.01, p < .01, than did non-PTSD children. No significant differences arose in the case of intrapersonal Distraction–Avoidance scores.

GSS scores

The MANOVA yielded significant results for GSS, F(2, 453) = 13.34, p < .001, and for the GSS × Coping interaction, F(8, 900) = 16.96, p < .001. Subsequent one-way ANOVAs revealed significant differences in checking, F(2, 453) = 26.78, p < .001, reassurance, F(2, 453) = 33.88, p < .001, and interpersonal distraction, F(2, 453) = 14.75, p < .001, but not in intrapersonal distraction. Post hoc Scheffé tests indicated significant differences (p < .05) between all three GSS levels in the case of checking and reassurance: Those with greater symptoms reported higher checking and reassurance seeking. In the case of interpersonal distraction, post hoc Scheffé tests revealed that the high and moderate symptom group used less interpersonal distraction than did the low symptom group.

Grade level

A MANOVA for Grade Level × Coping Factors yielded a significant grade-level main effect, F(2, 454) = 4.87, p < .01, and a significant Grade Level × Factor, F(8, 902) = 9.04, p < .001.

The one-way ANOVA for the Checking factor was significant, F(2, 454) = 3.39, p < .05. Although the means indicated an apparent reduction of checking activities with increasing grade level, Scheffé contrasts showed that only 5th graders differed (p < .05) from 10th graders.

The interpersonal Verbal Distraction factor also differed significantly across grade levels, F(2, 454) = 11.74, p < .001: Scheffé tests revealed that 5th graders engaged less in verbal distractions (p < .05) than did either the 7th or 10th graders, who did not differ from each other.

The Reassurance Request factor also varied across grade levels, F(2, 454) = 14.40, p < .001. Fifth graders engaged more in requests for reassurance (p < .05) than did either the 7th or 10th graders, who did not differ from each other.

The Distraction–Avoidance factor differed across grade levels, F(2, 454) = 9.77, p < .001. Fifth and 7th graders engaged less in intrapersonal activity (p < .05) than did the 10th graders.

Sex

The MANOVA for sex indicated a significant interaction of sex by coping, F(4, 445) = 2.66, p < .05. Subsequent one-way ANOVAs yielded significant results only for the Reassurance Request factor, F(1, 448) = 9.63, p < .001. Girls sought reassurance more than did boys.

Children's reports concerning the sealed room indicated that, despite a shared feeling of tension, the basic emotional stance was a positive form of detached optimism. The most prevalent forms of coping involved elements of information seeking, monitoring others, and wishful thinking. As expected, the type of coping was associated with postwar stress reactions: Greater emphasis on coping activities directed at the threat (i.e., more problem-focused coping) and reduced emphasis on avoidance and distraction (emotion-focused coping) were associated with more severe postwar stress reactions. The developmental findings indicate that the older adolescents were more likely to embrace a pattern of effective coping, using greater emotion-focused coping and lesser problem-focused coping than their younger counterparts.

Our findings on missile threat, like Cohen and Lazarus's (1973) work on surgery, indicate that there are times when avoidance is highly effective: Persisting in problem-focused coping in a situation that cannot be changed can lead to undesirable consequences. These findings are consistent with the theoretical work of Lazarus (1983) and his colleagues (Folkman et al., 1986; Lazarus & Folkman, 1984; Roskies & Lazarus, 1980). It is also consistent with a series of studies reviewed by Taylor and Brown (1988), all of which showed that accurate reality testing may in certain situations be associated with poor psychological adjustment, whereas the adoption of certain illusions may enhance mental health.

The reported appearance of positive feelings in the sealed room side by side with the admission of tension raises the question of denial. There is no doubt that the missile attacks aroused fright and apprehension in most people. The children found themselves in a passive, helpless position pervaded by a fear of the unknown (e.g., Would it hit? Was it conventional or unconventional?). Live coverage of wreckage and rescue workers evacuating casualties broadcast on TV made the threat even more tangible. In this light, the reports of primarily positive emotional reactions in the sealed room may be seen to reflect a denial of threat, an attempt to encapsulate and isolate scary feelings aroused by the war.

The issue of denial may also be connected with the finding of less effective coping in the younger children. Burke, Borus, and Burns (1982), studying children in flood areas, found that the ability to use denial tends to increase with age. The poorer coping of the younger children may thus be a result of cognitive developmental limitations in their ability to deny or distort external threats.

In evaluating the study findings, it is important to call attention to two methodological issues. The first relates to denial and the retrospective assessment of coping in the sealed room. Our study findings indicate that those who focused on emotional coping involving elements of denial also indicated less postwar stress reactions. On the basis of our a priori theoretical analysis and earlier findings on adults facing trauma, we interpret these findings to be consistent with the hypothesis that denial in the sealed room led to reduced postwar stress reactions. Yet, it is possible that denial took place at later periods, up to and including the time of data collection, and led some subjects to alter their recollections of both coping and symptoms. Further research involving multiple time sampling would aid in clarifying this possibility.

Another issue involves the reliance on self-report measures whose psychometric features are not fully established, although both the PTSD and the GSS measures were developed from highly regarded source works. These included the symptom list of the DSM–III–R (American Psychiatric Association, 1987) and the interview work of Frederick and Pynoos (1988), who have specialized in the systematic assessment of PTSD in children. Furthermore, initial work has suggested the psychometric merits of these measures. The use of self-report instruments for the assessment of PTSD has been shown to be a valid and reliable method with adults (Weisenberg, Solomon, Schwarzwald, & Mikulincer, 1987). In the present case, teachers' reports of decline in academic performance offered preliminary evidence for the validity of our stres measures. Finally, although clinical interviews might have been more sensitive to nuances or denial tendencies, their use would have been costly and unwieldy within the constraints in our study. The self-report instruments enabled an examination of a large sample within a short time in a period falling on the heels of the event itself.

1  The ethnicity distribution was based on estimation provided by the school principals.

2  It was legitimate to conduct this factor analysis, despite the high rate of endorsement of some of the coping items because the skewedness of the items ranged from −.77 to .68, with one exception (Item 10).

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author.

Burke, J. D., Borus, J. F., & Burns, B. J. (1982). Changes in children's behavior after a natural disaster. American Journal of Psychiatry, 139, 725–730.

Cohen, F., & Lazarus, R. S. (1973). Active coping processes, coping dispositions, and recovery from surgery. Psychosomatic Medicine, 35, 375–389.

Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a middle-aged community sample. Journal of Health and Social Behavior, 21, 219–239.

Folkman, S., Lazarus, R. S., Gruen, R. J., & DeLongis, A. (1986). Appraisal, coping, health status, and psychological symptoms. Journal of Personality and Social Psychology, 50, 571–579.

Frederick, C., & Pynoos, R. S. (1988). Child Post-Traumatic Stress Reaction Index. Los Angeles: UCLA Neuropsychiatric Institute and Hospital.

Hoffman, M. A., Levy-Shiff, R., Solberg, S. C., & Zarizki, J. (1992). The impact of stress and coping: Developmental changes in the transition to adolescence. Journal of Youth and Adolescence, 21, 451–469.

Lazarus, R. S. (1983). The costs and benefits of denial. In S.Breznitz (Ed.), Denial of stress (pp. 1–30). Madison, CT: International Universities Press.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer.

Monaco, N. M., & Gaier, E. L. (1987). Developmental level and children's responses to the explosion of the Space Shuttle Challenger. Early Childhood Research Quarterly, 2, 83–95.

Pynoos, R. S., & Eth, S. (1985). Developmental perspective on psychic trauma in childhood. In C. R.Figley (Ed.), Trauma and its wake (Vol. 2, pp. 36–52). New York: Brunner/Mazel.

Roskies, E., & Lazarus, R. S. (1980). Coping theory and the teaching of coping skills. In P. O.Davidson & S. M.Davidson (Eds.), Behavioral medicine: Changing health lifestyles (pp. 38–69). New York: Bruner/Mazel.

Solomon, Z., Schwarzwald, Y., Weisenberg, M., & Mikulincer, M. (1987). Post-traumatic stress disorder among front line soldiers with combat stress reaction: The 1982 Israeli experience. American Journal of Psychiatry, 144, 448–454.

Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective on mental health. Psychological Bulletin, 103, 193–210.

Vitaliano, P. P., Russo, J., Carr, J. E., Maiuro, R. D., & Becker, J. (1985). The Ways of Coping Checklist: Revision and psychometric properties. Multivariate Behavioral Research, 20, 3–26.

Weisenberg, M., Solomon, Z., Schwarzwald, J., & Mikulincer, M. (1987). Assessing severity of post-traumatic stress disorder: The relationship between dichotomous and continuous measures. Journal of Consulting and Clinical Psychology, 55, 432–434.

Submitted: October 25, 1991 Revised: May 16, 1992 Accepted: June 16, 1992