Introduction

Many present global conflicts take the form of violence and attacks directed against civilian communities, spreading danger, fear and insecurity in the population at large. Foremost at risk are children and youth who are forced to deal with current danger at a time when their entire routine is thrown into turmoil.

The majority of children exposed to war and conflict will experience substantial psychological morbidity (Shaw, 2003). A rapidly growing body of research has documented both short- and long-term negative consequences of exposure to political violence and war. Short-term effects include distress, shock, fear, phobic avoidance of public places, anger and emotional pain (El Zein & Ammar, 2011; Joshi & O’Donnel, 2003), and aggressive behavior (Guttmann-Steinmetz, Shoshani, Farhan, Aliman, & Hirschberger, 2012). Long-term effects include posttraumatic stress disorder (PTSD) (Finzi-Dottan, Dekel, Lavi, & Su’ali, 2006; Pat-Horenczyk et al., 2009; Rosner, Powell, & Buttollo, 2003), anxiety and depression (Dyregrov, Gupta, Gjestad, & Mukanoheli, 2000; Gupta & Zimmer, 2008; Lonigan, Phillips, & Richey, 2003), and sub-clinical symptoms (Slone & Shechner, 2009). In addition, there appears to be a dose–effect relation, insofar as frequency and severity of events are associated with the severity of the trauma response (Wolmer, Laor, & Yazgan, 2003).

Particularly in war-torn areas, therapeutic interventions have been developed for emergency situations when routine is disrupted by events such as war, armed attacks, rocket firings, and terrorist attacks (De Berry, 2004; Paardekooper, 2002). A crucially sensitive period also follows when the firing stops and youth are forced to return to routine life amidst surrounding destruction and remnants of war. The ceasefire produces an illusion of normality with return to school, leisure activities, and family routine; whereas, in fact, adolescents are in need of a period of psychological rehabilitation. This could be a high risk period for many children and youth since post-war periods represent a time for processing the trauma, interpreting experiences, and adjustment (Wessells & Monteiro, 2004).

The need to maximize youth resilience after exposure to war is critical in order to resolve the crisis successfully (Klingman & Ben Eli, 1981). Primary prevention during this critical period is aimed at reducing risk factors and enhancing protective factors among general population groups with various levels of risk for developing mental disorders because of their life experiences (Cowen, 1983; Offord, 2000). In this setting, common primary prevention programs comprise organizational and educational interventions rather than individual and clinical ones. Such programs typically utilize the two components of the positive mental-health perspective: helping adolescents adjust to the painful life experience and providing growth-producing experiences (Klingman, 2002).

Because of its central position in adolescents’ lives, schools are the optimal setting for implementing preventative intervention programs that are cost-effective and encompass large populations of children in their natural environment. Common programs of primary prevention are of an anticipatory nature and are carefully preplanned; but when the school is faced with unexpected externally-imposed stress, there is no time for psychological preparation for the events to come. Thus, new models must be put into practice, and different strategies should be employed that can resolve the crisis.

The present study evaluates a school-based primary prevention intervention to promote adolescents’ coping in the immediate aftermath of war exposure. The research was conducted in Ashkelon, a city in Israel that was heavily affected by missile attacks before and during Operation Cast Lead, immediately with the return to school at the conclusion of the military operation. This study evaluated the efficacy of a school-based intervention program instituted at the point of the ceasefire. The study employed a repeated measures design that assessed pre- to post-test modifications in resilience factors and psychological outcome measures, namely behavioral difficulties and psychological symptoms and distress, in an experimental versus a control condition. The two resilience factors promoted in the intervention program and assessed for pre- to post-test modification were mobilization of social support and self-efficacy.

In the Aftermath of War in Israel

Operation Cast Lead was a 3-week armed conflict that occurred in the Gaza Strip and Southern Israel in December 2008. The operation began as a military offensive by Israel following Hamas’s rocket attacks on Southern Israel and included a heavy air and ground offensive in the Gaza Strip. Hamas intensified its rocket and mortar attacks against Southern Israel, reaching Ashkelon and Sderot and also the major cities of Beersheba and Ashdod for the first time.

Ashkelon lies 15 km from the Gaza Strip and came under sporadic rocket fire for several months prior to Operation Cast Lead. Despite the imperfect aim of these rockets, they caused deaths and injuries as well as significant damage to homes and property and major community psychological distress. During Operation Cast Lead, the city came under heavy rocket and missile attack, which led to school closings and children spent long periods indoors in bomb shelters. The close proximity of Gaza to Ashkelon allows citizens only 15 s to reach a shelter following an alarm. Everyday activities in the city ground to a halt. The operation ended on January 18 when Israel declared a unilateral ceasefire, and when Hamas announced a 1-week ceasefire 12 h later. Almost immediately, children returned to routine life with the opening of schools and resumption of everyday activities.

School-Based Primary Prevention Intervention at Point of Ceasefire

Exposure to traumatic events associated with war and its aftermath affects mental health at both the individual and community levels. Studies have shown a relation between amount and severity of previous exposure to conflict, war, and political violence, and type and severity of psychiatric outcomes (Slone & Shechner, 2009).

The traumatic circumstances of the missile attacks and sharp changes from war to routine schedules emphasized the need to provide youth with a transitional environment during which they could process the experiences and rehabilitate. These conditions prompted the present study, in which we sought to administer and examine the efficacy of a school-based primary prevention intervention program immediately post ceasefire to assist adolescents in coping with their recently experienced crisis.

Primary prevention aims to forestall the development of dysfunction, disorder or pathology (Caplan, 1964). The advantage of primary prevention is its potential to be employed as a systematic school effort to maximize children’s adjustment after unfavorable exposure to traumatic conditions, to facilitate their prompt return to full participation in all aspects of everyday life.

In post-war and conflict circumstances, both universal and targeted mental health primary prevention programs have been used. Universal programs are offered to the general population in a setting considered high risk for development of disorder or dysfunction without reference to those at particular risk. Targeted or selective programs are directed toward individuals or groups at elevated risk for a disorder or condition of interest. Children and adolescents may be categorized as being at high risk on the basis of their own characteristics or on the basis of the high risk group to which they belong (Offord, 2000).

In line with the rationale of utilizing a natural school setting and educational staff following mass trauma, several school-based interventions have been developed, evaluated and proven effective (Berger & Gelkopf, 2009; Udwin, Boyle, Yule, Bolton, & O’Ryan, 2000). Several programs have been universally implemented to all students in the research population and these programs are relevant when an entire population has been exposed to traumatic occurrences. Other programs have targeted at-risk samples.

For example, a targeted intervention implemented in post-war Bosnia, which consisted of distributing psycho-education material, training in coping strategies and providing specialized consultation for high risk children, was successful in decreasing post-trauma symptoms (Layne et al., 2008). Targeting traumatized refugees and asylum-seekers from war-affected countries, Ehntholt, Smith, and Yule (2005) delivered a manualized cognitive behavioral therapy (CBT) classroom-based intervention that produced significant improvements in overall behavioral difficulties and emotional symptoms.

Using a targeted Classroom-Based Intervention (CBI) based on creative-expressive and experiential therapy, cooperative play and CBT for children exposed to armed conflict, Jordans et al. (2010) and Tol et al. (2008) demonstrated short-term improvement in social-behavioral and resilience indicators among students in Nepal and Indonesia, respectively.

In Israel, with its prolonged history of war and conflict, several interventions have been conducted and empirically evaluated. Following continuous terrorist attacks in Israel since September 2000, Gelkopf and Berger (2009) administered a universal school program to adolescents based on homeroom teachers’ delivery of psycho-educational material, skill training and resilience enhancing strategies. Three month post-intervention evaluations showed significant decreases in posttraumatic, depressive and somatic symptoms as compared to a wait-list control group. A similar universal teacher-delivered skill-oriented and present-focused intervention was found effective in preventing and reducing PTSD symptoms among adolescents exposed to continuous missile attacks in Sderot (Berger, Gelkopf, & Heineberg, 2012).

A program developed by the Israel Trauma Center for the Victims of Terror and War (NATAL), which was delivered universally to elementary school children, was effective in producing significant improvement in mental health and adjustment (Berger, Pat-Horenczyk, & Gelkopf, 2007). This program was based on guidance with psycho-educational material and skill training for stress reduction including meditation, bio-energy exercises, art therapy, and narrative techniques.

A universal teacher-delivered protocol focusing on enhancing personal resilience proved successful in enhancing resilience of children in the north of Israel after the 2006 Lebanon War, during which they had been exposed to daily massive rocket attacks (Wolmer, Hamiel, Barchas, Slone, & Laor, 2011a). Participating children showed significant improvements in mood and post-trauma symptoms 3 months after termination of the intervention. In addition, a preparatory teacher-based stress resistance intervention implemented in grade schools before the rocket attacks that occurred during Operation Cast Lead was successful in reducing adverse psychiatric outcomes (Wolmer, Hamiel, & Laor, 2011b).The common denominator of most of these teacher-led prevention efforts is based on psycho-education and modalities such as expressional and creativity techniques or emotional, cognitive and behavioral stress reduction techniques (Jordans, Tol, Komproe, & De Jong, 2009).

A different approach was instituted by Slone and Shoshani (2008) in a study showing the effectiveness of a resilience-promoting school intervention for adolescents during a peak of national terrorism in the Second Palestinian Intifada. The program was developed to enhance personal characteristics found to function as resilience variables, rather than attempting to reduce psychopathology directly.

In extreme war circumstances, the major therapeutic effort has been mitigation of psychological symptoms and prevention of subsequent fully developed disorders (Jordans et al., 2009). However, consideration should be paid as to when early symptomatology should be addressed. Preventive interventions are valuable, particularly when they do not directly address symptomatology. Despite some evidence of the success of post-trauma exposure and debriefing techniques, the effectiveness of these approaches is debatable because of the dangers inherent in the resurfacing of frightening material and the disruption of natural healing processes (Rose & Bisson, 1998). The present study used a resilience-enhancing strategy based on the rationale of the importance of promoting resilience as a preventive process for pathology development in the general school student population in a high-risk situation.

The innovation of the present primary prevention intervention lies in its instigation at a strategic period known to be critical for prevention of deterioration into chronic disorders (Betancourt & Williams, 2008). This universal school program aimed at reaching large numbers of adolescents during this critical period, complemented by individual referral for more specialized treatment in severely affected adolescents.

Enhancing Post-Crisis Resilience

Resilience refers to the process or capacity by which successful adaptation is achieved despite challenging or threatening circumstances (Werner, 2000). For children, resilience can be expressed by maintenance of competence despite high risk or stressful circumstances and in successful recovery from trauma (Masten, 2001). Detection of resilience factors that moderate the negative effects of trauma exposure would aid attempts to protect children against harmful outcomes (Rutter, 2000; Werner, 2000).

In this study, resilience was operationalized as factors that moderate the relation of trauma exposure and outcome. Detection of these factors is dependent on assessing the type and severity of risk exposure. This is imperative in environments in which populations are exposed to heterogeneous war or conflict-related traumatic events over a protracted period leading to a wide variety of mental health difficulties.

The complex issues of measurement of protracted political violence in a conflict environment has been addressed in several ways ranging from assessments of consequences of exposure to acute events (Brown & Goodman, 2005; McDermott, Duffy, & McGuiness, 2004) to attempts at evaluating the severity of cumulative exposure (Slone, 2006). In the present study, exposure was measured by means of a political life events (PLE) scale, and resilience factors were conceptualized as those constructs that moderate successfully between this type of traumatic exposure and adverse outcomes. The legitimacy of measuring political violence exposure quantitatively is increasingly being recognized and several pertinent inventories have been constructed and utilized (Hollifield et al., 2005; Jones & Kafetsios, 2005; Smith, Perrin, Yule, Hacam, & Stuvland, 2002). The PLE measure used here enabled assessment of the relation between exposure and psychological outcomes and detection of those variables that moderate this relation.

This conceptualization of resilience was tested in a large seven-year research project using a PLE, moderating factors and outcome paradigm that resulted in the specification of successful personality and social resilience factors (Slone, 2006). Detection of successful resilience factors in the particular context of war allows for the design and evaluation of intervention programs that can attenuate associations between stress and adverse outcomes. Protective factors can be translated into areas for intervention allowing efforts to be directed toward creating climates that encourage the development and nurturance of broad expressions of these characteristics and skills across various contexts. In the present study, the intervention was based on the two factors of social support and self-efficacy found to be the central resilience factors that mitigate negative psychological outcomes from war and conflict in a large seven-year research project (Slone, 2006).

Social support, referring to the perception that one is cared for and part of a social network of mutual assistance and obligations, is known to attenuate the experience of stress, enhance well-being and speed mental and physical recovery (Seeman, 1996). The decision to solicit and receive support depends both on individual tendencies and on the availability of support resources. Strong social networks are key predictors of psychiatric resilience since they facilitate processing of and coping with traumatic events, thereby acting as a safeguard against the development of psychopathology (Pina et al., 2008). In disaster situations, peer support can provide significant comfort on the basis of shared experiences (Moore & Varela, 2010).

Promoting the utilization of social support is one of the main objectives in teaching adolescents effective coping skills. The presence of a significant adult who acts as a role model, providing emotional support and promoting self-esteem and effective coping mechanisms can reduce the potential short- and long-term negative outcomes of terrorism-induced stress (Fremont, 2004). The absence of a competent, caring adult in the life of a child exposed to highly adverse conditions creates high risks for maladjustment (Masten & Coatsworth, 1998).

In addition, there are several personal characteristics known to improve prognosis after traumatic exposure, central among which is a strong sense of self-efficacy (Benight & Bandura, 2004; Charles et al., 1999). The second factor promoted in our intervention was self-efficacy, which has been defined as the belief in one’s self-competence and personal capability to solve problems and execute actions required to manage life situations (Bandura, 2001). Research has shown that children with below average problem solving skills and a low sense of their own competence have more difficulty in managing threatening circumstances (Masten & Coatsworth, 1998). This finding is congruent with findings that demonstrate that life satisfaction during youth co-occurs with high levels of self-efficacy (Fogle, Huebner, & Laughlin, 2002).

Hypotheses

The study posited two hypotheses and an exploratory question. In the first hypothesis, we predicted a main effect of exposure to political violence on self-efficacy, ability to mobilize support, psychological distress and behavioral difficulties, such that high PLE exposure would be associated with high levels of symptomatology and low levels of self-efficacy and perceptions of ability to mobilize support.

In the second hypothesis, we predicted that participants in the intervention group would exhibit a greater increase from pre- to post-intervention in self-efficacy and perception of the ability to mobilize support, lower levels of psychological distress and symptoms, and lower levels of behavioral difficulties than control group participants.

In order to examine whether intervention efficacy differed by level of exposure to political violence, we explored the effects of the interaction between level of political violence exposure and type of intervention on changes in self-efficacy, perceptions of mobilization of support, psychological distress and behavioral difficulties.

Methods

Participants

Participants were 179 adolescents (82 girls, 97 boys) aged 16.1–17.9 (M = 16.3, SD = 1.1) from grade 10 in two demographically similar high schools in Ashkelon in southern Israel. Allocation to study conditions followed a two-step procedure. First, two schools were selected for participation from four eligible high-schools in Ashkelon. Exclusion criteria for schools were ultra-orthodox religious schools and demographically incomparable schools. Second, in each school, classes were randomly allocated to the intervention or control condition by an experimenter who did not have any prior knowledge of the features of the classes and was not acquainted with any school staff. In one school, three of the six classes in the grade level were randomly assigned to the intervention condition (two classes) and to the control condition (one class). In the other school, three of the five classes in the grade level were randomly assigned to the intervention condition (one class) and to the control condition (two classes).

All 208 school students in the six classes across the two schools were eligible to participate in the study. All parents and school students complied with informed consent requirements. A total of 29 school students did not complete the study, 21 due to absence and eight due to refusal to complete questionnaires. Across schools, in total, the intervention group comprised 94 and the control 85 adolescents, approximately evenly divided by gender in each condition. A participant flow diagram is presented in Fig. 1.

Fig. 1
figure 1

Participant flow diagram

Full size image

Instruments

Brief Symptom Inventory

The Brief Symptom Inventory (BSI; Derogatis & Spencer, 1982), the abbreviated version of the SCL-90-R, comprises 53 self-report symptom items rated on a 4-point Likert scale to identify clinically relevant psychological symptoms in adolescents and adults. The BSI was designed for use with adolescents from the age of 13 and reports norms for that age. The inventory provides distress indices and symptom load assessment on 10 subscales—somatization, obsessive–compulsive, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, psychotic ideation and miscellaneous. For a single summary measure, Derogatis and Spencer (1982) recommend the Global Severity Index (GSI) calculated as the average of ratings assigned to symptoms. The BSI has yielded good psychometric properties: Cronbach’s alpha coefficients of 0.71–0.81, high test–retest reliability (correlations between 0.78 and 0.90) and high concurrent validity with the Minnesota Multiphasic Personality Inventory (Butcher et al., 1989). The inventory has been back-translated into Hebrew with good internal consistency and concurrent validity (Canneti, Shalev, & Kaplan de-Nour, 1994) and has been widely used in Israel (Slone, 2006). In the current study, Cronbach’s alpha coefficients were α = 0.97 for the GSI and α = 0.79–90 for the subscales.

Strengths and Difficulties Questionnaire (SDQ)

The SDQ (Goodman, Meltzer, & Bailey, 1998) is a 25-item self-report questionnaire assessing psychiatric disturbance. The questionnaire yields a total difficulties score and five scales comprised of five items each—Emotional symptoms (e.g., I am often unhappy, depressed or tearful), Conduct problems (e.g., I take things that are not mine from home, school or elsewhere), Hyperactivity scale (e.g., I am constantly fidgeting or squirming), Peer problems (e.g., I am usually on my own), and a Prosocial scale (e.g., I often volunteer to help others). In this study, the four problem scales were calculated as a sum to produce a Behavioral Difficulties score.

This study used the official Hebrew version of the SDQ as translated by the Israeli Ministry of Health. The instrument reports excellent criterion validity for community and clinic samples and high cross-informant self-report to parent- and teacher-rated correlations (Goodman et al., 1998). The Hebrew version reports acceptable to good internal consistency (α = 0.51–0.72) and good construct, concurrent, and discriminant validity (Mansbach-Kleinfeld, Apter, Farbstein, Levine, & Ponizovsky, 2010). In our sample, Cronbach’s alphas were acceptable for the total difficulties score (0.78), and for the four problem scales (0.68–0.70).

Social Support Matrix

The Social Support matrix was designed to assess adolescents’ self-perceptions of ability to mobilize support from different possible providers of support (Slone, Shoshani & Paltieli, 2009). The matrix is a chart with providers of support listed along the horizontal axis and provisions of support listed along the vertical axis. Provider categories were family, teachers, friends, community professionals, religious leaders and unspecified categories to be filled by the respondent, if relevant. Provision categories included providing a sense of security, perceived as being empathic, allowing for sharing, and offering practical advice.

In each cell produced by this matrix, respondents marked the extent of their ability to receive each support provision from each provider on a 0–3 scale. A composite support perception score was derived by summing ratings for providers across all provisions that yielded coefficients of α = 0.83 on the pretest and α = 0.86 on the posttest.

Self-Efficacy Questionnaire for Children

The 24-item self-report Self-Efficacy Questionnaire for Children (SEQ-C; Muris, 2001) measures adolescents’ beliefs about their competence on each of three 8-item subscales—social, academic and emotional. Social self-efficacy measures perceived capability for peer relationships (e.g., How well can you become friends with other children?), academic self-efficacy measures perceived capability for management of learning behavior and problem solving (e.g., How well can you study a chapter for a test?), and emotional self-efficacy measures perceptions of coping with personal emotions (e.g., How well can you give yourself a peptalk when you feel low?). High construct validity has been shown between SEQ-C scores and low scores on depression, anxiety and neuroticism scales (Muris, 2001). The present study yielded α = 0.85 on the pre-test and α = 0.86 on the post-test.

Political Life Events Scale (PLE)

The PLE scale (Slone, 2006) contains 20 event items that participants mark for exposure over the past 6 months. The PLE severity score is calculated by summing all items marked positive for exposure, weighted on the basis of previously determined assessments of severity by independent same-aged adolescent judges according to the formula: mild items (e.g., A security drill at school) multiplied by 1, moderate items (e.g., Harm to property as a result of terrorism, political violence or rocket attacks) multiplied by 2, and severe items (e.g., Injury to a close family member as a result of war or military circumstances) multiplied by 3. Generalized weighting based on judges’ personal or vicarious experience is an acceptable scoring system for life events (Paykel, Prusoff, & Uhlenhuth, 1971) with adequate predictive value for psychological adjustment. For statistical analysis purposes (Slone & Shechner, 2009), subjects were split above and below the median for PLE (Median = 10.00), yielding two groups reflecting high and low exposure.

There is no justification to calculate an internal consistency score for PLE since there is no theoretical rationale to expect consistency in exposure to discrete events. Validity studies assessing cross-nationality transferability for Jewish and Arab–Israeli youth, for Palestinian youth and for black and white South African adolescents have yielded excellent results and have shown predictive validity for these communities (Slone, 2006). Test–retest scores have ranged from r = 0.86 to r = 0.94 (Slone & Shechner, 2009).

Intervention and Control Conditions

Intervention Group

For each resilience factor—mobilization of support and self-efficacy—a handbook was developed comprising theoretical explanations of the concepts, a series of experiential activities aimed at strengthening the factor, and all the complementary material necessary for implementation of the activities in the classroom. The program had previously been developed on the basis of several pilot studies with different groups of children. The program functioned as a training workshop in which school counselors and class teachers participated in a series of three group seminars, each of 5 h duration, aimed to impart information and strategies for implementing the activities in the handbooks. Teachers were responsible for implementing activities in the classroom and school counselors were given responsibility for overseeing implementation of the program in the school. Information collected from the school principals regarding characteristics of participating teachers in both the intervention and control groups revealed that they were all experienced and highly trustworthy, and that their cooperation was assured.

The workshops adopted an experiential focus in which the educational staff both received didactic instruction about theoretical aspects of the two resilience factors, social support and self-efficacy, and participated in some of the actual activities themselves. Thus, in the first seminar, participants learned about mobilization of social support; for example, research evidence of the contribution of social support to adolescents’ coping with war stress and types of social support relevant to adolescents. In addition, teachers familiarized themselves with activities aimed to enhance this factor. The activities in this module included the following six topics: Identification of existing and new personal sources of support; Discovery of types of support available from each source; Mapping existing styles and developing new styles of giving and receiving support; Exploring times to be alone and times to be with others; Group cooperation and sharing activities; and Stressing the path forward with constructing group projects for social and community involvement. Thereafter, teachers were responsible for implementing the six sets of activities presented in the handbook in the classroom twice weekly over a 3-week period.

The second seminar began with sharing feedback from the experience of implementing the classroom activities and then moved to the second factor of self-efficacy, using the same format as the first seminar. In the second seminar, teachers and counselors were given the theoretical background to the concept of self-efficacy and its relation to different types of coping and strategies for problem solving. Thereafter, they participated in several activities presented in the handbook. The six components of this module included: Identifying personal and others’ coping patterns and strategies; Problem-solving activities in certain and uncertain circumstances; Constructing outdoor activities that meet emotional ventilation needs; Training in self-relaxation and self-control techniques; Exploring personal strengths and their day-to-day application; and Identifying ways to achieve a sense of self-empowerment.

Again, teachers administered the six sets of activities in this handbook during lessons twice weekly for a period of 3 weeks. In this way, the entire program was administered in the classroom twice weekly for a period of 6 weeks immediately on the students’ return to school after the ceasefire declaration.

In order to verify that the intervention was administered in full, teachers completed a report after each intervention lesson documenting their provision of the lesson, ease of administration of activities, difficulties that may have arisen, deviations from the manualized activities, and class cooperation. To increase intervention fidelity, school counselors checked randomly that the program was being implemented in the classrooms during the assigned lessons and completed a report noting their impressions of the administration and any problems that had arisen. All participating teachers reported full administration of the activities in the two handbooks and excellent class cooperation with the program. All teacher and counselor reports were carefully inspected by the research team and irregularities were discussed. No particular irregularities, difficulties or deviations from the program were reported that would warrant exclusion of any classroom from the study.

The cycle terminated with a seminar that included feedback from implementing classroom activities and a summary of the group process both for the seminar participants and for schoolchildren. Seminars were co-facilitated by two clinical psychologists trained in group dynamics.

The classroom program included activities, discussions, reading poems and stories and viewing movie clips of popular Hebrew songs dealing with self-efficacy and social support. An example activity for mobilization of support was The Support Map, in which schoolchildren produced a class map of support agents available to the group and the frequency with which they were utilized by individual students. An example activity for self-efficacy was individual children’s selections from a pool of picture cards those that typified their strategies for coping under stress. Students produced a ‘package’ that depicted their personal repertoire and that promoted a discussion of others’ coping repertoires.

Most activities stimulated a lively discussion and frequently generated emotional responses, intense and novel group processes, and new channels for communication among students and between educators and students. In this way, the circles of influence of the program extended from individual teachers to the educational staff as a group, from individual students to the classroom as a group, and finally to teacher–student interactions. The school counselor and teacher training workshops were considered to be of great value in allowing school staff to process their own difficult experiences during the war and aiding them in separating their own experiences and reactions from those of their students. In addition, workshops were reported to provide practical and useful tools to respond to adolescents’ stress and the emotionally loaded material that emerged in the class after they returned to school.

Control Group

The control program consisted of the same format as the intervention program insofar as teachers and school counselors participated in a series of three group seminars. However, seminars consisted of imparting information and discussing general issues relating to adolescence. Control seminars were conducted by the same two clinical psychologists who carried out the intervention seminars. During school hours, teachers continued with the regular curricula of social science lessons that were designed to impart information and to discuss issues pertinent to adolescent development. The control groups did not participate in any resilience-promoting experiential strategies or activities as utilized in the intervention groups. The intervention and control conditions ran parallel to each other over the same period.

Procedure

After receiving authorization for the study from the University Ethics committee, the Regional Director of School Psychological and Counseling Services, and the school principals, the schools secured passive consent from parents and written consent from the adolescents themselves. Parental passive consent was obtained by sending parents or guardians a letter informing about the intervention and research and, if they objected to their child’s participation in the study, parents were requested to return notification to this effect. No parents objected to their children’s participation in the study.

The intervention began 1 week after the ceasefire declaration. Control groups received the intervention at the termination of the study. Teacher and counselor seminars for both intervention and control groups were conducted concurrently by the same prevention specialists who were trained in clinical psychology and group process and who co-facilitated each seminar. Except for content, intervention and control seminars had the same formal format in terms of duration, combination of written, powerpoint, and discussion materials, and length of time spent in training. Teachers and counselors were asked not to discuss their experiences outside of the seminars, and many reported being too busy during the school day to do so. The duration of each seminar was approximately 5 h. The entire intervention program was delivered for a period of 6 weeks.

All questionnaires and intervention materials were tested in a series of pilot studies as part of an ongoing research program on prevention interventions. Before commencement of the program and after its termination, the test battery was administered to participants in the classroom concurrently in the intervention and control groups by four psychology students blind to the status of the class. In order to match pre- and post-test questionnaires, all questionnaires were coded by a project manager who was not related to the school in any way. In cases in which an adolescent demonstrated severe pathology or answered positively any of the questions relating to suicide ideation on the BSI, the school counselor was informed and assumed responsibility for referring the student to individual therapy. After termination of the program and data collection, a seminar was sponsored for each school in which the results of the study were presented and the schools were given a gift.

Results

Baseline characteristics were compared with χ2 tests and independent sample t tests to compare the mean scores of continuous variables. The sample at baseline consisted of 97 boys (54 %) and 82 girls (46 %) between the ages of 16.1 and 17.9, with a mean age of 16.3 (SD = 1.11) years. The study population consisted of Jewish youth of whom 8 % reported Orthodox adherence, 33 % traditional, and 59 % secular. As to their socioeconomic status, 69 % reported themselves as middle class, 18 % as low middle class, and 13 % as high middle class. Post hoc comparisons revealed no significant differences between the intervention and control groups on socioeconomic status, religious adherence, and the outcome measures. In addition, no interactions were found between the demographic factors and the research variables. Table 1 shows comparisons at baseline of demographic factors and scores on outcome measures.

Table 1 Demographic and sample characteristics at baseline
Full size table

The study included four within-subject dependent variables: perception of mobilization of support, self-efficacy, behavioral difficulties as measured by the SDQ, and psychological distress as measured by the BSI (Global Severity Index). In a repeated measures design, we conducted four factorial ANOVA analyses for the four main outcomes—using time with two levels (pre- and post-intervention), type of intervention with two levels (control and intervention group), and political violence exposure (high and low PLE) as factors.

The first hypothesis that predicted the influence of political violence exposure on pretest levels of self-efficacy, perceptions of ability to mobilize support, psychological distress, and behavioral difficulties was partly confirmed. In support of this hypothesis, participants reporting higher PLE exposure exhibited higher levels of Global Severity Index scores (psychological distress), F(1, 177) = 12.28, p = 0.001, η 2 = 0.06 (M = 1.26, SD = 0.54), than those reporting lower exposure (M = 1.03, SD = 0.52). Further, participants reporting higher PLE exposure exhibited lower levels of self-efficacy, F(1, 177) = 23.97, p < 0.0001, η 2 = 0.12 (M = 71.52, SD = 11.42), than those reporting lower exposure (M = 83.91, SD = 10.79). However, there were no significant differences between participants reporting high and low PLE in pretest levels of perception of ability to mobilize support (p = 0.39) and in behavioral difficulties (p = 0.34).

Regarding BSI symptomatology, a MANOVA analysis was conducted and yielded a significant main effect for PLE on the nine BSI subscales, Wilks’ Λ = 0.83, F(4, 174) = 9.3, p < 0.0001, η 2 = 0.17. For all the BSI subscales the effects were significant, p < 0.01. These results are displayed in Fig. 2.

Fig. 2
figure 2

Pre-test symptom means according to political life events (PLE) exposure. Low PLE n = 89, high PLE n = 90; 95 % confidence interval error bars; For all effects, p < 0.01. BSI Brief Symptom Inventory, GSI General Severity Index, SOM somatization, OC obsessive–compulsive, INT interpersonal sensitivity, DEP depression, ANX anxiety, HOS hostility, PHOB phobia, PAR paranoid ideation, PSY psychotic ideation

Full size image

The second hypothesis predicted an interaction effect between types of intervention and time (pre- and post-intervention) and stated that participants in the intervention relative to those in the control group would exhibit a greater increase over time in self-efficacy and perception of their ability to mobilize support, and a greater decrease over time in psychological distress and behavioral difficulties. Comparisons of mean changes between intervention and control groups are presented in Table 2.

Table 2 Comparisons of mean changes between intervention and control groups
Full size table

A significant effect emerged for mobilization of support, F(1, 177) = 46.93, p < 0.0001, η 2 = 0.21, with intervention group participants reporting significant increases in their ability to mobilize support (M = 5.26, SD = 5.92), and control group participants reporting significant decreases (M = −2.19, SD = 5.11). A significant effect emerged also for self-efficacy, F(1, 177) = 4.66, p = 0.03, η 2 = 0.03, with intervention group participants reporting significantly greater increases in self-efficacy (M = 13.21, SD = 10.91) than those in the control group (M = 3.36, SD = 10.51). In addition, intervention group participants reported decreases in psychological distress (M = −0.15, SD = 0.58) and control participants reported increases (M = 0.07, SD = 0.52), F(1, 177) = 18.37, p < 0.0001, η 2 = 0.06.

Finally, there was no significant difference between intervention and control groups in changes in the total behavioral difficulties measure, F(1, 177) = 0.13, p = 0.71, η2 = 0.00. However, a significant difference was found on the emotional symptoms scale, F(1, 177) = 19.16, p < 0.0001, η2 = 0.10, with decreases in emotional symptoms in the intervention group (M = −1.23, SD = 1.42) and no significant change in the control group (M = 0.12, SD = 1.65).

For the exploratory question, there was no significant interaction between level of PLE exposure and type of intervention on modifications in self-efficacy, p = 0.31, mobilization of support, p = 0.24, psychological distress, p = 0.14, and behavioral difficulties, p = 0.17.

Discussion

The ceasefire represented a unique period at which both adolescents and school staff reunited in the school setting following the trauma of war. In the sudden deceptive quiet after massive missile attacks directly onto civilian targets, concerns for children’s well-being emphasized the need for an appropriate preventive intervention. The schools constituted an excellent setting for this type of intervention since it facilitated accessibility to large groups of adolescents in their natural routine environment, where they could be monitored by experienced education specialists. This study tested and found empirical evidence to support a school-based program designed to promote adolescents’ coping and resilience during a period of high risk for the development of psychological and behavioral difficulties.

Our findings indicated that adolescents in the intervention group showed an increased ability to mobilize support, whereas those in the control group exhibited significant decreases. This confirms not only the efficacy of the intervention in promoting strategies to support mobilization but also the process of decreasing support following the crisis. It seems likely that in conditions when routine social interactions are suspended, when families are confined to close living quarters in bomb shelters for prolonged periods and municipal psychological services provide emergency support, children would feel enveloped by social support. This intensive social support may then dissipate rapidly with a return to routine life post-crisis and, if so, would necessitate greater investment in mobilizing support when it is no longer readily available.

War and conflict produce chaotic and threatening environments in which it is difficult for adolescents to maintain personal control and a sense of self-efficacy. In this study, intervention group participants reported significantly greater increases in self-efficacy than those in the control group. In addition, the intervention was effective in reducing psychological distress and emotional symptoms in contrast to the control group participants, who reported increases in these indices. These elevated levels of psychological distress in the control group would seem to further affirm the effectiveness of the primary prevention intervention in mitigating distress in this at-risk population.

Taken together, study findings provide validity for the effectiveness of the intervention in promoting its two targeted resilience factors of mobilization of support and self-efficacy, and reinforce the importance of implementing these interventions during ceasefire periods. These findings suggest the need to apply interventions to general populations of children in war conditions at the earliest possible time post-exposure. We were able to begin implementing this intervention several days after the cessation of the war, as opposed to most other studies that report school-based interventions from several months to years post-disaster. Study findings also suggest the benefits of enhancing resilience factors, rather than focusing on debriefing or reducing direct symptoms.

Study findings concerning differences between adolescents with high and low political violence exposure in symptomatology, self-efficacy and social support substantiate our hypothesis that cumulative exposure to these violent events represents a significant risk factor for adolescents. This emphasizes the importance of evaluating the effects of cumulative and prolonged exposure rather than restricting consideration to exposure at a single point in time. Chronic and prolonged conflict situations continue worldwide and it is necessary to account for their influence on children’s development and emotional status differently than in circumstances of acute or isolated traumatic events. The intervention was effective in producing significant improvement in psychological symptoms and in strengthening resilience factors for adolescents exposed to both high and low levels of political violence. This finding suggests the need to offer the intervention to highly exposed adolescents who are at greater risk for negative consequences.

Study Limitations

This school-based intervention should be further examined to test whether it can be generalized to other groups of adolescents affected by political violence and war. The intervention was offered to a general school population and should not be considered an alternative to psychotherapeutic or psychiatric treatment for traumatized adolescents with clinical indicators for such.

There were considerable logistic difficulties in administering a teacher-delivered school-based intervention at short notice immediately post-crisis, and this limited our sample size. In addition, the post-war environment produced difficulties in securing responses from multiple informants, which limited the study to adolescent self-report measures.

In an intervention study of this nature, it was impossible to completely avoid contagion and to monitor schoolchildren’s discussions among themselves. However, the major strategy of the intervention program was based on experiential and interactive activities guided by the teachers in the classroom that involved both personal and group dynamic change processes. It is unlikely that this experiential process could be shared verbally with control group participants. Study findings suggest that if there were contagion effects across the intervention to control groups, these were minimal.

Another limitation of this study, again related to practical difficulties in implementing the study in the immediate aftermath of war, was the existence of one control condition and the lack of an active alternate intervention control. Further, a delayed post-test to confirm long-term effectiveness was not feasible due to unavailability of participants who were approaching year end examinations at that time. These limitations suggest the need for further research to ascertain the generalizability of our findings to other contexts.

Counseling Implications

School staff occupies a pivotal position in introducing and implementing extra-curricular clinical programs during crises such as occurred in this post-war period. Teachers’ availability to schoolchildren positions them as the most accessible professional to meet the needs of the entire school population, especially in times of war. The non-trauma related nature of the program that was based on enhancing resilience factors provided an appropriate, easily implemented tool for teachers that could be monitored within the school setting. A ceasefire after the stress and turmoil of war will inevitably create a sense of relief. This may be deceptive by masking the need for restoring security and well-being to children after such a crisis. The importance of the current study lies in the effort to construct an evidence-based intervention for adolescents in a stressful context of immediate ceasefire, the need for which tends to be overlooked.