Pilot Evaluation of Setu,a School-Based Violence Prevention Program Among Indian Adolescents

Abstract

A quasi-experimental study was conducted in the eighth grade of two public schools located in Central Delhi. Two classes (95 participants) in the intervention school and three classes (108 participants) in the comparison school participated. The brief intervention was implemented across four 2-hour sessions for 1 month in the intervention school. Self-reported violence behavior (bullying, fighting, and victimization) was measured at baseline, posttest, and a 6-month follow-up in both schools. Students and teachers also provided feedback regarding feasibility and acceptability. The difference-in-difference analyses indicated nonphysical aggression, physical aggression, and victimization were significantly lower in the intervention group vis-à-vis comparison group at the 6-month follow-up, but not at baseline or immediate posttest, suggesting sleeper effects. Qualitative feedback indicated that the intervention was acceptable to students and suggested that implementation quality was a key to program effectiveness. These findings suggest that the intervention may curtail the increase in violent behavior in early adolescence.

Keywords

adolescence emotional intelligence social skills empathy implementation quality

Violence is a major public health problem. Worldwide, interpersonal violence is in the top five leading causes of death for adolescents (World Health Organization [WHO], 2018). In Western, postindustrial countries such as the United States, nationally representative surveys suggest that almost half of youth had experienced a physical assault in their lifetime (Finkelhor, Turner, Ormrod, & Hamby, 2009). Very little research has been done in the area of violence and school-based violence prevention in India, and there are currently no nationally representative surveys about the rates of victimization. However, regional, cross-sectional studies suggest that victimization is prevalent (10%-67%) across the country. Among 10- to 19-year-old adolescents in Chandigarh, 27% reported being victims and 13% reported being perpetrators of physical aggression (Munni & Malhi, 2006). In a sample of late adolescents in Delhi, about half of adolescents reported being in a physical fight in the past 12 months (Sharma, Grover, & Chaturvedi, 2008). The self-reported prevalence of violent behavior decreased from mid-adolescence to late-adolescence, with the highest prevalence (60%) among 14- and 15-year olds and the lowest prevalence (43%) among 18- and 19-year olds (Sharma et al., 2008). There is evidence to suggest that rates of violence are higher in urban compared to rural areas in India, with rates ranging from 46% to 67% in urban areas, and 12% to 51% in rural areas (Kishore, Grewal, Singh, & Roy, 1999; Samanta, Mukherjee, Ghosh, & Dasgupta, 2012). The wide variation of reported violent behavior within these studies could be attributed to the difference in measurement scales, the geographic differences, and the time elapsed between studies. India’s growth in the last 20 years has included rapid development of urban areas characterized by vast income inequality and increasing unemployment, which is an environment that engenders violence (Padmanabhan, 2015; WHO, 2015). Given the emerging evidence that violence is a significant problem among Indian adolescents, the need to establish effective prevention programming has become an urgent issue.

Violence among adolescents in its severe forms contributes significantly to the burden of premature death, injury, and disability (S. Singh & Gopalkrishan, 2014). It also has psychological effects such as depression, anxiety, suicidal tendencies, and increased health-risk behaviors, such as smoking, alcohol and drug use, and unsafe sex (Finkelhor, Turner, Ormrod, Hamby, & Kracke, 2009; WHO, 2015). The annual costs of physical, sexual, and psychological violence against children (measured indirectly as losses in future productivity) are 2%–5% of global gross domestic product (Butchart et al., 2008; Pereznieto, Montes, Routier, & Langston, 2014). Unfortunately, exposure to violence and victimization is also associated with an increased risk that adolescents will behave aggressively, thus continuing the cycle of violence (Fowler, Tompsett, Braciszewski, Jacques-Tiura, & Baltes, 2009).

Despite the high rates of violence globally, existing research on violence prevention interventions among adolescents has generally been limited to developed countries, with little work reported from the developing world (Hughes et al., 2014). For example, one review found that only 9.3% of all outcome evaluation studies in violence prevention were conducted in low- and middle-income countries, despite the fact that more than 85% of violent deaths occurred in these countries (Hughes et al., 2014). Furthermore, of almost 350 unique studies published on violence prevention programs, only two took place in India, and only one of those targeted adolescents’ nonsexual violent behavior (Hughes et al., 2014). Specifically, Kannappan and Lakshmi Bai (2008) compared the impact of parent management training combined with either a year-long yoga-based cognitive intervention or year-long social skills training among deviant boys. Both groups showed significant reductions in antisocial behavior that were maintained over time. However, there were significant limitations to this approach, including that the intervention was indicated, intensive (twice a week for 50 minutes, in addition to the parenting group), and resource-heavy, with a long duration. Thus, there is still a need to explore the effectiveness of brief, universal prevention programs in India and South Asia more generally.

School-based social-emotional learning (SEL) programs have established moderate effectiveness in the United States and in Western countries, according to multiple meta-analyses (Durlak, Weissberg, Dymnicki, Taylor, & Schellinger, 2011; Matjasko et al., 2012; Wilson, Lipsey, & Derzon, 2003). There is a median effect size of about 15% reduction in violent behavior, with the smallest effect sizes for middle school students, perhaps due to the increasing trajectory of violent behavior during that time period (Hahn et al., 2007). SEL is based on the premise that youth may have skill deficits that make it difficult for them to deal effectively and prosocially with interpersonal conflict (Boxer & Dubow, 2002). Thus, SEL focuses on improving skills such as emotion regulation, empathy, problem solving, and effective communication (Ananiandou & Claro, 2009; Boxer & Dubow, 2002; Scott, 2015).

SEL skills are only adaptive if they are appropriate to their context. Contextual fit is the main reason behind the success of SEL skills, highlighting the need to adapt the training to specific contexts of each country and each region (M. Singh, 2003). Owing to this, curriculum developers have a responsibility to ensure that skills being taught in interventions are ecologically valid (Farrell, Mehari, Mays, Sullivan, & Le, 2015). Multiple methods have been proposed to ensure the cultural proficiency of interventions (for a review, see Bernal & Domenech Rodríguez, 2012). One such method is an investigator-initiated approach, in which investigators identify a theoretical basis for their curriculum, and key components from existing evidence-based interventions are identified, and curriculum is designed to incorporate those key components in a culturally effective way (Barrera, Castro, Strycker, & Toobert, 2013). This approach allows researchers to take advantage of the wealth of existing research on evidence-based SEL programs in the United States and Europe while ensuring both deep-structural and surface-level cultural fit, similar to the Prevention Service Development Model (Sandler et al., 2005).

This article presents a pilot evaluation of Setu, a violence prevention program among urban adolescents in India. Setu (सेतु) means “bridge” and was a metaphor used throughout the program. We developed a theory-based program using evidence-based practices and designed it to suit the cultural milieu of children living in an urban metropolis in India. The brief program is school based, which can be the most efficient and organized way to reach the majority of the adolescent population (Hahn et al., 2007; WHO, 2009). Setu was based on the theory that mastering social-emotional competencies results in a developmental progression that marks a shift from predominantly reacting to external factors to acting increasingly in accord with internalized beliefs and values, responding deliberately, and taking responsibility for one’s choices and behaviors (Krug, Mercy, Dahlberg, & Zwi, 2002).

In the proposed logic model, the active ingredients include positive identity development, emotion regulation strategies including mindfulness, empathy and perspective-taking skills, and problem-solving skills. The healthy development of a positivity identity, in combination with social information processing and emotion regulation skills, is integral to responding to conflict situations in a prosocial manner. Furthermore, these intervention components were identified as particularly relevant to adolescence, given the tasks of adolescence and adolescents’ cognitive development. Erikson argued that forming one’s identity—in love, work, and ideology—is the central task of adolescence (Arnett, 2010). Identity formation likely guides the development and prioritization of social goals (e.g., social responsibility and intimacy vs. dominance; Kiefer & Ryan, 2008), which drives behavior choices. In addition, adolescents are able to think more abstractly about values, promoting their ability to develop their own value system (Piaget, 1972). For effective, prosocial social information processing, problem-solving skills and emotion regulation skills are needed as well. For example, although adolescents have the cognitive capacity to take others’ perspectives, they may need scaffolding to use their perspective-taking skills in a range of situations (Selman, 1977). They also may need support to generate and evaluate a range of response options based on their values and goals. In addition, they need emotion regulation skills so that they can modulate their emotional states in pursuit of their goals (Herts, McLaughlin, & Hatzenbuehler, 2012). According to our model, these active ingredients (positive identity development, emotion regulation skills, and social information processing skills) decrease rates of aggression and victimization. We hypothesized that the intervention group would demonstrate significantly lower rates of aggression and victimization following the intervention than the comparison group. We also hypothesized that Setu would be feasible and acceptable based on teachers’ and participants’ reports.

Method

Intervention

Curriculum development

An interdisciplinary, participatory approach was employed to develop Setu, with a team including university and community-based colleagues in public health and education (García, Pintor, & Lindgren, 2010; García Luis, Heckman, Leaf, & Prados, 2016). We reviewed the existing curricula on social-emotional violence prevention to identify evidence-based active ingredients that were culturally congruent (Central Board of Secondary Education [CBSE], 2009; Department of Education and Skills, 2007; Family Health International, 2007; Goleman, 1996; Illinois PBIS Network Technical Assistance Brief, 2010). However, some activities that were successful in a Western context were determined not to be culturally sensitive within the Indian context, decreasing the likelihood of acceptability (Gonzales, Lau, Murry, Pina, & Barrera, 2014; Resnicow, Baranowski, Ahluwalia, & Braithwaite, 1999). Some components like stories and videos were entirely replaced by local folklore, movie songs, and other references to contemporary culture. Through this, we tried to invoke the deep-rooted cultural, social, historical perspective in which participants grew up, thus addressing cultural sensitivity at a deeper level (Resnicow et al., 1999). This process resulted in an intervention consisting of four, 2-hour sessions focused on teaching mindfulness, affect identification and regulation (e.g., deep breathing and imagery), problem solving, empathy, and effective communication. Setu was piloted and then refined following the administration to a group of 30 students (Sharma, Kishore, Sharma, & Shukla, 2014).

The final version of the program included multiple culturally appropriate teaching methods. Table 1 describes the objectives and activities used during each meeting. Throughout the program, the image of building a bridge to cross a turbulent river was used, with each skill representing a new plank in the bridge. Interactive narratives were used as educational tools because they are believable, memorable, and entertaining (Rossiter, 2002). In addition, stories invite, even demand, active meaning making. Cultural literature was searched by the authors to pick stories which could encourage applications of the skills. Stories from Panchatantra were selected (Children’s Book Trust, 1979). The widely known animal fables of Panchatantra are hailed as “Niti Sastra,” or a treatise on “the wise conduct of life.” Children in India grow up listening to these tales and can easily relate to them. We made use of this age-old tradition of storytelling. The stories were narrated live by the trainer and were paused at strategic points to discuss the actions of the characters. The participants, in turn, generated alternative shapes to the stories by identifying emotions, motivations, and alternative behavioral choices of characters (Griffith, Griffith, Cobb, & Oge, 2016). Students practiced skills both directly in the class (e.g., deep breathing, muscle relaxation, and imagery), and through role-plays. Peer discussions were used to enhance concept formation, which was especially useful in this case, given that the skills were complicated and fairly abstract (Smith, Wood, Adams, Weiman, & Knight, 2009). Students discussed their answers first within the group and then presented them in the class as a team product. We also used videos of songs and advertisements to engage students and increase understanding of skills. Given the chaotic nature of the classroom setting, program implementers incorporated the use of incentives for positive behaviors. Students received decision dollars when they listened respectfully, acted prosocially toward peers, and actively participated in the content of the intervention. At the end of each session, participants received a leaflet that summarized the lesson and to practice the skills learned during the class. For homework assignments, it was expected that the participants would involve their family in discussions, monitor their own social interactions, reflect, and document the difficulties that arose in implementing the skills in real life.

Table 1.

Session by Session Summary of Intervention.

Session	Objectives	Activities	Homework
1	• Define aggression and victimization; and discuss its impact • Demonstrate prosocial skills • Promote perspective taking • Introduce mindfulness • Explore positive identity development	• Group discussion on the impact of victimization • Perspective-taking activity • Mindfulness activity using five senses • Positive self-visualization • Art/craft activity displaying students’ self-identified talents	Journal activity to describe self-identified talents
2	• Define aggression and victimization; and discuss its impact • Demonstrate prosocial skills • Promote affect identification and regulation strategies • Promote critical thinking	• Group discussion on the impact of victimization • Mirror exercise to identify different emotions • Interactive storytelling emphasizing the need of affect regulation • Mindfulness & deep breathing • Action song on affect regulation	Art/craft activity displaying different emotions
3	• Define aggression and victimization; and discuss its impact • Introduce empathy • Promote problem solving and perspective taking • Demonstrate effective communication skills	• Group discussion on the impact of victimization • Speaking object to identify the scale of emotion • Video demonstrating empathy • Group activity for problem solving, perspective taking • Group activity followed by discussion on effective communication skills	Journal activity to reframe a conflict using a problem-solving problem approach
4	• Define aggression and victimization; and discuss its impact • Demonstrate prosocial skills • Describe responsible decision making • Promote problem solving	• Group discussion on the impact of victimization • Interactive storytelling to introduce decision making • Role-play to demonstrate all skills learned over the four sessions while solving a complex problem	Writing a story to practice creativity, critical thinking, and problem solving

Adherence and dosage

The administration of both the schools supported the program, which is important for an intervention’s sustainability and the ability to scale up an intervention (Fagan & Mihalic, 2003). Administrators made time slots available well in advance, and lectures were adjusted to avoid any lag in the regular curriculum. Prior to Setu being implemented with students, 24 middle school teachers attended an orientation session for teachers. The average attendance for teacher orientation was robust in both classes of School A, Class P, and Class Q (42.5/46 and 44.25/49, 91.85% and 90.31%, respectively). There were no delays or rescheduling when implementing Setu with the two classes in School A. The intervention was delivered smoothly with no untoward incidents reported during the implementation and follow-up visits, suggesting that the intervention is feasible. To track adherence, facilitators logged the delivery of intervention components. All core components were delivered for each of the sessions. Related to dosage, 67.5% of students were present for all four sessions, and homework assignments were turned in by 30% of students.

Procedure

A quasi-experimental design was used to assess the effectiveness of the intervention. We selected two public schools in urban areas of central Delhi. These coeducational schools, managed by the same central administration and following the same curriculum, were randomly assigned to intervention (School A) and control (business as usual) groups (School B). Neither school had received or implemented any SEL program or any program designed to prevent problem behaviors prior to Setu. School B, the control group, continued with business as usual, which involved no SEL or prevention programming. The intervention was implemented in the two eighth grade classes of School A, and the three corresponding classes in School B served as the control group (see Figure 1 for study flow). The program started with educating School A teachers about the intervention in an hour-long meeting. After that, four 2-hour, weekly sessions were conducted at School A during the school day. The 2-hour sessions were generally held in the morning hours in place of the usual core subject areas. Two bilingual (English and Hindi) external facilitators, one woman and one man, conducted the sessions. Both received training in the implementation of the program. Prior to the intervention, participants in Schools A and B completed pretest measures. They completed the same measures 1 month after implementation (posttest) and 6 months after implementation (follow-up). After the follow-up measures were completed, all participants in School B received the intervention. We obtained assent from participants and written informed consent from parents. The ethics committee of the college approved the protocol.

Figure 1.

Flow of participants through phases of the study.

Participants

Participants included 95 Grade 8 students in School A and 108 Grade 8 students in School B (see Table 2 for full demographics). Participants’ ages ranged from 11 to 14 years old (M = 12.6 years). Approximately two thirds (69.5%) of the sample was male. The majority of the participants were categorized as lower middle class, with the average per capita income of the participants’ families equal to INR 8,356 (US$130) per month. The majority (90%) were Hindu, with a small minority of participants reporting their religious affiliation as Muslim, Sikh, Christian, Buddhist, Jain, or Jewish.

Table 2.

Baseline Demographics of Sample and Tests of Differences Across Intervention and Control Groups.

Characteristic	Intervention school	Control school	Difference test
Characteristic	M (SD)	M (SD)	T-test
Age	12.7 (0.6)	12.6 (0.7)	−1.5
Siblings	1.6 (0.9)	1.4 (0.7)	−1.0
Members in family	4.8 (1.2)	4.7 (1.1)	0.1
	Number (%)	Number (%)	χ² test
Sex: male	58 (72.5)	63 (67.0%)	0.4
Religion			8.7
Hindu	77 (96.3)	81 (89.0)
Muslim	1 (1.3)	3 (3.2)
Sikh	1 (1.3)	5 (5.3)
Christian	0	4 (4.3)
Jain	1 (1.3)	0
Jew	0	1 (1.1)
Socioeconomic status^a			8.7**
Upper	5 (6.85)	7 (7.45)
Upper middle	53 (72.6)	49 (52.1)
Lower middle	15 (20.6)	38 (40.4)

Note. N = 73 to 80 for the intervention school and 94 for the control school.

Calculated based on the modified Kuppuswamy scale (2012).

p < .01.

Measures

Participants completed a survey that included demographic data and frequency of aggression and victimization. The Illinois Bully Scale (Espelage & Holt, 2001) was used to measure violence. Participants indicated the frequency with which they experienced or enacted 18 items over the past 30 days on a 5-point rating scale from 0 = Never to 4 = seven or more times. The measure comprises three subscales: Bully (nine items), Fight (five items), and Victim (four items). Although the first subscale is labeled “Bully,” it simply assesses nonphysical aggression without considering the power imbalance, repetition, and chronicity that define bullying (Gladden, Vivolo-Kantor, Hamburger, & Lumpkin, 2014). For example, one item asks, “I teased other students.” It includes verbal aggression (e.g., “I started arguments or conflicts”) and relational aggression (e.g., “I spread rumors about other students). The Fight subscale was used to assess physical aggression (e.g., “I hit back when someone hit me first” and “I got into a fight”). The Victim subscale is a global measure of victimization that includes both physical victimization (e.g., “I got hit and pushed by other students”) and verbal and relational victimization (e.g., “Other students made fun of me”). The measure has established validity and reliability (Cronbach’s alpha ranging from 0.83 to 0.88). The measure was pilot-tested and then modified to fit the Indian context. Modification involved translation to Hindi and modification of the word “clique” to “group.” In this study, the subscales had adequate reliability as evidenced by Cronbach’s alpha (at baseline: nonphysical aggression subscale = 0.77; physical aggression subscale = 0.69; victim subscale = 0.66).

To assess feasibility and acceptability, interviews of teachers from the intervention group (n = 6) and administrative staff (n = 2) were conducted, and School A students completed feedback forms. These anonymous feedback forms were filled by students at the end of each session and included the following questions: “What went well about the session?”; “What didn’t go well?”; and “How could the session be improved?.” Teachers were asked about their views on the scenario of violence in their school, their perceptions regarding facilitators and barriers to the implementation of the prevention program, and their suggestions to improve the program.

Data Analysis

Intent-to-treat analysis was used to examine the intervention’s effectiveness. The primary analysis included all students who completed the surveys, regardless of the number of sessions attended. A difference-in-difference regression model was used to estimate the change in the frequency of physical and nonphysical aggression and victimization in the intervention group compared with the control group. We used separate multivariate regression models for each of the aggression subscales. In addition, separate models were used to compare the change across pretest and posttest measurements and across pretest and follow-up measurements. The estimate of change after SEL intervention was obtained from interaction indicators for intervention and follow-up (the difference-in-difference estimate). Furthermore, in separate models we measured effect modification by sex and age group (11-12 and 13-14 years). To the main model, we added the variable and interaction terms between the variable, intervention, and follow-up. The analysis was done using Stata version 13.

Written student feedback and teacher interviews were transcribed, translated into English, codes were generated using inductive and deductive approach, and similar codes were combined together into themes. The focus was to identify factors that may impede or facilitate the implementation of the intervention. Regular meetings were held to discuss the analysis, resolve coding discrepancies, and achieve consensus.

Results

Only the 174 participants who completed baseline assessments were included in the analyses (84.2% of eligible students from School A, and 87.0% from School B). There were no significant differences in demographic factors or in baseline frequency of physical and nonphysical aggression and victimization between the two groups except a slight difference in socioeconomic status (see Table 3). However, girls were less likely to endorse aggression and victimization (t-values from 3.0-5.1; p < 0.01; Table 3). There were no age differences in aggression or victimization. In each session, at least 90% of participants were present. About 67.5% of the intervention group was present for all four sessions. There were no differences in levels of aggression and victimization between participants who attended all the sessions and participants who attended three or fewer sessions at pretest, posttest, or follow-up (see Table 4).

Table 3.

Baseline Rates of Violence Across Intervention and Control Groups, Sex, and Age.

Violence subscale	M (SD)	M (SD)	T-test
Violence subscale	Intervention school	Control school	T-test
Victimization^a	4.4 (3.4)	4.1 (3.3)	0.5
Nonphysical aggression^b	5.2 (4.9)	4.6 (4.7)	−0.8
Physical aggression^c	3.9 (3.4)	3.8 (3.2)	−0.2
	Boys	Girls
Victimization^a	4.7 (3.6)	3.1 (2.3)	3.0**
Nonphysical aggression^b	6.0 (5.1)	2.3 (2.4)	5.1***
Physical aggression^c	4.6 (3.5)	2.1 (1.7)	4.8***
	Ages 11-12	Ages 13-14
Victimization^a	4.1 (3.3)	4.4 (3.4)	−0.5
Nonphysical aggression^b	5.0 (5.0)	4.7 (4.6)	0.4
Physical aggression^c	3.7 (3.5)	3.9 (3.1)	0.4

Note. N = 80 for the intervention school and 94 for the control school. N = 121 for boys and 53 for girls. N = 100 for ages 11 to 12 and 74 for ages 13 to 14.

Scale range 0 to 16.

Nonphysical aggression: scale range 0 to 32.

Scale range 0 to 20.

p < .01. ***p < .001.

Table 4.

Differences Between Participants Who Attended All Sessions and Participants Who Attended Three or Fewer Session.

Characteristic	Attended 4 sessions	Attended ≤3 sessions	Difference test
Characteristic	Number (%)	Number (%)	χ² Test
Male	36 (66.7)	22 (84.6)	2.8
	M (SD)	M (SD)	T-Test (df)
Age	12.7 (0.6)	12.9 (0.5)	1.6 (61)
Victimization—Pretest	4.3 (3.7)	4.6 (2.9)	0.5 (60)
Nonphysical aggression—Pretest	5.1 (5.1)	5.2 (4.5)	0.1 (56)
Physical aggression—Pretest	3.7 (3.3)	4.1 (3.5)	0.5 (47)
Victimization—Posttest	3.6 (3.5)	4.2 (3.0)	0.7 (50)
Nonphysical aggression—Posttest	4.1 (5.3)	3.1 (2.9)	−1.0 (70)
Fighting—Posttest	2.7 (3.2)	2.6 (2.5)	−0.2 (53)
Victimization—FU	3.1 (3.3)	3.8 (2.9)	0.9 (50)
Nonphysical aggression—FU	3.5 (4.4)	3.3 (3.5)	−0.2 (55)
Physical aggression—FU	2.3 (2.8)	2.3 (2.3)	−0.0 (53)

Note. No differences were statistically significant. N = 80 at pretest, 75 at posttest, and 77 at follow-up. FU = follow-up.

Intervention Effectiveness

At the 1-month posttest, there were no significant differences in nonphysical aggression or victimization between the intervention and control groups (see Table 5). The difference-in-differences estimate were −1.0 (95% confidence interval [CI] = [–2.4, 0.4]) and −0.2 (CI = [–1.2, 0.8]), respectively. However, there was a significant decrease in physical aggression for the intervention group compared to the control group, partially supporting our hypothesis (estimate = −1.3; CI = [–2.3, –0.2]). At the 6-month follow-up, there were significant differences between the intervention and control group for all three parameters, which fully supported our hypothesis. The difference-in-differences analysis showed that the frequency of nonphysical and physical aggression and victimization during the second follow-up changed by −3.8 (CI = [–5.3, –2.2]), –3.0 (CI = [–4.0, –2.0]) and −1.8 (CI = [–2.9, –0.7]) points, respectively, in the intervention group compared to control, which points toward a significant decrease (Table 5, Figure 2). To determine the effect size, Cohen’s d was calculated to compare the difference between the intervention and control group at follow-up. Effect sizes were medium to large for nonphysical aggression (d = –0.5 [–0.8, –0.2]), physical aggression (d = −0.9 [–1.2, –0.5]), and victimization (d = −0.4[–0.7, –0.1]).

Table 5.

Impact of the Intervention on Indicators of Violence and Sex and Age Moderation Based on a Difference-in-Differences Regression Model.

Violence subscale	M (SD)	M (SD)	Difference-in-differences estimate (confidence interval)^a
Posttest	Intervention school	Control school	Difference-in-differences estimate (confidence interval)^a
Victimization	3.8 (3.3)	3.8 (3.2)	−0.2 [–1.2, 0.8]
Sex moderation^b			−1.0 [–2.6, 0.5]
Age moderation^c			1.3 [–0.2, 2.8]
Nonphysical aggression	3.8 (4.7)	4.2 (5.7)	−1.0 [–2.4, 0.4]
Sex moderation			1.4 [–0.8, 3.7]
Age moderation			1.0 [–1.2, 3.3]
Physical aggression	2.7 (3.0)	3.8 (3.9)	−1.3 [–2.3, –0.2]*
Sex moderation			1.1 [–0.3, 2.6]
Age moderation			1.5 [–0.0, 3.0]
Follow-up	Intervention school	Control school
Victimization	3.3 (3.2)	4.9 (4.0)	−1.8 [–2.9, –0.7]*
Sex moderation			−1.0 [–2.7, 0.6]
Age moderation			0.6 [–1.0, 2.1]
Nonphysical aggression	3.4 (4.0)	6.5 (7.1)	−3.8 [–5.3, –2.2]**
Sex moderation			2.0 [–.4, 4.4]
Age moderation			2.0 [–0.4, 4.3]
Physical aggression	2.2 (2.6)	5.2 (4.1)	−3.0 [–4.0, –2.0]***
Sex moderation			1.9 [0.4, 3.3]*
Age moderation			2.6 [1.1, 4.1]**

Note. N = 80 for the intervention school and 94 for the control school. N = 155 for posttest and 161 for follow-up.

Comparisons to pretest.

Reference group is male.

Reference group is ages 11 to 12.

p < .05. **p < .01. ***p < .001.

Figure 2.

Impact of the intervention on nonphysical aggression, physical aggression, and victimization.

Participants’ age and gender appeared to moderate the effect of the intervention on physical aggression at the 6-month follow-up (see Table 5). Girls showed a greater change in physical aggression as compared to boys (estimate = 1.9; CI = [0.4, 3.3]). The intervention also was more effective for the older age group (13-14 years old) than the younger age group (11-12 years old; estimate: 2.6; CI = [1.1, 4.1]).

Feasibility and Acceptability

In the feedback forms, participants responded to the question, “Did the session achieve its said objectives?” on a 5-point Likert-type scale from 1 = I disagree to 5 = I agree. Both the classes rated the sessions high (M = 4.96; SD = 0.01; and M = 4.77; SD = 0.08). In their qualitative feedback, most of the students indicated that they liked the facilitators. For example, one student wrote, “I like the way Ms. X teaches us everything step-by-step.” Students also explicitly mentioned they liked the fact that the trainers were friendly and that they did not beat or scold any of the students.

Students also liked the incentives and activities. One student wrote, “When we gave the answers, we got decision dollars, and then children gave all the more answers; this, I liked the most.” Another student wrote, “All students play games and are happy.” Most of the students wrote “nothing” in response to the things that they disliked about the class. Some students indicated that they most disliked the behavior of their classmates, such as rowdiness, fighting, and laughing: “Please children don’t make noise and keep discipline.” When asked for suggestions to improve sessions, class size came up as an important issue: “There will be more discipline if we had a smaller group.” Participants also wanted more information on bullying in the form of videos or reading materials and to play more games.

Teachers reported a high need for the intervention. For example, one teacher said, “Children face violence every-day through media exposure (movies, songs and even news).” The teachers saw the presence of impulsivity, fights, verbal abuses, as well as both online and offline bullying among the pupils they taught. One teacher said, “Girls abuse other girls online using Facebook by creating fake accounts. They humiliate others by making personal information public.” Another teacher discussed a time in which “a boy hit another boy with a brick, and he bled profusely.” They believed that both parents and teachers both have an essential role in preventing risk behavior.

The teachers observed a perceptible change among the students in the study, particularly girls. One of the teachers remarked that the students appeared more self-aware: “Now, whenever we point-out if the students behave aggressively, they smile instead of defending themselves. It seems they realise it’s not good to be involved in such activities.” However, they also asserted that for some, the intervention might not have been useful as their negative behavior was deeply ingrained and difficult to change. When asked about suggestions to improve the sessions, the school administrator said, “The moment the children had started molding, the program got over,” suggesting that “a longer intervention would be better.” The teachers suggested starting the intervention earlier in development, prior to “bad habits” being developed. They also proposed to involve parents and to add refresher sessions.

Discussion

This study provides emerging evidence that a brief, universal, culturally proficient, school-based, violence prevention program can be effective among Indian adolescents in a high-risk environment. The intervention demonstrated quantifiable benefits in rates of aggressive behavior and victimization. Interestingly, the entire impact of the intervention was not apparent immediately following the intervention, with only differences in physical aggression noted at posttest. Instead, significant differences between the intervention and control group emerged at the 6-month follow-up. This suggests that the intervention might show a “sleeper effect”—a phenomenon that occurs when there is a delayed impact of a prevention program. Sleeper effects have been found for other universal, school-based violence prevention programs that involve SEL (Sklad, Diekstra, Ritter, Ben, & Gravesteijn, 2012). In this case, the control group experienced an increase in nonphysical aggression and victimization at the 6-month follow-up. One possible explanation is that this phenomenon is indicative of the developmental trajectory in Indian adolescents’ aggressive behavior (Atienza & King, 2002). Based on this hypothesis, it is possible that the intervention curtailed the naturally occurring increase in violent behavior in the intervention group. Thus, the effects of the intervention were not apparent until there was an increase in violent behavior in the comparison group.

Interestingly, the intervention was more effective for the older students (13- and 14-year olds) and for girls. It is unclear why the intervention was more effective for older students. It is possible that they may have benefited more from more abstract concepts (e.g., mindfulness, thinking toward the future) due to their stage of cognitive development compared to younger students. That is, they may have been able to grasp hypothetical reasoning, goal orientation, and affect regulation better than their younger peers. In the qualitative feedback, the teachers indicated the need for an SEL program for younger students; however, new curriculum for Setu will need to be developed to address this need. Future work should attempt to adapt the program to increase the effectiveness of the program for preadolescents (e.g., ages 11-12). Younger students may need more practice with emotion regulation and problem-solving skills for those skills to generalize to other contexts, and they may benefit less from positive identity development and discussion of abstract concepts.

The finding that girls benefited more from the intervention was unanticipated, especially since their base rates of involvement in violence were lower. One review of universal violence prevention programs found no consistent findings across studies and concluded that overall, there was insufficient evidence to support the idea that gender moderated the intervention effectiveness (Farrell, Henry, & Bettencourt, 2013). In this case, given that girls were the substantial minority, but the primary facilitator was a successful woman with a medical degree, it is possible that the girls saw a role model for their future that inspired behavior change. Future research on this intervention should identify whether this finding is consistent over time, and possible mediators that might explain this relation (e.g., increased hope or positive future orientation, increased empathy or perspective taking, improved problem solving).

Dosage did not seem to have a major role in determining effectiveness of the program for individual participants. This finding was unexpected. However, given that the majority of adolescents received all four sessions, there might have been a diffusion of intervention effects to adolescents who did not receive a particular session through adolescents who did receive that session. In addition, it is possible that the intervention impacted classroom climate and changed norms, exerting a widespread effect on adolescents’ behaviors. Therefore, even though dosage for individual students did not predict outcomes, it is possible that the dosage that the classroom as a whole received was important.

In addition to effectiveness, the SEL program was found to be feasible and acceptable. The intervention fit well within the normal school day. The positive student and teacher feedback imply high acceptability of the program. Combined, the emerging evidence of effectiveness, feasibility, and acceptability suggest that this investigator-initiated culturally proficient intervention development was successful. These findings open the door for continued research on school-based violence prevention programs in south Asia. It also suggests that culturally appropriate teaching methods, such as the use of storytelling, can be effective in behavior change interventions.

More research needs to be conducted on the sustainability of SEL programming in India. The characteristics of the trainers, as highlighted in the feedback, might have influenced many students’ decision to participate actively in the sessions. This is consistent with previous research suggesting that implementer competence is just as important as program content in predicting the impact of a violence prevention program (Low, Smolkowski, & Cook, 2016). A highly sustainable, effective program should be implemented by teachers or other school staff, following appropriate training. Implementer training should build skills for creating a positive classroom climate and for scaffolding adolescents’ social-emotional development (B. D. Singh & Menon, 2015).

However, such an implementation would require a commitment of resources (teacher and classroom time) from Indian schools. Currently, state-funded schools in India are hugely underfunded and inadequately staffed, especially in the poorer and rural regions where the ability of children in demonstrating core literacy and arithmetic skills is often disturbing. Moreover, the current stand-alone life-skills curriculum introduced by the Government of India is treated as an auxiliary course for which the teachers lack both the skills and the time (Behrani, 2016; CBSE, 2009; B. D. Singh & Menon, 2015; Srikala & Kishore, 2010). Currently, violence prevention may not be the main priority. However, as countries continue to develop, adolescent health and safety likely become more integral to national priorities, as suggested by the increased emphasis on SEL in schools in Western Europe and North America since the 1970s. Early research is vital to establishing the need for such programming in India, which will ideally motivate priority shifts.

Limitations and Future Directions

It is important to emphasize that this study was a pilot study using only one intervention school and one control school, so the results should be interpreted with caution. It is possible that other factors, such as changes brought in by the school administration or classroom teachers, could have also caused the change in aggression and victimization. Also, there is a risk of potential bias as the program was researcher-led. Reviews of prevention programs have shown that efficacy studies in which programs are implemented by researchers are generally more effective than programs implemented by end-users (such as schools; Wilson et al., 2003). However, other reviews have suggested that SEL curricula are actually more effective when teachers, rather than outside facilitators, implement them, perhaps because the classroom teacher can promote integration and generalization of skills (Durlak et al., 2011). To establish the effectiveness of Setu, it will be important to test its impact when implemented by classroom teachers rather than the research, especially with a larger sampling of schools, before definitive conclusions can be drawn. Nevertheless, this study provides emerging evidence that Setu is suitable for the target population of Indian students in early adolescence and is effective in curtailing the trajectory of violent behavior. It thus adds new information to the sparse existing literature on youth violence prevention programs in the developing world.

Other limitations included measurement of outcome and mediating factors. Dependence on student self-reports has the potential for biased responding, such as due to social desirability bias (Detels, Beaglehole, Lansang, & Gulliford, 2009). However, the fact that most significant differences did not emerge until the 6-month follow-up suggests that social desirability did not play a major role. Due to limitations in the length of the survey, we only assessed the desired distal outcomes (aggression and victimization) and did not measure the mechanisms of change proposed in our logic model. Assessment of mechanisms of change such as emotion regulation, problem solving, empathy, and perspective taking would have been beneficial, as these could help uncover the causative pathways through which the intervention had an impact (Kobor, 2009). In addition, we did not assess bullying as traditionally defined (Gladden et al., 2014). Currently, the construct of bullying has not been explored and operationalized in Indian contexts. Interestingly, there is no Hindi equivalent to the English word “bullying,” with the most approximate term equivalent to the idea of intimidating. In-depth qualitative research and rigorous measure development strategies must be used to operationalize bullying in Indian contexts.

Clinical and Policy Implications

There is an urgent need for violence prevention among youth in south Asia. Culturally proficient SEL programs may be a promising solution. However, the challenges posed by low-resource settings often make implementation through schools difficult. Adequate training and support of school teachers and parents will be paramount for sustainability. Nevertheless, short, innovative programs like this one, delivered on school premises, offer hope. Through programs such as this, health-promoting schools providing a safe environment for students could help build SEL skills that will ideally increase life expectancy and quality of life.

Footnotes

Authors’ Note

The research was conducted when D.S. was a Junior Resident, Department of Community Medicine, at Maulana Azad Medical College.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research was supported by grants from Indian Council of Medical Research.

ORCID iD

Drishti Sharma

Author Biographies

Drishti Sharma is Clinical Research Specialist at IAVI and an adjunct faculty at Indraprastha Institute of Information Technology, New Delhi. She did her MD in Community Medicine, and her thesis focused on the effect of social-emotional skills on aggressive behavior among adolescents. She has a keen interest in behavioral science and its application in Public Health.

Krista R. Mehari is an Assistant Professor in the Department of Psychology at University of South Alabama. Her research focuses on violence prevention in adolescence and on the impact of contextual risk factors on development and functioning.

Jugal Kishore is a Director Professor and Head of Department of Community Medicine at Vardhman Mahavir Medical College, New Delhi. His research interests lie in mental health, adolescent health, and occupational and environmental health. He is also the secretary general of the Indian Association for Adolescent Health.

Nandini Sharma is a Director Professor and Head of the Department of Community Medicine, Maulana Azad Medical College, New Delhi. Her area of interest is respiratory medicine and public health.

Mona Duggal is an Assistant Professor in Advanced Eye Centre, Post Graduate Institute of Medical Education and Research, Chandigarh. She has focused her research on the intersection of behavioral science, technology, and clinical intervention research.

References

Ananiandou

Claro

(2009). 21st century skills and competences for new millennium learners in OECD countries [OECD education working papers]. Paris, France: Organisation for Economic Co-operation and Development.

Arnett

J. J.

(2010). Adolescence and emerging adulthood: A cultural approach (4th ed.). Boston: Prentice Hal.

Atienza

A. A.

King

A. C.

(2002). Community-based health intervention trials: An overview of methodological issues. Epidemiologic Reviews, 24, 72-79.

Barrera

Jr. Castro

F. G.

Strycker

L. A.

Toobert

D. J.

(2013). Cultural adaptations of behavioral health interventions: A progress report. Journal of Consulting and Clinical Psychology, 81, 196-205. https://doi.org/10.1037/a0027085

Behrani

(2016). Implementation aspects of life skills education program in Central Board of Secondary Education Schools. International Education and Research Journal, 2(3), 68-71.

Bernal

Domenech Rodríguez

M. M.

(Eds.). (2012). Cultural adaptations: Tools for evidence-based practice with diverse populations. Washington, DC: American Psychological Association. https://dx-doi-org.web.bisu.edu.cn/10.1037/13752-000

Boxer

Dubow

E. F.

(2002). A social–cognitive information-processing model for school-based aggression reduction and prevention programs: Issues for research and practice. Applied and Preventive Psychology, 10, 177-192. doi:10.5993/AJHB.24.4.3

Butchart

Brown

Khahn

H. A.

Corso

Florquin

Muggah

(2008). Manual for estimating the economic costs of injuries due to interpersonal and self-directed violence. Geneva, Switzerland: World Health Organisation.

Central Board of Secondary Education. (2009). Facilitators handbook for training of resource persons nodal teachers. New Delhi, India: Author.

10.

Children’s Book Trust. (1979). Stories from Panchatantra. New Delhi, India: Author.

11.

Department of Education and Skills. (2007). Social and emotional aspects of learning for secondary schools (SEAL): Guidance booklet. Author. https://dera.ioe.ac.uk/6663/7/f988e14130f80a7ad23f337aa5160669_Redacted.pdf

12.

Detels

Beaglehole

Lansang

M. A.

Gulliford

(2009). Oxford textbook of public health. New York, NY: Oxford University Press.

13.

Durlak

J. A.

Weissberg

R. P.

Dymnicki

A. B.

Taylor

R. D.

Schellinger

K. B.

(2011). The impact of enhancing students’ social and emotional learning: A meta-analysis of school-based universal interventions. Child Development, 82, 405-432. doi:10.1111/j.1467-8624.2010.01564.x

14.

Espelage

D. L.

Holt

M. K.

(2001). Bullying and victimization during early adolescence: Peer influences and psychosocial correlates. Journal of Emotional Abuse, 2, 123-142.

15.

Fagan

A. A.

Mihalic

(2003). Strategies for enhancing the adoption of school-based prevention programs: Lessons learned from the blueprints for violence prevention replications of the life skills training program. Journal of Community Psychology, 31, 235-253.

16.

Family Health International. (2007). Life Skills Education Toolkit for orphans and vulnerable children in India. New Delhi, India: Author.

17.

Farrell

A. D.

Henry

D. B.

Bettencourt

. (2013). Methodological challenges examining subgroup differences: Examples from universal school-based youth violence prevention trials. Prevention Science, 14, 121-133. doi:10.1007/s11121-011-0200-2

18.

Farrell

A. D.

Mehari

K. R.

Mays

Sullivan

T. N.

A. T.

(2015). Participants’ perceptions of a violence prevention curriculum for middle school students: Was it relevant and useful? Journal of Primary Prevention, 36, 227-246. doi:10.1007/s10935-015-0391-6

19.

Finkelhor

Turner

Ormrod

Hamby

S. L.

(2009). Violence, abuse, and crime exposure in a national sample of children and youth. Pediatrics, 124,1411-1423. doi:10.1542/peds.2009-0467

20.

Finkelhor

Turner

Ormrod

Hamby

S. L.

Kracke

(2009). Children’s exposure to violence: A comprehensive national survey. Washington, DC: Office of Juvenile Justice and Delinquency Prevention.

21.

Fowler

P. J.

Tompsett

C. J.

Braciszewski

J. M.

Jacques-Tiura

A. J.

Baltes

B. B.

(2009). Community violence: A meta-analysis on the effect of exposure and mental health outcomes of children and adolescents. Development and Psychopathology, 21, 227-259. doi:10.1017/S0954579409000145

22.

García

Pintor

J. K.

Lindgren

(2010). Feasibility and acceptability of a school-based coping intervention for Latina adolescents. Journal of School Nursing, 26, 42-52.

23.

García Luis

Heckman

J. J.

Leaf

D. E.

Prados

M. J

. (2016). The case for investing in disadvantaged young children. Retrieved from https://heckmanequation.org/resource/the-case-for-investing-in-disadvantaged-young-children/

24.

Gladden

R. M.

Vivolo-Kantor

A. M.

Hamburger

M. E.

Lumpkin

C. D.

(2014). Bullying surveillance among youths: Uniform definitions for public health and recommended data elements (Version 1.0). Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.

25.

Goleman

(1996). Emotional intelligence, why it can matter more than IQ. London, England: Bloomsbury Publishing.

26.

Gonzales

N. A.

Lau

A. S.

Murry

V. M.

Pina

A. A.

Barrera

M. J.

(2014). Culturally adapted preventive interventions for children and adolescents. In Cicchetti

(Ed.), Developmental psychopathology (pp. 874-933). Hoboken, NJ: John Wiley.

27.

Griffith

Cobb

Oge

(2016). The use of narrative as a treatment approach for obesity: A storied educational program description. The Permanente Journal, 20(3), 15-29.

28.

Hahn

Fuqua-Whitley

Wethington

Lowy

Liberman

Crosby

. . . Dahlberg

(2007, August 10). The effectiveness of universal school-based programs for the prevention of violent and aggressive behavior. A report on recommendations of the task force on community preventive services. Morbidity and Mortality Weekly Report, 56, 1-12.

29.

Herts

K. L.

McLaughlin

K. A.

Hatzenbuehler

M. L.

(2012). Emotion dysregulation as a mechanism linking stress exposure to adolescent aggressive behavior. Journal of Abnormal Child Psychology, 40, 1111–1122. https://doi.org/10.1007/s10802-012-9629-4

30.

Hughes

Bellis

M. A.

Hardcastle

K. A.

Butchart

Dahlberg

L. L.

Mercy

J. A.

Mikton

(2014). Global development and diffusion of outcome evaluation research for interpersonal and self-directed violence prevention from 2007 to 2013: A systematic review. Aggression and Violent Behavior, 19, 655-662.

31.

Illinois PBIS Network Technical Assistance Brief. (2010). Effective Bullying Prevention (BP) within a school-wide system of positive behavior interventions supports. Positive Behavior Interventions & Supports. https://www.isbe.net/Documents/bully-prev-pbis-brief.pdf

32.

Kannappan

Lakshmi Bai

(2008). Efficacy of yoga: Cognitive and human relationship training for correcting maladjustment behavior in deviant school boys. Journal of the Indian Academy of Applied Psychology, 34, 60-65.

33.

Kiefer

S. M.

Ryan

A. M.

(2008). Striving for social dominance over peers: The implications for academic adjustment during early adolescence. Journal of Educational Psychology, 100, 417–428. https://doi.org/10.1037/0022-0663.100.2.417

34.

Kishore

Grewal

Singh

Roy

(1999). Risk behavior in an urban and a rural male adolescent population. National Medical Journal of India, 12, 107-110.

35.

Kobor

(2009, November 1). Government relations update: NIH initiative on “science of behavior change” beginning to take shape. Retrieved from http://www.apa.org/science/about/psa/2009/11/behavior.aspx

36.

Krug

E. G.

Mercy

J. A.

Dahlberg

L. L.

Zwi

A. B.

(2002). The world report on violence and health. The Lancet, 360, 1083-1088.

37.

Matjasko

J. L.

Vivolo-Kantor

A. M.

Massetti

G. M.

Holland

K. M.

Holt

M. K.

Dela Cruz

(2012). A systematic meta-review of evaluations of youth violence prevention programs: Common and divergent findings from 25 years of meta-analyses and systematic reviews. Aggression and Violent Behavior, 17, 540-552. doi:10.1016/j.avb.2012.06.006

38.

Low

Smolkowski

Cook

C. R.

(2016). What constitutes high-quality implementation of SEL programs? A latent class analysis of Second Step® implementation. Prevention Science, 17, 981-991. https://doi.org/10.1007/s11121-016-0670-3

39.

Munni

Malhi

(2006). Adolescent violence exposure, gender issues and impact. Indian Pediatrics, 43, 607-612.

40.

Padmanabhan

(2015, May 23). Urban inequality rising far more in India: François Bourguignon. Mint. https://www.livemint.com/Politics/5JVPBWc7VGX0AMt3U3YeQM/Urban-inequality-increasing-far-more-in-India-Franois-Bour.html

41.

Pereznieto

Montes

Routier

Langston

(2014). The costs and economic impact of violence against children. London, England: Child Fund Alliance.

42.

Piaget

(1972). Intellectual evolution from adolescence to adulthood. Human Development, 15(1), 1-12. https://doi.org/10.1159/000271225

43.

Resnicow

Baranowski

Ahluwalia

J. S.

Braithwaite

(1999). Cultural sensitivity in public health: Defined and demystified. Ethnicity & Disease, 9, 10-21.

44.

Rossiter

(2002). Narrative and stories in adult teaching and learning. ERIC Digest. https://eric.ed.gov/?id=ED473147

45.

Samanta

Mukherjee

Ghosh

Dasgupta

(2012). Mental health, protective factors and violence among male adolescents: A comparison between urban and rural school students in West Bengal. Indian Journal of Public Health, 56, 155-158.

46.

Sandler

I. N.

Ostrom

Bitner

M. J.

Ayers

T. S.

Wolchik

S. A.

Daniels

V. S.

(2005). Developing effective prevention services for the real world: A Prevention Service Development Model. American Journal of Community Psychology, 35, 127-142.

47.

Selman

R. L.

(1977). A structural-developmental model of social cognition: Implications for intervention research. The Counseling Psychologist, 6, 3-6.

48.

Scott

C. L.

(2015). The futures of learning 2: What kind of learning for the 21st century. Paris, France: UNESCO.

49.

Sharma

Kishore

Sharma

Shukla

(2014). Pilot Study for process evaluation of school-based Life-Skills Education Program for prevention of violence in adolescents. Indian Journal of Youth and Adolescent Health, 1, 12-26.

50.

Sharma

Grover

V. L.

Chaturvedi

(2008). Risk behaviors related to inter-personal violence among school and college-going adolescents in South Delhi. Indian Journal of Community Medicine, 33, 85-88.

51.

Singh

B. D.

Menon

(2015). Life skills in India: An overview of evidence and current practices in our education system. New Delhi, India: Central Square Foundation.

52.

Singh

(2003). Paper commissioned for the EFA Global Monitoring Report 2003/4, The leap to equality. Paris, France: UNESCO Institute of Education.

53.

Singh

Gopalkrishan

(2014). Health behaviors & problems among young people in India: Cause for concern & call for action. Indian Journal of Medical Research, 140, 185-208.

54.

Sklad

Diekstra

Ritter

M. D.

Ben

Gravesteijn

(2012). Effectiveness of school-based universal social, emotional, and behavioral programs: Do they enhance students’ development in the area of skill, behavior, and adjustment? Psychology in the Schools, 49, 892-909.

55.

Smith

M. K.

Wood

W. B.

Adams

W. K.

Weiman

Knight

J. K.

(2009, January). Why peer discussion improves student performance on in-class concept questions. Science, 323, 122-124.

56.

Srikala

Kishore

K. K.

(2010). Empowering adolescents with life skills education in schools—School mental health program: Does it work? Indian Journal of Psychiatry, 52, 344-349.

57.

Wilson

S. J.

Lipsey

M. W.

Derzon

J. H.

(2003). The effects of school-based intervention programs on aggressive behavior: A meta-analysis. Journal of Consulting and Clinical Psychology, 71, 136-149. doi:10.1037/0022-006X.71.1.136

58.

World Health Organization. (2009). Violence prevention: The evidence. Geneva, Switzerland: World Health Organization Press.

59.

World Health Organization. (2015). Preventing youth violence: An overview of the evidence. Geneva, Switzerland: Author.

60.

World Health Organization. (2018). Adolescent health epidemiology. Retrieved from https://www.who.int/maternal_child_adolescent/epidemiology/adolescence/en/