Introduction

Children spend a lot of time at school, which is a relevant context for their development. Learning is usually considered the main aim of educational systems, but school environments provide much more than cognitive achievement to children (Greenberg et al. 2003). Through the contact with peers and teachers and the sharing of social norms, children learn which behaviors are accepted and functional to have friends or to be popular, and they become increasingly more able in understanding and regulating their emotions. Spending leisure time with peers reduces disruptive behavior and is functional to the development of social and emotional competence (Veiga et al. 2016). Thus, psychological well-being at school is built on different layers, such as having positive relationships with peers and being able in regulating emotions, which were the focus of the present work.

As to peer relations, we addressed bullying dynamics. Bullying is defined as a form of proactive aggression, unprovoked and aimed at hurting peers, in which an imbalance of power or strength exists between bullies and victims (Olweus 1993; Salmivalli 2010; Smith et al. 1999). The literature agrees that bullying takes place in a group in which all peers are involved or come to know about it. The social framework developed by Salmivalli et al. (1996) about peers involved with different roles has been adopted in many studies and found reliable to explain bullying dynamics. According to this model, peers are implicated not only as bullies or victims but also as passive bystanders (or outsiders), who shy away without taking sides, defenders of the victims, who help or comfort the victimized peers, or followers of the bully, who support the bully by laughing or assisting more actively in harassing others. Whereas defending behavior has been associated with popularity, empathy, and other aspects of adaptive social and emotional competence, the pattern of outsider behavior seems to be characterized by low morality, personal distress, and low self-efficacy (Gini et al. 2008; Mazzone et al. 2016; Rieffe and Camodeca 2016; Salmivalli et al. 1996).

Although data on the incidence of bullying-related behaviors depend on the measures and criteria used, the majority of the classmates tend to witness as bystanders (Goossens et al. 2006; Pronk et al. 2013; Salmivalli et al. 1996). Naturalistic observations, which evaluate the real behaviors occurring, also found that peers are present during about 88% of harassing episodes at school, but they only intervene in 19% of them (Hawkins et al. 2001).

Children who assist to and witness violence can be hurt by aggressive episodes similarly to victims and bullies (Srabstein and Merrick 2013). Addressing bullying as a group phenomenon is, therefore, paramount to reduce adverse psychological effects among children. Several interventions have been implemented to tackle bullying and to improve socio-emotional well-being among school children, and some of them stress the importance to involve not only victims and bullies but also bystanders for effective prevention (Gaffney et al. 2018; Kärnä et al. 2011; Van der Ploeg et al. 2016). As a matter of fact, if bystanders were enabled to intervene efficaciously, or if they became more responsible of what happens among classmates, or even if they reduced their passive behavior, a clear message would reach the bullies, who often use harassment to improve their self-image in front of peers or to show power and dominance (Kärnä et al. 2011; Thornberg et al. 2012).

Besides peer relationships involving bullying patterns, emotion regulation also contributes to influence school well-being. It refers to the ability to manage how emotions are expressed and the intensity and duration of the level of arousal (Gross 2002). Some emotions can become lasting and overwhelming, and people employ different strategies to cope with them, such as avoidance of arousing situations, social support, and cognitive reappraisal. Having difficulties in emotion regulation could be a risk factor for developing internalizing and externalizing problems (Hernandez et al. 2015; Rydell et al. 2003). For instance, a child, who cannot manage the arousal due to a high level of anger, may misunderstand social situations and overreact in front of peers, acting out hostility and aggressive behaviors (Lemerise and Arsenio 2000). Children who bully their peers show proactive aggression, along with reactive and impulsive tendencies, which can be fostered by a deficit in regulating their emotional arousal (Camodeca and Coppola 2019; Camodeca et al. 2002; Shields and Cicchetti 2001; Garner and Stowe Hinton 2010). Overwhelming emotions or a labile mood are also associated with outsider behavior, because they hinder children from employing appropriate strategies to react to the environment, facilitating the use of avoidant strategies or feelings of personal distress in front of demanding situations (Camodeca and Coppola 2019; Rieffe and Camodeca 2016).

In contrast, good regulation skills are associated with social competence, popularity, prosociality, and with defending behavior (Camodeca and Coppola 2019; Denham et al. 2003; Hay et al. 2004). For instance, if a child can manage well his/her emotional reactions, he/she will be more likely to be accepted by peers than a child with emotion regulation difficulties and will be more able to correctly evaluate a given situation, to feel self-confident, and to react appropriately.

These findings underline the close association between emotional and social aspects in determining behaviors and relationships and the need to include both in intervention programs aimed at promoting personal well-being and a school climate of inclusion and respect (CASEL; 2009). In addition, based on the previous literature reviewed, we think that empowering emotion regulation skills may also have positive effects on bullying reduction.

The Social Theatre Model

In recent years, society is giving more and more importance to performing arts, which are considered a valid support for developing creative processes and improving relational and emotional well-being in various contexts, such as schools, workplace, and leisure time (Gjærum and Ramsdal 2008; Joronen et al. 2011). In particular, Social Theatre uses workshop activities to promote individual and community well-being, encouraging participants’ authorship and involvement in the creative and learning process, enhancing participants’ characteristics and resources, and strengthening relational and communicative skills (Bernardi 2014; Bernardi and Innocenti Malini 2015a, b; Boehm and Boehm 2003; Innocenti Malini 2017; Guerra and Militello 2012).

The origins of Social Theatre in Italy – the country in which this study was conducted – date back to 1947 (Bernardi 2014), when Mario Apollonio developed a new concept of theatre, aiming at transforming the passive receptivity of the spectators into commitment and responsibility. The public became an active protagonist of artistic creation, and performances could take place also in non-specific sites. In the same period, the “Experimental Theater” spread in Italy; it aimed at overcoming the conventions and limits of traditional theater, in the perspective of a creative process that could start from people of every social class and was not only a professionals’ prerogative (Bernardi 2014). Finally, the “Theatrical Animation”, developed in the sixties, aimed at changing traditional education into an active and dynamic training process, respecting and valuing children’s free expression.

These movements paved the way for the Social Theater, which employs theatrical and performative activities in extra-theatrical contexts with social, educational, and therapeutic aims (Innocenti Malini 2017; Saxton and Prendergast 2009).

One of the most used actions in Social Theatre projects is the workshop, which creates a protected context where all the participants may experiment new relational dynamics and develop expressive and artistic skills (Rossi Ghiglione and Pagliarino 2011). In the first meetings, the trainer presents many playful activities to help participants to know each other better and to create a welcoming and nonjudgmental context. Subsequently, he/she presents a set of exercises with the purpose of helping participants to improve their expressive abilities, to experience their action in the group and the action of the whole group as a system, and to interact with other participants in a one-to-one relationship or small groups. Finally, the trainer can introduce activities for the dramaturgical and scenic creation, which may include plays, events, or open workshops (Rossi Ghiglione and Pagliarino 2011). The whole work would be incomplete without sharing outside the group the knowledge and expression processes developed within the workshop. If performed in these terms, the openness to the community becomes an act of social dramaturgy: The participants share what they have learned with their spectators, who, at the same time, have the opportunity to feel part of the community (Rossi Ghiglione and Pagliarino 2011).

Theatre Interventions at School

Schools are more and more often including theatrical experiences and workshops in their educational programs. Although different projects use heterogeneous methods, the common value of the theatre is recognized in the possibility of introducing new ways of communicating and relating within the school environment (Bernardi 2014; Guerra and Militello 2012). Unfortunately, the few studies available on the application of Social Theatre to improve emotional and relational competence in primary school children only described the workshop or reported interviews, but they did not employ experimental methods (Beare and Belliveau 2007; Boehm and Boehm 2003; Douglas et al. 2000). The only study using an experimental research design did not find effects on self-concept, problem behavior, or social skills (Freeman et al. 2003).

However, some studies applied other theatrical methods in school contexts and indicated that these activities supported the emotional development of children and adolescents, increased social skills, and improved the relationships between students and with adults (Mavroudis and Bournelli 2016). For instance, a study conducted in Canada with adolescents showed a significant improvement in peer-to-peer relational skills, conflict management ability, and confident leadership style (Walsh-Bowers 1992). Moneta and Rousseau (2008) analyzed the effect of theatrical intervention on ethnic minority adolescents with behavioral difficulties. Employing improvisation and the analysis of emotions and life stories, the workshop helped the participants to reach enhanced levels of understanding of their own and others’ emotions and emotion regulation, and increased the amount and complexity of emotional expressions. Finally, a program including role-play and creative drama found a reduction of aggressive behaviors and the acquisition of effective strategies for conflict resolution in middle and high school students (Graves et al. 2007).

Theatrical activities were also employed to tackle bullying behavior. The study by Burton and O’Toole (2009) was built throughout 10 years and addressed adolescents. The intervention increased the awareness about bullying and the ability to understand the point of view of both the bullies and the victims. Teachers pointed out significant changes in the behavior of students identified as bullies, who became active promoters of the project by committing themselves to train students of other classes. Victims became aware of the successful actions to prevent and deal with bullying, and many students recognized the importance of acting against bullying and identified bystanders as those who can be effective in stopping aggression. Another study (Joronen et al. 2011) investigated the effectiveness of a theatrical project on the reduction of bullying episodes in a Finnish primary school. The activities, carried on at school and home, focused on the development of empathy, social competence, and emotional awareness. Findings showed a significant improvement in the quality of the relationships in the class, both among students and between children and adults.

The Present Study

Moving from the promising findings of studies which employed theatrical activities at school, the present work aims at implementing, in Italy, a workshop based on the method of Social Theatre. The activities proposed to children (like games in small groups, circle time, skits’ creation, and specific theatrical exercises, as detailed in the Method section) aimed at enhancing skills in communication and cooperative works, at promoting the capacity to understand others and to feel at ease in social contexts, and at reinforcing self-efficacy and self-awareness. They were thought to promote personal and interpersonal well-being, and, therefore, to have a positive impact also on emotional and relational dimensions, although they did not focus explicitly on emotion regulation or bullying dynamics. We hypothesized an increase in emotion regulation and in defending behavior, and a decrease in the rate of maladjustment, namely, in emotion lability, bullying, victimization, and passive bystanding behavior, among children participating in the workshops. In addition, considering the association between emotional and social aspects (Denham et al. 2003), we also investigated whether a change in emotion regulation or in emotion lability, due to the workshop participation, could affect a change in bullying behaviors. We hypothesized that an increase in emotion regulation and/or a decrease in emotion lability would be associated with an increase in defending behavior and a decrease in bullying and outsider behavior.

Considering that literature reported gender differences in bullying involvement (Salmivalli et al. 1996), whereas the few studies about Social Theatre did not report results on the participants’ gender (Beare and Belliveau 2007; Boehm and Boehm 2003; Douglas et al. 2000), we planned to include gender in the analyses, without making any assumption about the effectiveness of the workshops for boys and girls.

We employed a quasi-experimental design and an evidence-based approach, which states some criteria to evaluate an intervention. These include at least two measurement points (before and after the intervention), the use of sound and valid instruments and of rigorous statistical analyses, and a long-term follow-up measurement (Flay et al. 2005). Except for the latter, we followed these guidelines and employed a control and an experimental group, randomly chosen and tested twice, i.e., at T1, before any action was carried on, and at T2, after 4 months of theatrical workshop.

Method

Sample and Procedure

Participants were selected within the same school in an urban area near a big city in Northern Italy. A total of 96 children from four third-grade classes of primary school joined the research (45 boys and 51 girls); their age, at the beginning of the study, was 7 to 8 years (M = 7.75; SD = 0.43). Two classes were randomly assigned to the experimental group (EG), which attended the workshop (49 children, 22 boys and 27 girls), and the other two classes were therefore assigned to the control group (CG), which was not involved in any theatrical activity (47 children, 23 boys, and 24 girls).

The four classrooms involved were chosen within the school, because they had never participated in theatrical workshop experiences, whereas the other children had already experienced theatrical laboratories in the past. Therefore, we wanted to test the efficacy of the activities without the influences of previous similar activities, which could have biased the outcomes.

The research project was presented to the school principal, the teachers, and the families, explaining aims, methodology, duration, and the possibility of a final public performance. Parents were asked to provide their written informed consent for the participation of their children, which was obtained for 96 out of 97 families. Children and teachers were assured confidentiality and anonymity of the data provided, which was guaranteed using numeric codes instead of names.

A trained Master student in developmental psychology, also qualified as social theatre trainer, administered the instruments to pupils and teachers and carried on the theatrical activities. She was in a constant dialogue with the teachers, who were present during the sessions and were involved in specific meetings for analyzing the activities and preparing the final play, and, after the conclusion of the workshop, for evaluating strengths and weaknesses of the experience.

The research was structured into three phases. In a school day in the first week of October 2016, two different questionnaires were filled out by children (Bullying Self-Report) and teachers (Emotion Regulation Checklist) (T1). Afterwards, still in the same month, the two experimental classes began the Social Theatre workshop entitled “Body and Voice: Poetry in Action” (as described in the paragraph “The Workshop”). The workshop included 15 weekly meetings of 2 h in each class and ended at the beginning of February 2017, with a play for children’s families. Finally, in the 2 weeks following the performance, the questionnaires used in the first phase of the research were re-administered (T2).

Given that two teachers worked in each classroom, they were both asked to fill questionnaires for their pupils. In one class of the experimental group, one teacher left during the school year, and, therefore, in this class, the questionnaires at T2 were filled in by the new teacher.

Instruments

The Bullying Self-Report Questionnaire was used in the Italian version by Pozzoli et al. (2012). Before administration, children read a definition of bullying: “We say that a child is bullied when another child or group of children tell him bad and unpleasant things. It’s still bullying when a child is hit, punched, kicked, and threatened, when he or she is locked up in a room, receives notes with offensive and bad words, when nobody speaks to him or her and similar things. It’s also bullying when a child is mocked repeatedly and nastily.” The questionnaire consists of 12 items divided equally to assess the four behaviors related to bullying: bullying (e.g., “I make fun of my classmates, giving them mean nicknames, insulting them, or threatening them”), victimization (e.g., “I get hurt, beaten, or pushed”), defending the victim (e.g., “I try to help or cheer up those who have been excluded from the group and left alone”), and being an outsider (e.g., “I watch without doing anything and mind my own business when a classmate is beaten or pushed”). Each behavior is therefore described by 3 items covering physical, verbal, and relational bullying. Participants were asked to evaluate their degree of involvement in the situations described on a Likert scale from 1 (Never) to 4 (Almost always). Items scores were averaged across each scale to yield the four bullying variables. Descriptive statistics are displayed in Table 1. Reliabilities were calculated as the greater lower bound (glb) index, which represents the lowest value of the real reliability (ranging, therefore, in the interval from glb to 1) (Sijtsma 2009). Reliabilities values at T1/ T2 were as follows: bullying (α = 0.57/0.75), victimization (α = 0.80/0.67), defending behavior (α = 0.79/0.73), and outsider behavior (α = 0.44/0.64). A low reliability in the outsider scale is quite common in the literature (Mazzone et al. 2016; Sutton and Smith 1999); item-total correlations at T1 ranged from 0.26 to 0.28 (p’s < 0.05), which could be considered acceptable, given the low number of items and the small sample size.

Table 1 Descriptive statistics of study variables, divided by group (experimental and control) and time of assessment (pretest and posttest)
Full size table

The Emotion Regulation Checklist (ERC) (Shields and Cicchetti 1997, Italian adaptation by Molina et al. 2014), filled in by teachers, was used to evaluate emotion regulation and lability/negativity. The ERC consists of 24 items divided into two scales: positive regulation (or emotion regulation) and lability/negativity. The first scale (8 items) refers to emotion awareness and to the ability to adjust emotional activation in order to respond positively and adapt successfully to the context (e.g., “Can say when she/he feels sad, angry or mad, fearful or afraid”). The lability-negativity scale (16 items) measures the lability of the subject’s emotional states and dysregulated negative affect (e.g., “Is prone to angry outbursts/tantrums easily”). Each item is scored on a Likert scale from 1 to 4 (“Almost never,” “Sometimes,” “Often,” “Almost Always”). High scores in the first scale indicate a good emotion regulation competence and high scores in the second scale indicate great lability/negativity (Molina et al. 2014). Variables were created by averaging the items scores across each scale. Descriptive statistics are displayed in Table 1. Averages of greater lower bound (glb) indexes of reliability across teachers for T1/ T2 were as follows for emotion regulation (α = 0.86/0.86) and lability/negativity (α = 0.94/0.94).

Given that the two teachers were equally involved in each classroom and knew very well their pupils, they could provide their personal view on each child. The inter-correlations across teachers’ scores were moderate (r range between 0.34 and 0.64 for emotion regulation, and between 0.48 and 0.70 for lability/negativity), indicating that children could show different emotional patterns with different teachers, or also that teachers could have different standards when responding to the questionnaire. Therefore, to have a combined point of view for each child, which could be more reliable than a single measure or two separate measures, an average score between the two teachers’ scores was computed for each scale at T1 and T2.

The Workshop

As previously indicated, the workshop was structured with the purpose of promoting children’s well-being and good relationships in the class through activities aimed at enhancing social connections and cooperation, understanding others’ point of view, and controlling one’s behavior to reach goals and to meet group norms. The theme of poetry and natural elements was chosen, in agreement with the teachers, to integrate the activities in the curricular program of science and literature and to promote interdisciplinary connections. Children, in the theatrical fiction, would have become guardians of water, earth, air, and fire, experiencing them through body and words in poetry. We assumed that giving children the opportunity to explore these school themes from different points of view, such as through the body or tactile materials (e.g., fabric), could increase their interest in these subjects and enlarge their possibilities to work in group and collaborate also during curricular lessons.

The specific activities in each session followed the succession of different phases, as required by the Social Theatre method (Bernardi 2014). In the “separation phase,” a ritual set of gestures and actions mark a detachment from everyday life and, in the present workshop, each child in turn had to stand up and declare to be the guardian of a natural element. Afterwards, the “margin phase” was aimed at improving participants’ body awareness, relationships, and communications, through playful activities based on group and dyadic games, dances, meditation, and specific theatrical exercises. For instance, to stimulate reciprocal listening, the mirror game was used, in which one member of the couple had to copy the movements displayed by the other member. The interaction of each child with the whole group was facilitated by several exercises, for example, tableaux vivant, which consists of the creation of a living picture on a selected theme, involving all children. During the “reintegration phase,” the trainer proposed to develop a small product to synthesize the experience, through representations, drawings, or writings. For example, while dealing with the theme of water, the trainer asked children to work in small groups and think at agreed-upon movements for a poem being read before and then to perform in front of the classmates. Children were also helped to create individual and group poems, always about the theme of natural elements. In this way, participants could reflect on what happened, process the experience, and share their impressions and emotions with the group.

In addition, during every session, the trainer verified the impact of the activities on a socio-emotional level by asking a feedback to children. Finally, a closing ritual marked the end of each workshop session (Rossi Ghiglione and Pagliarino 2011). All the activities carried on during all the sessions were useful for children to set up the final play for their families, based on the poetic material collected and developed during the workshop, and on the improvisations.

Data Analysis

First, in order to check whether study variables had similar means in EG and CG at the baseline (T1), t-tests were run. Therefore, pretest and posttest assessments in experimental and control groups were compared by means of repeated measures general linear models, including pretest and posttest variables as within-subject factors, and group (experimental and control) and gender as between-subject factors. In this way, the multivariate test provides results about differences between boys and girls and between CG and EG (between-subject factors), about differences between T1 and T2 (within-subject factors), about the interactions between them (i.e., whether study variables changed between T1 and T2 in CG and EG, and in boys and girls), and about the 3-way interactions (i.e., time X group X gender). In this latter case, in order to disentangle the effects of the workshop participation for boys and for girls, repeated measure analyses of variance were run again, separately for boys and girls.

Finally, we run hierarchical regressions to investigate whether emotion regulation and/or emotion lability moderated the change of bullying-related behaviors. Bullying-related behaviors at T2 were the outcome variables; gender, group, corresponding bullying variable at T1, and emotion regulation/lability at T1 and T2 were controlled in the first step, and the interaction between group and emotion regulation/ lability at T2 entered in the second step.

Results

Results of the t-tests to investigate possible differences between CG and EG at T1 showed that variables’ means did not differ between the two groups, except for emotion regulation, which was higher in the control group than in the experimental group (t(94) = −6.39; p < 0.001). When boys and girls were compared separately, analyses confirmed a higher emotion regulation among both boys and girls in the CG than EG group (t(43) = −4.70; p < 0.001, for boys, and t(49) = −4.81; p < 0.001, for girls), and indicated that outsider behavior in boys was higher in EG than CG (t(43) = 2.68; p < 0.05). Descriptive statistics for EG and CG are shown in Table 1.

Findings of the repeated measures general linear models pointed to a significant between-subject effect for gender (independent from time of assessment and group): Boys, compared to girls, obtained higher scores in lability/negativity (F = 23.99; p < 0.001), bullying (F = 16.42; p < 0.001), outsider behavior (F = 12.77; p < 0.01), and victimization (F = 7.79; p < 0.01), and lower scores in emotion regulation (F = 5.92; p < 0.05). The between-subject effect for group (independently from time of assessment and gender) indicated that emotion regulation was higher in the CG than in the EG (F = 42.91; p < 0.001), and outsider behavior was higher in the EG than in the CG (F = 4.41; p < 0.05).

Within-subject results (independently from group and gender) showed a significant main effect of emotion regulation (Pillai’s Trace = 0.05; F(1,92) = 4.93; p < 0.05), which was higher at T1 than at T2. As indicated by the significant interaction effect between group and emotion regulation (Pillai’s Trace = 0.07; F(1,92) = 6.77; p < 0.05), this decrease was attributable to a significant decline in the control group (t(46) = 3.87; p < 0.001), whereas emotion regulation remained stable in the group participating in the theatre activities. A significant effect emerged also for the 3-way interaction between outsider behavior, group, and gender (Pillai’s Trace = 0.05; F(1,90) = 4.71; p < 0.05). To further investigate gender differences, two repeated measure analyses of variance were run separately for boys and girls, showing that outsider behavior in boys in the EG decreased after the workshop (Pillai’s Trace = 0.09; F(1,42) = 4.23; p < 0.05; M = 2.24 and M = 1.73 for T1 and T2 in EG; M = 1.59 and M = 1.62 for T1 and T2 in CG). Results were not significant for girls, who remained stable in their outsider behavior (M = 1.42 and M = 1.47 for T1 and T2 in EG; M = 1.40 and M = 1.33 for T1 and T2 in CG). Changes in emotion regulation and outsider behavior are depicted in Figs. 1 and 2.

Fig. 1
figure 1

Emotion regulation at T1 and T2 in experimental group (EG) and control group (CG)

Full size image
Fig. 2
figure 2

Outsider Behavior at T1 and T2 in Experimental Group (EG) and Control Group (CG) for Boys and Girls. Differences between T1 and T2 were Significant only for Boys (Solid Lines).

Full size image

Given that emotion regulation, but not lability/negativity, changed after the activities, we tested whether emotion regulation affected changes in bullying involvement. Hierarchical regressions indicated that the interaction between emotion regulation at T2 and group was significant only in predicting T2 outsider behavior (R2 = 0.22; p < 0.05; β = -3.74; p < 0.05). Post hoc regressions run separately for EG and CG showed that a lower score in T2 emotion regulation in the CG predicted a higher score in T2 outsider behavior (β = −0.32; p < 0.05). Equation was not significant for the EG.

Discussion

The study hypotheses have been partially satisfied, and findings are promising in showing the efficacy of Social Theatre activities to improve children’s well-being. Although emotion regulation remained stable in the experimental group, it diminished after a few months in the control group. It is difficult to find a plausible explanation for this finding. We can speculate that children become more tired and stressed in the second term, likely because of homework overload, which might have contributed to the decrease in emotion regulation. It is also possible that other factors (e.g., level of stress within the family; lack of support from parents and teachers), which accumulate over the school year, contribute to a decrease in emotion regulation skills. However, emotion regulation did not decrease among children participating in the workshop. We could surmise that the presence of several moments of reflection and sharing of emotional contents during the workshop sessions, together with the opportunity for children to express these contents in the language more congenial for them (verbal, nonverbal, written, drawn), could have helped in exercising self-control and emotion management. The trainer often focused on the legitimation of every positive and negative emotional expression, and on the necessity to share personal experiences and feelings with the group to reach the objectives of each session. It is also possible that the workshop constituted a pleasant and relaxing activity in the week routine, and that it contributed to keep a positive mood.

The decrease of emotion regulation in the control group was also partially responsible of an increase of outsider behavior in the same group. Children who have difficulties in managing their emotions may develop deficits in understanding social situations and in selecting proper responses (Lemerise and Arsenio 2000). In addition, a high level of emotional arousal may indicate personal distress in front of others’ sufferings and, therefore, contribute to keep these children away from bullying and unable to help their peers (Rieffe and Camodeca 2016).

We also found that outsider behavior decreased among boys who participated in the theatrical activities. As described in the Method section, the activities proposed during the workshop promoted the appreciation of each participant’s abilities and characteristics and contributed to a nonjudgmental and respectful context which allowed children to express themselves freely. Moreover, working in small groups to reach appealing aims encouraged the development of effective verbal and nonverbal communication skills, group collaboration, and caring attitudes toward classmates (Bhukhanwala 2014; Joronen et al. 2011; Mavroudis and Bournelli 2016; Walsh-Bowers 1992; Walsh-Bowers and Basso 1999), which may have strengthened bonds and helped outsider children to increase their sense of class belongingness. In addition, the theatrical laboratory supported the idea that every individual is responsible, in the group, for the success of each exercise and the whole final performance and can help to achieve results through cooperation and active participation. It is possible that increasing one’s own sense of participation and responsibility may have contributed to motivate some boys to reduce their passive behaviors (which may provide bullies with an important public) and to become more concerned for their peers. However, a decrease in passive observation behavior does not indicate an increase in active defending behavior or in intervening in conflicts resolution, which, as indicated in the next paragraph, may need more time.

In our sample, no significant changes emerged regarding defending behavior toward the victims. As highlighted by various authors (Burton and O’Toole 2009; Mavroudis and Bournelli 2016), to promote an effective and active change, it may be necessary to provide, during the workshop, explicit references to the dynamics underlining defending behaviors. It is also possible that fostering a direct intervention toward the victim is not easy and requires more time, because children need to change their competence beliefs and cognitions about their self-efficacy, as well as to learn conflict mediation skills, and improve social and physical abilities (Pronk et al. 2013).

Gender differences deserve a reflection. Girls, in their natural interactions with peers and compared to boys, are more acquainted to being involved in group discussions or in observing peers, employ a more cooperative style, and are more empathic and concerned for others (Garaigordobil 2009; Hay et al. 2004). It is possible that these features characterizing the workshop have been less salient for outsider girls than for outsider boys, who, in contrast, could have taken more advantage of exercises on communication, care for others, and emotion expression. The theatrical activities resulted, thus, are more effective for boys than for girls. Indeed, boys initially showed high values on the outsider scale, which, after the workshop, reached a similar rate as those of boys in CG and girls. Although we do not know whether the sense of responsibility of outsider boys really increased, we assume that the involvement in collaborative activities with a common aim contributed to the reduction of this form of disengaged behavior.

Limitations and Future Perspectives

One of the main limitations of this study lies in the reduced length of the theatrical lab, which could be more efficacious if it stretched throughout the entire school year and fostered the inclusion of the workshop activities and routines in the daily management of classes (Joronen et al. 2011). In fact, the benefit of Social Theatre method would lie in promoting lasting changes, which are facilitated through longer projects and the involvement of all the adults around children (Bernardi and Innocenti Malini 2015a, b; Innocenti Malini 2017). For this reason, before and during the workshop, it would be useful to organize training sessions on Social Theatre for teachers, educators, and parents, to provide them with means for understanding what happens during the workshop and with useful techniques to maintain and amplify the benefits obtained. Likewise, it could be useful to add a session, after the final play, to analyze the whole experience with the children and to allow them to share impressions and feelings with their classmates.

Similarly, stronger outcomes could have been yielded employing activities with a clear focus on bullying dynamics or emotion regulation. As a matter of fact, some previous studies structured the workshop duration similarly to the one described in this paper, but they explicitly addressed these aspects, obtaining direct and immediate effects (Graves et al. 2007; Joronen et al. 2011). In contrast, employing activities that were not specifically designed to tackle bullying or to improve emotional regulation may have accounted for the stability (rather than the increase) of emotion regulation or defending behavior, and for the stability (rather than the decrease) of bullying or victimization.

Another limit is the large number of participants to each session, which, combined with the presence of a single trainer, influenced both the amount of time planned for each exercise (reducing the number of activities) and the trainer’s attention to the relationships among participants. In addition, the questionnaires were administered by the same person who led the workshop, which could have influenced the answers of the participants involved in the activities, due to the relationship developed between the trainer and the children in the weeks spent together.

We realize that the reliability of outsider behavior at T1 was low, which has been found also in previous studies (Mazzone et al. 2016; Sutton and Smith 1999). This may be due to the small sample size or to the short length of the scale. It is also likely that the items grouped in the outsider scale measure different aspects of the construct, which are less easy to detect because they do not indicate clear actions, but may be perceived as “non-behaviors” (e.g., standing by, doing nothing, minding one’s own business) (Mazzone et al. 2016; Sutton and Smith 1999; Tavakol and Dennick 2011). It is therefore possible that children provided incoherent responses, in particular at T1, when they saw the items for the first time and had never thought about them, or roles were not defined yet, as it could be at the beginning of the year. We could surmise that, in the second administration, children recognized the questionnaire or had reflected upon items, and provided more consistent responses, or that outsider behaviors became more coherent within the same persons, because relationships and roles tended to stabilize throughout the school year. Although we do not think that different reliability values at T1 and T2 affected the outcomes, because results go in the expected direction, further studies are needed to confirm our findings with more robust measures and clarify the internal consistency of outsider behavior.

The small sample size, obtained from a single school, did not allow to generalize results. Future studies could extend the project to larger samples of children of different age groups. The effects of Social Theatre activities on other aspects of socio-emotional well-being could be also analyzed, including self-efficacy, which could motivate children to take action against bullying (Pronk et al. 2013), communication skills, creativity, and self-esteem, which have been found to improve with theatrical activities implemented with different methods (Boehm and Boehm 2003; Hui and Lau 2006; Kardaş and Koç 2017).

Finally, future studies could include the collection and analysis of both qualitative and quantitative data, integrating results from questionnaires with interviews, focus groups, and observations. This combination of different methods could be useful to understand what participants experience in each phase of the workshop and to achieve a better comprehension of the results of the questionnaires (Costa et al. 2014).

Although it is possible to identify several actions to improve the interventions based on Social Theatre and the evaluation of their efficacy, we think that the results obtained are promising, set a good starting point for future studies, and may contribute to increase research attention on the effects of the Social Theatre workshops in educational contexts. Analyzing the benefits of theatrical activities through an evidence-based, multi-informant, approach is a strength to be considered in future research or practical works aimed at implementing performing arts in schools.