School Improvement by Addressing Bullying: Α Case Study in Greece

The main purpose of this study case is to investigate the contribution of the school community to the improvement of the school. In that context, a two-phase research was conducted. In the first phase, the self-evaluation process was implemented during the first year of the research with the participation of the school community. An overall picture of the school was created, with its strong and weak points reflected in the school's final self-evaluation report. Upon the completion of the school self-evaluation process the school community decided on the implementation of actions in order to reduce a number of dysfunctional behaviors, such as bullying incidents that occurred in the school on the part of some students. The school actions and the relevant results constituted the second phase of the research work. The results showed that some of the dysfunctional behaviors were found to be decreased to a statistically significant level after action was taken by the school community.


Introduction
In many countries, the need of dealing with school bullying has only emerged after the death of children. More specifically, in Norway its importance was highlighted in 1982 after the death of three children (10-14 years), while in Japan a girl's suicide in 1986 triggered the recognition and discussion of the problem. In 1997 the torture and murder of a 14-year-old child by six girls and a boy was a notable incident of school bullying in Canada. In 1999 in Columbine, USA two high school students murdered twelve students and one teacher while injuring twenty others before committing a suicide, executing a three-hour revenge and insanity plan (Hymel et al., 2010). In Greece, the first officially recognized victim of bullying was Giakoymakis Vangelis who was pushed to suicide because of his constant and excruciating bullying by his fellow students in Ioannina in 2014 (Giakoymakis, 2020). Giakoymakis' suicide was preceded by the disappearance and murder of young Meschiville Alex in 2006, who had fallen victim of school bullying and whose body was never found (Murder of Alex Meschisvili, 2020).
Regarding research on bullying conducted in Greece, two studies are worth mentioning. The first is an older research study, implemented in schools in Athens, with the participation of 791 students. Its main finding was that nearly one out of three pupils is bullied (38.4%) (Andreadakis et al., 2007). The second one is a more recent study by Ioannides (2018) with the participation of 800 pupils of secondary education from the prefecture of Ioannina. According to the research observations, 16% of pupils reported that they had been bullied, and at the same time 35% stated that even when they realize that a classmate of them is being bullied they choose to stay uninvolved (Ioannides, 2018).

The school bullying phenomenon and its psychological impact
School bullying is defined as the deliberate and conscious desire to harm, threaten or intimidate someone, provided that there is a power imbalance between the offender and the victim and that the victim cannot defend himself or herself (Olweus, 1993;Volk et al., 2014). has a significant negative impact on its victims. At first people's self-esteem decreases, aggressive tendencies are created while both school dropout and the use of psychotropic substances increase. The individuals also exhibit depressed mood, social phobias, and social withdrawal (Dill et al., 2004;Kumpulainen & Räsänen, 2000;Nansel et al., 2003;National Academies of Sciences, Engineering and Medicine, 2016;Swearer et al., 2004). In addition, student performance is reduced and the ability to build healthy social relationships is ceased (Twemlow et al., 2009). In some cases suicidal thoughts are also observed (Honig & Zdunowski-Sjoblom, 2014).
What is not disputed is that experiencing bullying behavior (being a "victim"), as well as using it (being a "bully"), is highly problematic for young people and both bullying roles are robustly associated with a wide-range of indices of short-and long-term maladjustment. A number of meta-analyses highlight the impact in domains as diverse as internalizing and externalizing problems (Gini et al., 2017). A more recent definition of bulling is defined as unwanted aggressive behavior by other youth (not including siblings or dating partners), with an observed or perceived power imbalance, and is repeated multiple times (Center for Disease Control, 2019).

Intervention programs to face school bullying
In recent decades, more and more intervention programs have been implemented to face school bullying (Bradshaw, 2017;Evans et al., 2014;Nickerson, 2017), as the latter is the most common problem reported in schools in terms of behavioral problems (Zhang et al., 2016). Some indicative examples are the following: (1) The Olweus Program is the first holistic intervention program for dealing with school bullying (Olweus, 1993), (2) Bullying Prevention in Positive Behavioral Interventions and Supports (BP-PBIS) is about creating school rules, teaching social skills and monitoring students' behavior (Good et al., 2011), and (3) Social Emotional Learning (SEL) focuses on teaching emotional skills (recognizing and controlling emotions and building healthy social relationships) that contribute to improving school performance and school climate, reducing school dropout and penalties (Durlak et al., 2011).
(4) KiVa is a school-wide evidence-based program developed in Finland for children aged 7 to 15 years. Its primary focus is on changing the role of bystanders (fellow students who witness bullying events) as a means to prevent and stop bullying in schools. The program teaches children to recognize bullying and how to respond if they see bullying occur (Axford et al., 2020).
Bullying prevention programs such as Olweus Bullying Prevention Program and the KiVa anti-bullying program have been developed and scientifically evaluated to counteract bullying in school (Thornberg et al., 2020).

Criticism on intervention programs against school bullying
Existing literature provides a plethora of evidence regarding increasing rates of school bullying as well as the emerging type of cyber bullying. Specifically, Indicators of School Crime and Safety of 2011 suggested that more than 28% of adolescents 12-18 years of age (girls 31% and boys 25%) have been a victim of school bullying (Finkelhor et al., 2013).
Crisis management at school is an urgent need that the educational community is called to manage. The contribution of the wider school community (parents and institutions) is considered necessary for the creation of the appropriate framework for such improvement actions.
However, adopting a program designed in another country may not always be effective. Indicatively, the Olweus (1993) program is more effective in the Nordic countries than in the US, due to cultural differences or differences in the curriculum Limber et al., 2004;Smith et al., 2004). According to Thomson (in Axford & Morpeth, 2013), an important factor for the success of an intervention program is the quality of the relationship between the two parties involved. In other words, unless there is a positive relationship between the "experts" piloting an intervention and the educational community it is very likely for the intervention program not to deliver the expected results (Divecha & Brackett, 2019).
According to the US Department of Education report (Lessne & Yanez, 2016), the fact that school bullying cases have remained steady over the past decade and in some cases have even slightly decreased, indicates that school bullying is a social phenomenon. It is worth noting a global trend. In more detail, violence in schools has reached a climax that goes far beyond conflicts among students, while targeting teachers. In the US Centers for Disease Control 9% of teachers have asked for help because of facing their students' threats and another 5% due to being physically attacked. Indeed, in 2014 attacks on teachers were twice as high as in any other profession. Similar results have been found in countries such as Germany (Bauer et al., 2007), Canada (Wilson et al., 2010) and Taiwan (Anderman et al., 2018;Chen & Avi-Astor, 2009).
As a result, many intervention programs focus on the symptom of problems, such as aggression or depression, rather than on the social causes that cause them. In other words, interventions are not structural and do not focus on dealing with unemployment, economic poverty and social exclusion. Instead, through personalized intervention and by addressing social problems results will come for students and their families. It is no coincidence that countries with the worst child welfare are experiencing extreme socioeconomical inequalities and tend to be more competitive overall (Axford & Morpeth, 2013). The US can be indicatively mentioned as an example where, according to critics of intervention programs, the latter are created to cover at least partially the absence of public services (Axford & Morpeth, 2013). Unfortunately, most bullying prevention programs do not work, at least not in randomized controlled trials ). Other researchers also support this observation (Ferguson et al., 2007), also mentioning changes often only among younger children (Yeager et al., 2015). Most programs focus on remedying dynamics such as emotional dysregulation, poor conflict management, and empathy deficits, factors that may explain only a portion of aggressive behavior (Faris et al., 2020).

Interventions for school bullying based on school initiatives
In recent decades, there has been a conflict regarding the effectives of approaches adopted against school bullying. Ttofi and Farrington (2009) state the existence of: 1) a control group 2) students' self-report statements 3) the effect size of the sample and 4) a sample with more than 200 individuals (Ertesvåg, 2014) as important criteria for effective intervention. Therefore, various researchers conclude that limiting school bullying should be based on research based on Evidence-based practice (EBP) and in particular what Sackett et al. (1996) define as ''best available evidence" (Ertesvåg, 2014).
However, these data do not contribute much to the daily practice of teachers, though certainly being very important for the design of an educational policy. The educational community often wants to improve the school climate with actions that start on its own initiative. Therefore, these initiatives are neither part of a broader educational policy nor the result of pressure for accountability, through the process of external assessment of the school units, which calls for a teacher effectiveness proof.
According to Gottfredson and Gottfredson (as cited in Greenberg et al., 2005), teachers are often unable to put theory into practice. This difficulty is exacerbated when teachers are called upon to put these theoretical principles into practice in real school settings. Therefore, they are often unable to implement an intervention designed by an external body.
However, regardless of the implementation of an intervention program, research shows that the way in which the educational community applies teaching practices, or the behavior of the teacher in general, as well as the wider school climate significantly contribute to reducing school bullying incidences (Cohen et al., 2009). As a result, when students have positive attitudes toward the school climate, they are less likely to develop and externalize behavioral problems (Espelage et al., 2000;Goldweber et al., 2013;Totura et al., 2009). Therefore, if a school wants to reduce school bullying incidences, Pas et al. (2018) suggest teachers adopt the following behaviors: 1) create a positive school climate, 2) teach social skills, so as to reduce the feeling of rewarding bullying, 3) react immediately to bullying incidents (Smith et al., 2004).
Parents also play an important role in dealing with school bullying. Parents are the most important people in a child's life (Von-Salisch, 2001). Therefore, in addition to the practices and actions implemented by the educational community, parents' contribution is very important. In other words, they need to transfer the same message to their children (Ostrander et al., 2018;Sheridan et al., 2004). However, according to a number of surveys with primary school students, 53% of the students who are subject to bullying report it to their teachers, while 67% of the students report it to their parents (Fekkes et al., 2005). Therefore, the cooperation of the members of the school community is a necessary prerequisite for dealing with school bullying. It is indicative that many parents are unaware of the fact that their children are bullying their classmates at school (Holt et al., 2009).
In the international literature, the improvement of the school climate is mentioned as the main advantage for the schools that implement intervention programs, in which all students are included (Kyriakides & Creemers, 2012;Nickerson et al., 2014;). The same findings are found in the Friendly Schools programs (Cross et al., 2019) and the Creativing a Peaceful School Learning Environment (CAPSLE) Schools (Twemlow et al., 2009).

Research Goal
The main purpose of this study case is to investigate the contribution of the school community to the improvement of the school. Especially, to reduce a number of dysfunctional behaviors, such as bullying incidents that occurred in the school on the part of some students.

Sample and Data Collection
Out of a total of 257 students in the sample, 123 students participated in the 1st year of the survey and 134 students in the 2nd year. Also, two (2) classes were excluded from the statistical analysis. Specifically, the 6th grade (the last grade in Greek primary education) from the 1st year of the research and the 1st grade from the 2nd year of the research, because the students did not participate in both years of research, due to their change of Education Level. Moreover, ten teachers participated in the interviews (10) conducted after the action. Six of them were the 'leader' teachers of a school class (I1, I2, I3, I4, I5, I6), teaching the core subjects while four were teaching specialization subjects, such as English, special education etc. (I7, I8, I9, I10), according to the Greek school curriculum.

Instrument and data analysis
The main research instrument was the questionnaire, which was structured by the researcher who was also a teacher in the specific school. This questionnaire was based on the principles given by the Hellenic Ministry of Education and Religious Affairs in the context of the self-assessment of the school unit (Hellenic Ministry of Education and Religious Affair, 2021).
The statistically significant differences between the means of the two-year research were studied, which relate to students' responses concerning which incidents they consider as remarkable from those happening in their classroom, playground and at home. The statistical criterion selected for the classroom, the playground and the home environment was the Mann-Whitney U test. The non-parametric Mann-Whitney U test was chosen, since the data did not follow a normal distribution (Howitt & Cramer, 2011). The reason for selecting this statistical criterion for independent and not dependent samples (Wilcoxon) was that the teacher-researcher, in the context of the research ethics, committed that the questionnaire would be anonymous and the results would not be able to be identified. Therefore, students were asked to be completely honest when filling in the questionnaire. Finally, the way quality variables are encoded is mentioned. The higher the mean towards the value of 4.00 is, the higher the degree of agreement of the subjects with the specific view becomes. This fact indicates that this criterion is very important or that this behavior is always observed. In more detail: • I totally agree/ always = 3.51-4.00.
• I disagree = 1.00-1.50 Hinton et al. (2004) reported acceptable reliability from above 0.50 (Hinton et al., 2004). Regarding the current research sample (Ν=256) the reliability with Cronbach's alpha value of 0.54 which is considered to be moderate.

Procedure
Due to the need for self-assessment in the context of the school unit, phenomena of school bullying were realized against which the educational community had to take specific actions. The present intervention belongs to the intervention programs that are implemented in real life conditions. These programs take place in schools without the participation of specialized staff (Cross et al., 2019).
More specifically, the educational community of the school (Head Principal and Teachers) with the cooperation of the teacher-researcher decided to implement an action, which concerned raising the awareness of students and of the entire school community regarding school bullying. A survey was initially conducted at the end of the school year, as a part of the self-assessment of the school unit (1 st year of the study). Then, a series of actions were carried out by the teachers of the school in collaboration with the parents of the students (2 nd year of the study). The main purpose of the school action was to raise the awareness of the entire school community on school bullying issues. The action specifically concerned: 1. Teachers, 2. Parents, 3. Students. Regarding the teachers of the school, the main objectives of the action were: To be informed about the specific phenomenon, to make their lessons game-based, concerning issues of school bullying taught during the "Flexible Zone", to be vigil when on duty at breaks or during the lessons, so as to deal with such phenomena, to provide parents with children with dysfunctional behavior with opportunities for interpersonal counseling. Regarding the parents of the school, the objectives of the action were: To be informed on the issue, to implement what is proposed during the interpersonal counseling sessions. The cognitive goals for the students were: To be informed about school bullying, to get to know effective ways of dealing with bullying. The emotional goals were to: Be aware of school bullying, to understand the feeling of bullying behavior. The psychomotor goals for students were: Τo participate in dramatization activities concerning bullying. Upon the completion of the activities at the end of the 2nd year, a second measurement was carried out in order to estimate the impact of the action on the students' behavior, as well as the way in which the students themselves would choose to face school bullying, if it happened to them. After the implementation of the action, which concerned raising the awareness of students regarding school bullying, a semi structured interview was conducted with the participation of the 10 teachers of the school. The purpose of this question was twofold: on the one hand to determine whether teachers consider that they can manage school bullying and on the other hand in what ways they choose to deal with it. For further insights on the effect of the school's action on children see Appendix (children's drawings).

Ethical considerations
Dealing with sensitive issues such as children and dysfunctional behaviors requires high ethical awareness (Lund et al., 2015). To protect the identity and the confidentiality of the children and of the school community as well, there is no reference of the school or of the region, beside of the fact that the research was conducted in Greece. Since the school action was implemented by the school teachers there was no conflict during the implementation of the program. Lastly, the fact that the researcher was part of the school community it helped to avoid any possible problem that might occurred.

Findings
The impact of teachers' actions on the class Table 1 shows (in bold) statistically significant means and standard deviations of students' responses, regarding incidents that they consider occurring in their classrooms. According to the data presented in this table (Table 1) it appears that there are statistically significant differences in the mean of four out of the twelve items in the questionnaire. More specifically, teachers repeat the material taught to students, in case they did not understand it more frequently during the 2 nd year (2014-2015) to a statistically significant degree (p = .028). The students also stated that almost always (mean: > 2.50) the noise is made by the students themselves, however, the mean of the sample decreased statistically significantly during the 2 nd year of the survey (p = .006). Finally, during the 1 st year, students stated that they sometimes (mean: > 1.50) help their classmates (p = .000). However, after the implementation of the intervention, they stated that they almost always (mean: > 2.50) help their classmates. Finally, the teachers in the 2 nd year of the research almost always examine (mean: 3.25) all the students, while in the 1 st year of the research the students stated that they always examine them (mean: 3.25) (p. =. 003). The degree to which the phenomenon exists were measured trough the effect size equation times 100. A generally accepted minimum level of power is 0.80, and the proposed conventions or operational definitions of "small (r = .10)," "medium (r = .30)," and" large"(r = .50) effect sizes are defined across all the statistical tests (Cohen, 1988). Small to medium effect size (r = .14 to r = .25 -14% to 25 % of the variance explained) are noticed for all statistically significant items. The largest effect size is for the item 1 (The teacher tests all students). Besides, small effect size of r ≈ .10 is accounted as well for the items 4 and 11, while the items 6 and 12 have some minimum level of power (see Table  1). One possible interpretation of this is that teachers may have used teaching time to discuss and deal with dysfunctional behaviors that occurred either in the classroom or at break time. In summary, it can be said that the climate in the classroom improved and the noise decreased, while at the same time the cooperation between the students increased. Table 2 shows (in bold) the statistically significant means and standard deviations of students' responses, regarding incidents that they consider occurring in the playground. According to the data in Table 2, it is clear that there are three criteria that show statistically significant differences in the means. First of all, the students in the 2 nd year (mean: 2.57) of the research stated that they hang out with younger children more often than in the 1 st year of the research. As a result, it can be said that student relations have improved to some extent. What is worth noting, however, is that in the 2 nd year of the study, the children always (mean: 3.12) reported the dysfunctional behaviors they observed compared to the 1 st year of the study, when they sometimes reported such behaviors (mean: 2.01) (p. =. 000). In addition, the students stated that they have sometimes (mean: > 1.51) been hit both during the 1 st and the 2 nd year of the research study. However, the means of the 2 nd year (mean: 1.64) are lower than the corresponding means of the 1 st year (mean: 1.86) and the differences are statistically significant (p. =. 025). Small to almost large effect size (r = .16 to r = .44 -16% to 44% of the variance explained) are noticed for all statistically significant items. The largest effect size is for the item 9 (I tell someone when I see some children intimidating other children). Also, small effect size of r ≈ .10 is explained as well by the items 2 and 5, while item 13 has some minimum level of power (see Table 2). Therefore, it seems that the cooperation of all members of the school community has contributed, to a certain extent, to the improvement of the dysfunctional behaviors that took place at break time. Table 3 shows (in bold) the statistically significant means and standard deviations of students' responses to incidents that they consider occurring with their family during the 2013-2015 school years. There is only one statistically significant difference in the means concerning the five questions posed. Therefore, it is very interesting that there is a statistically significant difference is observed during the 1st year of the study, when the children stated that they almost always felt alone (mean: 1.71) at home, as opposed to the 2nd year, when they stated that this sometimes happens (mean: 1.41). Small to medium effect size (r = .20 -20% of the variance explained) could be noticed for the statistically significant item 5 (I feel all alone). Besides, small effect size of r = .12 is accounted as well for the item 2 (see Table 3). Although this improvement in parental practices may have occurred due to various reasons, perhaps to some extent, the cooperation of teachers with parents has also contributed. Table 4 below presents various functional and non-functional strategies for dealing with school bullying. Also, the distribution of students' answers regarding the way they would personally choose to face school bullying is recorded. I give them money 16 11.9 9 6.7 9 6.7 100 74.6 1.56 1.051 I do whatever those bothering me ask me to do 12 9.1 5 3.8 21 15.9 94 71.2 1.51 .937

Ways students can deal with school bullying
I start bothering other children too 9 6.9 6 4.6 10 7.6 106 80.9 1.37 .862 The control of the mean values of Table 4 shows that the majority of students totally agree with the four strategies, which relate to the confidentiality of bullying by the students themselves to individuals important to them. In particular, 91.8% of students agree or totally agree with the informing parents strategy (mean: 3.66). Also, the percentage of students (90.9%) who would inform the school principal (mean: 3.66) is particularly high. A significant percentage (90.9%) of students would confide in a person of trust (mean: 3.65), and 91.8% of students would report it to their teachers (mean: 3.64).
Also, with a mean of > 2.50, students agree with two of the strategies. More specifically, 83.8% of students show special confidence in their classmates, as they believe that the eyewitnesses (bystanders) of the incidents may help them (mean: 3.39). This element is of particular importance, as the importance of bystanders in dealing with the phenomenon of school bullying is becoming increasingly important. Nearly one in two students (55.7%) believes that their parents would realize this without mentioning it to them (average: 2.69).
The majority of students disagree with the following eight strategies. In particular, 57.9% of the students disagree or totally disagree with the strategy of inaction (mean: 2.17). 68% of students disagree (mean: 2.13) with attempts to impress classmates with bullying through the display of seemingly expensive clothes. However, this strategy would be adopted by one out of three students. This fact needs further investigation. Also, 72% of the students disagree with the choice of school change (mean: 1.95). However, one out of three students may have made such a change. Avoiding dangerous classmates at breaks is referred by 27% of the students who prefer to stay in class, while 73.1% of them disagree with this strategy (mean: 1.85). Also, 79.7% of students disagree with the strategy of retaliation against students showing bullying behavior (mean: 1.69). It should be noted, however, that one out of five students would hit a classmate in retaliation, an incident that would possibly trigger a vicious cycle of violence.
Moreover, 82% of the students disagree with the strategy of staying at home to avoid possible bullying at school (mean: 1.62). At the same time, however, one out of five students may have adopted such a strategy. If the money giving strategy was to stop bullying, 18.6% of students would resort to this option, while 81.3% of students disagree with this strategy (average: 1.56). The vast majority of students (87.1%) would not engage in blackmail, where the victim would do whatever asked by the perpetrator (mean: 1.51). Finally, 88.5% of students totally disagree with the adoption of the strategy of teasing other children in order to stop intimidating themselves (mean: 1.37).
It is therefore self-explained that while the majority of students would not resort to dysfunctional strategies for dealing with school bullying, one out five students may have adopted these behaviors in various cases.

Teachers' perceptions on school bullying
After the implementation of the action, which concerned raising the awareness of students regarding school bullying, a semi structured interview was conducted with the participation of the 10 teachers of the school. The purpose of this question was twofold: on the one hand to determine whether teachers consider that they can manage school bullying and on the other hand in what ways they choose to deal with it. For those cases maybe we will need the help of experts (I8). The category, it is not treatable, because teachers need further training was reported by one teacher with the following words: Because the school is a little mirror of society, school bullying has always existed in the old days and now just now it has come to a point that requires action to be taken, but without the appropriate training of the teaching staff and the local community I do not think that we can except much to happen (I5). One teacher stated: I cannot think of anything right now… (I9).

Discussion
The improvement of the school climate is mentioned as an advantage for the schools that implement intervention programs including all students (Kyriakides & Creemers, 2012;Nickerson et al., 2014;. The same findings are found in the curriculums of the Friendly Schools (Cross et al., 2019) and the Creative a Peaceful School Learning Environment (CAPSLE) Schools (Twemlow et al., 2009). Therefore, in the present research, it is obvious that the holistic actions aimed at reducing the cases of school bullying contributed to the improvement of the school climate. It was previously mentioned that in the 2 nd year of the research and upon the completion of the intervention by the school community, the students were given the same questionnaire as the one they had completed in the 1st year of the research. Statistically significant differences were observed in students' responses between the 1 st and the 2 nd year of the study regarding what they considered occurring in their classroom. In particular, the analysis of the questionnaires reveals a reduction in the noise in the 2 nd year of the survey and this reduction is statistically significant, while the assistance provided by children to their classmates has increased. Therefore, the school climate improved through the action (See Table 1).
During the 2 nd year of the study and upon the completion of the action by the school community, the students stated that they were hit less times and the difference was statistically significantly compared to the 1 st year of the study. This is especially important as it effectively reduces the number of dysfunctional behaviors at break time. Also in the 2 nd year of the intervention, dysfunctional behaviors reported by child witnesses increased. In other words, raising the awareness of the school community through school action contributed to improving children's school day-to-day life to a certain extent. Reduction of dysfunctional behaviors after the implementation of intervention programs is also confirmed in other studies (Olweus, 1993;Twemlow et al., 2009). Upon the completion of the intervention, the children stated that they feel less lonely at home compared to the 1 st year of the study (see table 3). In other words, the parents, through their cooperation with the teachers, improved their habits at home. This fact was also reflected in students' answers in the questionnaire. The importance of school-family collaboration has been extensively highlighted in Eipstein's work (2018,2011).
Finally, regarding the strategies they would adopt to deal with school bullying, in case it happened to them, the students stated the following. They totally agree with the choice of confiding in their parents concerning the incident of bullying or in the principal, or someone they trust or their teachers. They also agree with getting help from their classmates, but also think that their parents would realize it without telling them. Of particular interest is the fact that the majority of students would not adopt any kind of dysfunctional behaviors. In particular, children disagree a lot with the strategy of inaction, in other words, doing nothing until they stop bullying them, choosing to change school, or not being courted during the break. They also disagree a lot with retaliating against those who challenge them, staying at home, giving them money, giving in to their blackmail and starting to hurt other children (victim-perpetrator). However, although the majority of students seem to choose functional coping strategies, one out of five students may adopt one or more of these dysfunctional behaviors.
With the completion of the action against school bullying, teachers were asked whether they believed that school bullying can be tackled with the means available to the school. However, they were not asked to state their opinions on the school's action, due to the relationship developed with the teacher-researcher, who was as an equal member of the educational community. Therefore, it was considered very likely that the feedback would be positive. On the contrary, teachers' personal theory of whether the school really has the means to tackle school bullying in a more general context, seemed to be more accurate concerning the evaluation of the school action. Statistical data of the children's questionnaire also resulted in the evaluation of the school action.
It should be noted that the majority of teachers referred to school-based actions to address school bullying. No mention was made of what is indicated in other articles (Cohen et al., 2009), more specifically that a disorganized school environment contributes to cases of school bullying. In contrast, a school with elements of mental resilience reduces the occurrence of such incidents (Κourkoutas & Xavier, 2017). In conclusion, it is worth noting that the Greek teachers themselves emphasized the importance of extra training in subjects that they consider to be lagging behind such as school bullying.

Conclusion
Recent research has shown that resilience-promoting interventions should address the multiple levels of influence in children's lives, with coordinated policies and programs that target protective factors that span and interact across both home and school environments (Crozier et al., 2010).
In conclusion, it is possible to improve school reality through a number of actions initiated by the educational community. In this case, a school with resilience elements is gradually created to reduce bullying incidents as several researchers argue (Kourkoutas & Xavier, 2017). However, improving school reality also depends on factors beyond the school itself. These include violence, unemployment, social exclusion, poverty, etc. that significantly affect the work of school teachers. In other words, societies with extreme social inequalities are more likely to report more violent incidents of school bullying. In such cases, school teachers should not be considered liable or the scapegoat of society for structural problems that stem from the society itself. However, regardless of the structural difficulties that the educational community needs to overcome it should, to a certain extent, implement actions to improve school life.

Recommendations
Further research should be conducted by school teachers in many educational fields (psychology, ICT, etc.). Through further research the school community will be in the position to enhance its empowerment, utilizing the results as a tool for school improvement. In addition, advice provided to school practitioners includes the following: -Teachers should create a safe environment for all children.
-Teachers must stay alert at all times especially during school breaks.
-The school curriculum should teach respect to differences, assertiveness and self-confidence techniques.
-Teach all children that inactive observers are not innocent.
-Teachers should inform parents for unusual behavior.

Limitations
This study has several limitations: Firstly, it is worth mentioning that the results of the current survey can hardly be generalized as they rather relate to the specific context of the school where the action was taken. Secondly, the main instrument was constructed by the teacher-researcher, following the guidelines of the Hellenic Ministry of Education and Religious Affairs. Therefore, the Cronbach a (0.54), despite moderate, is at the acceptance limit. Thirdly, the fact that the researcher was also a teacher at the specific school may have affected, to a certain limited extent, the opinions of the school community.