Teacher Training for Prevention and Management of School Bullying Situations: A Systematic Literature Review

Abstract This article presents a systematic review of empirical studies on teacher training in school bullying through goal analysis, research design, participants, dependent and independent variables, taught content and skills, training results, and research limitations. Ten databases were consulted, with 12 articles being selected. Although the studies involve teachers as participants, teacher training in bullying has not yet been the subject of study itself, as only four of the analyzed studies evaluated the effect of training in relation to the teacher. The lack of a detailed description of training contents, study procedures and lack of baseline and follow-up were important limitations. Possibilities for future research are discussed.

Violence is a growing public health problem worldwide (Krug et al., 2002).Each year, more than one million people lose their lives, and just as many suffer nonfatal injuries resulting from self-harm, interpersonal aggression, or collective violence (World Health Organization;WHO, 2014).One of the most visible forms of violence is perpetrated by young people between the ages of 10 and 21 (youth violence) and, in general, the assaults are witnessed within the school institution (Lopes, 2005).The term school violence refers to all aggressive and antisocial behaviors, including damage to property, criminal acts, interpersonal conflicts, aggression between students and teachers, institutionalized practices of rights violation, and aggression between peers, better known as bullying (Lopes, 2005;Stelko-Pereira & Williams, 2010).
The definition of bullying was developed in the 1970s by Dan Olweus as behaviors involving intentional aggression, repeated over time, and involving an imbalance of power between the aggressor, perpetrator, and victim (Olweus, 1997).Bullying can be classified as direct, which involves physical and/or verbal aggression, and indirect, which presents itself through social exclusion of peers, slander, defamation, etc. (Fekkes et al., 2005).With the advent of

METHOD
A search for scientific articles was conducted in 10 databases: PsicINFO, PsicArticles, Scielo, Science Direct, ERIC, Wiley Online Library, Scopus, Lilacs, Springer, Web of Science, published until March 2022.
In the search fields, descriptors were entered, from the areas of Psychology, Education and Health, accompanied by the Boolean operators OR and AND: bullying OR bully AND "teacher training"; "bullying prevention" AND "teacher training" and "bullying prevention" and their corresponding terms in Portuguese: bullying, combined with the words: capacitação de professores; treinamento de professores; capacitação docente; treinamento docente (meaning teacher training); prevenção ao bullying (meaning bullying prevention).The descriptors were taken from the Thesaurus of Psychological Index Terms (APA), BVS Thesaurus Psychology, Thesaurus HSDs (health science descriptors) and Brazilian Thesaurus of Education (Brased) and from terms most commonly used in the literature to refer to bullying and teacher training programs.
The criteria for the selection of articles were: empirical studies related to teacher training/teacher education and training of other school community members that were peerreviewed.Literature review articles, empirical studies whose intervention was exclusively with children/adolescents and/ or their parents, and duplicate articles were excluded.Most of the excluded articles were characterized as theoretical studies, research studies on participants' perceptions of school bullying, or had children, adolescents, or parents as research participants.
A total of 2,118 articles were retrieved, of which 168 were removed as they were duplicates.The title and abstract of 1,106 articles were read, and 12 articles that met the inclusion and exclusion criteria previously defined for this systematic review were selected.These were read in full, extracting information such as: authors and year of publication of the study, participants, objectives, identification and form of evaluation of the dependent variable, design, training content, skills/behaviors taught, and limitations, including the results.Figure 1 schematically presents the path taken in the survey of articles identified in the research period.
Three categories were developed for data analysis: 1) Studies characteristics: such as authors/year of study publication, participants, study objective, dependent variable, dependent variable assessment, design; 2) Contents and skills/ behaviors taught in the training programs conducted with teachers as participants; 3) Studies limitations: identified by the authors of this article and/or by the studies' authors.
In the searched databases, there is lack of research carried out by Brazilians related to teacher training focused on knowledge and behavior development for the prevention and/or management of situations involving school bullying behaviors.Recent research on teachers' knowledge and intervention strategies adopted to deal with this type of violence has shown that educators have insufficient knowledge on the main characteristics of bullying, an inadequate behavioral repertoire on how to deal with this phenomenon, and most of them report to feel unprepared to identify and deal with aggression by students (Santos et al., 2015;Silva & Rosa, 2013;Silva et al., 2014).
The need for training programs became more evident in 2015 when Law No. 13,185 (2015) came into force establishing the program to Combat Systematic Bullying and one of its objectives provides for training teachers and educational teams to implement actions to discuss, prevent, guide and solve the problem.
Table 1 presents the authors, publication date, objective, and the studies' subjects.
Research data (Bauman et al., 2008;Bell & Wills, 2016;Burger et al., 2015) has highlighted that teachers still have many questions about bullying and the appropriate intervention strategies to effectively deal with this phenomenon.Misconceptions about knowledge and intervention strategies used can inhibit efforts related to prevention and management of the phenomenon (Hazler et al., 2001).
It was observed that in trainings on the subject of bullying, the teachers' behavior has not yet been considered an object of study in itself, for of the four studies that evaluated the effect of teacher training, only two investigated teacher ability/behavior (Goncy et al., 2015;Newman-Carlson & Horne, 2004).This highlights that, there is lack of training development for teachers to teach skills/behaviors to directly intervene on the phenomenon through prevention or management of situations involving bullying behaviors, even in studies conducted in other countries.Psic.: Teor.e Pesq., Brasília, 2023, v. 39, e39310 Teacher Training and Bullying The study by Burger et al. (2015) corroborates this observation by examining what intervention strategies teachers used in a hypothetical episode of school bullying.The results showed that the teachers' preferred strategies were authority-based interventions, followed by non-punitive work with the bully and the involvement of another adult.Also, the participants indicated that they were less likely to work with the victims of bullying demonstrating certain inability to deal with the phenomenon.The authors discuss the need to prioritize teacher training at universities as there is no systematic education on how to prevent and manage bullying incidents in most curricula, and for those already in school, that there should be intervention to increase self-efficacy and skills in alternative methods of dealing with bullying cases.
In addition to the need for further studies that have teacher knowledge/behavior as the dependent variable, it is also important to provide conditions for them to be effectively evaluated.In the retrieved studies, the dependent variable "Teacher knowledge on school bullying" was assessed by questionnaire (8%; n = 1) (Benitez et al., 2009), scale (8%; n = 1) (Newman-Carlson & Horne, 2004), and teacher adherence and instructional competence by direct observation (8%; n = 1) (Goncy et al., 2015).However, there were some limitations to the use of these measures.For example, although the study by Goncy et al. (2015) proposed to conduct constant measures regarding teacher behavior, an important condition to increase control over the experimental variables and, consequently, the internal validation of the study, it does not describe the use of a pretest.Using such a measure to check teachers' skills prior to training would contribute to increasing the internal validity of the overall study (Ttofi & Farington, 2012).The limitations of each study will be fully discussed below.
Table 2 presents the dependent variable, the assessment of the dependent variable performed, and the design employed in each study.It is observed that the studies presented a diversity in the evaluation form of the dependent variable as well as the use of a variety of designs.The majority of the designs employed in the research were Experimental Group and Control Group Design (50%; n = 6) (Boulton, 2014;Buils et al., 2020;Goncy et al., 2015;Ju et al., 2009;Leadbeater et al., 2016;Verseveld et al., 2021) and Quasi-experimental Design with pre-test and post-test (33%; n = 4) (Benitez et al., 2009;Mendes, 2011;Newman-Carlson & Horne, 2004;O'Moore & Minton, 2005).In a systematic review (Cantone et al., 2015) of international research that analyzed the outcomes of interventions to reduce or prevent bullying and cyberbullying, a wide variety of experimental design and a lack of common outcome measures were identified, indicating that there is no systematic approach to data collection in this area of research.
Despite the variety of designs and the form of results evaluation identified in the studies, it was observed that there was no use of a single-subject design with teachers as participants.Experiments with single-subject designs analyze data from individual behavior, that is, even though several subjects have been subjected to the same conditions, the results are treated individually (Sidman, 1960(Sidman, /1972)).Participants are exposed to a variety of conditions, their performance is repeatedly measured, and an orderly relationship between the conditions manipulated in the experiment and changes in these measures is verified (Matos, 1990).
The subject is said to be his or her own control because he or she is subjected to the experimental conditions and multiple behavioral measures are performed for the same subject, rather than statistical treatment of data, where the group average is verified (Andrey, 2010;Sidman, 1960Sidman, /1972)).Psic.: Teor.e Pesq., Brasília, 2023, v. 39, e39310 Teacher Training and Bullying One of the great advantages of single subject research is the employment of a rigorous, scientific methodology used to define basic behavior principles and establish evidence-based practices by documenting causal or functional relationships between independent and dependent variables (Horner et al., 2005).For future studies with the interest of evaluating the effect of training on teacher behavior, this design may be an interesting variable to more systematically measure the results.

Content and Skills/Behaviors Taught in the Training Programs Conducted with Teachers as Participants
The content delivered in the trainings of the reviewed studies were categorized into: information/facts about the phenomenon of bullying (contextualization about violence, identification of bullying, definition and characteristics, assessment of the phenomenon, research findings, feedback of the occurrence of bullying in the teachers' classroom) (50%; n = 5); content about intervention to the bullying phenomenon (40%, N=4); content on bullying prevention (30%; n = 3); knowledge about research (action research) (10%; n = 1); one of the studies did not present the content of training (10%; n = 1); positive reinforcement; social and material reinforcement; modeling; extinction and differential reinforcement of the students' target behavior; self-control techniques; role-playing and the directed game (10%; n = 1).
In Boulton (2014)'s study, although the author describes that in the training teachers were taught how to prepare students at risk of bullying or how to teach the perpetrators of bullying to build more effective alternatives for dealing with situations, the author does not describe the behaviors/ skills, besides describing in a few details the content taught, citing only the basic steps addressed by the I Decide Program, adopted in the training: thoughts, triggers, feelings, behaviors, and consequences.
The training of Ju et al. (2009) addressed knowledge about research procedure and methodology; knowledge about school bullying; action research; intervention skills including brainstorming, quality circle, self-confidence training and roleplay, intervention program with the bullying victim and perpetrator, and the implementation of the intervention program conducted from the action research model: planning-action-observation-reflection.Although the authors mentioned that the training presented themes related to school bullying, this content was not described operationally by them.Benitez et al. (2009) developed the content contextualization on school violence and introduction to the bullying phenomenon: characteristics and definition of the problem, etiological factors, analysis of the agents involved, effects of bullying, phenomenon evaluation, knowledge and practice for intervention in bullying prevention.Ross and Horne (2009) did not describe the content presented in teacher training.The study only shows that the training involved the implementation of the BP-PBS program in 2 steps: 1) training involving the whole school: institutional, administrative staff and supervisors and 2) use of BP-PBS in the curriculum, by the staff, school, to train students.School staff received a one-hour workshop using the BP-PBS program manual.The article only indicated the website where this program would be available, but did not present which skills and/or contents were taught to teachers.
The study by Leadbeater et al. (2016) involved students and teachers as participants and data collection was carried out in five phases.The one that involved teacher training was available online so that they could access the training.The research teams followed the online training of teachers, in groups of 20 to 40 people, as well as the fidelity of implementation.The article cites the main contents of the WITS program: resolving peer conflicts, increasing social responsibility (helping and caring) and promoting the use of teaching strategies in times of conflict.Professionals completed the training module questions individually.Each month, teachers were asked to select a book from a given list to use in their lesson plan by creating conversations to convey the messages of the WITS program.
O'Moore and Minton ( 2005) implemented an intervention program consisting of four key elements: training of trainers who would work with the school community, training with teachers, meeting with parents, and informational brochures and manuals to students and parents.In one day of training developed with teachers, the content of the program involved: information about bullying behavior with emphasis on classroom management, the development of a positive atmosphere in the classroom and at school, leadership of people, and cooperation between parents and teachers.Salmivalli et al. (2005)  Mendes' ( 2011) training for teachers addressed strategies and techniques for promoting social skills such as self-control, interpersonal relationships, and response repertoire to deal with bullying situations, aimed at reducing/ MG Panosso, N Kienen, & RF Brino preventing violence, to be used in the classroom with students.In their interventions with students, teachers used positive reinforcement, social and material reinforcement, modeling, extinction and differential reinforcement of the target behavior, self-control techniques, role-playing, and the directed game, teaching processes and strategies that they most likely learned during the training.
In the aforementioned studies, it is possible to identify that they focused on teaching teachers only content related to the issue of bullying at school, to the detriment of the concomitant teaching of important behaviors to deal with the social reality presented in schools.This still seems to be a limitation of the studies, especially when considering that the skills to deal with bullying situations involve much more than just "learning theoretically" about this phenomenon.From a behavior analytic perspective, for example, it is possible to define teaching as "an arrangement of contingencies under which students learn" (Skinner, 1968(Skinner, /1972, p. 62), p. 62), with learning understood as a change in behavior resulting from this arrangement (Kubo & Botomé, 2001;Skinner, 1968Skinner, /1972)).The contingencies are arranged by a trainer/ teacher in order to make the learner able not only to "know a knowledge" but to "know how to deal with the world in light of the knowledge" (Kienen, 2008;Kubo & Botomé, 2001).Teaching, therefore, is not only about "transmitting content" but about developing behaviors that enable the learner to deal with aspects of their professional context (Bordignon & Botomé, 2017;Kubo & Botomé, 2001).
Evidence suggests that universal, that is, school-wide interventions are the most effective means of reducing bullying behavior (Cross et al., 2011), and one factor determining the effectiveness of anti-bullying programs is the quality of training for school staff.However, research on training teachers and school staff on prevention and intervention in school bullying situations is still limited (Gorsek & Cunningham, 2014).
Although all of the studies presented teacher training in their descriptions, only two of them presented and described the skills/behaviors taught to these participants, in addition to content.Table 3 presents the skills/behaviors taught.
In the study by Goncy et al. (2015) teachers received training conducted by the School's Bullying Prevention Coordinating Committee (BPCC) and an Olweus certified trainer.The meetings were characterized by teaching content such as identification of bullying and skills such as anger management, respect, stress, and problem solving through role-play, small group activities, discussions, information presentation, and then assessed teachers' abilities to provide students with adequate instruction on how to handle conflict situations.The Psychoeducational intervention program Bully Busters: A Teacher's Manual for Helping bullies, victims and bystanders was used, in seven modules, which aimed to teach skills such as relaxation, coping and collaborative problem solving in addition to information on bullying and victimization, intervention strategies, stress management techniques and classroom activities.

Study
Skills/behaviors taught 1 The author only mentions that the workshop shows teachers how to prepare students at risk of bullying or how to teach the bullying perpetrators to build more effective alternatives for dealing with situations, but does not describe how it has done.
2 Intervention skills including brainstorming, quality circle, self-confidence training, and roleplay (as described by the authors).Identify, prevent and reduce bullying appropriately, model pro-social and positive behavior, and create and maintain anti-bullying policies.
Table 3 Description of the skills/behaviors taught.
Psic.: Teor.e Pesq., Brasília, 2023, v. 39, e39310 Teacher Training and Bullying In the study by Verseveld et al. (2021), training to teachers was aimed at teaching different skills that would enable them to identify, prevent, and reduce bullying.The e-learning was developed to increase teachers' and staff's awareness and their ability to respond to bullying; encourage them to model pro-social and positive behavior; help them actively create and maintain anti-bullying norms; and support them in implementing the program.
In the aforementioned studies, there was the teaching of skills/behaviors.Behavior is understood as the relationship between the subject's actions and aspects of the physical and social environment, which concern both what precedes his action and what results from it (Kubo & Botomé, 2001).The Experimental Behavior Analysis has teaching technology capable of effectively contributing to the teacher's education based on knowledge about behavior (Skinner, 1968(Skinner, /1972)).One of these technologies is the Programming of Conditions for Behavior Development (PCDC) or Teaching Programming (Kienen et al., 2013), which proposes the teaching of behaviors that are relevant to the learner and the environment of which he/she is a part (Kienen, 2008;Kubo & Botomé, 2001), contributing to the development of behaviors that will allow him/her to solve problems that he/she faces in his/her daily life.
Such training initially proposes to characterize the social needs of the target audience, in this case teachers, and to survey the possibilities of action derived from these needs (Kienen et al., 2013).In order to teach the relevant behaviors, the programming of the teaching conditions will take into account: (1) the problem situations that the learners will have to deal with in their natural environment, (2) the outcomes that will be produced by the learners' actions, both during the learning process (training) and in the natural environment (outside the classroom) and, finally, (3) the responses that will have to be emitted by them in order to produce such outcomes (Cortegoso & Coser, 2011).
In order to do so, it is necessary to identify the repertoire of entry of these learners, i.e., which behaviors they are not presenting on a daily basis, but should, and those that are presented, but should not (Cortegoso & Coser, 2011).From this, the trainer will be able to decide where to start teaching and which relevant behaviors should be developed for the teacher's performance in the school context.
In teaching by behavior development, in addition to the dimensions of "time" and "content", the behaviors that the learner should be able to present in their reality are considered (Botomé & Kubo, 2002).Therefore, unlike just presenting information about the bullying phenomenon, its characteristics or even content on prevention and management, as proposed by most of the studies reviewed in this article, training developed through PCDC may be an efficient and effective alternative (Carvalho, 2015;Gonçalves, 2015;Luca, 2013;Teixeira, 2010) for teaching relevant behaviors to teachers so that they can act in bullying situations that arise in the school environment.

Study Limitations
The study by Boulton (2014) presented limitations such as lack of detailed description of the procedure, making it difficult to replicate the study to future interested parties, as well as providing only the general topics addressed by the workshop developed with the teachers.
In the studies by Buils et al. (2020), Ju et al. (2009), and Verseveld et al. (2021), the measurement of the dependent variable was conducted only in relation to student behavior, even though they developed a training with teachers as participants.The study did not make clear how the intervention with teachers contributed to the results.The description of the study was very general about how the teachers actually participated in the intervention.Furthermore, there was no follow up to verify if the effect of the intervention was maintained over time.Studies such as those of Benitez et al. (2009) and Salmivalli et al. (2005) described the procedure without detail making replication difficult for future interested parties.Like the study by Ju et al. (2009) and Newman-Carlson and Horne (2004), there was no follow up to see if the effect of the intervention was maintained over time.
Regarding the evaluation of the intervention with teachers, in the study by Goncy et al. (2015), the authors indicated that too many observers in a small room may have influenced the behavior of the participants.Also, there was no baseline to verify how teachers carried out instruction to students before the intervention and, as the authors themselves indicate, it would be necessary to reevaluate the reliability and validity of the measures.
In Ross and Horner's (2009) study there was no evaluation of teachers' behavior before and after training, only from the students.The research by Leadbeater et al. (2016) was conducted in rural schools which limited the generalizability of the results, as indicated by the authors.Also, the study was a long-term study and, during the course of the study, there were no constant measures in order to minimize the possibility of extraneous variables' influence.Although the study by O' Moore and Minton (2005) presented a training program with interesting themes in the theoretical and practical spheres with the professionals responsible for training teachers, with enough time to cover issues of extreme relevance for those who live in environments where bullying situations recur, the training of teachers covered few contents and in a period of few hours.Finally, in the study by Mendes (2011) there was no evaluation of the behavior of teachers before and after the training, only of students.There was no detailed description of the contents presented to the teachers in the training sessions, nor of the procedure, making replication difficult.No follow up was performed.
Regarding the results, only 33% of the studies (Benitez et al., 2009;Boulton, 2014;Goncy et al., 2015;Newman-Carlson & Horne, 2004) presented the evaluation of knowledge, perception and/or behaviors of teachers after the training provided to them.It is observed that in all studies there was teacher training.However, in only 25% (Goncy et al., 2015;Newman-Carlson & Horne, 2004) the behaviors/skills taught were assessed and presented in the results.It is likely that the conditions arranged to assess the dependent variable interfered with the imprecise and detailed description of the study results that assessed teachers' knowledge and skills.
Table 4 shows the studies that evaluated and presented the results of teacher knowledge, perception and/or behavior.
In addition to the lack of description of skills/behaviors, another limitation presented by the studies was lack of detailed description of the content presented to teachers in the training sessions.In addition, in 83% of the studies, although there was teacher training in their interventions, there was no assessment of the participants' knowledge.Therefore, there is a difficulty in identifying whether, in fact, the contents addressed in the trainings, even if presented in a general way, contributed to the teachers' knowledge.
The few studies that evaluated teacher knowledge (Benitez et al., 2009;Newman-Carlson & Horne, 2004) did so with the use of questionnaires through which improvements in teacher knowledge on the subject could be verified.However, the lack of follow up did not allow us to identify whether the results of the intervention were maintained over time.Finally, one of the studies (Goncy et al., 2015) that had teacher behavior as the dependent variable, did not conduct baseline (pre-test) in order to verify teachers' skills (how they issued instructions) before training.When there are no measurements before the intervention the internal validity of the study is further compromised (Ttofi & Farington, 2012).Furthermore, when proposing a training, identifying and describing the learners' input repertoire is of fundamental importance for the trainer to be clear about what should be taught and where to begin teaching their program (Cortegoso & Coser, 2011).
In short, the most frequent limitations presented by the studies were: lack of description and assessment of the skills/ behaviors taught to the teachers in the trainings (83%, n = 10); lack of detailed description of the content presented to the teachers in the trainings (75%, n = 9); lack of follow up (75%, n = 9); and lack of detailed description of the procedure, which hinders the possibilities of replication (50%, n = 6).Furthermore, similar to the findings of Cantone et al. (2015) in a systematic review of studies that described interventions on bullying and cyberbullying in schools, it was noted in the present review that in the studies that conducted teacher training, there was a wide variability of experimental designs and lack of common standardized measures in outcome evaluation, which characterizes important limitations in this research field.
Therefore, it is suggested that future studies that aim to enable teachers to intervene in the school bullying phenomenon can include measures such as pre-test, posttest and follow-up in their research, in addition to constant measures of teacher behavior throughout the experiment, thus allowing a greater control of variables, increase of the internal and external validity of the study and, consequently, a more adequate evaluation of the intervention's effectiveness.
Also, developing training for teachers based on PCDC, which proposes from the elaboration of teaching objectives relevant to the learner and his social context, to the programming, application, evaluation and improvement of both learning processes and teaching procedures (Kienen et al., 2013), can be efficient and effective alternatives given the scientific evidence of training programs developed based on this area of study (Carvalho, 2015;Gonçalves, 2015;Luca, 2013;Teixeira, 2010).

Study Results
Studies that presented as results the evaluation of teachers' knowledge, perception and/or behaviors 1 In the pre-test, ¾ of the teachers considered the cognitive behavioral approach effective in combating bullying although only 54.2% of those who had participated in the I Decide Program at another moment reported using it rarely, and 26.2% reported using it frequently in their practice.Post-test: 52.4% of teachers who participated in the program indicated that they used the approach frequently.There was significant correlation between reported use and duration of training, but not in perceived effectiveness.The association between training duration and reported use of the approach was mediated by self-efficacy.

3
The EG members modified and improved their knowledge and perception about bullying (defining the phenomenon: 41.5% of the EG defined the phenomenon by listing at least three characteristics when compared to 1.2 of the CG).Regarding the identification of the phenomenon's incidence and the aggressors' and victims' personal characteristics, and regarding the selfefficacy in treating bullying, there were statistically significant differences in the EG and none in the CG.

4
Teacher's behaviors such as: instructional adherence, procedural adherence, instructional competence, and procedural competence were positively and significantly correlated with each other, with student engagement and in following rules.

9
The program increased teachers' knowledge and use of intervention skills, their personal self-efficacy and self-efficacy related to children with specific characteristics, and, the reduction of classroom bullying as measured by disciplinary references.

Table 4
Results about teachers' knowledge, perception and/or behaviors.
Psic.: Teor.e Pesq., Brasília, 2023, v. 39, e39310 Teacher Training and Bullying This study showed some limitations, such as the reduced number of selected and reviewed studies and their methodological diversity, which prevented the realization of a meta-analysis.Still, it is possible that, when carrying out searches in other databases, different results are found, recovering studies that were not included in this review.Despite these limitations, it is noteworthy that the present study seems to be the first systematic review, to the authors' knowledge, about teacher training on the subject of school bullying.

FINAL CONSIDERATIONS
Systematic literature reviews, such as this one, are fundamental to scientific research as they allow the identification of the knowledge produced in a particular field of study, identify relevant variables to be investigated, and possibilities for future research (Randolph, 2009).
The selection and characterization of the studies retrieved made it possible to identify that teacher training on the school bullying theme has not yet been considered an object of study itself, especially in Brazil.The methodology employed by the few studies that evaluated teachers' behavioral repertoire makes it difficult to identify whether the research participants actually developed a repertoire that the training proposed to teach.Furthermore, it was possible to identify that the focus of the training courses is more on providing theoretical knowledge to teachers than on promoting the development of behaviors that make them more capable of dealing with school bullying.Besides a change in perception, it is necessary that teachers are taught how to behave and use appropriate strategies to manage this problem.
One of the main determinant factors in the effectiveness of an anti-bullying program is the quality of staff training (Lund et al., 2012).However, researches on teacher training focused on school bullying prevention and intervention are still limited (Gorsek & Cummingam, 2014).It is suggested that future research that has as an independent variable training for teachers proposes to teach content and behaviors for the prevention and management of bullying situations in order to prepare them to deal with day-to-day situations that arise in the school environment.Furthermore, it is suggested to evaluate the participants' behavior with pre-test, post-test, follow-up and constant measures throughout the training, in addition to the use of a single-subject design.
This study made it possible to identify what has been produced about teacher training on school bullying, nationally and internationally, in terms of variables researched in such studies, identification of gaps present in these studies, and possibilities of improvement for future studies, with the intent of contributing to the advancement of scientific research, either by conducting systematic replications, or even, arranging contingencies for planning and researching new variables.

Figure 1 .
Figure 1.Articles retrieved and selected for analysis.

Table 2
Dependent variable evaluated and design applied in the studies.
1 Teachers' beliefs and perceptions about the effectiveness of a program based on the cognitive-behavioral approach (I decide Program) and its selfeffectiveness.Scale developed by the researchers to assess the effectiveness and perception of the cognitive-behavioral approach; self-efficacy; about the use of the I Decide program and, content analysis of the open-ended questions.Experimental group design.Experimental groups: 35 participantspart-time workshop; 31 participants: 1-day workshop; 30 participants: 2-day workshop; 28 participants: 3-day workshop.There was pre-test, post-test and follow up 2 Student's experience with bullying/ victimization on the way from school back home and in the school environment.The Anonymous Bullying/Victimization Questionnaire modified.Cheklist to assess children's experience with bullying and victimization, based on "My School Life" checklist.Group experimental design and, for the experimental group, action research (whose objective is to change the target behavior in a real situation).There was pre-test, posttest.There was no follow up 3 Teachers' knowledge about bullying in relation to the definition and incidence of the phenomenon; personal and family characteristics of bullies and victims; and teachers' perceptions of their ability to act in bullying situations.School Bullying Questionnaire adapted.Quasi-experimental.Pre-test and posttest with two non-equivalent groups (Experimental Group, n = 106 and Control Group, n = 93).There was pre-There was pre-test, post-test, and follow up.