Factors associated with interpersonal violence among children from public schools in Fortaleza , Ceará , Brazil

The school has witnessed the growth of manifestations of interpersonal violence among children, and in this perspective, this paper examines the prevalence of interpersonal violence in the relationship between students and the associated factors. This is a cross-sectional study developed with 874 public school students aged 10 and 11 years conducted in 2013. A questionnaire implemented the collection of data that were submitted to bivariate analysis with statistical significance calculation between associations. The results show that the child-perpetrator (83.2%) is associated with the male gender (PR=1.08), as well as being beaten at home (PR=1.13) and having a family that encourages retaliation (PR=1.17). The child-victim (89.5%) is associated with the family that encourages retaliation (PR=1.05), participation in conflicts (93.6%) and age. The 10-year-old child is up to 3.0% more likely to participate in conflicts. The “family that encourages retaliation” is positively associated with the situations of assault, victim condition and participation in conflicts, which entails the reframing of these parental practices.


Introduction
Analyzing the social place of the child in the family and at school contextualizes the practices inscribed in the scenarios that consider the historical, social and political aspects.In this logic, the school is a privileged social space where children can decide, show individual and autonomous attitudes in the face of conflicts, show divergent ideas among their peers and express their viewpoint, pondering on what has been said, taught or experienced.
The school has been institutionally and socially conceived and built to receive and contribute to the education of children, building on the interactions 1 among the leading roles and structuring in these relationships different forms and representational contents, showing the societal culture in which they are inserted 2 .However, the school has been shaped as an arena and locus of exponential growth of relationships eroded by behavioral incivilities and shattered values, understood in this study as the manifestations of interpersonal violence.
In the school context, the phenomenon leads to concerns and reflections that extrapolate the educational context and enter the field of collective health, in continuous construction and interdisciplinary character [2][3][4][5] .Despite its growing visibility, there is no consensus in the conception of scholars about theoretical constructs and operational concepts 6 , except that children and young people are involved as perpetrators or victims 1,7 .
In this polarity, it is understood that these institutions are influenced by and influence the child's historical-social context 8,9 , in which violent behavior results from the interaction of individual development with the triad family, school and community 1,7,10 .Considering this meaning, the fact that the family is the one to be blamed the most for the behavioral changes in children 11,12 has been recursive in the literature, which leads us to suppose that "showing" affectivity to the detriment of conflicting exacerbations in the family mirrors the construction of respectful, supportive and mediating relationships.
Thus, the set of social and cultural characteristics of the households should be considered, paying attention to the parental education models [13][14][15][16][17] .Reinforcing this assertion, individual and family factors may be at the root of school violent behavior 14,18 .
The Brazilian production on school violence 12,[19][20][21] and studies on interpersonal violence in children is worth mentioning, adding that studies on children interpersonal violence, especially that of a population character 22 still has discrepancies that may make debates on the subject more complicated.
Despite the increased knowledge about the issue, the unfolding and singularities that emanate from it require continuous investigations so that it can be understood and contribute to the elaboration of coping strategies, with the expanded knowledge that is found in the literature, marking the peculiarities of the field where the investigations are inscribed.
Thus, the decision by the municipality of Fortaleza as a field of analysis reiterates the importance of increasing investigations on violence due to the increased rates of this phenomenon [23][24][25] in this capital, which has also been widely publicized in the media network.
By approaching the school-family binomial and seizing it as an essential context in the behavioral construction of children, the research analyzes the prevalence of interpersonal violence in the relationship between schoolchildren and the associated factors.

Methods
This cross-sectional study is a selection from the dissertation "Modalities of child violence in the school context" 26 developed in the city of Fortaleza, State of Ceará.This municipality is divided into six Regional Coordination Offices (CORES) that are responsible for the management of the districts.This political-administrative division aims to organize the management and access of the population to the services provided by the Municipality 27 .
In this research, the selected area of coverage was CORES V, consisting of 18 districts with high levels of violence, high demographic density (the most populous area of the capital), mean income of one minimum wage and a young population (44%) 28,29 .The set of vulnerabilities justifies the chosen field of analysis (CORES V) because the literature indicates that vulnerable populations are more exposed to urban violence, which in turn can influence school violence 30 .
In the selection of the participants, the study requirements were: children aged 10 and 11 as per the age group established by the Statute of the Child and Adolescent 31 , attending the 5 th or 6 th grade during the first semester of 2013, meeting the age-for-school year adequacy crite-ria.This age group was chosen because, at this age, children are in the transition phase (childhood-adolescence) and seek independence in their attitudes, towards becoming active subjects in the construction of their social world 3,32 .
In response to these requirements, 5,264 students were identified, of which 539 from 51 CORES V schools were randomly sampled.Four schools that did not provide the selected school years in their didactic-pedagogical structure were excluded.
The collection included the following steps: visits to schools with prior scheduling; meeting with the manager to explain the objectives and provision of the Term of Assent for the children and Informed Consent Form for those responsible.Before completing the questionnaire, the Term of Assent was explained and accepted by the students; then the questions were read by one of the authors and the students answered the instrument in writing.In this logic, 1,203 questionnaires were collected so as not to exclude those who were in school discrepancy situation during the classroom collection.Thus, 874 questionnaires were eligible according to the participation criteria.
This questionnaire was adapted from the instrument validated by Orpinas 33 and used by Santos 34 .We refined the analysis process by selecting questions (among the blocks) that resulted in the dependent (children who were perpetrators, victims and participants in conflicts) and independent variables (school year, age, gender, family structure, number of brothers in the household, fear of being beaten at school, being beaten at home and family encouraging retaliation).
The variable child-perpetrator resulted from the affirmative answers to the questions: did you make fun of other colleagues to the point of irritating them?Did you strike back someone who hit you first?Did you say things about another person to make your classmates laugh?Did you encourage colleagues to fight?Have you ever pushed your colleagues?Also, did you curse other colleagues?
The variable child-victim resulted from the affirmative answers the questions: Did any colleague pick on you to make others laugh?Did other colleagues encourage you to fight?Has a colleague ever pushed you?Did some colleague call you to fight?Has a colleague cursed you or your family?Has a colleague threatened to hurt or beat you?
Moreover, the variable participant in conflicts resulted from the affirmative answers to any question of the inquiries that originated the dependent variables child-perpetrator or child-victim.
The family structure was recoded in nuclear (father, mother, and children) and non-nuclear family (mother/children, father/children, father/ mother/children/other relatives); fear of being beaten at school in the dichotomous variable in which the answers "rarely", "sometimes" and "always" originated the "yes"; and the answer "never", the "no".
For the variable being beaten at home, the answer "never" determined the item "no"; the item "yes" considered the statements for any of the answers: "rarely"; "once a month"; "once a week" or "almost daily".
The family that encourages retaliation encompassed the affirmative answers to the questions: "when someone hits you, hit him/her back"; "when someone curses you, hit him/her"; "when someone curses you, curse him/her back"; "when you cannot solve the problem by talking, you better solve the problem by fighting"; and, negative for the questions: "when someone curses you, don't be bothered"; "when someone calls you to fight, try to talk him/her out of it"; "when another colleague asks you to fight, you should talk to the teacher"; "no matter what happens, fighting is not good and there are other ways to solve the problem".
The data was organized in Excel v.7 and exported to STATA v.8 software (Stata Corp College Station, Texas), and univariate and bivariate frequency analyses were performed.The bivariate analysis was based on the cross-tabulation between dependent and independent variables with the calculation of statistical significance between the associations, with the use of Pearson's Chi-Square test.All analyses had a significance level of 5% (p ≤ 0.05) and a 95% confidence interval.The Research Ethics Committee of the University of Fortaleza (UNIFOR) approved the study.

Results
In total, 874 primary school 5 th graders (60.6%) and 6 th graders (39.4%) participated in the study.Table 1 describes school and sociodemographic data, school and family relationship related to interpersonal violence among children in the 5 th -6 th primary school years of public schools, in the city of Fortaleza, Ceará.
Regarding the occurrence of interpersonal violence, the following scenario was observed: stated that they were afraid of being caught in school (48.9%); assured that they were beaten at home (65.3%) and families encouraging retaliation (58.7%).Regarding the polarity of self-perception as perpetrator-victim, the study showed that 83.2% of the children assumed the role of perpetrators; the number of victims prevailed in 89.5%, and 93.6% of students had participated in conflicts during the last seven days before the survey.
Table 2 shows the association between school year, demographic data, school and family dynamics related to interpersonal violence with the child-perpetrator variable.Associations were found between gender, being beaten at home and family encouraging retaliation, with p-value < 0.01.It is observed that being male increases [PR (95% CI) = 1.08 (1.02-1.15)]by up to 8.0% the probability of the child becoming the perpetrator, as well as being beaten at home and having a family encouraging retaliation increases [PR (95% CI) = 1.13 (1.05-1.21)] up to 13.0% and [PR (95% CI) = 1.17 (1.10-1.25)]17.0%, respectively, the probability of the child being the perpetrator.
Table 3 shows the results of the association between the variables studied and the child-victim, with a significant difference between the variable family that encourages retaliation (p=0.030),resulting in an increase [PR (95% CI) = 1.05 (1.00-1.10)] of up to 5.0% the probability compared to children who do not have families encouraging them to retaliate.
Table 4 shows the relationship between the factors associated with participation in conflicts.A statistically significant difference in the independent variables was found: age (p=0.026),family that encourages retaliation (p=0.005).Regarding gender (p=0.063) and school year (p=0.088),borderline significance was identified, as attested by the respective prevalence ratios -[PR (95% CI) = 0.96 (0.93-0.

Discussion
We can perceive that the theme of interpersonal violence in schools, choosing the fine line between being a child and becoming an adolescent, requires investigations that point to theoretical constructs that can establish new practices, whether in the dynamics of families or the reshaping that encompasses the educational system.
This research establishes the predominance of females among the participating schools, which leads us to believe that it is related to the proper age-school year criterion since the school census of the city of Fortaleza confirms this profile.However, the literature reiterates a higher proportion of females in classrooms [35][36][37] .
Regarding the family structure, the nuclear family prevailed in the responses of the participants, according to the last national census, in which this conformation represented 49.4% of Brazilian households in 2010 38 .
The significant percentage of respondents who assumed the roles of perpetrators, victims or participated in conflicts in school settings causes a stir.Possibly, in this study, these results reflect the non-requirement of the repetition criteria and regular periodicity of some situation of violence so that they can call themselves perpetrators or victims.
In this logic of reasoning, anchored in the discussions that confront the distortion of educational action -diverted to a context that en-ables the exercise of power, reinforcement of fear and submission between peers -, the results require critical reflections on these frequent manifestations of violent attitudes between the participants in the 47 schools investigated.It is assumed that we are also witnessing the naturalization of these manifestations before the resolution of conflicts in the daily life of social settings, whether within families, schools, and regardless of the subjects involved.
Regarding the complexity of this phenomenon, Goergen 39 argues that although we recognize the importance of the relationship between ethics/morals and education, both in families, in social institutions, in the media and also in the school itself, the ethical "lens" evidences despise rather than esteem.The author reiterates 39 that the multicultural society, strengthened by globalization and social mobility, in which they share a space of multiple views of life and world, has further aggravated this bewildering in education and school (...) with so many disparities and relativism.Schools that must serve and respect everyone face a difficult challenge.
According to a survey conducted by Plan Brasil 12 with 5,168 students in five public and private schools in the country's regions, 70.0% of the students said they had witnessed aggressive scenes among their colleagues, while 30.0% said they had lived at least one violent situation in the year before the survey.
The Latin American Faculty of Social Sciences (FLACSO), in partnership with the Ministry of Education (MEC) and the Organization of Inter-American States (OEI), in 2016, from the Brazilian school census and analyzing 6,709 6 th graders to the last year of secondary school students pointed out that 42.0% suffered some violence at school.When considering the city of Fortaleza 37 , the survey puts it first with 67.0%, approaching the findings of this study, in which 89.5% predominated.This scenario of spiraling peer violence follows other locations.A study conducted in Esteio (RS) with the participation of 161 5 th -8 th graders, also adapted from the Orpinas questionnaire 33 , showed that 56.9% of the students self-referred as victims and 38.5% as perpetrators 40 .In Brasília, Federal District, an investigation conducted with 288 students aged 11-15 years identified a physical violence rate of 85.4% 41 .
It is important to point out that the studies mentioned above analyze ages above those chosen in this study.Thus, the findings of the high prevalence of the phenomenon occurring in the transition from childhood to adolescence, in which 48.3% refer to fear of being beaten at school, show their power towards contributing to a broader discussion of the topic, emphasizing its occurrence as early as in childhood.
Returning to the Plan Brasil study 12 , this feeling of fear related to the fact of being a victim or witness of school violence was identified, and could be responsible for the creation and existence of a particular configuration of the world in which it is socialized and operates as a social control mechanism 42 .
In this reasoning, the study under discussion presupposes that minimum levels of tolerance and openness to dialogue in the face of conflicts have been established in daily school life, suggesting weak mechanisms of sociability and coping with adverse situations.
In this case, the child-perpetrator variable showed a statistically significant association with the independent variables gender, being beaten at home and family encouraging retaliation.Regarding child-victims, the variable family encouraging retaliation was significant; the variable participation in conflicts has been associated with age, gender and, once again, family encouraging retaliation.
We observed that children aged 10 years are less likely to participate in conflicts, which agrees with systematic reviews on the subject 43,44 .Inter-est in this subject is growing in the literature, justified by the recurrence of the findings that the child (or adolescent) who suffers violence would be at higher risk of experiencing a subsequent or simultaneous episode, termed as a revictimization or cycle of violence 45 , making the future rupture of these experiences a complicated process.
Likewise, the engagement of children in acts of violence, specifically as perpetrators, is pointed out as a risk factor for the development of antisocial patterns with the possibility of entering into the field of crime in adolescence or adulthood 46 .
Regarding the gender variable, male children are more likely to be perpetrators and participate in conflicts; however, they were not more likely to be victims, contrary to the results of previous studies, in which the highest incidence is still observed in boys, both in the roles of perpetrators and victims 20,44,47,48 .
Family structure issues drive the discussion to another context associated with victimization among peers.In the data presented, approximate- ly half of the children (46.6%) belong to non-nuclear (single-parent, extended, reconstructed...) families.However, there were no significant associations with the variables child-perpetrator, child-victim or participant in conflicts.Although the variables do not have a significant association, we cannot disregard their relevance in this context, since the literature carries discrepancies concerning associating the family structure with the child being a perpetrator or victim.
A research carried out in Portugal with 242 7 th -9 th graders evidenced an association between the student being a victim and having a single-parent family or living with the simultaneous absence of the paternal and maternal figures; the student-perpetrator showed association with nuclear families 13 .This association (perpetrator-nuclear family) is upheld in other studies conducted in different countries 16,33,40 .Attesting this divergence in the literature, one Portuguese research with 5 th -6 th graders indicates that 46.6% of the perpetrators come from single-parent families and 58.6% from nuclear families 36 .
The analyses reveal that the association of the phenomenon with the family does not cease before its demographic characteristics, since high proportions of perpetrators and victims can represent the reproduction of parental values and practice 18 , in which the child internalizes the family relationships and establishes them for others contexts, in a perspective of historical, social and subjective construction of the social processes and the experienced interactions 49 .
In this study, parental practices with coercive measures as an educational model in the analyzed context were found to be recurrent.The variable being beaten at home is strongly associated with the child being the perpetrator (13.0%greater likelihood) evidences the possibilities of behavioral reproductions in other settings.In a compilation of studies, Minayo 1 shows that children/adolescents who are physically battered at home are four times more likely to be identified as aggressive at school.
The study at hand evidenced the prevalence of the variable family encouraging retaliation associated with all dependent variables, namely, perpetrator (17.0%more likely), victim (5.0% more likely) and participation in conflicts (5.0% more likely).
The association of the family encouraging retaliation to the events of interpersonal violence in the schools investigated unleashes a complex and broad discussion before the educational practices that are rooted in the families.One can understand the great challenge to articulate adequately the partnerships between families and schools that can understand the family structures and dynamics built in their historicity and, together, identify resolutive propositions.
Another research 15 concludes with these results when it verifies the correlation between increasing reports of violent behavior at school such as cursing peers; participating in fights; suffering threats at school; feeling scorned, and having pleasure in fighting with family environments perceived by the child as violent or non-affective.It is highlighted that the more recurrent the family violence, the higher the likelihood of the students reporting involvement in interpersonal violence, with no difference between genders 20,33 .
The literature advocates that the parental educational practice gives rise to or steps up violent behavior.Cultural and social aspects, especially in abusive families and practicing punitive, authoritarian and affection-devoid practices can have repercussions on child behavior 1 and the development of antisocial attitudes 18,50,51 .
During this study, we considered the potentialities of building and strengthening an articulated intersectoral action regarding welfare, education and health, weighing the associations identified with the manifestations of interpersonal violence in the school space.It is challenging, but not impossible, to analyze strategies that resume dialogue, solidarity, respect for the diversity and the uniqueness of the other, and call on the public power to take responsibility for the problem.
In this incomplete state, but assuming the propulsive spring contour, the data point to the capillary complexity, which is associated spirally-wise with interpersonal violence.This setting undertakes continuous investigations that resume old debates, foment new clashes that result in current solutions.
We should elucidate some limitations of the investigative scopes.In the first allusion, due to the population studied in its transition from childhood to adolescence, the results cannot be generalized to the other regional ones of the city of Fortaleza, nor to another regional and national setting.Nevertheless, respecting socioeconomic and cultural diversities, the study portrays similarities with other findings when confronting national and international literature.

Conclusion
Concerning the uncovered picture about interpersonal violence in children of municipal public schools and the identification of associated factors, the investigation showed that they were related to gender, age and parental practice, signaling that the school context is shaped like a space for the reproduction of internalized violence, living with value distortions which are important references to fostering other ways of resolving conflicts.
Families who follow coercive measures and instruct peer retaliation have been significantly associated with interpersonal violence in children, recognizing, in this logic, that the meanings of school experiences are also based on parental education practices.We suggest that the study supports projects and strategies to address interpersonal violence, strengthening the sectoral integrality, the conflicts and debates in other collective venues.

Collaborations
Nobre CS participated in the elaboration and development of the research, analysis of the data and the drafting of the paper; Vieira LJES participated in the analysis of the data and the drafting of the paper; Noronha CV participated in the drafting of the paper and Frota MA participated in the elaboration and development of the research.

Table 1 .
Distribution of school year, demographic data, school and household dynamics related to interpersonal violence of children from public schools.Fortaleza, Ceará, Brazil, 2013.

Table 2 .
Association between school year, demographic data, school and household dynamics related to interpersonal violence with the child perpetrator variable.Fortaleza, Ceará, Brazil, 2013.

Table 3 .
Association between school year, demographic data, school and household dynamics related to interpersonal violence with the child victim variable.Fortaleza, Ceará, Brazil, 2013.

Table 4 .
Association between school year, demographic data, school and household dynamics related to interpersonal violence with involvement in conflicts.Fortaleza, Ceará, Brazil, 2013.(n = 874).