When is Peer Aggression 'Bullying?' An Analysis of Elementary and Middle School Student Discourse on Bullying at School

Many forms of peer aggression are referred to as "bullying" by students, parents and adults, and this can be a source of confusion in schools. The main purpose of this study is to explore the circumstances under which students characterize peer aggression as “bullying” incidents. A secondary goal is to examine the feelings students have about the effectiveness of reporting peer aggression to adults. Both objectives are intended to reveal information that will enhance communication about peer aggression and bullying between students and adults. Six focus groups with 54 students in grades three through eight were conducted. The groups were organized in patterns based on grade level and gender, and qualitative methods were used to analyze the results. The findings showed that although the students defined bullying in ways that are similar to the criteria in the literature, they chose different words to describe them. Younger students also expressed greater faith in the ability of adults to respond effectively to bullying situations. Older students preferred to confront a bully with equal force or to reason with a bully to stop the aggression.

Palabras clave: agresión entre iguales, acoso escolar, Escuela Primaria, Instituto de Secundaria, grupos de discusión he past decade has been marked by a dramatic rise in interest in bullying as it has become more widely recognized that students who fear being harassed or degraded at school cannot achieve or function to their potential.Research shows long term negative effects for bullies, victims, those who are both bullies and victims, and also observers (Nansel, Haynie, & Simonsmorton, 2003;Rivers, Poteat, Noret, & Ashurst, 2009).In addition, maladaptive social behavioral patterns can increase the occurrence of various other forms of peer rejection and victimization in school, and societal pressure can lead uninvolved students to become impassive bystanders that are unwilling to help the victim or report the incident to an adult.
Qualitative Approaches to Bullying Olweus (1993) defines bullying as a deliberate and repeated long term exposure to negative acts performed by an individual or group with either higher status or greater strength than the victim.It may involve verbal acts such as threats or insults, physical acts such as assault, and social isolation as in deliberate exclusion of an individual from a group (Due et al., 2005).Survey research is the dominant methodology used to measure bullying prevalence, and school administrators are advised to carry out surveys as an initial way of determining the extent of bullying in their schools.In most student questionnaires, the actors and behaviors thought to be associated with bullying incidents are defined by researchers and presented to students on paper, on a computer screen or read aloud.Despite the abundance of this research, doubts may be raised about the accuracy of quantitative estimates, considering the age and developmental level of the research subjects, the degree of cultural variation across school settings, the varying definitions of bullying used in the questionnaires, and the fear of reprisal if confidentiality is breached.Overreliance on surveys may also contribute to a void in what the research community knows about the ways that children understand and interpret their own actions (Teräsahjo & Salmivalli, 2003).In addition, the use of generalized terms for actors (e.g.victim) and incidents can have a lasting effect on children and their behaviors, that will impact their future interactions and identities (Ringrose & Renold, 2010).
Ethnographic studies of middle school and high school students on peer aggression are less common than surveys, but they are notable for the ways T in which they reveal aspects of adolescent culture.In School Talk, for example, Eder (1995) exposed ritualistic aspects of verbal abuse and sexual harassment that were a very normal part of everyday life at school.Similarly, in Freaks, Geeks and Cool Kids, Milner (2004) draws an analogy between high school peer behavior and the traditional Indian caste system by demonstrating how victimized students can be reduced to the status of one who is so low that he or she is considered "untouchable." Other qualitative studies show that children talk about bullying using different terms than educators and the research community.For example, Mishna (2004) found that 4th and 5th grade students tended to define bullying as a situation in which an older or a stronger person is hurting someone who is smaller or younger, whereas teachers and parents defined bullying as an individual or group exerting power over others, or taking advantage of them.Vaillancourt et al.'s (2008) research suggests that children use different criteria to identify personal experiences with bullying when they are primed with a standardized definition of what bullying means.In this study, a sample of 1,767 students between the ages of 8 and 18 were divided into groups, one that was read a standardized definition of bullying and the other that was given the opportunity to write freely about what bullying meant to them.The students who were allowed to write freely reported more victimization than those given the standardized definition, and their definitions included very few of the terms that researchers use, such as intention, repetition, and power imbalance.Instead, the students who wrote freely about bullying reported incidents that were in line with expectations about specific forms of harm for their grade level (younger students reporting more physical aggression and older students reporting more relational aggression).One evident policy implication drawn by the authors of this study is that clear and standardized definitions yield more conservative estimates of prevalence.
There is also evidence that young people describe bullying situations using predictable speech patterns that minimize the impact of bullying.Teräsahjo & Salmivalli's (2003) case study of children in classes with noticeable bullying problems identified interpretative repertoires that the subjects used to explain things that adults defined as bullying.In general, the students were found to be downplaying the harm that a bully's actions have on their victims by using discursive devices that suggested that things were not as bad as they seemed to adults, or that the victim was to blame because of his or her differences from the others.Similarly, Guerra, Williams, & Sadek (2011) found that young people viewed bullying as "as a next step in the continuum after teasing, somewhat normative at low levels but problematic when really hurtful" (p.303).
Child discourse on the motivations for bullying can also be understood in an interpretive fashion.Thornberg's (2010) qualitative interviews of children aged 10 to 13 years positioned child representations under the microscope by permitting the subjects to feel as though they were the experts in defining bullying.Thornberg's approach draws upon the symbolic interactionist, new sociology of childhood, and grounded theory traditions by resisting the structural constraints imposed by authoritative categorizations of behaviors, and electing instead for an interpretivist approach that is attentive to the ways in which actors themselves describe their cognitions and activity.The results indicate that most students believe that bullying is a response to social deviance, such as the possession of an unfavorable characteristic or membership in a minority group.Ranking second is social positioning, such as an attempt to improve one's popularity or to simply "fit in."Both of these representations were made by children who were not primed with a standard definition of bullying, yet they reflect established definitions of behaviors such as biased-based bullying, or that which is associated with undesirable victim characteristics (Greene, 2006;Rigby, 2002), and the commonly referenced "power imbalance," which may be maintained or reinforced through harmful behavior toward others perceived as vulnerable.
These qualitative results point to complexities in the nature of child behavior and cognition which may be nearly impossible to detect in survey research.The illusiveness of this phenomena is exemplified in Crosnoe's (2011) ethnography of a large public high school.Interview data from this school show that students identified particular "looks" or facial expressions as conveying the message that they did not fit in.If these behaviors are correctly perceived by the subjects to be incidents of social exclusion, they might meet Rigby's (2002) criteria for bullying since they are intended by the perpetrator to inflict harm and they are experienced by the victim as a form of unjust treatment.Yet paradoxically, one student interviewed by Crosnoe felt she did not fit in because she reported receiving no looks from others at all.Although the feeling of injustice might be experienced in this case by the apparent victim, it is unclear whether the intention to do harm was really present on the part of the believed perpetrators.In these cases, qualitative interviewing has value in its potential to reveal forms of bullying that are unlikely to be discovered in a survey.One of the goals of the current study was to learn about peer aggression and bullying from the actors themselves, by enabling students to freely describe and interpret their own experiences.
In the current study, we examine the ways that young people define and respond to situations involving peer aggression.We also explore the congruence and dissonance between established, published adult definitions of bullying and those of the study participants.The data are derived from a series of focus groups conducted with students in grades 3 through 8 using deliberate naiveté (Kvale & Brinkmann, 2008) on the part of the moderators as a way of stimulating discussion about incidents that the students define as bullying on their own, regardless of whether they would meet the criteria for a standardized definition of bullying.The results provide information that can be used by school social workers, counselors, psychologists and educators to better identify with the language that young people use to describe bullying and their willingness to report it to an adult.

Method
The data for this study came from a larger Needs Assessment for an antibullying program in a parochial elementary and middle school in the Northeast region of the United States.At the time when the study took place the school did not have an anti-bullying curriculum.All of the 179 students in grades three through eight were invited to participate in a confidential computerized survey on bullying in their school by sending home notices and consent forms.161 students (or 89.9%) agreed to participate in the surveys.This subsample was 51.6% female, 45.3% White, 18% Asian, 2.5% Black or African American, and 31.7%"Other" or mixed race.50% of the students who selected the "Other" race option, or 26.1% of the overall subsample, also identified as Hispanic or Latino.A random sample of 54 students was then selected to participate in a series of six focus groups.In order to ensure an equal number of girls and boys and an equal number of students from each grade, the sample of 161 survey respondents was first stratified by age and gender, and then a systematic random sample was drawn.
Among the 54 randomly selected students, parental consent was obtained for 46 (85%), and all but two of the subjects provided child assent to participate.The consent form made it clear that although the focus groups would be tape recorded, the names of the children would not appear in any publications or reports, and the transcriptions would be destroyed after a five year period.The students were divided into three focus groups for each of the two grade levels, based on their gender and the order in which they were randomly drawn.The elementary set (grades 3-5) included an all boys group, an all girls group, and a mixed group of boys and girls.The middle school set (grades 6-8) included an all boys group, an all girls group, and a mixed group of boys and girls.The final sample was 50% female, 36.4% White, 25% Asian, 2.3% Black or African American, and 34.1% "Other" or mixed race.34.1% of the subsample also identified as Hispanic or Latino.Each grade level was represented by 18-20% of the final sample, except for the third grade which represented 11.4%.Five of the six focus groups ranged from 7 to 9 students and one focus group was comprised of 5 students.Each session lasted between 55 and 65 minutes.
A female researcher moderated all three of the focus groups in grades six through eight with the assistance of a female graduate student, and a male researcher moderated the same in grades three through five with the assistance of a different female graduate student.The first two focus groups were used as norming sessions, enabling the moderators and assistants to silently observe one another.One norming session was run by the male moderator and his assistant, while the other moderator and assistant observed.The second was run by the female moderator and her assistant, while the other moderator and assistant observed.The remaining four sessions were conducted with just a single moderator and a single assistant.
The focus groups were held in a comfortable space (all participants sat in couches) in an activity room at the school.Each session began with the moderator reading a description of how the session would take place, and advising them that their words would be tape recorded.The students were told that they would be asked to talk about times when they witnessed peer aggression, or they were involved in it, being careful not to reveal the names of the people involved.The students were then given an opportunity to indicate whether they wanted to participate or not on an assent form.It was viewed by the research team as essential that the students were asked to assent after they entered the room so that they could see the other students first and use that information to help them decide whether they wanted to participate.In the event that any students decided not to assent (this happened with two students), they were walked back to class.
At the beginning of the focus groups the moderators asked the students to express a form of bullying that came to their minds by drawing a picture and sharing it with one or two other students seated near them.After they discussed the drawings, the moderator started posing questions to the group by asking about the common actors in a bullying situation, namely the bystander, the victim, and the bully.In reference to each common actor, the groups were asked, "If you were this kid, what would you do?" Normally this led the students to talk about being a bully, a victim, or a bystander.It also led them to bring up experiences they have had in the past.When this happened, the moderator treated all of their contributions as plausible and worthy of consideration.No presuppositions about the appropriateness of defining an incident as "bullying" were made by the moderators.Complementary probes were used to invoke further comments by the responders, and the group as a whole, such as, "Good," "I understand," "That was a good idea," etc.During the interviews, the moderators exercised deliberate naiveté, which Kvale and Brinkmann (2008) define as, "openness to new and unexpected phenomena, rather than having readymade categories and schemes of interpretation" (p.28).
The interviews were analyzed using open coding to identify themes in the student responses and discussions.The process began with a researcher who was not present during the focus groups listening to all of the audio recordings.Then the recordings were transcribed, and examined again by the same researcher.After each word-processed transcription was carefully read, common strings of text were compared and analyzed following Corbin & Strauss' (1990) open coding technique.This led to the identification of code segments which were grouped and labeled by their focus group session, the gender of the speakers, the gender mix of the group, and the grade level of the group.
The code segments were used to develop one event category, classified as a "bullying incident," and several other themes and subthemes, as recommended by Creswell (2012).A bullying incident was coded as any situation in which the students told a story about physical violence, teasing or relational aggression.All stories told by the students were considered, regardless of whether they fit popular definitions of bullying in the literature (e.g.power imbalance or minimum duration).The coded bullying incidents included one-time events, recurring events, fights, insults, rumor spreading, etc. Incidents were further coded as acts of physical aggression if they involved physical contact, such as hitting, kicking, pushing or punching; verbal aggression if they involved shouting, cursing, name calling or other forms of teasing; relational aggression if they involved spreading rumors, telling lies to hurt someone, exclusion, or deliberately damaging someone's friendships; and cyber aggression if they took place on a computer, a cellular phone, or another technological device.
During the probing and discussion, the moderators also asked prepared questions on how they defined a situation as a case of bullying.For example, the moderator would listen to a story and then ask the speaker if they thought the situation was a case of bullying, and why.The responses to these questions, and the discussions that ensued, contained many "definitions of bullying" which were coded.Similarly, when students reported having witnessed or been a victim of bullying, the moderator would ask how they responded to it.The dialogue that followed these questions contained many "responses to bullying" which were also coded.During the coding process, two subthemes on responses to bullying among the older students were also identified and labeled, one on "confronting the bully," and the other on "reasoning with the bully."Once all of the data were coded, the quotations were examined again, classified by the grade level and gender mix of the group, and summarized in writing.

Bullying Incidents
During the interviews, the students made references to 49 incidents of peer aggression.Fifty one percent of the incidents involved verbal aggression.In one instance, a boy named Jimmy said, "...one day I saw a couple of kids playing basketball and a kid missed a shot and they were all like making fun of him and calling him stupid."In 29% of the incidents physical aggression was reported.For example, Eric, a boy in grades 3 through 5 said, "I was in the park and um, I was like going down the stairs to um play baseball and then I saw a guy punched another guy in the face and his glasses like flew off."Only one case of cyberbullying was reported and in a small number of other incidents, the nature of the conflict was unclear.In 18% of all incidents (or 64% of all cases of physical aggression), it was determined that a physical fight had occurred, meaning that both students were engaging in the physical contact.In 60% of the incidents mentioned, the students reported that they had been a victim and in 36% of the incidents, an observer.Their role was unclear in the remaining incidents.No child admitting that he or she had bullied anyone else.
Forms of relational aggression were reported in 18% of the incidents.For example, several girls in grades 3 through 5 responded in turn to a moderator's question about whether or not they had witnessed bullying: Alicia: People kept telling me to like somebody, but I didn't really like the person.So then I told my mom and she told me to ignore them and they will probably find somebody else to bully.Mary Anne: Rumors…a rumor that I like someone else, but then I said "No I don't."But then I just ignored them entirely.Brittany: Rumors… people kept saying that I did something and I really didn't, and they kept saying it.I told my mom and my mom knew the person that spread the rumor.She told them to stop.

Definitions of Bullying
Students in all grades included a variety of physical encounters (mostly fights), lying, spreading rumors and hurting feelings as forms of bullying.In their definitions, no children in either the younger groups or the older groups made a specific reference to a power differential between the bully and victim, but they often described situations in which one child was harming another child, by hurting them physically, verbally harassing them, or damaging their reputation.
Students in grades 3 through 5 tended to use broader definitions of bullying than the older students.The following exchange provides a representative example of a student from one of the younger groups trying to capture a form of relational aggression into her definition of bullying: Samantha: Lies…once, one of my friends was talking to her other friend and her other friend told that it was me who did it, when it was actually her friend because she lied to her and doesn't want to tell on her.
Moderator: Do you think all that stuff is bullying or is that just someone saying some things that aren't true about you?Samantha: It's bullying...I think it is when others hurt another person.
Another student at the same grade level attempted to use physical aggression in his definition of bullying: Moderator: Is there a difference between someone being bullied and being mean to each other?So what is the difference, Bobby?Boy 1: The difference is when two people are fighting, they are fighting but they are just staying stuff back and forth to each other.When somebody is bullying, they are actually being mean and making someone feel bad.
Among the students in the older grades, there was a tendency to identify an incident as bullying when a situation moved from bad to worse, or when it occurred with a certain degree of regularity.An example of this sentiment was expressed in an exchange between a few of the girls in one of the older focus groups: Sharon: If it is like in that scenario with the phone it's just like teasing.Saying you are better than me, you're not better than me just because you have a different phone.Moderator: It's interesting; you are talking about the difference between teasing and bullying.How do you know?What to you makes the difference between teasing and bullying?Sharon: Bullying is like something really serious.Lacey: When they are teasing, I know they are just joking.Sharon: Exactly.Lacey: And usually it's with a close group of friends.Sharon: But then when you kind of get really hurt instead and you're kind of like "ok it's enough," then it is bullying.Lacey: And when it happens day after day.

Responding to Bullying Situations
Students in grades 6 through 8 typically said they would try to work the situation out with the bully, but students in grades 3 through 5 were more likely to report the incident to an adult.Among the older students who said they would try to stop the bully, two subthemes were evident in the data.In the first, students said they would try to stop the bully by presenting a countering force, either physical or verbal.For example, Joseph (grade 6-8) said, "I would just go up to the kid and be like wow you're really making fun of him?Why don't you think you could pick on some one you could actually take.What makes you think you could pick on little kids.I would say look why don't you take him?He's your size."In the second subtheme, students made appeals to the bully's sense of reasoning.In reference to an actual incident, Daniel (grade 6-8) reported, "Well somebody was calling somebody else names and I said you really shouldn't do that because you're also one.The kid was saying that he was ugly and weird and wasn't smart.But the other guy was smarter than him.Which was like really stupid because he kind of knew his self so I just waited there and said you're dumber than him." Reporting the incident to a teacher was common among the younger students but very uncommon among the older students.When asked hypothetically what adults should do in response to bullying situations, the younger students and the older students suggested punishments, shouting and forcing the children to work out their problems.One difference between the two age groups was noticeable in their expectations for results.Younger students tended to think these actions would resolve the problem, yet the older students tended to think this would do no good.For example, a common response among the younger students was expressed by Sylvia (grade 3-5), who simply said, "Walk up to the two who are talking, break them apart, and say it's not right to talk behind the person's back.And give them reasons."In contrast, Dylan (grade 6-8) said: Dylan: If the victim goes to the teacher and says this person is bullying me...they say you should (be) mature and work it out yourself.I think they should at least give them advice to help you.The victim has a problem and he is going to the teacher for help.So I think if you want them to work it out on their own, at least help them a little bit because obviously they don't know what to do.So I think that is what the teacher should say to the victim to help them.
When the moderators probed with the younger students, however, it was evident that reporting incidents of bullying often did not lead to the intended results.
Carolina: This one kid was picking on me and I told him to stop but he said no.Then I told the teacher but the teacher talked to the person and he didn't stop.So I told my mom and he still didn't stop, so I just stayed away from him.
Many of the older students seemed to be recalling experiences like these when they said that they felt telling an adult would accomplish very little.For example, an older student named Michael had the following exchange with the moderator: Moderator: Is telling the bully to apologize, something that works usually?Michael gives body language, indicating "no."Michael, do you think no? Michael: No, they don't actually mean it; they're just doing it so that the adult would just walk away and another time do it.Moderator: Does that sound like it would happen?A couple of students respond affirmatively.Moderator: Mark you think so? Mark: Yeah Other students in grades 6 through 8, expressed the sentiment that telling was something they would do only if they felt they had no other choice.This feeling is evident in the comments made by Angela and Jessica: Angela: I wouldn't just let them just bully me...I would say something back.I would say leave me alone, or I would kind of defend myself.And if that didn't work, I would tell an adult.If it really got bad and I was really upset, then I would tell an adult.Jessica: I would just tell them to stop and if they don't, I would just tell them again.I wouldn't go to an adult, unless it was really serious.

Discussion
Qualitative methodologies can be used to obtain "a more naturalistic, contextual and holistic understanding of human beings" (Todd, Nerlich, McKeown, & Clarke, 2004, p. 3).Although qualitative studies remain outside of the mainstream literature on bullying, their findings provide illustration of the ways that young people understand various forms of peer aggression, such as relational aggression.They also demonstrate areas of weakness in the generalized conceptualizations of bullying used in the literature, and in schools, by bringing attention to discrepancies between young people's definitions of bullying and those used on questionnaires and in official school codes of conduct (deLara, 2012).
Like other focus group studies on this topic (Cunningham, Cunningham, Ratcliffe, & Vaillancourt, 2010;deLara, 2012) the interviewing technique in this research allowed the respondents to engage freely with the moderators and their fellow students in a way that led them to new discoveries and conclusions about their past experiences and the events they have witnessed.School social workers can also benefit from this approach by enabling students to gain support from their peers through productive interaction in group settings (Cowie, 2011).
A socio-ecological systems perspective views the phenomena of bullying as a result of complex relationships between the individual, family, peer group, school, and community (Benbenishty & Astor, 2005).Yet when describing bullying situations, the respondents in this study made very little mention of their family members or school community members other than teachers.This may be explained by the belief expressed among the older students that adults are incapable of helping effectively.
Although both age groups reported that adults tend to punish bullies, the older students had little faith that this would change the bully's behavior.In result, reporting tended to be viewed as a last resort.More often, the older students said they would prefer to take on the bully all by themselves either by challenging the bully to a fight or by returning a verbal assault to defend another student.Others tried to convince the bully that what they were doing was wrong, or that they should stop because they would not want it done to them.The younger students, on the other hand, exhibited a greater faith in the ability of adults to respond effectively, but it was evident from their accounts that the behavior often did not stop after they reported it.These findings point to age-related differences in the ways in which adolescents view bullying and adult responses to bullying, and emphasize the need for tailored approaches to anti-bullying curricula.
The differences in the opinions between the older and younger students about reporting might be explained by the older students' higher awareness of a failing reporting cycle.If their experiences tell them that reporting does not bring about a positive change, then their decision to employ other means may be adaptive.For this reason, school interventions that are intended to encourage a greater willingness to report cases of bullying may not bring positive results.Instead, school social workers might attempt to better educate parents and adults at school about more appropriate and effective ways of helping children who are victims and bystanders, and offer strategies that will enable them to better obtain student trust.
The results also showed that the children did not use the specific terms to define bullying situations that are common in the literature, such as the presence of a power differential, the intention to cause harm, or a minimum duration of the aggression.Instead, they used their own terms to capture many of the same ideas.For example, it was evident that the children saw bullies as more powerful than victims when they described the direct physical or psychological harm they inflicted upon them.The children also described situations as bullying when they were more hurtful or when they worsened, indicating that they may relate to the inclusion of criteria such as intent, severity, and harm, in the definition of bullying even though they preferred an alternative language for describing them.It is also possible, however, that the students were not prepared to talk about power or a minimum duration of aggression in the context of a discussion on bullying, especially since the school did not have an anti-bullying curriculum where these topics might have been introduced.
The results of this study demonstrate the value in encouraging children to speak about bullying and their feelings about reporting it.By engaging in open group discussions, school social workers and psychologists can enable young people to express themselves about bullying using the words that they prefer.This can lead to an enhanced understanding of the congruence between the terms in the literature and the words that young people use themselves and perhaps enable children to inform research and practice on this topic in a more direct and meaningful way.It can also enable school personnel to learn more about how they can be more receptive to student needs when writing curricula and developing new strategies to help victims and encourage incident reporting.

Limitations
Although this study features several desirable conditions for focus group research, such as randomization, and groups characterized by grade homogeneity and varying gender compositions, a main limitation is that it is set in only one parochial school.Focus groups are also prone to conversation digressions, especially among children.In the current study, this was minimized by having the moderators utilize a highly structured script of questions and activities.Focus group interviews also require discussion in a public place, which in this study included closely related peer group members.This may have inhibited the willingness of the students to discuss sensitive information or to reveal their most inner thoughts on some subjects.To counteract this problem, the moderators used the drawing exercise to break the ice and build rapport.This may have enabled the conversations to take place with greater trust.They also promised them confidentiality and required them not to use names when discussing real situations.Finally, parents were informed about the availability of free mental health services in case their child became upset during or after participation, and the moderators and assistants were prepared to facilitate this if needed.

Conclusion
As school social workers, psychologists and counselors learn about the fears, anxieties, and difficulties that students face when dealing with aggression, there is a benefit they can bring to a school community by sharing the information with others in a sensitive and productive manner.This can serve a dual purpose as it not only enables adults at school to better respond to student needs, but also creates the potential for students to perceive greater advantage to reporting incidents of peer aggression when they see it or when they are victimized by it.Furthermore, by developing a more in depth understanding of the cultural cues and contexts that young