Bullying Prevention in Schools: The Need for a Multiple Stakeholder Approach

Since bullying is an ecological and systemic phenomenon that occurs in multiple contexts with multiple actors, it makes sense to consider the perceptions of multiple stakeholders and their relationships with one another and in relation to bullying prevention in schools. Using a non-probability, purposive sample, this study examined the perspectives of 45 school stakeholders, namely, principals, school social workers, bus drivers, and parents from an urban school district in the Midwest. The study unveils some of the implicit and explicit challenges associated with bullying prevention efforts. For example, bullying can be quite nebulous because people tend to look at the issue through the prism of their own experiences and positions, limiting their understanding of other stakeholders. Some stakeholders’ perspectives may be muted when bullying behaviors are discussed or reported. Overall, the findings support the use of multi-stakeholder approaches in developing a more holistic view of bullying. Recommendations include avoiding the reification of the views and voices of a select few and having a more open system of dialogue among stakeholders to create inclusion when addressing bullying.

In view of the ominous effects of bullying, including student suicides, school bullying remains a major social problem and public health concern (Cornell & Limber, 2015;NASEM, 2016). Despite the pervasiveness of bullying, its manifestations may vary from one context to another (Hong & Espelage, 2012). People's views of bullying are likely to be shaped by factors such as their social and cultural orientation as well as their perception of the world around them. In addition, both personal and professional experiences serve as filters for how people may explain or interpret bullying behavior (Bourdieu, 1984;Reed-Danahay, 2005). These multiple contexts contribute to the likelihood of both deliberate and unintended bias in bullying discourse. Therefore, multiple and counter-perspectives, are essential to consider in creating a comprehensive understanding of the complexity of bullying (Bradshaw et al., 2013;Flygare et al., 2013;Hong & Espelage, 2012). Further, research involving different stakeholders and the synthesis of multiple perspectives can aid in the critical examination of bullying to facilitate the development of viable solutions (Hong & Espelage, 2012;Hymel & Swearer, 2015;Kim, 2017). Thus far, such research appears to be limited. Therefore, this study examined key stakeholders' perceptions of school bullying and prevention efforts.

Literature Review
Bullying is described as deliberate acts of repetitive abuse intentionally aimed at an individual or group by another individual or group  and may occur in a number of forms, including physical, verbal, or written acts of aggression. Bullying can also be relational, and is often characterized by power differentials (Oliveira et al., 2018;Olweus & Limber, 2010). School bullying is a complex phenomenon that defies spatial boundaries because it is not confined to any one particular place within the school environment (Richard et al., 2012;Vaillancourt et al., 2010). For instance, in a study of 2,766 Dutch schoolchildren, Fekkes, Pijpers, and Verloove-Vanhorick (2005) found that bullying occurred in almost every conceivable space within the school environment including the playground, classrooms, school canteens (cafeteria), gym, and bathrooms. The study participants also noted that bullying occurred by bicycle racks outside the school and near their homes. In Robers and colleagues (2015) Indicators of School Crime and Safety report, students in the U.S. reported bullying occurring in similar areas as well as on school buses. Several studies, domestic and international, have corroborated these findings (e.g. Oliveira et al., 2018;Richard et al., 2012;Vaillancourt et al., 2010).
In an attempt to understand and address bullying, there has been a significant focus on the perspectives of students; specifically, as aggressors, victims, and bystanders (Ross et al., 2017). Teachers are fairly represented in bullying research literature as well, and most perceive bullying to be a major problem (Bradshaw et al., 2013;Hayes, 2017). However, the extent to which teachers consider bullying to be a problem tends to vary based on factors such as the type of bullying, the characteristics of the teachers and the students in their care (Bell & Willis, 2016;Eden et al., 2013;Espelage et al., 2014). Some teachers have considered physical and cyberbullying to be more serious problems than other forms of bullying (Bell & Willis, 2016;Eden et al., 2013). Male teachers, special education teachers, and teachers of younger students have expressed greater concerns about cyberbullying being a problem than female and general classroom teachers, and teachers of older students. Teachers' perceptions have been reflective of students concerns about the incidence of bullying suggesting some degree of consensus (Eden et al., 2013;Espelage et al., 2014).
In comparison, scant attention has been paid to the views of school principals (Foody et al., 2018), social workers (Slovak & Singer, 2011), and bus drivers de Lara, 2008). However, it is important to note that these are key actors in the school ecology who constitute an essential resource when assessing bullying, seeking to understand students' behavior, and intervening as needed.
In the immediate school context, principals' serve as primary gatekeepers with oversight of programming and general administration, and ensuring a supportive climate is provided for students (Melik-Stepanyan, 2014). Social workers also have critical roles in addressing students' behavioral and mental health problems (Whitley et al., 2013), while parents play a primary role in socializing children about social norms and expectations (Fekkes et al., 2005). Bus drivers (and attendants/monitors) are also covert and overt observers, as well as critical informants of students' behavior outside the primary school environment (de Lara, 2008;McNamee & Mercurio, 2008). Considering the multiple interactional contexts in the ecology of schools, a multi-informant or multi-stakeholder approach, as used in this study, enhances the representation of stakeholders and promotes a wider understanding of a phenomenon as complex as school bullying.
Previous research suggests principals and parents acknowledge bullying does occur, but differ in what they consider bullying. Further, principals are more likely to underestimate the incidence of bullying, while parents are more likely to perceive bullying victimization (Newgent et al., 2009). Previous research also indicates that bus drivers report bullying to be a systemic problem, and express concerns about the lack of parental accountability and responsibility, and lack of support from school administrators including inconsistent responses to student misconduct de Lara, 2008). In Soliman's (2017) exploratory study in Illinois, school social workers identified limited communication with students and parents as a problem and expressed the view that the inclusive engagement of constituents was necessary in promoting a prosocial school climate. In comparison to other stakeholders, bus drivers are rarely included or prioritized in most discourses on bullying and bullying prevention (de Lara, 2008;Putnam et al., 2003). Similarly, though previous research highlights the key role social workers can play in addressing school bullying, it appears they are often homogenized under the label of school staff and for that reason their unique experiences and perspectives become unclear or get lost (Slovak & Singer, 2011).

Theoretical Framework: Ecological Systems Theory and Habitus
This study uses Bronfenbrenner's (1979Bronfenbrenner's ( , 1986) ecological systems (eco-systems) theory and Bourdieu's (1984) concept of habitus to develop an understanding of bullying that is grounded in critical perspectives. Bronfenbrenner's eco-systems theory posits that the interplay of factors across multiple environments (or systems), namely, the micro, meso, exo, and macrosystem, provides a framework for understanding human behavior (Bronfenbrenner, 1979(Bronfenbrenner, , 1986. The microsystem encompasses primary or direct relationships and interactions, such as what happens within a person's family. The mesosystem refers to interconnections that occur with multiple microsystems-for example, how a school may be linked to a family. The exosystem refers to a sphere within which an individual may not have direct engagement yet is profoundly impacted by the interactions or actions that occur there. Examples of exosystem factors in school bullying include neighborhood characteristics and the urbanicity of schools (i.e., the impact of a school's urban location; NASEM, 2016). The macro system is another indirect or distant sphere that also has a profound impact on an individual. The macro system encompasses things such as politics, social and cultural norms, economic structures, and religious beliefs (Bronfenbrenner, 1979(Bronfenbrenner, , 1986NASEM, 2016). For instance, legislative efforts at the federal and state level include civil rights laws aimed at guiding how schools respond to bullying and protect students (Cornell & Limber, 2015). Within eco-systems theory, distal and proximal relationships, interactions, and actions are noted to impact social phenomena including individual development (NASEM, 2016). Through the lens of the eco-systems framework, factors such as parent-teacher/staff relations, school-community partnerships, policies and legislative provisions, and even the socioeconomic conditions of families are linked to bullying (Hong & Espelage, 2012;Hymel & Swearer, 2015).
Complementing the application of eco-systems theory in the current study is Bourdieu's (1984) concept of habitus. Habitus encapsulates the entire gamut of people's interactions in formal and informal, social and professional networks, and how these experiences become embedded or ingrained in an individual (Bourdieu, 1984;Shimoni, 2017). Further, habitus reveals the intuitions and schemas people develop over time through those experiences. As experiences accumulate, they become part of a person's worldview, influencing how they perceive and interpret events both consciously and unconsciously (subconsciously; Bourdieu, 1984). The application of habitus in this study is helpful in showing how the positionality of these stakeholders -parents, social workers, principals, bus drivers and their attendants -shapes their views on bullying (Bourdieu, 1984). For example, principals' response to bullying is likely to be shaped by bullying policies from the school district, state, or federal government (a macro system factor in eco-systems theory), in cases where such provisions exist (Cornell & Limber, 2015). Having to interpret and apply the policies may mean making the determination of whether a particular report qualifies as bullying. With that being said, the reports that come to principals are often secondary reports from teachers, parents, students, and bus drivers and attendants (Newgent et al., 2009). Depending on the extent to which those making or bringing the reports know or understand the policies, their views about what constitutes bullying may be different.
Further, not everyone may know or understand the investigative process (where such mechanisms exist) used to establish if an incident qualifies as bullying or not. On the other hand, some stakeholders, like parents and even principals, may not have a complete view of bullying on the buses or in the classrooms. By applying the concept of habitus, these researchers seek to highlight the distinctions in how people view bullying. Though the stakeholders may be in the same ecology (i.e., school community), that does not mean that they will share the same views. In the present study, the researchers believe that exploring the perspectives of these stakeholders provides useful information for a comprehensive understanding of bullying to further inform the development of effective anti-bullying programs and interventions.

Method Participants
To interview bus drivers, parents, principals, and social workers, a non-probability sampling method (purposive and maximum variation sampling) was used (Padgett, 2008). Table 1 shows the demographic characteristics of the 45 study participants. All the participants were from the same urban school district where an empathy-based bullying prevention program was being implemented in the elementary and middle schools. All four parents were affiliated with the same school. There were 64 schools in this school district and at the time of the study, 36 schools had received the empathy-based bullying prevention program.

Recruitment
Recruitment of participants took place after receiving approval from the Institutional Review Board of the researchers' institution. The researchers specifically sought constituents of the school district who may have insight on student behavior and events in the schools. Again, the goal was to include people with different positions/roles in the school environment. This was intended to enhance the chance of having a sample with varied experiences. For parents, the invitation for participation was extended through the leaders of parent associations of schools that had received the empathy-based bullying prevention program. For school principals, invitations were extended to principals of schools that had received or were receiving the program. For bus drivers and attendants, an email was sent to the bus garage. For social workers, an email was sent to the coordinator for support services for the school district. The invitations for participation included a description of the study and an explanation of the study procedures. Details presented included: informed consent processes, confidentiality, and the option for participants to discontinue their participation at any time, the mode of data collection (audio-recorded focus group discussions), the duration (approximately 50 minutes), and a sample of the questions/topics to be discussed.

Procedure
Focus Groups and Transcription. Overall, there were seven focus group discussions-two separate focus group discussions with social workers, two with school principals, one with parents, and two with bus drivers and bus attendants. Each focus group discussion lasted approximately 90 minutes, exceeding the initially estimated time of 50 minutes. The focus group discussions were audio-recorded and subsequently transcribed. Study participants self-selected a color or spice as a pseudonym to ensure confidentiality. In line with the objective of the study, focus group discussions supported the examination of diverse experiences and contrasting viewpoints in a time-saving and cost-effective manner. Through moderated interactions, the focus groups inherently allowed for a triangulation of sources and an opportunity for gaining critical insights (Nyumba et al., 2018;Padgett, 2008).
There was some overlap between the collection and analysis of data. We adapted the theoretical sampling process from grounded theory. This was done in a manner that could be best described as quasi-theoretical sampling, because we were not necessarily guided by an emerging theory (see Glaser & Strauss, 1967, for more on "classical" theoretical sampling). Some of the data analysis steps such as memoing and coding were conducted during the intervening periods of data collection with the different stakeholders. As a result, preliminary findings from the data collected from early focus group discussions informed subsequent ones (Glaser & Strauss, 1967). For instance, some of the questions posed to principals were informed by insights gained from the focus group discussions with bus drivers and attendants and social workers.
Positionality. The two lead authors were hired independently by an outside public health agency to measure changes in schools' climate related to bullying. Although no money was paid to the researchers, the two-year project did buy out two university courses. The two lead researchers acknowledged biases and subjectivity within the frame of this research continuously throughout the research process. Each of the lead authors believe that school bully prevention efforts, especially regarding empathy training, can be helpful in developing supports for students who bully. However, both lead authors felt that the limited dosage of the programing (twice per classroom) would not significantly impact incidences of bullying. The lead authors were also aware that certain groups of people who provide services to schools and students are underrepresented in bullying prevention and bullying research. This explains the engagement of bus drivers and attendants in order to learn more about their perspectives on school bullying.

Data Analysis
The transcribed interviews were deductively and inductively coded using MAXQDA, a computer-assisted qualitative software. The analysis of the data was not linear or hierarchical, but more cyclical and iterative. To ensure the trustworthiness of the findings, the following steps were taken. First, data were independently coded, using analytic and reflexive memos to keep track of the process and ensure auditability. Second, notes were compared to determine the points of convergence and divergence in the conclusions. This second step constituted a form of triangulation (analyst/investigator triangulation) which allowed for prolonged engagement with the data and served as a good way to test the credibility of findings (Lincoln & Guba, 1985;Padgett, 2008). Member checking with participants was also completed, which included initial debriefings with participants immediately after the focus groups. This approach was used to give study participants an opportunity to quickly clarify points they had made. The transcribed data were shared with participants who provided their email addresses to be reviewed for accuracy, and to allow an opportunity to provide feedback if changes were needed (Lincoln & Guba, 1985). Based on participants' interest and availability, additional conversations took place about the study and interpretations. Though invitations for member-checking were extended to all participants, participation in these processes were voluntary, and not all the study participants took part.
For the next level of analysis, thematic analysis (Braun & Clarke, 2006;Dixon-Woods et al., 2005) of the transcript data was conducted. Using both a semantic (descriptive) and latent (critically interpretive) lens (Braun & Clarke, 2006;Dixon-Woods et al., 2005), major recurring themes were identified. To examine how these themes aligned with each other, and which of those themes was better suited in conveying the main point, we adopted the constant comparative method, which is common in grounded theory (Glaser & Strauss, 1967). In addition, elements of meta-ethnography, namely, reciprocal translational analysis and refutational synthesis were applied (Dixon-Woods et al., 2005;Noblit & Hare, 1988). In reviewing the independent accounts and experiences of the four groups of stakeholders, the meta-ethnography approach (specifically, reciprocal translational analysis and refutational synthesis) enabled us to examine the extent to which those accounts were similar and the extent to which they stood in direct opposition to each other (Dixon-Woods et al., 2005;Noblit & Hare, 1988). Though the focus groups were conducted separately, with this approach to analysis, we were able to examine the narrative data as though the participants were in direct conversation with each other. Overall, the eclectic approach we adopted facilitated the comparison and contrast of the views expressed by the various stakeholders. This made it possible for us to see arguments and counterarguments related to the incidence of bullying in schools, efforts to address the issue, and the stakeholders' respective roles or lack thereof. Finally, synthesized conclusions were distilled.

Findings
This research examined the perspectives of school principals, social workers, bus drivers /attendants, and parents on bullying. Specifically, the study focused on their views of what constitutes bullying and its prevalence; forms, causes, and enablers; the dynamics of stakeholder involvement or inclusion in interventions; and variations in applying bullying policies.

Perspectives on What Constitutes Bullying and its Prevalence
The participants unanimously noted that bullying occurs in their respective schools (including the buses) and the school district as a whole. However, they differed on the prevalence of bullying and how other stakeholders understand the phenomena. For instance, there were principals who did not try to quantify the prevalence of bullying in their schools. Nevertheless, they suggested that the level of bullying was low: Ms. Pepper's comment about "age-appropriate misconduct" may be reflective of the notion that there are certain behaviors that are linked to a person's developmental stage. Ms. Sage also added that, "about 95% of [her school's] reported bullying is peer conflict." In the same vein, some parents and social workers noted that the overestimation of bullying was quite likely because, on the part of both students and parents, "sometimes peer conflict gets confused with bullying." Ms. Pink, a social worker, mentioned that investigations of reported bullying cases in the past year in her school showed that, "99% were peer conflicts; kids being mean to each other [and] its rarely been an act of bullying." A parent explained that the confusion surrounding what bullying means or entails was common with younger children who interpret "everything like, when somebody is just a little mean to them or maybe doesn't want to play with them on the playground" to be bullying. It is quite evident that there is some framing of the perceived bullying behavior in a developmental context. Though it is important to avoid misclassifying behaviors and wrongly using the term bullying, overly subscribing to developmental frameworks could result in the normalizing of problem behaviors.
In contrast, the bus drivers and attendants noted that acts of aggression were rife on the buses. There was some support for this assertion from other participants. Ms. Clove, a principal, acknowledged, "the buses are very problematic." Ms. Yellow, a social worker, also shared that, "the bus is a hotbed for bullying." However, there were some contrary opinions, some of which seemed to question the veracity of the bus drivers' and attendants' reports. Ms. Rosemary, a principal, made it clear she did not think "the bus drivers see much of anything" and they were "just kind of oblivious." She buttressed her point with a situation involving four students who were spitting on each other, however, "the driver…had no idea." It appears the fact that "they [i.e. students] were in the back, he's driving he can't see" was not given careful consideration in her assessment of the bus drivers. To the bus drivers' defense, Ms. Clove elaborated on the challenges bus drivers face: It's hard on the buses because many of our buses are full, there is only the driver and not a monitor [attendant] on the bus to help address some of the situations and so the driver can't necessarily stop what they are doing in order to address the situation that is happening.
For Mrs. Mint, a parent like many others that live in a mile or less radius from their school suggests the "buses can be a real great incubator of bullying because you've got such a range of kids…proximity of travel is so small, even if there was bullying it certainly wouldn't happen very much." Within bus drivers and attendants' discourse about bullying, there was no reference to students and parents' difficulties in distinguishing between conflicts and bullying. Further, in discussing the occurrence and prevalence of bullying, some of the accounts bus drivers and attendants gave were of situations that may not be classified as bullying. For example, Ms. Coal, a bus attendant, shared that, "I have a little boy [on her bus]…one little girl was singing, he say, he didn't want to hear no singing, he took his belt off and whopped her face." This was clearly an assault that could meet a higher threshold of aggression. Yet using the definition of bullying which emphasizes the repetitiveness of actions Olweus, 1993), this particular incident would not qualify as bullying.
The principals shared that "repetition" was a critical element in determining whether an action was bullying. To provide some clarity about how bullying is defined, Ms. Pepper, a principal, said, "repeated behaviors that are making another person feel uncomfortable, intimidated or unsafe," and Mr. Basil, another principal, added, "with the emphasis on 'repeated' we have to constantly remind our students." With this in mind, if a bus driver or attendant reported a one-time incident like Ms. Coal's to a principal, the attempt to make the appropriate classification could potentially put them at odds with each other.

Forms, Causes, and Enablers of Bullying
The participants noted that bullying occurred in different forms, and shared examples of physical, verbal, and non-verbal acts of bullying. They also indicated that bullying was not a strictly gendered, male-dominated practice. Though Mr. Blue, a bus driver, shared an example that involved "…the guys…intimidating the girls on the bus," another bus driver noted that girls bully as much as boys do. A principal also mentioned that contrary to the belief that the "mean girl behavior" mostly occurs in middle school, some of it "starts as early as second or third grade." This was corroborated by a parent, Mrs. Red. Based on her children's experiences, she shared that, "with the girls, there is even probably some bullying…the beginning stages of it...at a younger age...With my son, I have seen more of that in the middle school." These views also suggest the participants perceive that bullying occurs across grade levels and may be occurring earlier than one may typically assume. The social workers and principals also noted the incidence of cyberbullying. For example, Ms. Clove, a principal, noted that for the last two years, in contrast to traditional bullying, her school dealt "…with just an insane amount of social media bullying." Concerning the causes of bullying, the participants made some sociological, ecological, and psychosocial attributions. On the sociological and ecological level, the participants talked with passion about the impact of the communities and neighborhoods students lived in, and more specifically, the home environment. Some participants expressed the view that bullying in the school was possibly a reflection of what occurred in the neighborhoods and homes of the students. A social worker, Mrs. Lime, shared that, "I have a neighborhood that has a lot of challenges and that comes into the school with the kids." Mrs. Gold, another social worker, shared a similar experience about her "old school" where the students' "neighborhood was such that it was so violent" that it negated efforts to promote positive behaviors because "there was a different mentality and culture, [and] different outlook." In contrast, she compared the old school to her new school, noting that, "they [students in her new school] have the support at home basically…and [in] their neighborhood." The parents also provided a contrast of the effects of neighborhoods and the home. They attributed the low prevalence of bullying in their children's school to the neighborhood they resided in and active support from other parents. Mrs. Oro, a parent, conveyed this point : I think there is a really strong neighborhood community here and that's not to be in anyway undervalued because I think the value of having just a strong parent community is huge in terms of helping our kids...I think what helps bring it all together is the fact that they share this common language that they all understand, practice and its re-enforced here [in the school], and then the kids take it home and its re-enforced at home.
Additionally, on the importance of the home (and parents), principals and parents shared that efforts to address bullying involved enlisting the support of the parents in changing the behavior of students. However, Mr. Parsley, a principal, added that, "it's even more challenging when that [bullying] behavior is learned from the parent." In line with this thought, the bus drivers revealed that, they get bullied by both the kids and their parents, and the "school system gets bullied by the parents." These findings suggest variances within ecologies and an association between school bullying and the homes or neighborhoods of students. Behavior that may be normative at home or in neighborhoods may not be acceptable in the schools (NASEM, 2016). Therefore, problems emerge when students are unable to successfully conform to the code of conduct at their schools.
Additional views were shared about the school environment and climate. Mrs. Salmon, a social worker, shared that her school experienced "a significant increase in bullying issues" following a change in leadership (the principal). She added that following the change, "the discipline and climate is not as consistent as it was previously…and things are not well communicated in that arena." The bus drivers also expressed concern about principals' handling of student misconduct, because the students continued to act with impunity. Per Ms. White, a bus driver, they wondered why the policies spelt out certain consequences "but none of that is happening." Ms. Yellow, a social worker, also shared that based on the district's policy on bullying, they documented reports and followed the steps such as "…calling parents and so forth. But...the end of that policy rests with the administration." These points appear to highlight the lack of resolute action from some principals resulting in an environment where student misconduct and bullying became prevalent.
Some of the principals stated that the school environment was sometimes a confluence of tensions, dysfunctions, and several unmet needs, and they had to deal with several issues that extended beyond providing education. For example, Ms. Pepper mentioned that her population of students was "high with domestic violence, and alcohol and drug abuse" and she had to deal with "Title IXs, and CPS [Child welfare cases with Child Protection Services]." This state of affairs often impacted principals' ability to deal with bullying in a manner they desired. Ms. Sage summed it up this way: We have a huge population of very very high need mental health issues in our parents as well as in our students…so while we would like to have more [support] groups with bullying, and more...supportive kind of things, a lot of the times it's kind of putting out fires.
On the psychosocial level, some principals, social workers, and bus drivers and attendants mentioned or alluded to the mental and emotional health of students. The bus drivers and attendants talked about some students "being on medicine." The social workers shared that the emotionally handicapped (EH) kids were "just a real big problem," and on the bus, they were "instigating trouble with others, and picking on the kindergartners [and] first graders." They, therefore, recommended that instead of having them on the "regular bus…those kids need to come on their own bus." Some of the principals also suggested there was an association between mental health and bullying, with Ms. Pepper, one of the principals, saying, "typically those kids that are bullying have mental health issues."

Alienation and Exclusion of Several Stakeholders in Bullying Interventions
Bus Drivers/ Bus Attendants. Despite the occurrence of bullying on the buses, the bus drivers and attendants stated that they were generally alienated from bullying prevention programs/interventions. Most of the bus drivers shared that there was a lack of alliances or partnerships with the schools (and principals). One bus driver, Mrs. Turquoise, highlighted the issue of alienation:

…I think you guys [Researchers]…and you're gonna work with the schools. You see, that's what the schools won't do with us and that's where the problem is. You guys can probably go in there and work with them but as far as [the schools] helping us out -Dead end.
Only one bus driver indicated that bullying was not a problem on her bus because her school treated her reports with the seriousness they deserved and responded with students being "suspended from the bus." Based on these submissions we inquired of the principals if they received reports of bullying on the buses from the drivers. The principals responded in the affirmative, with one of them adding that they did so "constantly." Further, in what appeared to be a striking contrast to virtually all of the bus drivers and attendants' reports of being on an island and having essentially no support from schools and the district in bullying prevention, Mr. Basil, a principal, shared that as employees of the school district, "…the bus drivers have to go through the [bullying] presentation anyway...and they are trained in bullying prevention and awareness." Mrs. Blue, a social worker, shared that the bus drivers in her school were trained annually and were included in bullying prevention efforts. She added that the training focused on her school's programming and it's distinctiveness from other schools "…so that they are also hopefully using the same language and carrying that over onto the bus so that they [students] are hearing it no matter where they are once their school day starts." For another school, the social worker, Ms. Teal, shared the principal's initiative of teachers volunteering as "bus buddy person" to promote more prosocial behaviors. Regarding how reports from bus drivers were handled she stated: I understood from our training for bullying at the beginning of the school year that all staff were to be using the same forms. But transportation [bus drivers], they still use their own forms...then in my building those are given to our in-school suspension person...who works with the buses and he handles some of it, some of it he refers to me, it really varies. I'm not sure how consistent it is.
Not only does Ms. Teal's account highlight variations in the approaches adopted by schools, it also makes one wonder the extent to which the use of different forms impacts how bus drivers' reports are handled. Such a lack of consistency could be detrimental to efforts to address bullying.
In what may be a nuanced suggestion of other factors contributing to problems on the buses, Ms. Pepper, a principal, added that, "I think [bus drivers] need more intense training on building relationships with kids and behavior management [Ms. Sage, another principal chimed in: "with kids and parents"] That's a huge issue." This seems to suggest that bus drivers are partially responsible for the challenges they have with students on the buses due to a deficit of skills. This view appears to be corroborated by Ms. Lime, a social worker, who noted that when incidents happen on the buses: The children feel more comfortable bringing it to a staff member than discussing it with the bus drivers or the attendants…they don't feel listened to when they talk to the bus drivers or attendants so they bring it in to the school. School Social Workers. Some social workers shared their own experiences with alienation and exclusion. However, their experience appeared to be qualitatively different from the bus drivers, because unlike the bus drivers who largely operate outside of the schools (albeit still within the school's ecology), the social workers operated within the schools. For these social workers, it was more like a betrayal by their schools. They expressed concerns about their lack of inclusion regarding how their schools' interface with non-profits and outside agencies that provide anti-bullying and prosocial programs. In some cases, they were not introduced to the representatives of the outside agencies. In direct reference to a representative of the agency running the empathy-based program, a social worker told the representative, "I didn't know who you were but you come in my building." In addition to feeling side-lined, they suggested that the decision-making processes about bullying prevention was sometimes power-and clout-based. Mrs. Green shared, "I don't feel like I have any power because when I go to my administrators, sometimes my suggestions fall on deaf ears." In contrast to the feelings of alienation and exclusion shared by some social workers and bus drivers, the parents shared that they had a great relationship with their children's school. The school was responsive and open to their input, and some of the efforts to support bullying prevention through music and entertainment had even been initiated by parents.

Variations in Applying Bullying Policies
All the participants pointed out that the school district had a "step-by-step" policy for addressing bullying. Per the policy, when a report or allegation of bullying is made, an investigation must be conducted. If substantiated, sanctions were to be applied. In some schools, the reports or allegations of bullying were "investigated by a social worker, or by another staff person" (shared by Mrs. Gray-1, a social worker). Mr. Basil, a principal, shared that in his school, "usually, it's me [him] and my social worker interviewing witnesses or targets, and bullies." Some of the social workers felt that "schools are pretty uniform on that policy." However, others indicated that, "there's probably variance throughout the whole process." The principals also indicated that there were some variations and subjectivity (due to differences in interpretation of the policy) in how the policy was applied across the district. The bus drivers and attendants demonstrated awareness of the policy. However, they noted that it was not effectively applied when they reported bullying on the bus. The parents also demonstrated an awareness of their school's policy. Mrs. Red, one of the parents, referenced the application of the policy in a situation that involved her family and that of another student. She stated that "the principal, the teacher, our social workers" worked with both families "to document everything and make sure that nothing was falling through."

Discussion
This study examined and synthesized the perspectives of multiple stakeholdersprincipals, social workers, bus drivers and attendants, and parents -on the incidence of bullying and bullying prevention in an urban school district. The findings demonstrate some striking similarities as well as differences in the perspectives of stakeholders. Overall, the findings support the use of multi-stakeholder approaches in developing a more holistic view of bullying.
Incidents of bullying cause consternation in schools and with parents, yet how stakeholders respond to bullying-actual or perceived-can be a source of further concern (Mehta et al., 2013). In the ecology of schools, where several stakeholders are present, differences in perspective are not uncommon, especially when people have different roles and functions in their unique spaces (Adelman & Taylor, 2000;Beddoe, 2019;Newgent et al., 2009). Considering different perspectives can contribute to building a stronger case about the nature of bullying, especially when perspectives are confirmatory and complementary. On the other hand, such differences can result in tensions and potentially undermine the effectiveness of bullying prevention efforts.
In line with the concept of habitus (Bourdieu, 1984), these differences may be linked to ways stakeholders perceive and understand bullying relative to their positions and experiences. As suggested by some respondents, having a more open system of dialogue among stakeholders could create a healthier way for addressing bullying. With this in mind, even contrasting views can generate a comprehensive and holistic picture (Kim, 2017).
When the study participants discussed bullying, they shared their experiences from different contexts, yet confirmed the ubiquity of bullying-occurring across grade levels, in a variety of settings, and being non-gendered (NASEM, 2016;Robers et al., 2015). However, contrasts in how bullying is perceived also became apparent. The extent to which study participants emphasized the technicalities of bullying and its definition varied.
Our findings show the existence of a district policy that includes a framework for determining if an incident is actual bullying. Through the lens of that policy and the framework, some participants-i.e., principals, social workers, and parents -emphasized bullying as a complex and technical issue. From the principals' administrative standpoint, the district policy shaped how they responded to reports of bullying from parents. Further, from the principals' accounts, which are complemented by some of the social workers and parents, there is an indication that the response and efforts to address bullying are systematic and that an objective outcome is sought. Yet, there is also some indication that the interpretation and application of the policy is characterized by some degree of subjectivity across schools. This is evident in the principals' discourse and what some of the social workers and bus drivers and attendants shared. This finding suggests fidelity in the implementation of the district's policy may be lacking. Fretwell (2015) emphasizes the need for fidelity in the implementation of anti-bullying policies, noting that their effectiveness is hindered when they are not properly implemented.
There may be other motivations for the emphasis on the technicalities of bullying by principals, some social workers, and the parents. The emphasis could be rooted in a deep concern about students and an attempt to avoid or minimize situations where they are wrongly labeled as bullies especially with the increasing shift towards criminalization which could result in involvement with the juvenile justice system (Cornell & Limber, 2015). Also, it is important to recognize that student suspensions could put a strain on school/family relations. Another motivation could be that, for these participants, having a low bullying record may be incentivized because a high prevalence of bullying compromises the image of their schools (Cornell et al., 2013). In comparison, the same cannot be said of the bus drivers and attendants. This could possibly explain why there was limited emphasis on the "technicalities" of bullying or concerns about overestimation in their discourse. For bus drivers and attendants, there may be more of a primary concern with stopping any form of aggressive and disruptive conduct on the bus. Hence, the "technicalities" of conflict or bullying may be secondary considerations in how they discuss bullying.
The findings also show variations in opportunities for stakeholder inclusion in bullying prevention. The inclusion or exclusion of stakeholders in bullying prevention may be determined by several factors including dynamics that may be unique to each school. Of importance is the disposition of the administrative team. Some of the accounts of noninclusion seem to be illustrative of the phenomenon of working in host settings or engaging with external systems (Beddoe, 2019;Dane & Simon, 1991). In reference to social workers, Dane and Simon (1991) define host settings as "organizations whose mission and decision making are defined and dominated by people who are not social workers," and the social workers are seen as "resident guests" (p. 208). This definition also applies to the bus drivers and attendants' engagement with the schools.
Beddoe (2019) notes that marginalization sometimes occurs in host settings. There may also be the delegitimizing of the "guests" experiences and views or attempts by the host agency to highlight the inadequacies of the guests instead of acknowledging their own failings. In this study, one principal felt bus drivers were oblivious to what happened on the buses, raising questions about the credibility of their accounts. It can be argued that principals with views like this may be less receptive to or limit the inclusion of bus drivers in bullying prevention. Other principals suggested that the bus drivers contributed to the problems on the bus due to a lack of skills in engaging with parents and children.
On the contrary, some social workers and most of the bus drivers and attendants perceived principals to be complicit in the prevalence of bullying. This was based on their experiences of alienation and non-inclusion and the view that principals were not taking resolute action resulting in a climate that fomented bullying. Despite this view, it is possible that principals may be working to address bullying; however, they may be constrained by other demands and pressures including limited resources (Lloyd, 2018). Several principals alluded to how they had to provide or coordinate social services for students and their families. Hence, the perceived complicity may be a result of a failure to appreciate the challenges principals and schools face. There may also be real cases where principals are complicit. However, that cannot be considered true for all schools.
Though most of the bus drivers indicated the schools were unwilling to work with them to address bullying, the social workers reported more varied experiences. Some social workers felt included and played an integral role in bullying prevention in their schools while others felt side-lined. The parents interviewed in this study also shared positive experiences of inclusion in their children's school.
From our findings, we conclude that the school district can be conceptualized as a large eco-system, and the individual schools constitute unique eco-systems embedded within a larger one. Thus, the dynamics of each school may be different. For that reason, differences in the experiences of stakeholders are not unusual (Fretwell, 2015;NASEM, 2016). The sample of parents in this study, for instance, had no complaints to make about their school. Though they acknowledged their school was not perfect, they stated that they themselves "live kind of inside a bubble." Again, some social workers reported good experiences, whereas others had challenges in their schools. Likewise, the principals provided some indications that they differed in how they facilitated processes. We also conclude that each group of stakeholders can be viewed as a subsystem of a school's ecology. Some of the interactions between these subsystems are characterized by a lack of openness. That lack of openness can result in fragmented relationships, and consequently, a fragmented understanding of and response to bullying. Conversely, the lack of openness could be the result of fragmented relationships (Fretwell, 2015).
The application of eco-systems theory in this study engenders an understanding of the multilevel factors that impact bullying in terms of its occurrence, as well as responses to bullying. In this study, the impact of the district's policy (a macrosystem factor) on schools and school-family interactions (the mesosystem level) was evident. Though definitive causal inferences about bullying cannot be made in this study, most of the participants believed that bullying was associated with neighborhood and home factors illustrating the perceived impact of the exosystem, mesosystem, and microsystem on student behavior (NASEM, 2016). Using Bourdieu's concept of habitus was also essential in highlighting the unique positions and experiences of the stakeholders (Bourdieu, 1984;Shimoni, 2017). This concept is a helpful guide on how to avoid overgeneralizing and identifying fault lines in bully prevention efforts.

Limitations and Future Research
Some limitations were recognized in this study. Though the study involved different groups of stakeholders which allowed for different perspectives to be garnered, the actual sample size of some of the groups was limiting. For instance, there were only four parents, all from the same school. There was also a lack of gender and racial/ethnic diversity among the different groups of stakeholders. Of a total sample of 45 participants, only four were males. The relative homogeneity of the study's sample has implications for its transferability to other contexts or situations. Lastly, teachers and students' voices were not included in this inquiry. Nonetheless, this study provided an avenue for the inclusion of stakeholders such as bus drivers and attendants and social workers whose perspectives are often not heard.
Another possible limitation involves our positionality in this study. The two lead authors conducted the study as independent evaluators for an agency that was providing an empathy-based bullying prevention program to schools. Though we maintained our professionalism and objectivity, it is possible our affiliation with the agency may have resulted in a desirability bias. It is conceivable that some participants may have provided responses to project a positive image of themselves and/or their schools because of the stigma attached to school bullying (Kaminska & Foulsham, 2013).
Research and evaluation of bullying prevention efforts need to continue as part of initiatives aimed at developing viable solutions (Espelage et al., 2014;Gentle-Genitty, et al., 2017). It may also be necessary to review our notions and definitions of who we consider to be a stakeholder in our bullying prevention efforts (Anderson-Butcher et al., 2006). Future research should further examine the impact of policies on how bullying is defined and the resulting interventions. Finally, future research can offer an in-depth look at the nature of stakeholder interactions relative to bullying.

Conclusion
These findings provide depth over breadth on bullying prevention and may be transferable regarding readers' experiences within urban schools. The study revealed that bullying can be quite nebulous because people tend to look at the issue through the prism of their own experiences and positions. Competing interests through roles and tasks may be experienced, even when trying to solve the same general problem like bullying. Therefore, a contribution from this study is the use of an eclectic and multi-stakeholder approach in the attempt to engender a balanced perspective of bullying and bullying prevention. In working with multiple stakeholders, care must be taken in order not to reify the views and voices of a select few. It is important that schools stay attuned to the unique needs and concerns of different constituents within the school environment to increase buy-in and enhance the chances of success in bullying prevention. Greater inclusivity within bullying prevention must be elevated to better serve stakeholders and the students they support.
The school social workers in this study were integrated in delivering schools' bully prevention activities and therefore must assert to their building principals the importance of being introduced to outside agencies who are invited to provide bully prevention programs. Such introductions create a platform that would enhance an outside agency's ability to understand the language of the school building, what is being done already, and what could be targeted to address needed service gaps.
Though social work practice often involves multiple systems, it is not uncommon to find social workers primarily focusing on direct or micro practice at the individual and small group levels. In the context of schools, this is evident in the attention given to individual and group counseling, behavioral management, etc. (Anderson-Butcher et al., 2006;Soliman, 2017). Greater emphasis on systems-thinking is needed in social work education and practice. School social workers should harness their knowledge of organizational dynamics and their advocacy skills to create open systems of dialogue not only for themselves but other stakeholders. Having open systems of dialogue can help in achieving greater consensus among stakeholders and capitalizing on their knowledge and experiences to ensure better bully prevention outcomes (Anderson-Butcher et al., 2006;Kim, 2017). prevention: Findings from a national education association study. School Psychology Review,42(3), 280-297.