DON'T HATE, JUST MEDIATE UNDERSTANDING THE IMPACT OF A CONFLICT MEDIATION PROGRAM ON ADOLESCENTS' EXPERIENCES WITH CONFLICT, SAFETY AND HELP SEEKING IN URBAN SCHOOLS.

DON'T HATE, JUST MEDIATE: UNDERSTANDING THE IMPACT OF A CONFLICT

MEDIATION PROGRAM ON

ADOLESCENTS' EXPERIENCES WITH CONFLICT,

SAFETY AND HELP SEEKING IN URBAN SCHOOLS

Introduction

School violence including conflict, bullying and peer harassment is a serious

problem for educators and research has consistently documented its harmful effects on

students’ emotional and physical health as well as their ability to learn in school. Federal

education policy highlights the seriousness of the problem by funding extensive research

and data collection on school violence as well as ensuring that schools gather and report

on their safety statistics (Merrell, Gueldner, Ross, & Isava, 2008). While acute incidents of

violence in school have steadily declined since the 1990s, research also indicates that

bullying, intimidation and verbal and physical aggression still represent major threats to

students’ safety especially since they may be more difficult to detect in schools than overt

violence (Mayer & Furlong, 2010). These covert kinds of aggression appear to peak during

middle school and adolescents generally report high exposure to these incidents because,

in part, peers relationships become increasingly important. Furthermore, the

psychological, emotional and physical costs of threats to students’ safety are also serious,

and, if not adequately addressed, may escalate into even more serious conflicts, violence or

even suicide. The present study focuses middle and high school students’ experiences with

the relatively less severe, but more common types of conflict and the extent to which a

conflict mediation intervention may reduce the burden of conflict and violence in schools.

Effective school interventions that reduce violence and improve safety in school

must consider multiple influences on students’ behaviors and responses to conflict. This

study adopts a socialYecological perspective on adolescent development, which provides a

useful framework for organizing the environmental and individual factors that contribute

to students’ choices about how they respond to conflict in schools. Students may respond

to conflict in either aggressive or prosocial ways as well as seek help from adults for peer

related problems. For adolescents in particular, a school’s safety climate, relationships with

teachers, peer and family influences and their own emotions and skills all represent

potential factors that may affect the strategies they adopt for resolving conflicts. Programs

at the middle and high school level need to consider the social context of adolescence in

order to be most effective. While research has contributed to our understanding of school

violence interventions, one criticism of these programs is that they are not adapted to the

local context or the perspectives of the communities that experience the problem (Swearer,

Espelage, Vaillancourt, & Hymel, 2010; Ringeisen, Henderson, & Hoagwood, 2003).

Schools need effective solutions for responding to conflicts and promoting a safe

climate for students. SchoolYwide strategies as well as targeted interventions can reduce

aggressive behaviors and conflicts in school by establishing clear discipline and antiY

bullying guidelines, creating a positive school culture, and teaching students prosocial

skills. However, evidence on the effectiveness of many schoolYwide programs is mixed

and interventions proven in one context may fail in others. Conflict resolution and

mediation programs are one type of antiYviolence programming that has been widely

adopted in schools. Mediation programs in particular help students respond to conflicts in

nonviolent ways by providing them with a safe space in schools where they can work out

problems with their peers and learn new skills for resolving conflicts. While there is

already a large body of research on conflict mediation programs, not enough attention has

been given to the contextual factors that influence the experiences of student disputants

especially in ethnically diverse, urban schools where the potential for conflict is high.

Furthermore, the role of emotions in students’ decisions for handling conflict needs

further exploration.

The Current Study

This study investigated the experiences of middle and high school students in a

conflict mediation program across four schools in a large, urban school district. To my

knowledge, this study is unique because it is the first to investigate the impact of a

universityYschool district, collaborative mediation program that trains undergraduate

students as volunteer mediators and deploys them in schools to help resolve conflicts. In

previous studies, crossYage mediators generally included middle or high school students

serving as mentors to peer mediators (Jones, 2004). However, some research has included

university students who provided mentorship to middle school mediators in a lowYincome

school district in California, but still the focus was on a peer mediation model (LaneYGaron,

1998; Cassinerio & LaneYGaron, 2006). In addition, this study focused on the perspectives

of the disputants rather than mediators including contextual or environmental factors that

influence outcomes, which needs further attention in research on conflict mediation

programs (LaneYGaron & Richardson, 2003). Lastly, student’s emotional experiences have

been neglected in conflict mediation research (Johnson & Johnson, 1996), and this study

represents an attempt to fill some of these aforementioned gaps in the literature.

This study uses a mixedYmethods, multiple case study design in order to

understand students’ exposure to conflict and their experiences in a mediation program

across several school sites. The defining element of a case study in comparison to other

research designs is the bounded nature of the unit of analysis rather than the subject area

of investigation (Merriam, 2002). For this study, the conflict mediation program’s

implementation across four, typical school sites determined the selection of the

participants as well as the study’s focus on student experiences across multiple topic areas

within this bounded system. In

addition, the case study provides several advantages over other research designs for

examining program effects. For example, the case study allows for the mixedYmethods

collection of multiple sources of data in a real world context and provides an inYdepth

understanding of both process and outcomes (Yin, 1992). A case study approach may also

improve the utilization of research findings because it gathers both context and

implementation data, which may be useful for program administrators (Yin, 1992). The

present study attempted to understand a conflict mediation program in a multiple school

context and utilized three, distinct sources of data including classroom and program

questionnaires and participant interviews. The rationale for a mixedYmethods approach

included the ability to compare and enhance the results of one method with another and

to develop quantitative variables from qualitative, contextual information (Schensul,

Schensul, & LeCompte, 2013a). This study had two primary phases that included, first, the

collection of classroomYlevel data on the school context and, second, the collection of preY

post and interview data on the program outcomes and the context of implementation.

Research Questions

This study attempted to understand the experiences of students who participated in

a conflict mediation program and included two main areas of investigation: 1) the school

context and students’ exposure to conflict and 2) the program context and student

perspectives on the impact of the conflict mediation program. Two guiding research

questions represented this study’s primary areas of inquiry:

1. How frequently do students experience conflict or violence and what are their

attitudes about help seeking?

a. Which groups of students might be most at risk for problems associated

with exposure to violence?

b. What differences are there, if any, among the schools?

2. What is the impact of a conflict mediation program on students’ brief, emotional

adjustment, approach to conflict resolution and attitudes about help seeking?

a. How do students describe their experiences in the program and their levels

of program satisfaction?

b. What contextual factors contribute to students’ decisions to resolve

conflicts nonviolently and their willingness to seek help?

This study’s findings are especially relevant to the field of conflict resolution

because of the unique characteristics of the student populations, the school sites, and the

mentoring aspects of the intervention. The mixed methods approach should contribute

important contextual information for multiple audiences including the school site staff,

program administrators, and other potential coordinators who may want to develop

similar mediation programs in other schools sites that face similar challenges. In

summary, this study’s findings support three main conclusions about factors that influence

students’ responses to conflict. First, reductions in feelings of negative emotions including

anger are related to students’ abilities to adopt nonviolent, conflict resolution strategies.

Second, the use of undergraduates as program facilitators allows students to form positive

relationships with older students who provide important mentorship to them. Third, the

environmental context in terms of school or adult response to conflict and peer culture

both strongly influence student’s decisions about how they resolve conflict nonviolently.

Finally, I have shared some of these findings with the current cohort of UCLA mediators

and will share results with each of the school sites as well as next year’s class of mediators.

Literature Review

Violence, Bullying and Conflict in School

Definition of the problem. This study draws on several traditions in the literature

of examining student safety and school violence. For example, three different but related

perspectives in school safety research include investigations of school violence, bullying

and conflicts, which all represent various threats to student safety and range in severity

from high to low. School violence falls along a continuum of negative behaviors that range

from rare, “highYlevel” incidents such as school shootings, homicides, or other serious,

violent crimes on one end to chronic, “lowYlevel” behaviors like fighting, bullying,

intimidation, harassment and other kinds of incivility on the other side (MeyerYAdams &

Conner, 2008). Researchers have also defined violence in school as the act or intent of

harming, damaging, threatening, or intimidating another person through the use of

physical force (Yoon, Barton, & Taiariol, 2004). While physical violence represents a

severe threat to safety, lowYlevel incidents like bullying and other types of conflicts are the

major concern of this study. The concept of bullying, which researchers believe captures

the majority of lowYlevel incidents of violence, is defined as frequent, aggressive behavior

directed toward a peer over time that is often unprovoked, but also includes an imbalance

of power between the bully and the victim (Johnson & Johnson, 2007; Olweus, 1996;

Olweus & Limber, 2010). While bullying may also represent a type of conflict, conflict

appears distinct from bullying in that it is a natural part of all social life and appears in

positive relationships as well as negative ones (Johnson & Johnson, 1996). However, there

is also considerable overlap between the two and researchers have generally viewed

conflicts in schools as incidents of verbal and physical aggression, which can include

arguments, gossip, fighting, harassment, dating and various kinds of relationship issues

(Johnson & Johnson, 1996). Similar to physical violence, the consequences of bullying and

conflict are significant and appear to lead to more severe conflicts or violent acts.

Researchers have attempted to understand the relationship between highY and

lowY level violence and found some evidence that less severe, but more common types of

aggressive behaviors increase the odds of committing more serious and violent crimes

(MeyerYAdams & Conner, 2008; Nansel, Overpeck, Haynie, Ruan, & Scheidt, 2003). Even

among lowYlevel conflict, social aggression can precede and lead to physical violence

(Talbott, Celinska, Simpson, & Coe, 2002). While lowYlevel violence may not be as harmful

or traumatic as highYlevel crime, researchers believe aggressive behaviors such as bullying

and peer victimization have lasting, negative effects on students’ mental health and

academic functioning (Cornell & Mayer, 2010). In fact, recent research indicates that these

lowYlevel conflicts are responsible for a greater share of psychological burden such as

anxiety, fear or avoidance, and are even harder for schools to address than more visible

types of violence such as assaults (Mayer & Furlong, 2010). Schools and educators require

effective approaches for reducing all kinds of violence in schools, but may need to be

especially vigilant when it comes to elusive, lowYlevel violence such as bullying and other

forms of peer harassment, victimization and conflict. For the present study, “bullying” is

sometimes used interchangeably with “conflict” as it has received considerable attention in

the literature and both terms share the lower end of the school violence continuum.

Prevalence rates. Just how frequent is lowYlevel violence in schools? Bullying,

which includes verbal, physical and psychological aggression, attacks, threats, teasing,

personal theft and social exclusion, represents one of the most common forms of lowYlevel

violence. Several national surveys as well as many smaller scale studies suggest that these

behaviors are common in schools (Farrington & Ttofi, 2009; MeyerYAdams & Conner,

2008). Nansel and colleagues (2001) reported one of the earliest estimates on the

prevalence of bullying using World Health Organization survey data from 1998. They found

that 30% of US students in grades 6 – 10 reported some involvement in bullying as a

victim, bully, or both. In their study, middle school students also reported even higher rates

of bullying than high school students. Similarly, survey data from the Indicators of School

Crime and Safety (ISCS) in 2005 showed that about 32% of students, ages 12 – 18, reported

being bullied at school while 43% of middle schools reported weekly incidents of bullying

(Dinkes, Cataldi, & LinYKelly, 2008). Some recent research has found alarming high levels

of bullying with rates reaching nearly 70% for any type of involvement (bully or victim) in

one sample of 1,100 students from nine elementary and middle schools (Swearer, Wang,

Maag, Siebecker, & Frerichs, 2012).

Recent national survey data also find similar rates although these surveys included

only victim reports of bullying (e.g., called names, insulted, target of rumors, teased,

pushed, threatened, excluded, property stolen or destroyed, etc.) rather than being a bully.

The 2011 Youth Risk Behavior Survey (YRBS) found that 20% of high school students

reported bullying at school and 16% experienced bullying on the Internet, or

cyberbullying, a newly added survey item (Centers for Disease Control and Prevention,

2012a). The YRBS with middle school students found that 44% of them reported bullying

at school and 23%

reported cyberbullying (Centers for Disease Control and Prevention, 2012b). In 2009, the

ISCS survey indicated that approximately 28% of students, ages 12Y18, reported

experiences of bullying, which peaked in 6th grade around 39% and decreased each year

through high school to around 20% in 12th grade (Robers, Zhang, & Truman, 2012). Across

studies, lowYlevel, violent behaviors and conflicts appear to be widespread, affecting

anywhere from oneYthird to threeYfourths of all students, and reach their highest point

during middle school (Bauer, Lozano, & Rivara, 2007; Goldstein, Young, & Boyd, 2007;

Johnson & Johnson, 2007; Nansel et al., 2001; Nansel, Haynie, & SimonsYMorton, 2003).

Consequences of conflict. The accumulation of research evidence suggests that the

experience of peer victimization has negative and persistent consequences for adolescents.

Involvement as a bully, victim or both has been associated with various problems such as

depression, anxiety, low academic achievement, distress and substance use (Hertz, Donato,

& Wright, 2013). Early research also found negative effects of bullying on students’ social

functioning. For example, involvement with bullying had associations with difficulties

making friends, poorer relationships, increased fighting, and feelings of loneliness or

isolation (Nansel et al., 2001). Other researchers also documented similar consequences

for bullies and victims such as reporting less life satisfaction overall and receiving less

social support from their teachers and peers (Flaspohler, Elfstrom, Vanderzee, Sink, &

Birchmeier, 2009). Moreover, these researchers found that students who reported both

victim and bully status had worse outcomes than either one alone. Researchers also found

that student bullies were at greater risk than bully victims for exhibiting more serious

forms of violence such as physical fights and carrying weapons to school, although, on the

other hand, they also reported less difficulty with making friends (Nansel, Overpeck, et al.,

2003). In addition to the negative effects on social functioning, researchers have also

examined the effects of bullying on student perceptions of school climate. Students

involved in bullying or relational aggression report more negative views about the climate

including lowered perceptions of safety (Goldstein et al., 2007; Nansel, Haynie, et al.,

2003).

While most studies on the consequences of bullying are crossYsectional and capture

single points in time, longitudinal findings also support the longYterm impact of these

harmful experiences. For example, Nansel, Haynie and colleagues (2003) found evidence in

their middle school sample that 6th graders involved in bullying continued to report lower

levels of psychosocial adjustment well into the end of 7th grade. Similarly, Juvonen, Wang,

and Espinoza (2010) found that being a victim of high levels of bullying behaviors was

significantly associated with lower levels of academic engagement and grades over the

course of the three years of middle school measured during the study. Given the robust

findings of negative consequences across psychological and academic domains for both

bullies and victims that persist over time, schools would certainly be remiss and could even

be considered legally neglectful of children’s wellbeing if they ignore these problems of

lowYlevel violence, bullying and harassment in schools (Mayer & Furlong, 2010).

Contributing Factors to Students’ Responses to Conflict

Theoretical framework: socialKecological model of conflict. In order to

understand the various influences on youth and conflict in schools, researchers

recommend the adoption of a socialYecological perspective on student behavior, which

highlights the interactions in the environment between students and their peers, teachers,

schools, parents and community members (Swearer et al., 2010). According to this

framework, students’ perceptions of the surrounding context, such as the family,

classroom, school or community, will have an effect on their behavior, motivations and

overall adaptation in the larger social setting (Eccles & Roesser, 2009). For example, if

students perceive the school environment as one where adults are responsive to incidents

of violence, they are more likely to feel safe, act less aggressively, and perceive the climate

more positively. Of particular interest in this study are the social and individual factors

that influence how adolescents choose from a range of potentially equally valid responses

to conflict. According to various theories of conflict strategies, individuals have several

different ways they can respond to conflicts, which fall along a continuum of competitive

behaviors on one end to cooperative actions on the other (Johnson & Johnson, 1996; LaneY

Garon, 1998). On the competitive side, youth may use verbal or physical force, coercion,

manipulation, withdrawal or avoidance, and on the cooperative side, they may exhibit

appeasing, compromising, problemYsolving or negotiating kinds of activities. In theory,

youth confront situations when any of these strategies could represent a valid response to

the conflict. However, an explicit goal of conflict education programs is to assist youth with

developing skills in problem solving and negotiation, as “untrained” students tend to

gravitate towards more competitive responses (Johnson & Johnson, 1996). An integration

of socialYecological and conflict strategy theories would then suggest that multiple factors

influence students’ decisions to choose an appropriate strategy for handling conflict, which

may vary depending on the context. For example, previous research has examined

individual and environmental factors that influence urban, adolescents decisions about

how to respond to conflict nonviolently (Farrell et al., 2008; Farrell et al., 2010). At the

individual level, youth reported that problemYsolving and emotional regulation skills,

beliefs, values, goals, perceived consequences and attributions are all factors that

contribute to their conflict choices. Youth also reported multiple environmental factors that

affected their decisions such as family endorsement and messages about fighting, parental

modeling, peer pressures, supervision at school and exposure to neighborhood violence.

Taken together, this research suggests that while cooperative strategies for conflict are the

most desirable for reducing violence from a school’s perspective, it also indicates that

various influences in the environment may contribute to or even endorse competitive or

aggressive approaches as well (Anderson, 1999). This expanded view of development has

implications not only for understanding the transactional relationship between student

behaviors and the surrounding context, but it also suggests the necessity of intervening at

different levels of the environment as well as the individual student level in order to reduce

conflict and violence in schools (Mayer & Furlong, 2010).

School climate and adolescent outcomes. School climate is a key aspect of the

social environment for students and has a range of effects on various outcomes related to

learning, socialYemotional adjustment, behavior and physical safety. School climate may be

difficult to define, but some researchers have described it as the nature of the relationships

among school staff and students including how often and how well they interact with one

another (Kuperminc, Leadbeater, Emmons, & Blatt, 1997). Other research has identified

the most important dimensions of school climate, which include academics (teaching and

learning), safety (physical and emotional), relationships (students and teachers), the

physical environment and the school improvement process (Cohen, Mccabe, Michelli, &

Pickeral, 2009; Thapa, Cohen, Guffey, & HigginsYD'Alessandro, 2013). All of these

dimensions have some implications for conflicts in schools, but the most relevant aspects

for the current study include the safety and relationships. Within these two domains, a

school’s conflict resolution “climate” can be thought of as the conflict dynamics within a

school, which includes the amount of conflict among students, between students and

teachers as well as the school staff’s response to conflict in each of these relationships

(LaRusso & Selman, 2011). From a socialYecological perspective, school climate should

play a significant role in influencing student approaches to conflict. However, despite the

evidence linking positive climate to reductions in violence well as improved learning

outcomes, the application of these research findings to practice has been slow and schoolY

based interventions still lack climate components (Cohen et al., 2009).

Research evidence points to a relationship between a school’s safety climate and

various adolescent outcomes such as learning, socialYemotional adjustment and risk

behaviors. For example, in a large study of middle school students, Brand, Felner, Shim,

Seitsinger and Dumas (2003) found that student reported indicators of mental health such

as low selfYesteem, depression and anxiety were significantly associated with negative

views of school climate after controlling for schoolYlevel differences in socioYeconomic

status (SES). In terms of school climate and learning, the authors also found that student

academic achievement was related to specific subYdimensions of school climate including

safety and the quality of peer relationships. Other research also supports a relationship

between school safety and learning. Using 10th grade data from the Educational

Longitudinal Study of 2002, Ripski and Gregory (2009) found that student reported

perceptions of hostility or lack of safety in the school’s climate were associated with lower

academic engagement levels and reading achievement. Other research has also found

correlations between negative perceptions of school climate and internalizing and

externalizing problems and increases in riskYtaking behaviors (Kuperminc et al., 1997;

LaRusso & Selman, 2011). School climate in general and safety in particular appear to be an

important part of the social environment not only for adolescents’ adjustment and

learning, but also as a potential contributing factor to their response to conflict.

The effects of school climate on student responses to conflict. Other important

research has investigated the relationship between school climate and how students

choose to respond to their own conflicts. Certain aspects of a school’s climate such as levels

of aggression, interpersonal relationships and school policies or norms about conflict may

have important effects on how students decide on a strategy for resolving conflict. In

addition, while students can respond to conflicts in either competitive or cooperative

ways, they may also choose to seek help from adults in situations when the stressors of the

conflict exceed their capacity to handle it on their own (Newman, 2008). Importantly, if

students are willing to seek help for conflicts, this may also contribute to an atmosphere

with less tolerance for low level violence and an improved climate of safety overall.

Therefore, this next section will consider how school climate influences students’

aggressive or nonviolent responses and their attitudes about seeking help for conflicts. I

will then end with a discussion of how a school’s response to conflict, particularly in terms

of its cultural norms and discipline policies, affects these student behaviors.

School climate and students’ responses to conflict. School climate appears to have

an effect on student’s behaviors and a negative climate may predict or increase the

likelihood that students will engage in more acts of aggression or competitive responses to

conflict. Many studies looked at the relationship between school climate and student risk

behaviors such as aggression or delinquency and a few exceptions also considered how

climate might influence a student’s response to conflict more specifically. Early research

found significant associations between student perceptions of negative school climates

and behavioral problems such as defiance, acting out, aggression, and delinquency (Kasen,

Johnson, & Cohen, 1990; Kuperminc et al., 1997). More recent studies have also found

similar results where a positive school climate was associated with fewer risk behaviors

and better emotional adjustment (LaRusso & Selman, 2011; Way, Reddy, & Rhodes, 2007).

However, while the measures of risky behaviors included aggressive behaviors and

fighting, they were not disaggregated from other kinds of risk taking such as breaking

rules, drugs or sex. Furthermore, the direction of effects in many of these studies must be

considered “tentative” since most of the data is crossYsectional even though some studies

attempted to determine directionality with advanced, multiYlevel models.

Other research has examined the relationship between school climate and students’

approaches to conflict more directly. One study considered how several factors including a

school’s norms about aggressive behavior, interpersonal relationship climate among

teachers and students, and the school’s response to violence might affect student’s

individual levels of aggression and selfYefficacy for nonviolent responses to conflict

(Henry, Farrell, Schoeny, Tolan, & Dymnicki, 2011). The results indicated that school norms

against aggression, positive interpersonal climate and greater teacher awareness of and

reporting of conflict all predicted lower physical aggression and greater selfYefficacy for

nonviolent responses at the individual, student level. The authors also indicated that these

school level effects were stronger for girls than boys and became weaker from 6th to 8th

grade, which suggests that effects of school climate may decrease over time. Similarly,

other research has also found an interactive relationship between student perceptions of

school climate and students’ responses to conflict that appears to change over time.

MeyerYAdams and Conner

(2008) looked at how school climate, or the psychosocial environment, was related to

middle school student perceptions of safety and their aggressive or avoidant responses to

bullying. They found that a negative school climate was associated with a greater likelihood

of aggressive and avoidant responses to conflict, which also appeared to be cyclical in

nature. As students became increasingly bullied or victimized at school, they reported

more negative views of school climate. As student perceptions of school climate decreased,

in turn, the students were more likely to respond to harassment in aggressive or violent

ways, which also further compromised students’ ability to learn in school. Consistent

across several of these studies is that student perceptions of climate are not fixed, but,

beginning in 6th grade, decrease over time across the middle school years (Way et al.,

2007). While school climate appears to exert an influence on students’ response to conflict,

it also seems important for schools to implement approaches to improve their safety and

relationship climates as early as possible and at multiple levels of the environment in order

to reduce or prevent lowYlevel violence in schools.

School climate and help seeking. Help seeking represents an alterative response

to peer conflicts that not only is nonviolent, but also may be constructive if it helps

students reduce stress and preserve their relationships with peers or adults. In general,

anywhere from a quarter to one third of students have sought help for bullying conflicts

and about 10% of students report they would seek help for personal problems (Eliot,

Cornell, Gregory, & Fan, 2010). Students may be reluctant to seek help for a variety of

reasons and there are a host of potential factors that influence their decisions to seek help

including the severity and chronicity of the conflict, the availability of supportive helpers

or teachers, popularity status, social competence and past experiences with seeking help

(Newman,

2008). In early adolescence, girls and younger students are also more likely to seek help

than boys and older students and there appears to be some disagreement about whether

more victimization is also associated with increased help seeking (Hunter, Boyle, &

Warden, 2004; Unnever & Cornell, 2004; Williams & Cornell, 2006; Yablon, 2010). Very few

studies have considered race and help seeking explicitly, and when they do discuss race, it

is usually in a postYhoc manner with no consistent findings among studies, although one

study suggested that AfricanY and NativeYAmerican students might have the lowest scores

(Eliot et al, 2010; Unnever & Cornell, 2004; Williams & Cornell, 2006). Despite individual

differences in help seeking, students are more likely than adults to be reliable informants

of bullying and threats of violence in school, although they may be much less likely to seek

help when they perceive their school climate and relationships with teachers as

unsupportive (Eliot et al., 2010; Greene, 2003). Thus, in addition to individual factors,

school climate can have a direct impact on students’ willingness to report conflict to adults.

Similar to student strategies for resolving conflicts, researchers have also

investigated how school climate may affect students’ willingness to seek help for exposure

to conflict, threats or victimization in schools. A school’s safety climate in terms of the

prevalence of bullying and other forms of conflict appears to have a relationship to

students’ willingness to seek help. For example, various studies have found that the

presence of bullying and teasing as well as aggressive attitudes at the school level are

negatively correlated with students’ help seeking attitudes (Unnever & Cornell, 2004;

Bandyopadhyay, Cornell, & Konold, 2009; Klein, Cornell, & Konold, 2012). In addition, low

scores on a measure of school disorder, which included disciplinary incidents, reports of

bullying and violence and students’ aggressive attitudes, was associated with high scores in

willingness to seek help among 9th grade students (Bandyopadhyay et al., 2009). Other

research examined school climate in terms of its tolerance for bullying, which included

norms about the acceptability of conflict and teacher awareness of and response to conflict,

and found that students were more willing to report problems when schools had climates

less tolerant of bullying (Unnever & Cornell, 2004; Williams & Cornell, 2006). Lastly, a

supportive school climate, as defined by student perceptions that teachers are caring,

respectful and interested, had a significant relationship with students’ willingness to seek

help for bullying and threats (Elliot et al., 2011). Taken together, these results suggest that

an important part of the relationship between school climate and help seeking is how

schools establish expectations about conflict. Considering that less than 25% of students

seek assistance with their experiences of victimization, schools that change how they

respond to conflict by improving their collective responsibility for violence and bullying

may improve help seeking attitudes and reduce the amount of aggression or bullying that

goes unreported in schools (Eliot et al., 2010; Henry et al., 2011).

School responses to conflict. Schools have a variety of means at their disposal for

responding to student conflicts and lowYlevel violence, which certainly has an impact on

students’ decisions about how to handle conflict, whether they seek help for problems and

their overall perceptions of the climate of the school. Seemingly common, but

inappropriate behaviors like nameYcalling or spreading rumors, if ignored by adults or

other students, can escalate into more severe conflicts with even greater consequences

(Talbott, 2002; Theberge & Karan, 2004). Schools have traditionally responded to

disruptions with punishments that exclude students from the learning process such as

office referrals, suspensions or expulsions, which may exacerbate student delinquency

rather than address

the underlying causes of the behaviors (Osher, Bear, Sprague, & Doyle, 2010). For example,

repeated suspensions can increase the risk of repeating a grade, dropping out of school,

and entering the juvenile justice system for atYrisk students (Fabelo, 2011). However,

some studies find that, in certain situations, students approve of these kinds of discipline

policies and see them as fair or effective punishment for aggressive behaviors, although

this may only be true in schools with already high suspension or expulsion rates (Talbott,

2002).

Furthermore, students also acknowledge that knowing about the consequences for

breaking rules acts as a real disincentive for fighting or responding to conflict aggressively

(Farrell et al., 2010). In addition, exclusionary discipline policies may discourage rather

than support students’ efforts to seek help for conflicts. Given these effects of school

discipline on student behaviors, it is important for schools to consider the impact of their

responses to conflict on students’ own decisionYmaking processes for resolving problems

especially when the majority of conflicts never reach the attention of adults.

As noted earlier, researchers have found that cultural norms against violence and

effectively responding to violent incidents reduces the levels of aggression and violence at

school (Henry et al., 2011). Schools appear to communicate and establish their culture and

norms about conflict through the actions of their teachers and, most importantly, their

relationships with students. Just as peer bystanders play a role in bullying and harassment,

adult bystanders who do not actively intervene in conflicts also convey an implicit message

of tolerance for aggressive behaviors (LaRusso & Selman, 2011). Teachers who receive

training in awareness of conflict and how to respond to it are more likely to intervene in an

effective manner, which also promotes a more positive and safe school culture (Greene,

2003). Other kinds of teacher behaviors such as forming close relationships with students,

providing support for nonviolence and close supervision of spaces in schools also helps

students with positive conflict resolution (Farrell et al, 2010). Furthermore students are

more likely to endorse appropriate conflict resolution strategies in schools with positive

climates where teachers listen and negotiate with students rather than coerce, intimidate

or engage in power struggles with them (LaRusso & Selman, 2011). When students

primarily view their teachers as evaluators or disciplinarians, they are also much less

likely to seek help for problems from them (Lindsay, 1998). In conclusion, school discipline

policies and school climates appear to influence how students respond to and seek help for

conflicts. Teacher behaviors and relationships with students also appear to be critical

components for establishing a positive school climate that supports students’ use of

prosocial conflict strategies.

Peer & family influences. Peers and families represent significant relationships in

the social milieu for adolescents and can potentially influence their responses to conflict in

positive or negative ways. Adolescence is a developmental time when the peer group in

particular becomes increasingly important for social status, feeling accepted and

establishing independence from adults (Espelage, 2002; Espelage & Swearer, 2003).

Because of the greater prominence of peers, interpersonal conflicts also increases and

youth learn to acquire new developmental and social cognitive skills such as managing

peer conflicts, problem solving, perspective taking, emotional regulation, goal formation

and negotiation (Yoon et al., 2004; LaneYGaron, 1998; Gerrard & Lipsey, 2007). However,

peers also appear to strongly influence aggressive and bullying behaviors of adolescents.

One theory on social dominance suggests that aggressive students or bullies do not lack

social skills, but seek to increase their status amongst their peers by harassing other

students, which may occur most often in transitions to new settings such as middle school,

where students attempt to reestablish dominant or popular peer groups (Espelage &

Swearer, 2003; Juvonen & Graham, 2014). Another theory based on the stage fit of

adolescence suggests that students form peer groups based on similarity and, as a result,

aggressors or bullies would negatively influence one another and engage in similar levels

of aggressive behaviors (Espelage & Swearer, 2003). Furthermore, conflict is often a group

activity where peer bystanders play a central role in influencing students’ aggressive

behaviors by either aligning themselves with an aggressor or deliberately choosing not to

aid a victim (Greene, 2003; Oliver & Candappa, 2007). Peer bystanders may also be

interested in their own status or reputation and associate themselves with a more popular

aggressor especially if the victim is perceived as weak and submissive (Juvonen & Graham,

2014). Peers also influence students’ willingness to seek help for conflicts and often in

negative ways. Qualitative research findings reveal student perspectives on how peers

norms make it difficult for them to seek help for conflicts at school. Adolescents report

elevated concerns over their image and reputation and feel considerable pressure to

demonstrate autonomy and selfYreliance while avoiding the appearance of weakness. As a

result, students tend to hold somewhat negative attitudes towards seeking help for

problems and may feel shame for turning to adults for help (deLara, 2012). However, in

one study, adolescents did not report any drawbacks for speaking with a confidential

source of help, which, when available, respects their need for privacy and provides

reassurance that adults will not become overinvolved (Oliver & Candappa, 2007).

While family influences are not a central part of this study, parental attitudes about

conflict and aggression represent another aspect of the social environment that may affect

adolescents’ responses to conflict. There is large literature on how high risk neighborhoods

influence parental socialization and childYrearing practices. In terms of messages about

conflict and aggression in these settings, again, qualitative studies help illuminate the

various ways parents may influence their adolescent children’s behaviors. Parents

implement several strategies to help protect their children from dangerous situations and

avoid violence. Parents actively monitor their children’s activities and friendship groups,

model nonviolent strategies for resolving peer conflicts, provide support or guidance for

how to deescalate and avoid conflict, and set disciplinary guidelines for involvement in

violence (Johnson, Cheng, Finigan, Bradshaw, & Haynie, 2012; Farrell et al., 2010). In urban

communities with high levels of violence, parents may also view conflict and violence as

unavoidable and endorse aggressive or physical responses to conflict when their child was

the victim of violence first, although avoidance was almost always the preferred response

(Johnson et al., 2012). In one study, as many as 25% of adolescent respondents indicated

that they received explicit messages from family members sanctioning aggression as a valid

response to conflict (Farrell et al., 2010). Furthermore, youth report that a lack of family

support could also negatively influence their response to conflict if they believed their

parents were not concerned about them or would not get involved (Farrell et al., 2010).

These familial messages about how to respond to conflict are important influences to

consider from an ecological perspective on adolescence and conflict resolution.

Individual factors in student responses to conflict. There are a number of

individual factors that influence students’ responses to conflict situations and whether they

seek out help for these problems. Furthermore, depending on the theoretical lens through

which conflict dynamics are examined, any number of these factors may surface as more

important or central to how students ultimately go about responding to and negotiating

conflicts with their peers. While an exhaustive review of the various psychological theories

on cognitive development and social information processing is beyond the scope of this

study, taking a closer look at some implications of theory from the literature on conflict

resolution in schools may be more instructive for highlighting important, individual

characteristics. Conflicts may be resolved competitively or cooperatively through a

continuum of lowY to highYlevel strategies that includes impulsive actions, unilateral

decisionYmaking, reciprocal exchanges, or mutual collaboration (LaneYGaron, 1998). In

addition, two major motivations for resolving conflict include achieving one’s goals or

interests and preserving some degree or quality of a relationship with the other person

(Johnson & Johnson, 1996). The importance each of these motivations and the level of

development of the disputants both contribute to the way an individual student would

choose to respond to the conflict. For example, older students, students with higher

cognitive or social skills or students with a history of friendship are more likely to use

higherYlevel strategies for resolving their conflicts. In various ways, gender also appears to

be related to the underlying motivations and selection of the conflict approach (Johnson &

Johnson, 1996). Other empirical work seems to support this theory of conflict resolution

strategies with adolescents. In their own words, youth indicate that many different factors

influence whether or not they would respond to conflicts nonviolently including their

problemYsolving and emotional regulation skills, selfYefficacy for a prosocial strategy,

goals and values, perceived consequences of an aggressive response, relationship history

and attributions of the other person’s behavior (Farrell et al., 2008). Out of the variety of

potential factors that may contribute to a prosocial response, problemYsolving and

emotional regulation skills and relationship preservation goals or values were some of the

most influential factors for the students in that study.

In fact, research has established an empirical link between conflict and emotions. In

a recent study with adolescents in urban, atYrisk settings, researchers found that low

coping skills for anger and sadness had associations with higher rates of both physical and

relational types of aggression (Sullivan, Helms, Kliewer, & Goodman, 2010). In addition, the

authors also found that youth who were most atYrisk for aggressive behaviors struggled

with both low coping skills and low expression of their emotions. In Farrell and colleagues’

(2008) interviews with urban, AfricanYAmerican teens about their conflict choices, they

observed that being able to regulate emotions was an important precursor for avoiding

fights and physical violence. While this also supports early research on the empirical

relationship between emotion and aggression, other researchers have examined how

responding to peer conflict or harassment by seeking help may also alleviate negative

feelings. For example, the act of seeking help is one way of coping with an external stressor,

such as a peer conflict, which in itself may reduce anger, anxiety or fear (Newman, 2008;

Oliver & Candappa, 2007). Students also appear to be more likely to seek help in situations

when they have high emotional responses to conflict, which could also be partially

responsible for noticeable changes in feelings after they confide in another person (Hunter

et al., 2004). In an effort to build on these research findings, the current study attempts to

measure changes in student’s emotions as well as their endorsement of or belief in a

nonviolent response to conflict as important indicators of the impact of a conflict mediation

program on middle and high school students.

Interventions to Reduce Conflict in Schools

Interventions that attempt to reduce school violence and bullying can also be

organized around a socialYecological framework. According to this heuristic, in order to

maximize effectiveness, interventions should incorporate components from each level of

the social environment such as individual skill building; peer involvement; teacher training

or classroomYbased approaches; schoolYwide policies and norms including discipline,

parent and community involvement strategies; and culturally specific adaptations

(Espelage & Swearer, 2003; Swearer et al., 2010). However, as one might expect, the extent

to which interventions are designed to include multiple levels or more than one program

element is fairly limited. Most antiYviolence programs primarily target individuals or

groups with skill development, but others may attempt to change the culture of an entire

school. In individualYbased approaches, teachers, counselors, or trained facilitators may

use antiYviolence curriculums or other methods to teach empathy, proYsocial or conflict

resolution skills (Espelage & Swearer, 2003). SchoolYwide approaches can focus on

universal prevention programs, discipline policies, awareness campaigns and behavior

modeling such as the Olweus Bullying Prevention Program (OBPP) or the PeaceYBuilders

Program (Flannery et al., 2003; Olweus & Limber, 2010).

SchoolKwide, antiKviolence/bullying programs. The research base on schoolY

wide, antiYbullying programs has reached sufficient size in order to justify several metaY

analyses on program effectiveness. Across most studies, the evidence appears to be mixed,

with many studies reporting null effects or little reduction in student reports of

victimization (Farrington & Ttofi, 2009; Merrell et al., 2008; Smith, Schneider, Smith, &

Ananiadou, 2004; Swearer et al., 2010). Most studies report the largest changes in

students’

attitudinal measures regarding bullying, conflict, or violence, and show the least amount of

evidence of change on measures of behavior (Merrell et al., 2008). At best, schoolYwide

antiYbullying approaches were found to reduce incidents of bullying by as much as 20%

(Farrington & Ttofi, 2009). However, even if programs had positive effects on some

measures, effect sizes were generally too low to deserve recommendation (Merrell et al.,

2008). Effective elements of programs seemed to include parent and teacher trainings,

classroom management intervention, increased supervision, and effective discipline and

antiYbullying policies. The involvement of the peer group, which was recommended in the

literature by some researchers, could even have possible iatrogenic effects (Farrington &

Ttofi, 2009). In a randomized control trial of the OBPP, oneYthird of the sample reported

experiences of bullying; however, researchers did not find reductions in reports of

victimization except among white students nor any significant, positive changes in

indicators of school culture such as safety, support or engagement (Bauer et al., 2007). Of

course, it is concerning that schoolYwide approaches, which in theory should adhere to an

ecological view of student development, do not demonstrate encouraging results. More

recent reviews suggest that the latest iterations of these programs include methodological

improvements and show some potential for reducing bullying or aggressive behaviors,

although several of them have not yet entered US schools (Juvonen & Graham, 2014).

Researchers have several hypotheses about what may contribute to the inconclusive

evidence on schoolYbased, antiYbullying programs including barriers to successful program

implementation and methodological problems in the evaluation research. One of the most

common implementation barriers observed across studies was fidelity to the original

intervention model. When researchers attempted to measure fidelity with mixedYmethods,

observations, or qualitative interviews, they often found that program implementation and

integrity were not always consistent among schools (Bauer et al., 2007; Smith et al., 2004).

Also, many of the schoolYwide, antiYbullying interventions targeted all students when, in

reality, a much smaller percentage (10 – 20%) of students were responsible for most of the

aggressive or bullying behaviors (Swearer et al., 2012). Other researchers have noted that

a student’s risk level for involvement in antisocial behaviors at school including violence

can moderate the effects of the intervention (Wright, John, Livingstone, Shepherd, & Duku,

2007). In other words, lowYrisk students may benefit more from the program than highY

risk students. While not an entirely surprising finding by itself, these researchers also

noted that future interventions should include strategies for engaging highYrisk students,

or those more likely to engage in aggressive behavior. Lastly, methodological limitations

often hamper researchers’ ability to detect real effects. For example, the majority of studies

measured outcomes with student selfYreport and neglect to include other methods or

informants such as parents, family members, adult staff at school or other members of the

community (Coyle, 2012). Other criticisms include the failure of interventions to faithfully

adopt a socioYecological perspective by not considering the cultural backgrounds of

vulnerable groups and not measuring effects at other levels of the environment such as the

school climate or neighborhood (Swearer et al., 2012). These difficulties with the

implementation and assessment of schoolYwide programs also raise concerns about the

reliability and effectiveness of these programs across multiple settings (Chambless &

Hollon, 1998). In an attempt to address some of these methodological problems, the

current study uses a mixedYmethods, case study approach to understand the context or

conditions that could potentially moderate the impact of the program.

Conflict resolution education landscape. Conflict Resolution Education (CRE)

programs may offer some advantages over schoolYwide, antiYbullying programs. Not only

can conflict programs help reduce incidents of low level violence and aggression in schools,

but they can also support a schoolYwide strategy for implementing positive alternatives to

traditional discipline practices (Smith, Daunic, Miller, & Robinson, 2002). CRE offers many

of the same core features as positive behavior supports (PBS) and socialYemotional

learning (SEL) approaches, which have garnered significant research and community

support over the last several years (Smith et al., 2002; Osher et al, 2010). For example, CRE

may offer both preventative and targeted components for addressing behavioral problems

similar to PBS as well as teach students important affective and cognitive skills to improve

selfYdiscipline as in SEL approaches (Osher et al, 2010). CRE as a schoolYbased

intervention has grown in popularity in the last decade as researchers now estimate that as

many as 25% of all schools (or, about 20,000 out of 85,000 schools) across the country

utilize some form of CRE to address the concerns of violence and conflict among students

(Jones, 2004; Garrard & Lipsey, 2007). CRE can be divided into three, broad approaches to

teaching conflict resolution skills including 1) direct skills instruction, 2) curriculum

integration, and

3) conflict mediation programs (Garrard & Lipsey, 2007). In addition, it may be helpful to

identify the four, core characteristics of CRE programs, which include one or more of the

program strategies as outlined above, the grade level, the target population and the

program implementation features of fidelity, maintenance, and coordination (Jones, 2004).

Direct skills and conflict mediation approaches may be universal or targeted, while

curriculum approaches are generally universal. The strategies may involve teaching

students socialYemotional or conflict management skills either through a specialized

curriculum or by diffusing conflict resolution principles into general education classes.

The conflict mediation program in this study falls under the third area of

mediationYbased programs. However, regardless of the exact intervention model, all CRE

programs share four broad objectives, which include creating a 1) safe and 2)

constructive learning

environment, 3) improving the socialYemotional development of students, and 4) building a

larger community based on principles of resolving conflict constructively (Jones, 2004).

In general, CRE programs have received considerable attention in the literature as a

way of combatting violence, pervasive bullying, and peer conflicts in schools (Johnson &

Johnson, 1996; Garrard & Lipsey, 2007). Several benefits have been associated with

schoolY based, CRE programs including improved interpersonal interactions, conflict

management skills, and school climate; as well as reductions in incidents of fighting,

bullying, and discipline referrals (Johnson & Johnson, 1996; Garrard & Lipsey, 2007; Smith

et al., 2002). In addition, when students are able to resolve their own problems, teachers

spend less time managing conflicts in the classroom and student perceptions of school

safety improve (Heydenberk, Heydenberk, & Tzenova, 2006). However, not all research

findings on CRE are positive and many studies show mixed or even null results (Garrard &

Lipsey, 2007).

Unfortunately, though US schools are somewhat saturated with schoolYwide conflict

resolution programs, they appear to suffer from low rates of utilization despite their

widespread adoption. Some research indicates that only 8% of students used conflict

mediation when as many as 95% of students in the school reported they were aware of its

existence (Theberge & Karan, 2004). Much like antiYbullying efforts, CRE programs are

also inherently plagued by issues common to all universal prevention programs such as

low

treatment dosages, failure to move beyond classroom based skills, and inability to engage

highYrisk students or shift cultural norms (Shuval et al., 2010). Other researchers have

been even more scathing in their criticisms of CRE and have attacked basic assumptions of

these programs such as: all students involved in conflicts have the same needs regardless of

risk level and all students have deficits in social information processing that need to be

fixed (Webster, 1993). In other words, many CRE programs may fail because they do not

take into consideration how the surrounding context may influence students’ decisions

about how to resolve to conflict. For example, a few studies found that students in

suburban schools reported conflicts primarily over control of resources and preferences

compared to relationship conflicts in urban schools (Johnson & Johnson, 1996). However,

program improvement research on CRE can address some of these limitations by paying

special attention to the core characteristics outlined by Jones (2004) such as the target

grade, population and implementation barriers.

Conflict mediation and program outcomes. Schools have implemented mediation

programs at all grade levels; however, research indicates that mediation is most effective

for high school followed by middle school youth in comparison to elementary school

students (Garrard & Lipsey, 2007). Many conflict mediation programs make use of

peer/student volunteers, who are often the same age or developmental level as the

disputants, or those students directly involved in the conflict. Less frequently, in a mentorY

mediator model, older students or adults assist younger, peer mediators with facilitation of

the mediation sessions and may conduct their own mediations (Cassinerio & LaneYGaron,

2006). Traditionally, mediation programs consist of “cadres” of volunteer, peer mediators

who facilitate a scripted mediation process for two disputants over a wide range of

potential conflicts including relational aggression, physical or verbal bullying or

interpersonal problems (Johnson & Johnson, 2007). The mediators receive formal,

mediation training in facilitation and conflict resolution skills. A mediation program may

also be universal or classYlinked, but it can also be selective or indicated for students who

are at risk for social or physical conflicts or were already involved in a serious conflict and

referred by an administrator.

Conflict mediation programs with peers in particular represent a large portion of

CRE interventions reported in the literature. In a review of a forty year period that

included only research studies with control groups and preYpost measures, conflict

mediation programs demonstrated nearly twice the effect size than direct skills instruction

and curriculum integration approaches (Garrard & Lipsey, 2007). However, peer mediation

models also have their own strengths and weaknesses. Previous research has found that

conflict mediation programs teach students perspective taking, empathy, and emotional

regulation skills, which are all associated with positive adjustment outcomes for youth

(LaneYGaron, 1998; Cassinerio & LaneYGaron, 2006; Heydenberk et al., 2006). Deficiencies

in these skills also have associations with worse academic performance and mental health

outcomes for students. In fact, many of the characteristics of resilient students such as

social competence, problemYsolving skills and a sense of autonomy are also the building

blocks of being able to resolve conflicts successfully (Crawford & Bodine, 1996). In

addition, researchers have found that curricularYbased, conflict resolution education as

well as peer mediation both demonstrate positive effects on student learning. Heyndeberk

and colleagues (2006), who looked at conflict resolution education in elementary

classrooms over the course of one school year, found improvements in conflictYrelated

skills including selfYregulation, problem solving, and empathy. Steps in the conflict

mediation process may also help slow down Other research on peer mediation programs

with elementary and middle school students found similar evidence for student gains in

cognitive and affective forms of perspective taking and empathy based skills (LaneYGaron,

1998; Cassinerio & LaneYGaron, 2006). However, the researchers also noted that disputant

participants did not benefit as much as their peer mediators in perceptions of school

climate, and, similar to the current study, participants did not report any change on a

measure of conflict resolution strategies. Researchers conclude that, in general, peer

mediation is more effective than classroomYbased approaches at teaching conflict

resolution skills, which might be explained by the high correlations found between

modeling and learning new skills (Garrard & Lipsey, 2007; Harris, 2005).

One measure common to almost all research on mediation programs is participant

satisfaction. Satisfaction is often a gauge of several different aspects of the mediation

experience including mediator effectiveness, fidelity and helpfulness of the program from

the participants’ perspective. Traditionally, disputants have often equated fairness in their

mediations with program satisfaction (Nix & Hale, 2007), and some students report that

one of the most helpful mediator skills is impartiality (Harris, 2005). Researchers also see

program satisfaction as a type of social validity, or the extent to which students (or adults)

accept a program as credible or helpful, which is important to consider because it

influences how likely students are to use the program and recommend it to others (Smith

et al., 2002). If the target population is highYrisk and does not view mediation with some

degree of social validity, they will most likely underutilize the program or avoid it

altogether. Researchers recommend that this should be assessed in future research or in

evaluations of conflict mediation programs (Smith et al., 2002).

Conflict Mediation and the Current Study

Jones’ (2004) framework for identifying the four core characteristics of CRE

programs is useful for placing the current case study in the context of previous research

on conflict mediation. Again, these characteristics include the program model, the grade

level, target population and implementation factors of fidelity, longevity and coordination

or integration within the school system. This study investigated the impact of a mentor

mediation program and may be the first to include college students instead of peers as

mediators in school settings. This study also included both middle and high school

students, who were mostly Latino, but included a mix of other groups as well, in a large,

urban school district. The target population at each school varied somewhat, but was

mostly available to all students in the school and, in some cases, included referrals from

administrators for disciplinary incidents. In terms of implementation, the mediation

program has a multiYyear presence in all three schools. Two schools integrated the

program into their discipline systems and another school integrated the program into an

existing, peerYbased, conflict education program. This study also focused on the impact of

the mediation program from the perspectives of the disputants rather than the mediators,

which has been understudied in previous research.

These program and study characteristics are fairly unique within the existing

research on CRE programs. The majority of research on conflict mediation focused on

white, elementary school students of relatively low risk; curricular components in addition

to peer mediation; and the experiences of peer mediators rather than the disputants

(Jones,

2004; Garrard & Lipsey, 2007). Some studies, however, were more similar and included

middle and high school students of color and focused on the perspectives of the disputants

rather than the mediators (Harris, 2005; Nix & Hale, 2007). Because the majority of

research on conflict mediation has focused on peer mediators, much less is known about

the impact of the program from the perspectives of the disputants including reasons for

program satisfaction and dissatisfaction (Nix & Hale, 2007). In addition, I only found one

other researcher who also used university students as mentors to peer mediators, but they

did not mediate the conflicts themselves (LaneYGaron, 1998; LaneYGaron & Richardson,

2003; Cassinerio & LaneYGaron, 2006). In particular, LaneYGaron’s work on mentor

mediation with students from Cal State University, Fresno has included schools that are

predominately Latino and lowYincome and, therefore, her research figures prominently in

the discussion of this study’s results.

While this study does not explicitly attempt to address gaps in the knowledge base,

it should contribute to some areas where further research is needed. Some have argued

that sufficient quantitative research already exists on the effectiveness of conflict

mediation, but the field needs more qualitative work to determine the contextual factors

that influence outcomes (LaneYGaron & Richardson, 2003). One of this study’s research

questions attempts to meet this need by soliciting students’ opinions about various

environmental factors that may facilitate or inhibit their use of conflict mediation.

Furthermore, previous research has not always yielded consistent results on how

programs differentially affect mediators, disputants, and nonYparticipants, or those who

never entered any mediation (Cassinerio & LaneYGaron, 2006). The first research question

of this study attempted to understand the experiences of conflict and help seeking from the

perspective of all students in the program’s school and not just from the students who

participated in the program. In addition, several student participants in this study were not

involved in the conflict directly, but were included as witnesses or third parties to the

conflict. There are also additional gaps that are not explicitly or only partially addressed in

this study, but should be considered more extensively in future research on conflict

mediation. These include an assessment of the types of students who benefit the most or

least from the program, the program elements that mediate specific outcomes, and ways

that conflict resolution skills may generalize to settings outside of scripted mediation

sessions (Jones, 2004; Daunic, Smith, Robinson, Miller, & Landry, 2000; Nix & Hale, 2007).

Methods

Description of UCLA Mediation Program

A collaborative effort among UCLA’s School of Education, the Institute for

Nonviolence in Los Angeles, and the Los Angeles Unified School District (LAUSD) is

responsible for implementing a conflict mediation program across five secondary schools

in LAUSD including three of the schools in this study. Professor Wilms, the faculty advisor

of this research study, teaches an undergraduate, Education course at UCLA called

“Restoring Civility: Understanding, Using and Resolving Conflict”, which trains

undergraduates as volunteer mediators and places them in several partner schools as part

of the class’s practicum component. The undergraduates receive training in facilitating

mediation sessions during Fall Quarter and go out to their school sites to conduct

mediations in Winter Quarter of the academic year. Some students volunteer to stay with

the program for an additional quarter in the Spring of the same school year.

The UCLA conflict mediation program is based upon the ITUNA approach to dispute

resolution, which is an acronym for a 5Ystep process that mediators use for facilitating a

conflict mediation session: 1) Information, 2) Talking and being heard, 3) Understanding,

4) Negotiating, and 5) Agreement. In the first step, information, the mediators share

information with the disputants about what will happen during the mediation, review the

ground rules, and explain the goals of the session. In the second step, talking and being

heard, the mediators facilitate a dialogue among the disputants where they share their

sides of the story and listen to the other person share his or hers. In the third step,

understanding, the mediators attempt to have the disputants understand the positions of

the other parties, but, in theory, this outcome can happen on its own after disputants listen

to the other person’s side of the story. In the fourth step, negotiation, the mediators

facilitate a conversation about what each party is willing to do in order to resolve the

conflict. Mediators encourage disputants to come up with their own agreements, but may

also suggest certain kinds of actions or solutions during this step. In the fifth and final step,

agreement, the disputants reach a specific agreement about their conflict and generally put

this in writing and sign the document as a way of staying accountable to the mediation. In

theory, ITUNA provides a kind of scripted process for mediators to follow, but in reality,

mediations often include many deviations from these specified steps.

One of the strengths of the mediation program is that each school site may tailor the

implementation of the intervention in a way that best suits their needs. For example,

students may be referred by a variety of sources for equally varied reasons. Teachers may

make referrals or students may refer themselves to the conflict mediation program

depending on the school’s implementation. Examples of appropriate referrals generally

include student conflicts such as gossiping, verbal altercations, teasing/bullying, fighting,

harassment and other kinds of lowYlevel incidents. While teachers or even the students

themselves may make referrals for mediation, participation in the mediation program is

usually voluntary; however, some schools have mandated mediation in certain situations.

Teams of undergraduate students serve as volunteer mediators and are assigned to each of

the participating schools sites. Once the mediators receive a referral, they meet with the

students together or separately, if necessary, and follow the ITUNA process for arriving at

an agreement between two or more student disputants. Conflict mediation sessions often

include only two students, but they can include several more people especially when the

incident involves multiple people such as other witnesses or third parties to the conflict.

Dyadic mediation sessions are often less complex than three, four, or five or more parties

in a single conflict. However, mediators receive their ITUNA training in their

undergraduate course during the Fall Quarter at UCLA before they enter the field and

facilitate mediations. Participants

School site selection. This study took place in two separate phases in four schools

in a large, urban school district in Los Angeles. The four sites in the current study included

two middle schools: Tom Joad Middle School (Joad), a regular public school; and Jim Casy

Middle School (Casy), a small charter school; and two high schools: Rose of Sharon High

School (Sharon), a small pilot school; and George Milton High School (Milton), a regular

public school1. These schools are located in different areas of the city including West, South

and Northeast LA. While the study design changed several times over the course of the

2012Y‘13 school year, the final study in its current form included all four schools for phase

1 and only three schools for phase 2. Phase 1 of the study included recruitment and data

collection for the classroom questionnaires in which all four schools participated. In phase

2 of the study, I collected preYpost measurement data and interviews from students who

participated in the conflict mediation program in all of the schools except for Joad. I was

not able to finish phase 2 at Joad and made the decision to continue phase 2 in Casy,

Sharon and Milton because the conflict mediation program already had an established

history in these schools and there was a higher likelihood of reaching an adequate sample

size2.

Phase 1: classroom sample. A total of four schools and 848 students participated

in the first phase of the study, which occurred in Winter and Spring Quarter of the 2012 –

2013 academic school year (see Table 1). There is a fairly even distribution of students

across all grades except for 12th because not as many students participated from Milton.

The classroom sample has slightly more females than males. It is also majority Latino at all

schools, but almost exclusively so at the small schools, Casy and Sharon. Milton was the

most ethnically diverse school. I also looked for any statistical differences among the four

schools. The two middle schools did not differ by percentage of students per grade in the

sample; however, the two high schools had significant differences in proportions because

of convenience sampling at Milton. The schools did not significantly differ by gender, but a

visual inspection of the data revealed that there were differences by race. However, in

terms of majority groups, there were no significant racial differences between Casy and

Sharon for Latinos, or between Milton and Joad for blacks and Latinos.

Table 1

Characteristics of Classroom Sample in Four Participating Schools (N = 848)

Casy Sharon Milton Joad Total

n (%) n (%) n (%) n (%) n (%)

Type Charter Pilot Regular Regular

Sample size 220 (26) 240 (28) 212 (25) 176 (21) 848

# of classrooms 11 12 8 8 39

Grade

Middle school

6 129 (37)

YY YY

47 (27) 176 (21)

7 133 (33)

YY YY

61 (35) 194 (23)

8 134 (30)

YY YY

68 (39) 202 (24)

High school*

66 (28) 61 (29) 127 (15)

74 (31) 46 (22) 120 (14)

46 (19) 70 (33) 116 (14)

54 (23) 35 (17) 89 (10)

Gender

Female 122 (55) 138 (58) 115 (55) 92 (52) 467 (55)

Male 98 (45) 100 (42) 95 (45) 84 (48) 377 (44)

Race/Ethnicity

Latino 204 (94) 216 (90) 116 (55) 118 (67) 654 (77)

Black 4 (2) 2 (1) 37 (17) 38 (22) 81 (10)

Asian 3 (1) 12 (5) 17 (8) 0 (0) 64 (08)

White 2 (1) 4 (2) 23 (11) 2 (1) 31 (04)

Other/mixed 5 (2) 5 (2) 19 (9) 17 (10) 46 (05)

*Significant chiYsquared test value < .01

Phase 2: intervention sample. A total of 52 students participated in 21 mediations

across the three program schools in Spring Quarter of 2013. Out of these 52 students, I

included 40 (77%) in the quantitative analysis who had a complete set of both preY and

postYquestionnaires (see Table 2). 12 students did not complete either a preY or postY

questionnaire for various reasons including my own inability to followYup with students

because of time or logistical problems, being expelled from school or not retuning consent

forms. For the qualitative analysis, I included 45 student and 11 undergraduate mediator

interviews. In the analytic sample (n = 40), about 45% of participants were from the

middle school, Casy, and the remaining 55% were from the two high schools, Sharon (25%)

and Milton (30%). The majority of participants were in 6th (23%) and 9th (28%) grades and

the least represented grade was 11th (5%). The sample was also predominantly female

(73%). Casy and Sharon were the most unbalanced while Milton had an even split. The

participant sample was also predominately Latino (73%) with blacks (18%) and Asians

(10%)

comprising the remainder. Race was similar to gender in that both Casy and Sharon were

nearly all Latino while Milton contained most of the ethnic diversity.

Table 2

Characteristics of Intervention Sample in Three Participating Schools (N = 40 matched pairs)

Casy

n (%)

Sharon

n (%)

Milton

n (%)

Total

n (%)

# of mediations 10 (48) 4

(19) 7

(33) 21

# of participants 25 (48) 13 (25) 14 (27) 52

Matched pairs

18 (45) 10 (25) 12 (30) 40

Student Interviews

22 (49) 13 (29) 10 (22) 45

Mediator Interviews

Grade

Middle school

(33)

18 (45)

(22) 4

(44) 9

6 9

(23)

YY YY YY

7 5

(13)

YY YY YY

High school

(10)

10 (25)

12 (30)

22 (55)

(10) 10 (83) 11 (28)

(40) 0 (0) 4

(10)

(10) 1 (08) 2

(05)

(40) 1 (08) 5

(13)

Gender

Female

15 (83) 8 (80) 6 (50) 29 (73)

Male 3

(17) 2 (20) 6 (50) 11 (28)

Race/Ethnicity

Latino 18 (100) 8 (80) 3 (25) 29 (73)

Black 0 (0) 0 (0) 7 (58) 7

(18)

Asian 0 (0) 2 (20) 2 (17) 4

(10)

White

YY YY YY YY

Other/mixed

YY YY YY YY

Procedures

Phase 1: classroom sample. Classroom nomination and recruitment in each of the

four schools worked slightly differently. Because Casy and Sharon were small schools, I

was able to collect data from all the classrooms in the schools with the exception of one

class at Casy that chose not to participate. Joad and Milton were much larger schools and I

had to selectively sample classrooms from each grade in order to make the sample as

representative of the school as possible. For these two schools, the classrooms were not

selected randomly, but at the recommendation of the conflict mediation program

coordinator. At Joad, I collected data from two classrooms (high and low ability) in each

grade level (6th – 8th) in addition to two special education classes for a total of 8 classrooms.

The special education classes were resource classrooms, which meant that the students

were integrated into the general education classes for the rest of the school day. In these

classes, I took special care to explain the questionnaires and answer student questions

about the meaning of each survey item. At Milton, the situation was more complicated since

there were six individual learning academies within the school. I attempted to collect data

from each grade in each academy, but this was not always possible because of logistic and

timing difficulties with the school site coordinator. Unfortunately, school informants at

Milton also suggested that there were in fact significant behavioral and academic

differences among the individual academies. In the end, I was able to recruit a total of 8

classrooms from 5 different academies, which included two 12th grade classrooms a

humanities academy; one 11th grade class from a science academy and another from an arts

academy; one 10th grade class from an arts academy and another from a humanities

academy; and two 9th grade classrooms from a science academy.

To recruit students into this phase of the study, the program coordinator and I first

spoke to each teacher about the purpose of the research. The teachers provided verbal

consent to include their class and recommended a period to return and speak to the class

as a whole about the study. I then visited each classroom that agreed to participate and

made an announcement to the students about the study’s purpose and its voluntary and

confidential nature. I passed out youth and parent assent forms and students had the

opportunity to take them home and obtain a parent or guardian signature before filling

out the classroom questionnaire. Students who returned to school with assent forms

completed were able to take the classroom questionnaire. Casy had a 76% average,

student response rate in classrooms that choose to participate with all required consent

forms; Sharon had a rate of 71%; and Milton had a rate of 66%. Despite the variations in

classroom recruitment and participation rates among the sites, each school represented

about ¼ of the total sample (see Table 1).

Phase 2: intervention sample. I recruited both undergraduate mediators and

students from the three program schools to participate in phase 2 of the study. As part of

their class requirements at UCLA, the undergraduates kept a fixed schedule and usually

went to the school one day per week to conduct mediations in the schools. Most of the time,

all the mediators assigned to a school went there on the same day. I selected each mediator

who completed at least one mediation to participate in the study, which included everyone

in the three schools except one mediator from Milton who did not conduct any mediations.

Before interviewing them, I reviewed the purpose of the study, solicited each mediator’s

consent to participate and had them sign the adult assent form. In their interviews, I asked

the mediators to refer to one specific mediation they conducted that quarter in their

answers to the questions from the interview schedule. In order to thank the mediators for

their participation, I gave each of them a toy stress ball.

I recruited student participants from a pool of referrals to the mediation program.

The site coordinators at each school had a slightly different system for handling program

referrals. At Casy, the site coordinator was the Dean of Students who kept a list of student

referrals each week for the mediators, which mostly came from discipline or behavior

issues in class, but also included selfYreferrals. At Sharon, the site coordinator was the

Principal, who also kept a list of students who needed new mediations each week. These

students were referred mostly because of behavior issues, but occasionally they included

selfYreferrals as well. At Milton, the referral process was not as straightforward. The site

coordinator was the Assistant Principal of a humanities academy, but for some reason,

there were very few mediation referrals to the program. After some discussion and

negotiation, the referral process was expanded to the Deans of Students, who had more

access to students involved in conflicts at the school. The Deans also maintained a list of

students who were referred to them for disciplinary issues and there were generally no

selfYreferrals. It was common at all three schools to have a waiting period of several days

from time of the conflict and referral until the actual mediation took place since the

mediators only came to the site about one day per week.

All middle and high school students who were referred to the mediation program

were part of a potential pool of participants for phase 2 of the study. Students always had

the option of participating in the mediation program while maintaining their right to not to

participate in the study. Casy was responsible for the bulk of the student (49%) and

mediation (46%) referrals while Sharon and Milton referred far less students and

conducted fewer mediations (about 25% respectively) to the program (see Table 3).

Ultimately, I was able to follow 21 mediation sessions across all three schools, which

included, on average, about 2.6 students per mediation. Mediation sessions at Sharon

tended to include more students while mediations at Milton usually only included about 2

students each on average. Before the mediators met with the students for the mediation

sessions, I was able to retrieve a list of referrals from the site coordinators. I reviewed the

study information with the students and sent a written assent form home with them for

their parents’ signature, which they returned before participating in the study. Out of 74

students who were referred for mediations, I recruited 52 students to participate in the

study with consent (70%), which resulted in 40 pairs of matching preY and postYtests and

45 student interviews. Students left the selection pool at any point for a variety of reasons

including no mediation ever occurring after the referral, inappropriate referrals for

mediation, being expelled from school, refusing to participate, missing a certain component

of the study or not returning assent forms. I administered preYquestionnaires to eligible

students before the mediators met with them for their mediation sessions. I also

administered postYquestionnaires and conducted individual interviews with students after

the mediation sessions, but, unfortunately, the time it took me to followYup with them was

not always consistent. On average, it took me about 2 weeks to followYup with students,

although this varied slightly by school (see Table 3). I conducted all interviews in a

confidential space in the students’ schools and offered them a choice of stickers,

pencils/pens, or toy stress balls as a gift to thank them for their time in the study.

Table 3

Program Characteristics in Three Participating Schools, Spring 2013

Casy Sharon Milton Total

n (%) n (%) n (%) n (%)

# of students referred 36 (49) 19 (26) 19 (26) 74

# of student participants 25 (48) 13 (25) 14 (27) 52

# of mediations referred 12 (46) 5 (19) 9 (35) 26

# of mediations conducted 10 (48) 4 (19) 7 (33) 21

# of mediators 3 2 4 9

Avg # of students/mediation

Avg # of days between

2.5 3.3 2 2.6

mediation and postYtest

14.5 21.8 10.1 15.5

Measures

There were three different components to the study including the 1) classroom and

2) preYpost phases and 3) student and mediator interviews. In this section, I review the

measures I used for each of these three parts. Tables 4 and 5 have a complete list of the

psychometric properties of each measure in phases 1 and 2 of the study (see Appendix A

for the classroom questionnaire and Appendix B for the preYpost questionnaire). The third

component of the study includes student and mediator interviews schedules (see

Appendix C for the student interview and Appendix D for the mediator interview). I also

created several independent, or moderator, variables based on students’ interviews and

then applied these codes to the data. I review these variables at the end of the section.

Phase 1: classroom questionnaires. All measures on the classroom questionnaire

were taken almost exclusively from existing surveys or studies on school violence and

bullying (except as indicated). The two primary sources for the measures include the

School Crime Supplement (SCS) to the National Crime Victimization Survey (NCVS)

released by the National Center for Education Statistics (DeVoe & Bauer, 2011), and the

School Climate Bullying Survey (SCBS) published by Cornell (2012), which was built on

decades of research by other notable people in the field such as D. Olweus.

Exposure to conflict measures.

Physical fighting. This question asked students if they were in a physical fight at

school in the last year. I coded students’ answers as yes (=1) or no (=0). 84% of students

responded no to this question and 16% of students responded yes. This question was

specifically taken from the SCS portion of the NCVS (DeVoe, & Bauer, 2011).

Witness fighting. This question asked students how often they witness physical

fights at school. Students reported their answers on a frequency scale that included: never

(=0), once/year (=1), once/month (=2), once/week (=3), and every day (=4). The mode, or

most commonly reported answer, was once/year (M = 1.3, SD = 1.1). I developed this

question for the study.

Witness bullying. This question asked students how often they witness any kind of

bullying or harassment that happens to others at school. Students reported their answers

on a frequency scale that included: never (=0), once/year (=1), once/month (=2),

once/week (=3), and every day (=4). The mode was never (M = 1.6, SD = 1.5). I developed

this question for the study.

Witness any bullying/harassment composite. I also created a composite measure for

witnessing any bullying or harassment that combined students’ answers from the two

witness questions. The composite variable was based on the highest reported frequency of

witnessing either fighting or bullying. For example, if a student witnessed weekly bullying

but no fighting, then this student would receive a weekly score for the composite. Scores

were recorded on the same scale as the other questions: never (=0), once/year (=1),

once/month (=2), once/week (=3), and every day (=4). The mode was once/year (M = 2.0,

SD = 1.4).

Bullying and harassment victimization. A series of four questions asked students

how often they were victims of four different types of bullying and harassment in school

including physical, verbal and cyber bullying and racial harassment. I provided a brief

description of what bullying/harassment meant and included a definition for each type that

I instructed students to read before they responded. Students reported their answers on a

frequency scale that included: never (=0), once/year (=1), once/month (=2), once/week

(=3), and every day (=4). For physical (M = 0.4, SD = 1.0), verbal (M = 0.7, SD = 1.3), and

cyber bullying (M = 0.2, SD = 0.6), and racial harassment (M = 0.4, SD = 1.0), the mode was

never for each of these types. These four questions were specifically taken from SCBS

(Cornell, 2012), although I adapted the frequency scale from similar questions on bullying

in the SCS portion of the NCVS (DeVoe, & Bauer, 2011).

Experience any bullying/harassment composite. I also created a composite measure

for the experience of bullying or harassment that combined students’ reports across the

various types. The composite variable was based on the highest reported frequency from

any of the four subtypes. For example, if a student reported weekly, verbal bullying but no

other kinds, this student would receive a weekly score for the composite. Scores were

recorded according to the same frequency scale as the other measures: never (=0),

once/year (=1), once/month (=2), once/week (=3), and every day (=4). This composite

provided a more consistent way of comparing student reports of any bullying or

victimization experiences to other research that did not specify the exact type. The mode

for this composite was never (M = 1.0, SD = 1.4). I developed this composite for the study.

Help seeking measures.

Seeking help. This question asked students who reported any experience of bullying

or harassment if they went to anyone for help at school for the last time it happened to

them. I coded students’ answers as yes (=1) or no (=0). 52% of students responded no to

this question and 48% of students responded yes. This question was adapted from the SCS

portion of the NCVS (DeVoe, & Bauer, 2011).

Snitching. This question asked students if they could report a threat or conflict in

school without being considered a “snitch”. I coded students’ answers as yes (=1) or no

(=0). 63% of students responded yes to this question and 37% of students responded no.

This question was adapted from the SCS portion of the NCVS (DeVoe, & Bauer, 2011).

Willingness to seek help scale. This 9Yitem measure asked students how much they

agreed or disagreed with statements about getting help for problems or conflicts in school.

The composite captured several areas of help seeking attitudes including student

perceptions of acceptability of getting help, teachers’ responsiveness to help seeking and

peer norms about seeking help. Examples of statements in the measure included “I would

tell someone if another student was…bullying or harassing me”, or “brought a weapon to

school”, or “was going to hurt another student”; “It’s okay for students talk to someone if

they’re being bullied or harassed”; and “teachers care about me” or “teachers would do

something about it if I told them I was being bullied or harassed”. Students reported their

answers on a 4Ypoint agreement scale that included: strongly agree (=1), somewhat agree

(=2), somewhat disagree (=3), and strongly disagree (=4). Scores were reverseYcoded so a

higher score indicated a greater willingness to seek help (M = 3.1, SD = 0.6). The reliability

coefficient (alpha) for this measure was .84 with the classroom sample. The help seeking

scale was specifically taken from SCBS (Cornell, 2012) and a large number of other studies

have also used and validated the measure (Williams & Cornell, 2006; Eliot et al., 2010;

Bandyopadhyay et al., 2009).

Table 4

Descriptive Statistics of Exposure to Violence and Help Seeking variables from Classroom

Questionnaires (N = 848)

n % M SD Min Max % missing

Exposure to violence

Physically fought 843

YY YY YY YY

Yes 139 16

YY YY YY YY YY

No 704 84

YY YY YY YY YY

Witness fights 844

1.3 1.1 0 4 <1

Witness bullying 833

1.6 1.5 0 4 02

Witness any conflict 848

2.0 1.4 0 4 0

Physically bullied 844

0.4 1.0 0 4 <1

Verbally bullied 843

0.7 1.3 0 4 <1

Cyber bullieda664

0.2 0.6 0 4 01

Racially harassed 847

0.4 1.0 0 4 <1

Experience any bullying 848

1.0 1.4 0 4 0

Help seeking

Sought help 348

30b

Yes 181 48

YY YY YY YY YY

No 167 52

YY YY YY YY YY

Not snitching 751

11b

Yes 472 63

YY YY YY YY YY

No 279 37

YY YY YY YY YY

Help seeking scale 833

3.1 0.6 1 4 02

a not included at Joad

b high % missing because of poor placement on questionnaire

Phase 2: intervention preKpost questionnaires.

Positive/Negative Affect Scale (PANAS). For negative and positive affect, students

responded to a list of words that captured negative and positive feelings before and after

their mediations. I prompted students to report on how much they felt each of the words

when they thought about the primary person from their conflict. The scale has 5 words for

positive affect and 5 for negative affect for a total of 10 items, which were distributed in a

random order on the questionnaire. Examples of words included “upset”, “sad”, “afraid”,

“strong”, “interested”, “excited”, etc. Students reported their answers on a 5Ypoint scale

from “not at all” (=1) to “extremely” (=5). I averaged the negative and positive items

together to create negative (pre: M = 2.1, SD = 1.0; post: M = 1.4, SD = 0.5) and positive

(pre: M = 2.2, SD = 1.1; post: M = 2.1, SD = 0.9) composite scores. For negative affect, the

reliability coefficients (alphas) for this measure were .80 (pre) and .77 (post) with the

intervention sample. For negative affect, the reliability coefficients (alphas) for this

measure were .65 (pre) and .61 (post) with the intervention sample. This scale was

adapted from previous validation research on the PANAS, which included both a short

(10Y

item) version as well as a version validated for use with children (Laurent & Catanzaro,

1999; Thompson, 2007).

Endorsement of nonviolent strategies to conflict and anger. In order to assess

beliefs or attitudes about conflict strategies, students answered 8 questions about how

they would respond to a potential conflict when they were angry with someone. The

measure asked students how likely they were to adopt a series of nonviolent behaviors.

Examples of questions include how likely are you to “ignore the situation?”, “try to talk it

out?”, “suggest a mediation?”, “try to see the other person’s point of view?”, etc. Students

reported their answers on a 4Ypoint scale from “Very unlikely” (=1) to “Very likely” (=4).

One item in the measure was reverse coded and then all items were averaged together to

create a composite measure (pre: M = 2.4, SD = 0.5; post: M = 2.5, SD = 0.5). The reliability

coefficients (alphas) for this measure were .69 (pre) and .52 (post) with the intervention

sample. Higher scores on the measure would indicate a greater endorsement of nonviolent

behaviors to resolve conflicts when students are angry. This scale was taken directly from

the Youth Violence Compendium, which was adapted from a previous study on middle

school students in grades 6 – 8 and showed excellent internal consistency, alpha = .84

(Dahlberg, Toal, Swahn, & Behrens, 2005).

Willingness to seek help scale. This was the same scale used in the classroom

questionnaire. Please refer to Phase 1 of the Measures section for a complete description.

A higher score indicates greater willingness to seek help (pre: M = 3.1, SD = 0.5; post: M =

3.2, SD = 0.5). The reliability coefficients (alphas) for this measure were .67 (pre) and .77

(post) with the intervention sample.

Mediation satisfaction. In order to assess satisfaction, I used a 5Yitem measure

that asked students how they felt about participating in the mediation program. Examples

of items in this measure asked students if they liked mediation, were treated fairly, would

return again or recommend it to their friends. Students reported their answers on a

4Ypoint agreement scale that ranged from strongly agree (=1) to strongly disagree (=4). A

higher score on this scale indicated higher satisfaction with various aspects of the

mediation program. Items were averaged together to create a single, composite satisfaction

score (M

= 3.2, SD = 0.7). The reliability coefficient (alpha) for this measure was .88 with the

intervention sample. This measure was adapted from two previous studies on conflict

mediation that also assessed participants’ satisfaction with their participation in the

program and reported excellent internal consistency (Harris, 2005; Long, Fabricius,

Musheno, & Palumbo, 1998).

Table 5

Descriptive Statistics of Measures from PreTPost Intervention Questionnaires (N = 40)

n M SD Min Max % missing

PreYtest

Negative Affect subscale 40 2.1 1.0 1 5 0

Positive Affect subscale 37 2.2 1.1 1 4.4 8

Nonviolent Response Endorsement 39 2.4 0.5 1.4 3.4 3

Help seeking scale 40 3.1 0.5 1.7 4 0

PostYtest

Negative Affect subscale 40 1.4 0.5 1 3 0

Positive Affect subscale 40 2.1 0.9 1 4 0

Nonviolent Response Endorsement 39 2.5 0.5 1.3 3.5 3

Help seeking scale 37 3.2 0.5 2.1 4 8

Mediation Satisfaction a39 3.2 0.7 1 4 3

a PostYtest only

Intervention followKup interview schedules. I used two interview schedules to

gather additional data from the students and the undergraduate mediators about their

perspectives on peer conflicts and their experiences in the mediation program. These

interviews were meant to supplement the questionnaire data and provide a deeper look

into the many contextual factors that could potentially influence students’ experiences with

resolving conflicts nonviolently. Both the student and mediator interview schedules went

through similar phases of development. I adapted both interview protocols from a

previously published study that focused specifically on disputant perspectives of the

mediation process including learning new skills, satisfaction and other kinds of outcomes

(Harris, 2005). Initially, I modified these interview schedules in order to create more

overlap with this study’s outcomes on the preYpost questionnaires. In addition, I

broadened the focus of Harris’ (2005) interviews from an emphasis on the mediation

process itself to encompass various socialYecological factors that might also contribute to

students’ decisions of how to respond to conflict. In this section, I will review the content of

each interview schedule as well as some moderator factors that arose from the student

interviews, which appeared to have an effect on the program outcomes.

Student interview schedule. The student interviews allowed me to gather

additional perspectives on several important aspects of the mediation process including

changes in key outcomes areas, their relationships with the mediators, program

satisfaction and various kinds of barriers to responding to conflict nonviolently as well as

accessing the program. Using the interview schedules from Harris’ (2005) study as a

starting point, I developed a 7Yquestion interview that asked students about their

experience in mediation from the point of referral to how the mediation agreement had

been working. After interviewing some participants, I modified the schedule several times

by eliminating or rewording questions that elicited only short or oneYword answers. I also

had initially decided not to ask students about the details of their conflict, but soon

realized

this made the rest of the interview extremely difficult because it was hard to ask any

follow up questions. The final version of the interview included seven thematic areas with

additional prompts under each of these categories. These topic areas included: 1) a

students’ history of participation in mediation; 2) details of the current conflict that

brought the students to mediation; 3) the mediation process including students’

relationships with the mediators; 4) mediation outcomes mostly related to the disputants’

relationships with each other after the mediation; 5) the process of learning conflict

resolution skills; 6) program satisfaction; and 7) barriers to accessing program or seeking

help and areas in need of improvement. I asked questions about mediation outcomes that

specifically overlapped with the preY and postYtest measures in order to triangulate

results and enhance the interpretation of the quantitative findings. The final version of the

interview protocol appears in Appendix C.

Mediator interview schedule. Interviews with the undergraduate mediators also

provided me with an additional perspective on students’ experiences with conflict and the

program. Students who are in the same conflict often have different perspectives of what

happened and may not always reach a resolution successfully. My interviews with the

mediators were sometimes able to resolve these kinds of discrepancies with students’

accounts of the conflict. Mediators were often able to provide a third party, neutral

perspective of the situation, which was also useful for adding an additional layer of

information about the conflict that students may not have been comfortable sharing with

me as an unfamiliar adult or interviewer. I also developed the mediator interviews in a

fashion similar as the student ones by taking Harris’ (2005) schedule as a starting point,

which focused on mediator modeling and disputant learning outcomes. For this interview, I

focused instead on two major goals including soliciting mediator perspectives on the

conflict in order to validate students’ reports and gathering contextual information about

their relationships with students and program implementation. Similar to the student

interviews, the mediator schedules went through one or two revisions such as modifying

questions that did not generate sufficient or useful information. The final interview

protocol had five thematic areas including 1) the mediation context, or details about who

was involved and what happened; 2) the mediation process including how they engaged

with students and facilitated the session; 3) mediation outcomes including their own

perspective on how successful they felt it was; 4) learning outcomes for both students and

the mediators themselves; and 5) areas for improvement including their own abilities as

mediators and the program itself (see Appendix D for the complete interview protocol).

Independent variables derived from interviews. After reviewing and coding the

student and mediator interviews, several factors arose from the data that seemed to

contribute to students’ experiences in a systematic way. These contributing factors

appeared to act as potential moderators (or independent variables) because students’

comments suggested they had an impact on outcomes including their emotions, learning,

resolutions or sense of satisfaction with the program. I developed three of these

independent variables from the qualitative data and I review each of them here as I tested

them in the quantitative, preYpost analysis.

Relationship history. Students’ comments suggested that having a prior history of

friendship affected their experience in mediation. I created a dichotomous, relationship

status variable and coded students whenever possible as either 0) not friends, or just

acquaintances with no previous relationship among disputants; or 1) friends who had an

existing friendship or relationship with the disputant prior to the conflict. I coded this

variable using information from both the student and mediator interviews. According to

this coding scheme, 58% of students (n = 23) had a relationship history and 42% of

students (n = 17) did not have a relationship history.

Role in conflict. Students’ comments suggested that their role in the conflict may

have resulted in different outcomes or influenced their experiences in mediation. I created

a role in conflict variable that labeled students as 1) an instigator/aggressor when the

disputant appeared to be a bully or the primary aggressor; 2) a victim when the student

primarily appeared to be the target of aggression in the conflict; or 3) a disputant when

there was not enough information or the students’ roles in the conflict were either equal or

unclear. I also used information from both the student and mediator interviews to code

this variable. According to this coding scheme, 52% of students (n = 21) were disputants,

25% of students were aggressors (n = 10), and 23% of students were victims (n = 9).

Type of conflict. Students’ comments suggested that the type of the conflict may

have influenced their experiences or key outcomes in the mediation. I created a type of

conflict variable and assigned codes to students as being either in 1) a social conflict when

it was primarily related to friendship problems or arguments, or included only verbal

aggression such as rumors, gossip or verbal harassment; 2) a physical conflict when the

dispute included aspects of physical aggression such as threats, violence, or physical

bullying. I coded this variable in a mutually exclusive way so that conflicts could not be

both social and physical, although the majority of conflicts began primarily as social. I also

used information from both the student and mediator interviews. I initially developed this

variable as an indicator of level of severity, ranging from low to high, but ultimately

decided that this method was a more objective way of labeling the conflicts. According to

this coding scheme, 53% of students (n = 21) were in conflicts that were primarily social in

nature while 47% of students were in conflicts with physical aspects (n = 19).

Data Analysis

Phase 1: classroom sample. In the first phase of the study, data analysis of the

classroom questionnaires consisted of several parts including data entry, preparation and

statistical analysis of relationships among the variables of interest. First, I assigned coding

schemes to all the questions from questionnaires and entered all of the data anonymously

into Excel. I cleaned the data as much as possible by flagging entries that either had blanks

or the same answers across the entire questionnaire. Lastly, I imported the data and

performed all statistical procedures in Stata 12. I looked at the classroom data in three

different ways including 1) correlations among all the variables, 2) between school

differences, and 3) within school, group differences for the whole sample and within two of

the four schools.

First, I calculated correlations among the classroom variables using Spearman’s

rankYorder correlation coefficient (rho) because of the ordinal and nonYnormal nature of

many of the variables (see Table 6 in Results). Most of the questions are either categorical

or ordinal because they ask about frequency and, in general, the results were highly and

positively skewed towards “never”, or less frequent exposure.

Second, I looked for school differences in demographics and most of the variables in

the classroom sample (see Table 7 in Results). In order to determine whether a specific

school had a significant relationship with any of the variables, I used two different tests

depending on whether they were categorical, ordinal or interval. Thus, I used a

chiYsquared

test of independence when both the independent (e.g., school) variable (IV) and dependent

(e.g., physical fighting) variable (DV) were categorical; and I used the KruskalYWallis rank

test when the DV was ordinal rather than categorical, i.e., when questions asked how often

something happened. The help seeking scale was the only interval DV in the classroom

sample; however, in this case, the KruskalYWallis test was still preferable to ANOVA

because the schools did not have equal variances according to a significant Bartlett’s test

statistic (p < .01), which failed the assumption of homoscedasticity for the ANOVA test.

Third, I looked for group differences within the classroom sample as a whole in

addition to two of the four schools (see Tables 8, 9 and 10 in Results). For the two schools,

I chose Joad and Milton for closer analysis because they reported the highest levels of

conflict and had the most ethnic diversity among the four schools. Again, in order to

examine significant mean differences among groups, I used several variations of chiY

squared distribution tests as well as rank order tests depending on the nature of the data,

e.g., categorical or ordinal, and the number of grouping levels in the IV’s, e.g., two for

gender, five for ethnicity, etc. For tests with categorical variables, I used a chiYsquared test

of independence for gender (categorical) and physical fighting (categorical), but I used

Fisher’s exact test for when any of the cell sizes were less than 5 such as in the case of race

(categorical) and physical fighting (categorical). For tests with ordinal DV’s, I either used

the WilcoxonYMannYWhitney test for twoYlevel IV’s such as gender and the

KruskalYWallis rank test for IV’s with three or more levels such as race. Again, the data

failed the assumption of homoscedasticity, which precluded the use of the ANOVA family of

tests and required me to use alternative, nonparametric tests for analyzing group mean

differences. In cases when I detected significant group differences (e.g., with race) and a

visual

inspection of the data did not reveal an obvious group difference, I conducted a simple,

linear regression with Bonferroni adjustments for multiple comparisons in order to

determine which groups were responsible for the statistically significant differences.

Phase 2: intervention sample. In the second phase of the study, I conducted mixed

methods analyses of the intervention samples, which included the preYpost questionnaire

sample and the student and mediator interviews. The quantitative procedures in this

phase consisted of several parts including data preparation, analysis of preYpost changes

and examination of additional factors, or IV’s (soYcalled moderators), related to changes in

outcomes, if any. I also conducted the analysis of qualitative data in several steps including

transcription of interviews, thematic content coding and comparison or integration of

major themes with the quantitative findings. I review each of these treatments of the

quantitative and qualitative data in turn here.

Quantitative data analysis. For the preYpost measures, I first developed and

assigned codes to all the measures from questionnaires and entered the data into Excel

without student identifiers. I cleaned the data by assigning missing values to empty cells

and eliminating suspicious entries. Lastly, I imported the data and performed all statistical

procedures using Stata 12. For an analysis of program effects, I included only students with

matching pairs of preY and postYtests (n = 40) and excluded students who only had either

one preY or postYtest. First, I calculated mean scores with standard deviations for all

outcome measures in the sample (see Table 11 in Results). In order to look for statistically

significant changes in outcomes from preY to postYtest, I used a Wilcoxon signed rank sum

test, which is a special version (nonparametric) of the paired samples tYtest for use with

ordinal (e.g., Likert or frequency scales) rather than interval or scale variables. I measured

all outcomes either as ranks or an expression of agreement with statements.

Next, I examined additional factors, sometimes called independent or moderator

variables, which also may have been related to the changes in the outcomes. For example,

certain groups of students may report a greater degree of change than others in a

consistent or systematic way. I tested a series of these variables to see if any of them had a

significant association with the change in outcome scores from preY to postYtest. In order

to make this assessment, I created an additional variable for each outcome that recorded

the amount and direction of a change for each student. I categorized students into two

groups based on their direction of change, which included 1) positive change and 2)

negative or no change. I then conducted two different kinds of statistical tests of

association to determine whether the outcomes had a significant relationship with this

change variable or grouping. I used a Pearson correlation (r) coefficient value when the IV

and DV were both interval in nature: for example, with measurement time (IV) and the

change in negative outcome (DV). In most cases, however, the IV and DV were categorical.

For example, the analyses included IV’s like gender, school, or role in conflict and the DV

was transformed to represent either positive or no/negative change as described earlier. In

this case, I used Fisher’s exact test, which is similar to a chiYsquared test for examining the

significance of the association between two categorical variables in a contingency table.

However, Fisher’s test is preferable to chiYsquared when any of the cell sizes in the table

are less than 10.

Dichotomizing the outcome variables into groups of change or no change and presenting

them in a table was also helpful for understanding the data in a visual way.

Qualitative data analysis. The last procedure for data analysis in this study was the

review of student and mediator interviews. This data analysis procedure was a multiYstage

process that started with recording and transcribing all the interviews. Even as I began to

transcribe, it became apparent that certain themes were present across multiple

interviews and I took notes on my observations of patterns in students’ comments during

the transcription process. After I completed the transcriptions, I had an unorganized list of

themes, which I needed to further revise in the subsequent steps of the data analysis. I used

Microsoft Word rather than a qualitative data analysis software package for the analysis

because I wanted to preserve the flexibility of the coding process, and not make it either

overly rigid or distract myself from the meaning of the data. In addition, I also believe I did

not have the volume of qualitative data to justify the use of a specialized program in terms

of a utility, time or cost investment. In Word, I first created a document with a list of major

theme areas and I then listed representative quotes underneath each of these major

headings. I counted the number of students who referenced these theme areas and

developed definitions for each code after reaching a certain level of saturation. In order to

ensure the evidence was representative, I also noted the gender, grade level and school for

all sources of the quotes.

Because of the mixedYmethods nature of this study, I did not necessarily approach

the data here as a “blank slate”, i.e., take a grounded approach to the analysis, but rather I

attempted to organize students comments around this study’s theoretical framework,

which provides a way of structuring and interpreting the data according to outcome areas I

had already determined to be significant (Sipe & Ghiso, 2004). While this type of horizontal

analysis may impose a structure onto the data at higher levels of inference such as preY

selecting important domains or factors, I was able adopt a more inductive, or “bottom up”,

approach with the data as I identified various subYfactors or themes within these major

domain and factor areas (Schensul et al., 2013a). For example, based on my literature

review of conflict resolution, I had already identified emotional affect as an important area

to study and then used semiYstructured interviews in order to further clarify and draw out

examples of students’ emotional experiences during the mediation process (Schensul,

Scheunsul, & LeCompte, 2013b).

Most of my coding and interpreting of the data occurred at this midYlevel location

between formative theory and the experiences of the participants. I read through the

students’ and mediators’ transcripts several times in order to draw inferences from the

data. First, I read all the transcripts by school starting with the middle school and moving

through each of the high schools. I pulled out key quotes from students and mediators that

seemed to represent various factors or variables in each of the major outcome areas.

However, the quotes often touched on different, but related aspects of the major domain

areas, which required me to begin the process of assigning codes to the quotes and

organizing them based on their shared qualities (Schensul, Scheunsul, & LeCompte, 2013c).

Next, I was able to focus on groupings of quotes within a code and make assertions about

how the code was significant or related to the highYlevel outcome area (Erickson, 1986). In

order to support my assertions with appropriate evidence, I included events that occurred

both frequently and rarely in the data and also made use of disconfirming or contradictory

examples when possible to strengthen the plausibility of the conclusions (Erickson, 1986).

Also at the advice of Erickson (1986), in order to demonstrate this process to the reader, I

presented quotes as evidence or support for my assertions in the narrative and I attempted

to introduce each quote and offer an interpretation afterwards. Of course, this recursive

process of building and consolidating thematic categories is never completely objective

and it may hide from the reader the extent to which I have introduced my own biases or

positions into the coding scheme (Sipe & Ghiso, 2004). Lastly, however, I attempted to offer

explanations of the quantitative results with evidence from the qualitative data, whenever

possible, as a way of providing a more complete picture of the meaning of the data. The

consideration of multiple sources together may hopefully convince the reader that my

assertions are at least somewhat credible in the absence of alternative ways of establishing

trustworthiness such member checking or peer review of my data (Kiyama, 2010).

Quantitative Results

This study included two primary sources of quantitative data in order to

understand students’ exposure to conflict at the classroom level and participants’

experiences in the conflict mediation program. The first data source was classroom

questionnaires where students reported on the frequency of exposure to violence and their

attitudes about help seeking. I collected this data at four school sites including two middle

schools, Casy and Joad, and two high schools, Sharon and Milton. The second data source

was preY and postY questionnaires that assessed students’ experiences in the conflict

mediation program. I collected this data at three schools including Casy, Sharon and Milton.

Phase 1: Classroom Questionnaire Results

The classroom questionnaires contained rich sources of data about students’

exposure to aggressive incidents like fighting, bullying and harassment as well as their help

seeking behaviors and attitudes for these kinds of problems in their schools. The classroom

data help answer the first research question about the prevalence of conflict in the

program schools, the nature of students’ help seeking attitudes or behaviors, and any group

differences among the students or the schools. Results indicated that there were several

significant group differences among students and schools. Male students consistently

reported higher levels of exposure to violence and lower levels of help seeking than female

students. Black students consistently reported the highest levels of exposure to violence

and the lowest levels of help seeking. Grade level, and by proxy, age, was negatively

associated with exposure to violence and positively associated with help seeking attitudes

and behaviors. For differences among schools, Joad had the highest levels of conflict while

Sharon had the lowest levels of conflict, but the highest levels of help seeking scores. I

present these findings in the next sections in the following order: 1) correlations among all

the variables from the classroom questionnaires, 2) significant differences among the four

schools, and 3) group differences by gender, ethnicity and grade level within the entire

sample and then between Joad and Milton because students reported the highest levels of

conflict in these two schools.

Correlational analyses. There were relatively few surprises among the

correlations and most were in the expected direction; however, some noteworthy patterns

emerged among the demographic, exposure to violence, and help seeking variables (see

Table 6). Among the demographic variables, gender had highest correlation with reports of

physical fighting followed by racial harassment and help seeking attitudes. Male gender

was positively correlated with these aggressive behaviors and negatively correlated with

help seeking, as one might expect, but there was no correlation between gender and any

specific type of bullying. Perhaps expectedly, race had the highest correlations with

reports of racial harassment. As expected, grade level was also negatively correlated with

all types

of exposure to violence. In other words, students reported lower levels of conflict as their

grade level increased. Grade level was also positively correlated with help seeking, which

indicates that older students report more favorable help seeking attitudes.

Both witnessing and experiencing bullying or physical fights were positively

correlated with each other while they were also negatively correlated with help seeking

attitudes, which would be in the expected directions. Cyber bullying was the only type of

aggressive behavior that did not have a significant correlation with any demographic

variable, which is likely the result of few reports for this type of behavior. While help

seeking was negatively correlated with all types of exposure to violence, it had the highest

correlation with witnessing violence and the lowest correlations with actual experiences of

bullying or aggression. Help seeking was also positively correlated with the other

questions about getting help for conflicts. Student who felt they could get help without

being a “snitch” had the highest, positive correlation with general attitudes about help

seeking.

Among actual students who reported getting help for their last conflict in school, the only

significant correlation was with witnessing bullying, though this relationship was not very

strong. Lastly, looking at the composite variables for exposure to violence, students who

reported any experiences with bullying were most often verbally bullied, and students who

witnessed conflicts were more likely to witness bullying compared to physical fights.

Between school differences in the classroom sample. I also examined school

differences in student reports of exposure to violence and help seeking in the classroom

questionnaires. Looking across the entire sample, there were significant school differences

for almost every indicator of exposure to violence and help seeking (see Table 7). These

stark findings suggest that there may be important differences in the environment or

Table 6

Correlation Matrix for Classroom Surveys Across Schools (N = 799) a

Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Demographics

1. Gender

2. Race

3. Grade

.02

@.00

.08*

Exposure to violence

4. Physically fought

5. Witness fights

.16**

.09**

.09

.16**

@.13**

@.16**

.26**

6. Witness bullying .02 .08* @.17** .19** .38**

7. Physically bullied .05 .02 @.20** .22** .24** .34**

8. Verbally bullied .06 .10** @.17** .19** .25** .42** .54**

9. Cyber bullied b

@.04

.06

@.02

.15** .25** .21** .42** .37**

10. Racially harassed .12** .18**

@.05

.20** .18** .31** .27** .36** .25**

Help seeking

11. Help seeking scale @.12** @.10** .09* @.18** @.30** @.31** @.19** @.20** @.13** @.22**

12. Sought help c

@.10 @.02 @.04 @.06 @.05

@.16* .03

@.08 @.02 @.10

.18**

13. Not snitching @.09* .01 .06 @.12** @.12** @.11**

@.03

.00

@.01

@.10** .32** 0.19**

Composites

14. Witness Any Conflict .02 .10** @.24** .23** .66** .84** .34** .40** .24** .29** @.34** @.13* @.12**

15. Experience Any Bullying .09* .14** @.18** .25** .31** .46** .60** .83** .43** .59** @.26**

@.07 @.04

.46**

* p < .05

** p < .01

a Spearman’s rank@order correlations (ρ)

b Only includes 3 schools: Casy, Sharon & Milton (N = 623)

c Only answered by students who reported any kind of bullying or harassment (n = 329)

culture among the four schools especially as they relate to school violence or safety. For

example, Joad clearly had the highest mean score for every indicator of exposure to

violence while Sharon had the lowest scores. Casy and Milton ranked similarly in the

middle of these two schools. Milton and Casy’s similar exposure to violence scores is

somewhat concerning because of the negative correlation between conflict and grade level,

which indicates that high school grades should report lower scores than middle school

grades. For help seeking attitudes, Milton and Joad had the lowest scores and Sharon had

the highest score; however, given the positive correlation between grade level and help

seeking, Milton appeared to have lower scores than expected.

Out of the three program intervention schools (not including Joad), Casy had the

highest mean scores for all types of bullying, which is consistent with the correlational

findings between grade level and bullying. Milton also had the highest reports of racial

harassment out of these three schools (on par with Joad), which is noteworthy given the

greater racial diversity of its student body. In terms of help seeking variables, the only

significant difference among the schools was on the help seeking scale with Sharon having

the highest mean score. Though not statistically significant, students at Sharon were most

likely to report that they would not be considered “snitches” for reporting problems at

school and students at Casy were most likely to report getting help for their last school

conflict. Cyber bullying was also the only exposure to violence indicator with no significant

school differences probably because of low frequencies reported for this type of behavior.

In summary, I found significant school differences for almost all exposure to violence

variables and the measure of help seeking attitudes. Joad had the highest mean exposure

to violence scores, which sometimes approached two to three times the levels of other

schools. Sharon had the lowest levels of conflict and the highest level of help seeking

attitudes reported by students. These findings suggest real differences among the schools,

which could have some effect on students’ experiences in the mediation program as well.

Table 7

School Differences in Exposure to Violence and Help Seeking Variables in Classroom Surveys

(N = 848)

Classroom Items Casy Sharon Milton Joad Total

% or M(SD)d% or M(SD) % or M(SD) % or M(SD)

M(SD

)

Sample size, n220 240 212 176 848

Witness any conflict* a2.0 (1.4) 1.2 (1.2) 2.1 (1.1) 2.8 (1.1)

2.0

(1.4)

Any amount 81 63 94 96 82

Weekly or more 37 18 39 65 37

Witness fights 1.0 (1.0) 0.7 (0.9) 1.5 (0.9) 2.3 (1.1)

1.3

(1.1)

Any amount 58 50 89 92 71

Weekly or more 09 03 06 43 14

Witness bullying 1.8 (1.5) 1.0 (1.3) 1.6 (1.4) 2.2 (1.6)

1.6

(1.5)

Any amount 71 46 72 72 64

Weekly or more 36 16 28 49 31

Experience any bullying* a1.1 (1.4) 0.5 (1.1) 1.0 (1.4) 1.5 (1.6)

1.0

(1.4)

Any amount 49 24 43 57 42

Weekly or more 21 09 21 33 20

Physically bullied 0.5 (1.0) 0.2 (0.7) 0.3 (0.8) 0.8 (1.3)

0.4

(1.0)

Any amount 26 10 13 28 19

Weekly or more 09 03 06 17 08

Verbally bullied 0.9 (1.3) 0.4 (1.0) 0.7 (1.2) 1.2 (1.5)

.07

(1.3)

Any amount 37 18 32 44 32

Weekly or more 17 06 11 26 14

Cyber bullied 0.3 (0.7) 0.1 (0.5) 0.2 (0.7)

0.2

(.06)

Any amount 14 05 12

Weekly or more 04 01 01

Racially harassed 0.4 (0.8) 0.2 (0.7) 0.6 (1.1) 0.6 (1.1)

0.4

(1.0)

Any amount 20 09 27 24 19

Weekly or more 04 03 11 09 06

Physically fought*b16 09 15 29 16

Help seeking scale*a3.1 (0.6) 3.3 (0.6) 3.0 (0.7) 3.0 (0.5)

3.1

(0.5)

Sought help b59 51 43 53 52

Report without snitching b57 70 62 63 63

* Significant school difference (p < .05)

a Kruskal^Wallis rank test

b Chi^squared test

c Item not included at Joad

d Range: 0^never, 1^yearly, 2^monthly, 3^weekly, 4^daily

Within school and between group differences. I also considered differences by

gender, race and grade level in the whole classroom sample and at two schools, Joad and

Milton, because they tended to report the highest levels of conflict and had the most ethnic

diversity. Looking at the classroom sample as a whole, almost all exposure to violence and

help seeking variables had significant mean differences for gender, race, and grade level

with only two exceptions (see Table 8). There were no significant grade level differences

for racial harassment and no gender differences for witnessing any conflict including

fighting or bullying. Consistent with the correlation findings, male students reported

higher levels of fighting and experienced more bullying and racial harassment than female

students. Female students also had higher help seeking scores than male students. Also

consistent with correlation findings, there were significant grade level differences for all

variables except racial harassment, which indicated that middle school students reported

exposure to violence with greater frequency and lower levels of help seeking compared to

high school students.

There were significant racial differences among all of the measures including

exposure to violence and help seeking. Black students consistently reported higher

frequencies of exposure to violence across all categories sometimes at levels of two to three

times higher than other racial groups. They also reported the lowest help seeking scores

compared to other racial groups, which is somewhat concerning given that they also

reported the highest levels of exposure to conflict. There were also significant gender and

race differences in reports of racial harassment. Interestingly, Latino students reported the

lowest levels of racial harassment while also representing the largest, single ethnic group

among all the schools. Although black students were the second largest ethnic group in the

sample, they were still a small minority compared to Latinos; however, they reported

equally high levels of racial harassment as other minority groups in the sample including

whites and other/mixed racial groups.

Table 8

Group Differences in Exposure to Violence and Help Seeking in Classroom Surveys Across the

Whole Sample (N = 848)

Physically

Fought

Racial

Harassment

Witness

Any Conflict

Experience

Any Bullying

Help

Seeking

Independent Vars. n (%) M (SD) M (SD) M (SD) M (SD)

Gender .000* a.001* c.469 c.005* c.000* c

Female 51 (11) 0.3 (0.8) 1.9 (1.3) 0.9 (1.3) 3.2 (0.6)

Male 86 (23) 0.5 (1.1) 2.0 (1.4) 1.2 (1.5) 3.0 (0.7)

Race .003* b.014* d.001* d.001* d.000* d

Latino 95 (15) 0.3 (0.8) 1.9 (1.4) 0.9 (1.4) 3.2 (0.6)

Black 25 (31) 0.8 (1.3) 2.7 (1.3) e1.7 (1.7) e2.9 (0.6) e

Asian 03 (09) 0.4 (0.8) 1.5 (1.1) 0.6 (0.9) 3.3 (0.5)

White 04 (13) 0.8 (1.4) 2.0 (1.2) 1.2 (1.5) 3.0 (0.8)

Other/mixed 11 (24) 0.8 (1.3) 2.1 (1.4) 1.4 (1.6) 3.0 (0.8)

Grade Level .000* a.138 c.000* c.000* c.001* c

Middle School 86 (22) 0.4 (1.0) 2.3 (1.4) 1.3 (1.5) 3.1 (0.6)

High School 53 (12) 0.4 (0.9) 1.6 (1.3) 0.8 (1.3) 3.2 (0.7)

* Significant test value (p < .05)

a Pearson chi^square

b Fisher’s exact test

c Wilcoxon^Mann^Whitney test

d Kruskal^Wallis rank test

e Significant differences w/ Latinos and Asians

I also looked for similar group differences at Joad and Milton, which had the

greatest ethnic diversity and the highest levels of conflict among the four schools. The most

striking difference in these within^school analyses compared to the whole sample was the

lack of significant group differences, which may have resulted from the reduced sample

size. In other words, there may have been insufficient power to detect significant

differences also, in part, because of the reduced range of the data that results from highly

skewed distributions in the frequency variables.

At Joad, significant gender differences persisted for physical fighting, racial

harassment and help seeking, but there were no longer any differences in witnessing

conflict or experiencing any kind of bullying (see Table 9). Racial group differences also

changed on a few indicators, although differences in the experience of any bullying and

help seeking remained significant. For any type of bullying, there were no significant

differences between black and Latino students, but the few white students reported

extremely high levels. A sub^analysis of only blacks and Latinos on help seeking revealed

a significant difference between these two groups (p < .05). There were no significant

grade level differences among any of the variables at Joad.

Students at Milton, by contrast, appeared to more closely resemble the overall

trends in the entire sample; however, there were also far fewer significant group

differences (see Table 10). Males reported greater exposure to bullying including racial

harassment and lower levels of help seeking compared to females. Despite Milton’s ethnic

diversity, there were no significant racial group differences on any of the measures even

though black students reported the highest mean levels of exposure to violence. Grade

level differences were also consistent with the findings from the whole sample. There were

significant grade level differences for witnessing conflict and help seeking, but not for the

other variables. In general, grade level was negatively correlated with exposure to violence

and positively correlated with help seeking, which was not the case for Joad, but was the

case for the whole sample.

Table 9

Group Differences in Exposure to Violence and Help Seeking in Classroom Surveys at Joad (N =

176)

Physically Racial Witness Experience Help

Fought Harassment Any Conflict Any Bullying Seeking

Independent Vars. n (%) M (SD) M (SD) M (SD) M (SD)

Gender .001* a.006* c.946 c.143 c.001* c

Female 17 (18) 0.3 (0.9) 2.9 (1.1) 1.4 (1.5) 3.1 (0.5)

Male 34 (40) 0.8 (1.3) 2.8 (1.2) 1.7 (1.6) 2.9 (0.6)

Race .284 b.283 d.064 d.043* d.024* d

Latino 30 (25) 0.4 (0.9) 2.7 (1.1) 1.3 (1.5) 3.0 (0.6)

Black 14 (37) 0.7 (1.2) 3.2 (1.1) 1.8 (1.7) 2.8 (0.5)

Asian

(^^)

White 0 (0) 2.0 (2.8) 3.5 (0.7) 3.5 (0.7) 3.7 (0.3)

Other/mixed 07 (41) 1.1 (1.6) 2.8 (1.2) 2.2 (1.7) 3.0 (0.5)

Grade .097 a.867 d.840 d.823 d.064 d

6th 18 (38) 0.5 (1.0) 2.8 (1.2) 1.7 (1.6) 3.1 (0.6)

7th 12 (20) 0.5 (1.0) 2.9 (1.0) 1.5 (1.6) 3.1 (0.5)

8th 21 (31) 0.7 (1.3) 2.9 (1.2) 1.5 (1.6) 2.9 (0.6)

* Significant test value (p < .05)

a Pearson chi^square

b Fisher’s exact test

c Wilcoxon^Mann^Whitney test

d Kruskal^Wallis rank test

Table 10

Group Differences in Exposure to Violence and Help Seeking in Classroom Surveys at Milton (N

= 212)

Physically

Fought

Racial

Harassment

Witness

Any Conflict

Experience

Any Bullying

Help

Seeking

Independent Vars. n (%) M (SD) M (SD) M (SD) M (SD)

Gender .567 a.042* c.146 c.011* c.500 c

Female 15 (13) 0.4 (1.0) 1.9 (1.1) 0.8 (1.2) 3.0 (0.6)

Male 15 (16) 0.8 (1.3) 2.2 (1.2) 1.3 (1.5) 2.9 (0.8)

Race .425 b.538 d.775 d.211 d.823 d

Latino 14 (12) 0.5 (1.0) 2.1 (1.1) 0.9 (1.3) 3.0 (0.7)

Black 09 (25) 0.9 (1.5) 2.2 (1.1) 1.6 (1.7) 2.9 (0.6)

Asian 03 (18) 0.7 (0.9) 2.1 (1.0) 1.1 (1.0) 3.1 (0.5)

White 03 (13) 0.6 (1.2) 2.0 (1.1) 1.0 (1.4) 2.9 (0.9)

Other/mixed 03 (16) 0.6 (1.1) 1.8 (1.2) 0.7 (1.1) 2.8 (1.0)

Grade .052 a.334 d.001* d.175 d.020* d

9th 15 (25) 0.7 (1.2) 2.3 (1.1) 1.3 (1.5) 2.9 (0.5)

10th 07 (16) 0.8 (1.4) 2.5 (1.1) 1.2 (1.5) 2.6 (0.9)

11th 05 (07) 0.4 (1.0) 1.7 (1.0) 0.8 (1.3) 3.1 (0.6)

12th 05 (14) 0.5 (1.1) 1.8 (1.1) 0.8 (1.4) 3.1 (0.8)

* Significant test value (p < .05)

a Pearson chi^square

b Fisher’s exact test

c Wilcoxon^Mann^Whitney test

d Kruskal^Wallis rank test

Phase 2: Pre-Post Mediation Program Results

The data from the program questionnaires helped answer the second research

question about whether students experienced any changes in major outcome areas

including program satisfaction. In this section, I report on each of the five main outcome

areas including negative affect, positive affect, nonviolent response endorsement, help

seeking attitudes and program satisfaction. In summary, only negative affect showed a

significant change from pre^ to post^test, while there were no significant differences for

the other outcomes (see Table 11). However, this may be a particularly meaningful finding

for this study because of how negative feelings like anger influence students’ decisions or

efforts to resolve conflict in peaceful ways.

Table 11

Primary Outcome Scores from Pre to Post Intervention (n = 40) a

Outcome Measure (DV)

Pre^test

M (SD)

Post^test

M (SD) n

Negative affect * 2.1 (1.0) 1.4 (0.5) 40

Positive affect 2.2 (1.1) 2.1 (0.9) 37

Nonviolent response 2.4 (0.5) 2.5 (0.5) 38

Help seeking 3.1 (0.5) 3.2 (0.5) 37

Satisfaction

3.3 (0.7) 46

* p < .001

a For a breakdown of outcomes by school, see Appendix E

b Included in post^test only

Negative affect. In the first outcome area, negative affect, I asked students to rank

how much they felt a list of negative emotions when they thought about the main person

with whom they had a conflict. Students reported a substantial decrease from 2.1 to 1.4

(pre: SD = 1.0; post: SD = .5) in negative emotions after they participated in the mediation,

which was statistically significant (p < .001) and in the expected direction. I also found

three potential factors that influenced the change in negative emotions including 1)

measurement time, or the number of days that passed from the date of the intervention to

the date of the post^test; 2) grade level defined by middle and high school; and 3) the

school where the intervention took place (see Table 12). I excluded other potential,

independent variables from the discussion here because they did not have a statistically

significant relationship with the change in negative affect.

First, I considered measurement time as one possible influence on student reports

of a change in negative emotions after the mediation. Time appeared to be moderately and

negatively correlated (r = ^.31; p < .05) with a decrease in students’ negative affect scores

from pre^ to post^test. More students (76%) reported decreases in negative affect scores

who took the post^test two weeks or more after their mediation session compared with

students (57%) who took the test within two weeks. In other words, students who took

the post^test more than two weeks after the mediation were more likely to report a drop

in feelings like anger or sadness than students who took it less than two weeks after their

mediation. Next, I found that a student’s grade level or school was significantly related to

the change in negative affect (p < .05). More high school students (82%) reported drops in

negative emotions than middle school students (44%) by an apparently large margin. This

was true for both high schools even though one of them had slightly more students who

reported decreases in negative affect. By contrast, more middle school students (56%)

reported either no change or an increase in negative emotions compared to the number of

middle school students (44%) who reported a drop in anger or other negative feelings. In

summary, I found three positive findings for student reports of negative affect: 1) negative

affect decreased from pre^ to post^test for most students; 2) measurement time was

associated with the change in negative affect; and 3) grade level and school were also

associated with the change. Though significant, the quantitative findings are somewhat

limited because they do not necessarily reveal how or why negative emotions decreased

after the mediation. However, findings do indicate which groups of students from the

sample might be more likely to experience this drop in affect.

Table 12

Numbers of Participants Reporting Change in Negative Affect Scores from PreN to PostNtest (n

= 40)

Change in Negative Affect

Independent

Variable (IV)

Down (%) No change

or up (%)

Total p^value

Measurement time

< 2 weeks 13 (57) 10

(43)

.049* a

> 2 weeks

Grade Level

Middle

13 (76)

8 (44)

4 (24)

10 (56)

.02* b

High

School

Casy

18 (82)

8 (44)

4 (18)

10 (56)

.04* b

Sharon 9 (90) 1

(10)

Milton 9 (75) 3

(25)

Total Sample 26 (65) 14

(35)

* significant (p < .05)

a correlation (r) significance value used

b Fisher’s exact test

Positive affect. In the second outcome area, positive affect, I asked participants to

rank how much they felt “alert, excited, strong, interested and energetic” when they

thought about the main person in their conflict. Students reported almost no differences in

mean positive affect scores, which changed from 2.2 to 2.1 (pre: SD = 1.1; post: SD = 0.9)

after they participated in the mediation. This change was very small and non^significant.

However, two additional factors, or independent variables, revealed more about the nature

of change in positive affect (see Table 13). Results indicated that having a history of

friendship was significantly correlated with students’ reports of change in positive affect (p

< .05). 71% of students who said they were friends with the other person prior to the

conflict reported an increase in positive feelings compared to only 29% of students who

were not friends. I also found that gender approached significance even with a relatively

small number of boys in the sample (p = .06). 59% of girls reported increases in positive

affect compared to only 20% of the boys. All other independent variables from the study

were tested, but none were significantly associated with the change in positive affect. In

summary, while there was no real change in positive affect, there was a significant

association between a change in positive affect and a history of friendship. Gender also

appeared to be related to a change in positive affect and a larger sample of boys may have

resulted in a significant finding.

Table 13

Numbers of Participants Reporting Change in Positive Affect Scores from PreN to PostNtest (N

= 40)

Change in Positive Affect

Independent

Variable (IV)

Up (%) No change

or down (%)

Total p^value

History of friendship .02* a

Yes 12 (71) 5 (29) 17

No 6 (30) 14

(70)

Gender .06 a b

Female 16 (59) 11

(41)

Male 2 (20) 8 (80) 10

* significant (p < .05)

a Fisher’s exact test

b approaching significance

Nonviolent response endorsement. In the third outcome area, nonviolent

response endorsement, I asked participants to rank on a scale from one to four how likely

they were to respond with various prosocial actions when angry at someone such as

“ignore the person, talk it out, suggest a mediation, calm down, see the other person’s point

of view”, or etc. Students reported almost no difference in nonviolent response

endorsement from pre^ to post^test, which changed from 2.4 to 2.5 (SD = .05, pre and

post). All other independent variables from the study were tested, but none were

significantly associated with a change in nonviolent response. In an analysis of individual

scores, 17 students (44%) reported a lower score, 16 students (42%) reported a higher

score, and 5 students (13%) reported no change at all from pre^ to post^test (see Table

14). This general breakdown does not necessarily suggest any observable patterns in

changes in students’ attitudes about nonviolent response.

Table 14

Direction of Change in Students’ Nonviolent Response Scores from PreN to PostNtest (N = 38)

Direction of Change Frequency (%)

No change (=) 5

(13)

Down (^) 17 (44)

Up (+) 16 (42)

Total 38 (100)

Help seeking attitudes. In the fourth outcome area, attitudes about seeking help, I

asked students to indicate on a scale from one to four how much they agreed or disagreed

with a series of statements about getting help for various kinds of problems or conflicts in

school. Students reported almost no difference in help seeking from pre^ to post^test,

which changed from 3.1 to 3.2 (SD = .05, pre and post). I also tested other independent

variables, but none were significantly associated with a change in students’ beliefs about

help seeking. In an analysis of individual scores, 16 students (43%) reported lower scores,

20 students (54%) reported higher scores and 1 student (13%) reported no change from

pre^

to post^test (see Table 15). These results do not necessarily suggest any observable

patterns in students’ attitudes about help seeking after participating in the mediation

except for a slight majority of students who reported increases in their scores.

Table 15

Direction of Change in Help Seeking Scores from PreN to PostNtest (n = 37)

Direction of change Frequency (%)

No Change (=) 1 (3)

Down (^) 16 (43)

Up (+) 20 (54)

Total 37 (100)

Program satisfaction. In the fifth outcome area, program satisfaction, I asked

students to indicate on a scale from one to four how much they agreed or disagreed with

several statements about their experiences in mediation. Students reported a fairly high

degree of program satisfaction with a mean score of 3.3 (SD = .7) for the sample (n = 46).

About 24% of students had scores of less than three and 76% had scores of three or higher.

I also tested students’ scores of program satisfaction for associations with other

independent variables and found that gender was significant (see Table 16). More female

students (86%) reported average satisfaction scores of three or greater compared to male

students (45%) with similarly high levels of satisfaction (p < .05). All other independent

variables from the study were tested, but none were significantly associated with

satisfaction. Unlike the other outcome areas, program satisfaction did not have any

pre^test scores so there was no measurement of change over time. Although the sample

size of boys was small, it was somewhat disconcerting, but not necessarily surprising, to

learn that they reported less satisfaction than girls.

Table 16

Number of Participants Reporting More or Less Program Satisfaction at PostNtest (N = 46)

Program Satisfaction

More satisfieda (%) Less Satisfiedb (%) Total p^value

Gender

Female 30 (86) 5

(14) 35

.01 c

Male 5

(45) 6

(55) 11

Total 35 (76) 11 (24) 46

a avg. satisfaction score >= 3

b avg. satisfaction score < 3

c Fisher’s exact test

Qualitative Results

I also conducted follow^up interviews with all students who took the pre^ and

post^ questionnaires at the three school sites including Casy, Sharon and Milton. I used

this data along with interviews with the mediators to create a more complete picture of

students’ experiences in the program. The qualitative data from these two sets of

interviews helped answer the second research question about students’ experiences in

the mediation program and the contextual factors that influence students’ responses to

conflict. In fact, these data revealed a rich description of students’ experiences in the

program that the quantitative data was not able to provide. Many students were candid in

their responses and discussed a wide range of themes in their interviews (see Table 17).

Some students, particularly the younger ones, however, had more concise comments. In

this chapter, I present quotes as evidence to support the major themes that characterize

students’ experiences in the program. I was less likely to include exemplary quotes from

terse students, although I was purposeful my attempt to provide a balanced description of

the salient themes, which meant including as much disconfirming evidence as possible.

The qualitative findings here suggest that the mediation program has significant

potential for reducing conflict in schools as well as teaching students alternative ways of

resolving conflicts before they escalate into greater violence. Three major findings follow

from the results of the coding analysis presented in Table 17. First, students described

how feelings of anger and their ability to manage these difficult emotions played a large

role in their sense of self^efficacy for peaceful conflict resolution, or, in other words,

whether they felt they could respond to conflicts nonviolently. Second, contextual factors

including peer culture and students’ relationships with adults also had a significant

influence on their decisions to respond to conflict or seek help for their problems. Third,

students described how the undergraduate students not only mediated their conflicts, but

they also acted as mentors to them and provided them with personal guidance and

motivation. To be fair, not all students had positive experiences in the program and several

other factors contributed to their varying levels of satisfaction with the program. Despite

significant room for improvement, students’ comments suggested that the mediation

program provided them with a unique way of resolving problems, learning new skills and

developing positive relationships that would not be available to them in their schools

otherwise.

Table 17

List of Themes from Student and Mediator Interviews

Major Themes Subtheme 1 Subtheme 2

Changes in negative

Direct effects

Expressing emotions

emotions Indirect effects ^> Passive relief of feelings

Passage of time

Learning nonviolent

Learning to manage anger

responses to conflict Perspective^taking

during mediation Recommending mediation

Exceptions to learning (or no learning)

Contextual influences on

Peer norms ^> Maintaining an image

responses to conflict & Consequences for snitching

help seeking Adult relationships

^> Time available

& response to conflict Confidentiality

Responsiveness

Mediators as Mentors

Shared backgrounds

More than just mediation

Mediator self^disclosure

Program Satisfaction

Helps with self^expression

Unique process

Helps avoid fighting

Dissatisfaction ^> Lack of investment

Lack of confidentiality

Lack of neutrality

Emotionally challenging

Poor facilitation

Changes in Negative Emotions

Students’ individual comments about their participation in the program may help

explain some of the reduction in negative affect. In the follow^up interviews, I asked

students to recall how they felt before and after the mediation regarding the main person

from their conflict. As expected, students gave a range of answers about their feelings. I

coded their responses for negative emotions and identified several subthemes. First,

students described how participating in the mediation session helped them express

themselves including their emotions. Second, their comments suggested that the

mediation partially reduced the severity of their negative feelings in some indirect ways.

Third, the passage of time appeared to influence students’ feelings of anger, sadness or

fear, although this did not always occur in a linear fashion as the quantitative results might

suggest.

Direct effects: expressing emotions. In several cases, students attributed the

change in their feelings, such as anger, to having the opportunity to express their emotions

to someone else during the mediation. One middle school student who was involved in a

conflict as a victim of bullying described how her anger seemed to just ‘go away’ after

talking it out:

It [the conflict] made me feel sad…I felt like I must’ve thought wrong…that we could be

friends, and that we could do things we wanted to do from the very first. Because, before, I

was really angry. Now, I’m not that angry. I felt better than I was before when the fight or

problem happened. I talked it out. It just went out and no problems are stuck inside

anymore.

Similarly, other students also described how their negative emotions changed after talking

about them. In the next example, a high school student described how expressing herself

helped alleviate some of her negative feelings that preceded the mediation session:

It [the conflict] got me mad and upset because it didn’t only affect them, it affected me too. I

was sad too. At first, I had nothing to do with it, but then they brought me into it. [After the

mediation], it did change. It was good. I had a chance to talk and tell them how I was feeling

and tell them my side of the story. I think it helped me out because I talk to both of them

now. I have no problems with them anymore. I still talk to them, but we’re not as close as

before probably because of the problems and what was being said.

While this student experienced a decrease in negative emotions after talking about how she

felt, her example also suggests that restoring the friendship to pre^conflict status might not

be very realistic. In the last example, two middle students in a dating relationship came to a

mediation session because of rumors students had spread about them. Independently, both

students acknowledged feeling better after talking about the problem. When I asked the

first student about her feelings toward her boyfriend both before and after, she replied:

Sad because we got in an argument, but then when I found out he called the mediation, I

knew that he cared. I felt a little bit better… They [my feelings] changed a little. I heard

what he had to say and I said my feelings. I knew how he felt and he knew I how I felt, so

they changed a little bit. I felt a little better since he heard what I had to say and I heard

what he had to say.

I also asked the second student how he felt about his girlfriend both before and after the

mediation session and he responded similarly:

I was mad at her believing it [the rumors] even though it wasn’t true. I was sad to know

that she was sad. [After the mediation], I felt better. We felt better because we had to talk it

out. I told her that it wasn’t true what people were saying, the rumors. She understood me

so that was pretty much it.

For both of these students, having the opportunity to express and validate each other’s

feelings appeared to play a significant role in reducing their negative emotions. The

comments here suggest that, at a minimum, some students find mediation helpful for

alleviating negative or unwanted emotions even if they do not always remain friends.

Indirect effects.

Passive relief of feelings. The change in negative feelings can also happen more

passively as a residual effect of simply going through the mediation process. As the last

example illustrated, the student reported feeling better after discovering that the other

student initiated a mediation session. Furthermore, even a less than ideal outcome to a

conflict can help mitigate students’ negative emotions. For example, I asked a middle school

student about her feelings before and after her mediation for a verbal conflict with a group

of her friends and she responded that she initially felt:

Upset. I was shocked. Angry because they don’t have a reason to be mad. [afterwards]

Relief. Now I know who my true friends are and I can stay away from them.

While she didn’t attribute the change in her feelings to the mediation itself, learning where

she stood with her friends brought her some degree of comfort even if the outcome wasn’t

the most desirable. Gaining more information or insight about the conflict itself appears to

help students cope with negative feelings even if the quality of their friendships never

returns to pre^conflict levels. In the next example, consider a high school student’s

comments about an especially difficult conflict involving rumors and multiple people:

It all started out with rumors from guys. I kinda despised her for what she said because I

thought she was a friend. [I felt] like appalled and vivid. [Afterwards], it changed

dramatically because I saw the truth. It wasn’t as I thought and it wasn’t as she thought on

both sides. We kind of cleared it up. So now it’s over and it’s neutral. There’s the

awkwardness of what happened, so we can’t get to that level that we used to be as friends.

Several students commented similarly that they felt less angry after hearing the “truth”

about what transpired between them and the other parties in a conflict. One conclusion

from this might be that the mediation process does not eliminate all negative emotions, but

instead partially alleviates very strong feelings of anger or fear. For example, one high

school student who was bullied and physically threatened by another group of students to

the point where she did not want to come to school shared the following:

I was angry because, me and the other girl, we never really talked bad about them to other

people. When we heard that these girls were talking about us and they were telling other

people our business, it got us angry because we didn’t want anyone to know what was

going on. We were just angry that they knew we never said anything, but they did say

something to other people. [Afterwards] I still kind of have that anger towards them, but I

try to distance myself from it because I don’t want it to affect me, my personality, or my

attitude towards people. I don’t have a problem with them, it’s just the whole knowing

what they said still gets to me. Other than that, it’s not a big deal. It’s not where I was

feeling weird coming to school sometimes because I didn’t want to see them.

While this student recognized that she still felt angry after the mediation, she also

acknowledged that she no longer felt as uncomfortable going to school as she did before

the mediation.

Taking these students’ comments together, negative emotions appear to change not

only as a direct result of expressing one’s emotions during the sessions, but also indirectly

as the mediation process unfolds and students gain a deeper understanding of their

relationships or the conflict itself comes to a resolution. Lastly, in some cases, participating

in the mediation process may partially reduce the intensity of students’ negative emotions

even if they persist to a lesser degree after the mediation session.

The passage of time. Measurement time was a significant factor associated with the

drop in students’ negative emotions after the mediation. If the adage that time heals all

wounds is true, several of the students I interviewed more than two weeks after their

mediations made comments that seemed to support this saying. However, negative affect

did not always decrease steadily, but rather some students reported fluctuations in their

emotions over time. First, as an example of the linear case, I interviewed one middle school

student more than two weeks after her mediation took place. Not unexpectedly, this

student’s emotions appeared to succumb to the dulling effect of time:

Yeah, because, as far as I know, I don’t feel any hatred or jealousy towards her. Honestly, I

don’t even know what were fighting about anymore. I just forgot about it. So everything’s

good.

Another student who I met with, unfortunately, more than one month after the mediation,

remarked similarly that she was also “over” it:

It doesn’t bother. It’s just like ‘whatevers’ now. [It used to bother me] in the first two

days, but not anymore.

Comments like these two examples seem to suggest that emotions associated with the

conflict become less intense and could be forgotten entirely after a certain amount of time.

Contrary to these examples, negative affect does not always change in a linear

fashion, i.e., from high to low, or vice versa. The original conflict or a related problem could

flare up again at any time after the mediation, which may change the trajectory of students’

feelings. For example, two high school girls with a history of friendship had used mediation

to address a conflict in their relationship. One of these students described how her feelings

of anger fluctuated up and down for weeks after the mediation:

I was really upset at her and it was frustrating that I wasn’t getting answers from her and I

was getting answers from other people… After the mediation, for a couple hours, er for a

day or so, it made me feel much better…But with this friend, she had told our good friend a

couple of things about…she wasn’t being very truthful or very genuine in mediation. So

after I found that out, it made me feel totally upset again and enraged with her. Other than

that, we’re okay.

The other student in this conflict also commented on how her own feelings went in

different directions for at least two weeks after the mediation:

I was ready to talk to her. I think all I was trying to do was tell her how I’ve been feeling. I

was upset. I wasn’t angry. I was displeased because I expected a lot from her…

Yeah it [how I felt] did [change]. It took awhile. I told her I needed space and eventually

when I talk to you, that’s when it changes. So it did take awhile, about two weeks, because it

was on and off. Yeah, it’s not the same. We don’t hang out, but we talk and she tried to open

up. And I tried too also, but it’s not the same.

In these examples and others, the amount of time that passed after a mediation session

appeared to have an effect on how students rated and talked about their negative feelings.

While the majority of cases seemed to follow a linear, decreasing trend in feelings of anger,

sadness or fear, it was not completely unusual for students’ emotions to ebb and flow over

time especially if the original conflict resurfaced or one student discovered another was

being disingenuous in the mediation.

Learning to Respond to Conflict Nonviolently

During their individual interviews, students talked at length about how they would

respond to potential conflicts in the future. Some students discussed specific techniques

they learned from the mediators about responding to conflicts more peacefully or

effectively. Other times, students admitted to not learning anything new or helpful from

participating in the mediations. Students often described how they wanted to be more

successful at managing their anger, but they also recognized how difficult it was for them to

control their emotions in these situations. While not all comments were positive, it was

encouraging to hear students talk about learning practical skills from the mediators and

how they might apply them to future conflicts. I organized students’ learning and attitudes

about nonviolent responses to conflict into four main areas. First, I discuss how students

learned more effective ways of managing their anger from participating in the mediations.

Second, I review their comments about learning to see another person’s perspective during

a conflict as well as some exceptions or limitations to this skill area. Third, I discuss

students’ use of mediation itself as a tool for dealing with conflict. Finally, of course, not all

students spoke positively about learning nonviolent responses to conflict, so I conclude

this section with examples of students’ divergent opinions or experiences in this area.

Learning to manage anger. Students’ comments revealed that the ability to control

anger has a large influence on how likely they are to respond to conflict nonviolently.

When prompted, most of the students could parrot the appropriate course of action to take

in a conflict, but they were also keenly aware of the difficulty of practicing these ideals

especially in the heat of the moment. Despite the challenges, students still attributed some

of what they learned about responding to conflict to the mediation program. In the first

example, a middle school student described how she was going to physically fight another

student because she was so angry. After I asked her what she thought she had learned from

the mediation, she said that the mediators taught her:

How to move on. They taught me how to hold back from being so angry and letting it all

out. They taught me how to hold it in and just be strong. I’d either come to mediation or

just talk to them one^on^one and make an agreement that I don’t want conflict

anymore. And if they [the other person] don’t want to, then I’d just give them their

space.

As this student points out, an important first step to effectively responding to conflict

includes controlling difficult emotions like anger. Mediators are also in a position to talk

about how to control anger more effectively even if students aren’t always successful at it.

One high school student involved in a conflict about rumors with a group of three other

girls described how the mediators helped her with this, but also acknowledged how hard it

was for her to put these skills in practice:

They gave me ideas about how I can calm myself down, but sometimes it doesn’t work. So

they showed me new things to do, and I should join sports and stuff, and it’ll help. And it

does… [I learned] I shouldn’t listen to what other people tell me. If things are brought to

my attention like that, I should sit down with her calmly and not just react in a second. I

said it was a good idea, but I have a hard time. I really do have anger issues. When I find

something out or when I’m mad, I snap.

In these examples, the mediators helped teach students new ways of controlling their

anger, which would allow them to respond more effectively to their conflicts with other

students. However, some students, especially those with difficulties controlling their

anger, may need more practice than one meeting with the mediators in order to

adequately internalize the skills that allow them to respond to conflicts nonviolently.

Besides teaching practical skills, participating in mediation sessions also seems to

help students develop a greater awareness of how anger affects the way they respond to

conflict. For some students, the mediations provided an opportunity for them to reflect on

how their emotions affected their decisions and responses to conflict. For example, one

high school student was in a conflict that began with rumors and escalated into threats of

physical violence against her. This student described in detail how she learned that her

emotions were a real barrier to effectively working with another person in a conflict:

It’s hard. It’s really hard because it doesn’t feel good to have these things said about you or

said towards you to your face and react in a calm way. Because there were times when you

wanted to say something in a mean way because you think that that’s how your point is

going to get across, but I think that’s the hardest part; to keep your cool and just respect

them because if you’re angry with them, you don’t really care about how they feel. You just

want to attack them. [laughing] I think that’s the hardest part, but it’s [the mediation] a

good experience because it helps you with other situations related to that…I think just the

way that we handle things: not jumping to conclusions when somebody says something

towards you, or not reacting in a certain way because people want that from you; they

want a reaction to start something bigger; not to tell everybody your business either. I

think

that’s a really big one because then people twist what you say and they tell people your

business especially in a small school. So that’s not a good thing to do.

While this student recognized the difficultly of controlling her anger, she also appeared to

gain a deeper perspective on the consequences of her actions on others. The student then

went on to describe how she would approach a similar conflict about rumors in the future:

I think I would confront the person directly. I would be a little scared. I don’t like physical

violence. It kind of scares me a little. I don’t think anything’s going to get resolved out of

physical violence. I think if you talk to the person and just kind of clear what happened,

then that’s a better way to solve the situation.

Another student, who was part of the group of girls that instigated the threats against the

student from the last example, described how the mediation taught her the importance

of gathering more information before her anger takes over:

Looking back, I was like, ‘what the hell? That was stupid. Why did I do that?’ Talking it out

was such a better way to get the truth out and it was so much calmer. We had both sides

from each other and it was like, okay, okay, this was it. We just brainstormed things out

that we had thought we had heard. It was just so much more helpful than doing it on your

own. [I learned] to question people. I’m not going to rely…I told them, ‘this is an experience

I’m taking to college.’ Don’t believe everything you hear. If I do hear something, I’m going

to go up to them and question them calmly...because I’m an angry person! I’m an angry

person.

This student recognized that her anger could interfere with her ability to remain calm,

which is often an important prerequisite for peacefully resolving conflicts with others.

In these examples, the mediation process seemed to represent a unique experience for

students. It helped them learn not only about alternative responses to conflict, but

also about how their own emotions can influence their reactions in undesirable ways.

The

mediation sessions also appeared to provide a certain level of structure and safety, which

may further facilitate the development of these kinds of insights.

While the students in these examples appeared to learn from their mediations,

it’s difficult, or impossible rather, to know whether their future behavior would actually

change without the additional scaffolding provided by the mediators. Students who are

more likely to be involved in conflicts may need multiple opportunities to practice. When it

comes to self^regulation, as another high school student put it, learning how to manage

anger, especially under duress, is likely to be a long^term process:

I might like forget the whole self^control thing and I might go back at it, but then I might

think about it and be like, alright, whatever, and just leave it alone… so I would try to take

it into the right hands and not the wrong hands and just be like, okay, whatever, they’re

gonna be petty, just do what you got to do…because I do want to go to college and I want to

have a successful life. If it does get to the point where it’s just like, ‘okay, I’ve had enough,’ I

would go to the Deans and have a mediation. My mom even tells me like, ‘you talk back and

you do this!’ And I’m like, ‘okay, I have to work on self^control even more,’ because I’m just

now getting self^control. I think I might just snap back and be like, ‘okay, I shouldn’t have

gone that way. I should’ve went the right way.’

Across multiple conflicts and mediations sessions, students described how their ability to

control anger plays a large role in how well they respond to conflict. Students could also

point to specific suggestions from the mediators and parts of the mediation process that

taught them valuable lessons about how to control difficult emotions in future conflicts.

However, in their interviews, the students also reminded us that this is not easy to do – in

fact, for many students, participating in the mediation may only represent the first step in a

much longer process of learning how to respond to conflict nonviolently.

Perspective taking. During their individual interviews, students repeatedly

commented on how taking another person’s perspective was an important skill they

learned from participating in mediation. Hearing from an opposing party seemed to give

them a better grasp of the current conflict that brought them to mediation and how they

might respond to future conflicts in less aggressive way. A mediator at one of the high

schools explained remarkably well how this process unfolds during a mediation session:

I think what helps is when they’re honest, especially when the victim is honest about how

she feels. I think also seeing the human emotions…when they see someone else tearing up,

having a hard time talking about the situation. I think it almost creates some kind of, not

just guilt, but they sympathize with that person. They don’t see the actual damage they’re

causing to someone else until it’s in their face and they see them crying or having a hard

time dealing with it. At least, in that situation, that’s how I felt. We were able to have them

apologize because they saw the other person crying.

Being able to actually see the consequences of their actions on another person from a

conflict had a profound impact on students. In the next example, one middle school student

went to a mediation session with one of her “former” friends who had insulted her in front

of her peers. She commented on how the mediators helped her see things differently:

They understood us. They kept on like, ‘yeah I get you.’ It wasn’t “if” and “but”. They were

trying to pull out both of our points of view, not just mine or hers. I got to see why she was

mad and if I was in her shoes, I would be upset too. That helped because I think she’s mad

just to be mad, and if I was in her shoes, I understand why she’s mad.

When students hear about the other person’s experience in a controlled setting, they have

an opportunity to look behind the externalized emotions and understand the problem on a

deeper level. This ability to see things from another’s perspective may be an important

antecedent to resolving these conflicts nonviolently.

A bullying incident from the middle school illustrates how perspective taking is a

significant part of learning about how to resolve or prevent conflicts for both “bullies”, or

perpetrators of the aggression, and the victims. In this example, a group of three girls was

harassing another student who had a reputation for being a victim in other incidents of

bullying at their school. A notable, but certainly all too common aspect of this incident was

that, to some extent, the aggressors seemed to lack awareness of the full impact of their

behaviors on the victim. One of the bullies described how she felt when she heard the

victim share where she was coming from:

Everyone told their part of the story, which helped. [Victim’s name] said that because she

doesn’t really have anyone who likes her, she says she has a lot of problems. She said that

she had overreacted because no one was on her side. I felt bad because I turned against her

and she didn’t have anyone.

In a similar fashion, one of the other aggressors disclosed that talking about the conflict

made her reflect on a time in her past when she was also the victim of bullying:

[Mediator’s name] was asking us, ‘do you think you can be friends?’ Even [the victim] was

sharing stories of her past and we knew what it was like for her so we felt sorry for her.

We kept telling stories back and forth and we realized that our stories are kind of the same

because [friend’s name] and me were kind of getting bullied when we were younger and so

was [victim’s name]. Nobody liked her. It made us feel that we saw the other person’s point

of view.

In addition, the mediator facilitated the process of helping the students to reflect on their

involvement by sharing one of her own experiences. One of the so^called bullies

commented on how this helped her to see the victim’s point of view:

I felt like, ‘oh I never knew this was going to happen.’ I thought she was going to smile. I

thought she was going to laugh. I never wanted that to happen… She [the mediator] talked

about it, and also that she went back to her stories about when she was bullied, when she

was bossed around, and we can see what [the victim] felt.

According to this student, the mediator helped the students empathize with the victim by

disclosing her own personal experiences with bullying. No matter what their role was in

the conflict, students frequently made comments about perspective taking similar to the

ones presented here. After participating in the mediations, students appeared to benefit

from an increased awareness of the emotional impact of their actions on others. Often

times, the mediators were instrumental in facilitating this unique learning opportunity.

While there were several examples like these where students benefited from seeing

another person’s point of view, they were not always able to make connections between

their own behaviors and their effects on the other person. For example, in a conflict

involving two high school boys in the locker room, one of the students seemed to initially

understand where the other one was coming from:

It’s understandable, because, honestly, if I’m in a bad mood especially…he didn’t really say

why he was in a bad mood, but I guess it was just that day because he usually messes

around in class and I was like, ‘oh okay. I guess it’s just his bad mood, so he took it a little

bit too far that day.’

The student’s comments seemed to imply that he understood why the other student

reacted angrily; however, later in the interview, the student also described how he didn’t

quite understand why his response of laughing at the student had escalated the problem:

He just said it ticked him off a little bit and it got him annoyed because it’s like I’m just

trying to test him because you know how he said, ‘laugh one more time, and I’m going to

beat you up.’ And I just kept doing it and I guess I was testing him…That’s my way of

‘retaliating’ the mediators said: ‘oh is this your way of retaliating?’ I was like, oh that’s how

I do it…I purposely did that. I don’t see why is it so bad when you’re just laughing because I

was just genuinely cracking up when he’s threatening me because I just think it’s pretty

funny when people make threats at me especially since I’m a pretty positive person in

school.

The mediator in this conflict also made a similar observation about this student implying

that he didn’t really grasp the full impact of his actions. At the other high school, one of the

mediators similarly commented that a student from her session had difficulties with seeing

the problem from the other person’s perspective:

I think she understood what mediation was, but I don’t think she understood that you

shouldn’t just blame the other person and not look at things from the other person’s

perspective, which is important in mediation for it to work. It was not going to be possible

to get her to admit to things or to really look at things from the other person’s perspective.

She was not really listening…as in she didn’t care. It was going in one ear and out the other.

But if someone’s 15, it’s not uncommon to not really understand the situation.

While perspective taking is an important part of resolving conflicts, these examples suggest

that several barriers that can interfere with this learning process. In the first conflict from

the locker room, the mediators may have neglected to surface the underlying emotions

behind the boy’s anger. In the second example, the student may have lacked a sufficient

level of investment, or, even, maturity for making this type of empathic connection. These

outcomes are not always guaranteed and may require a certain level of skill in facilitation.

Recommending mediation. Students who participate in mediations sessions may

learn new, prosocial ways of handling conflict, but they may not always be successful at

implementing them in the moment without some assistance or scaffolding from a third

party. As an additional approach to resolving conflicts nonviolently, I asked students if they

would use mediation themselves or recommend it to others for similar conflicts in the

future. The majority of students seemed to endorse this strategy while some acknowledged

important limitations or caveats. One high school student who was involved with spreading

rumors about another student said she would recommend this approach:

Yes. The way I would’ve taken it and the way I know my friends are, it’s just like, ‘yeah,

let’s go get them. Let’s go talk shit.’ Here it’s like, you’re not going to get a conclusion out of

that there. In here, you get to talk it out. You get to tell your story to someone. The other

person gets to tell their story. There’s a bystander [mediator] who will always point out

the things you would never catch on your own.

Unfortunately, there is no way to know if the student would actually follow her own

endorsement of this nonviolent approach; however, the student recognizes that mediation

has certain advantages over aggressive responses such as reaching an actual resolution.

Another high school student, who had used mediation the previous school year, described

how her past experience with the mediators actually helped with her present conflict:

It definitely opened my eyes a little more. I had dealt with something similar to this before

and I didn’t react in a good way. By them telling me how to react was really helpful for me

for future situations. Because in this situation, I didn’t really say anything. I just ignored it,

but it got to the point where it escalated in me. And with the previous situation, I would

backfire or say something and they would say something and it was just a back and forth

argument. But now, we resourced to mediation so the back and forth wouldn’t get bigger. I

feel that if there wasn’t a mediator, it would’ve gotten physical so they did help a lot.

Not only did the mediation prevent possible violence between her and the other girls, but it

also helped her refrain from “backfiring”, as she put it, or letting anger dictate her response

to the conflict. Consistent with previous themes on nonviolent responses from the data,

the mediation session allowed her to express herself in a more constructive, peaceful

manner.

While many students endorsed mediation as an alternative strategy for conflict

resolution, not all of them said they would return to mediation or recommend it to others.

Students who had more unsatisfactory experiences or a more pessimistic outlook about

their peers’ adoption of mediation did not recommend this approach. For example, one

high school student who was the victim of rumors and called inappropriate sexual names

by other girls did not view mediation very favorably:

I wouldn’t go back to this thing [mediation] because I already went through it. Just for

talking again, I wouldn’t want to go through the process, like calling me and calling the girls

and having a meeting. I just wouldn’t want to go through it again. [I would feel] more

frustrated because people won’t be completely honest. I would probably just ignore it.

This student’s frustration with her experience is not entirely unexpected especially in these

situations that involve very sensitive conflicts and multiple students. Also, dissatisfaction

with the mediation process may discourage students’ endorsement of mediation as a

realistic, nonviolent strategy to resolve conflict. For example, another high school student

described how he wouldn’t necessarily recommend mediation to his peers because, in his

view, it’s just not a good fit for how they choose to handle their conflicts:

I guess people who are more understandable would take mediation more serious and

other people who are just more reckless and are like, ‘oh I’m going to beat him up no

matter what…’, and I actually know people like that. Nothing’s going to stop them. Not even

a little ticket or consequences will stop them. They’re just going to do it. Even after a

mediation, they’re going to get even more pissed because they have to see that person.

Everyone’s different.

This student accurately notes that no single strategy will work for all students or all types

of conflicts. Similar to the examples presented here, other students also shared the feeling

that mediation may not be very effective in some extreme situations, although most

students viewed mediation as an acceptable approach for resolving conflict.

Exceptions to learning nonviolent strategies. In addition to these comments

about learning new, nonviolent responses to conflict, students and mediators also provided

counter examples of not gaining any new, prosocial skills from their mediations. For

example, I asked one of the high school students I interviewed if she had learned anything

from her participation in mediation and she responded:

Not really, not more than I already know…I didn’t want to have anything to do with her. I

was just done. There was no reason for her to say anything to me, and if she does, I

would just either ignore it or I would say something back…

While the student’s honesty here was at least refreshing, in general, students may not

always report positive learning experiences especially if the outcome of the mediation was

negative or the intensity of the conflict was high. In another example, one high school

student was skeptical about taking anything of value from his mediation experience:

They [the mediators] talk a lot. They give a lot of information and whatnot, but yeah I

guess. I probably learned to talk one^on^one. I don’t know honestly. It’s just the whole

process. The mediators are just there, the middleman. Honestly, we didn’t need the

mediators.

In fact, another middle school boy also complained that the mediators talked too much,

although the sample was not large enough to know if these were isolated incidents. When I

asked another high school student how she would respond if a boy from class teased her in

in the future, she spoke candidly about how she would handle it:

I would tell him to chill out or just to stop joking around with me. At first, I would try to

talk about it, but I’m saying…let’s say we’re talking about it, and if he like touches me and if

it hurts me, of course I’m going to do something back. I’m not going to necessarily let

myself, but I’m going to do something back. I would try to solve it by telling him, like don’t

play with me like this or something because I’m going to end up hurting you or something

like that. I think I would just be like, oh maybe you should calm down because you know

that you and I are friends.

100

In this example, trying to assess the student’s learning is difficult because she referenced

both nonviolent and aggressive responses to the conflict. In fact, the student also claimed

that she would revert to physical aggression out of self^defense if her verbal rebuffs failed.

Not only is further probing necessary to determine whether this student learned anything

from the mediation directly, her comments also raise additional questions about situations

when aggressive or violent responses to conflict might be warranted.

Students may also be less likely to report learning new ways of responding to

conflict because of poor facilitation of the mediation sessions. For example, novice

mediators could lack the adequate experience, training or skills necessary for addressing

the sensitive and complex nature of conflicts. I asked one of the mediators at a high school

if she thought her students learned anything from her mediation for an intense, verbal

altercation that started online:

No, not really. I try to say it whenever I have a mediation, especially to seniors, the older

students: what happens in high school is petty drama and it’s not going to affect you for the

rest of your life unless it’s really something big. So I had that Twitter one [conflict]. I was

just like, ‘you guys, really think about this. This is nothing. You’re graduating in a month.

Get over it.’ It’s kind of harsh, but it’s the truth. I try to say that, is it really worth it? Is it

really worth this fight? Maybe that’s what they learned [laughing].

The mediator’s apparently insensitive remarks about how she viewed the students’ conflict

may have limited their ability to learn new ways of responding to conflict nonviolently.

However, the mediator also emphasized that students just ignore the problem, which could

be effective in some situations, but, unfortunately, may not have been the appropriate

response for these students here. These less positive examples of students’ experiences

suggest that multiple factors are likely to contribute to student learning such as

satisfaction with the outcome, the intensity of the conflict or the experience of the

mediators.

101

The lack of uniformity in students’ reports of learning nonviolent responses to

conflict seemed to be consistent with the quantitative findings, which did not support a

single trend in this outcome area. For many students, it was easier to verbally state the

“correct” way to handle conflict than to put these beliefs into practice because of

difficulties with anger. In this regard, mediations may only be a first step in otherwise

lengthy learning process. However, students also reported positive learning experiences

related to nonviolent approaches including the importance of managing anger, taking

others’ perspectives and recommending mediation for future conflicts. Despite the

complexity of measuring this outcome area, it was clear that some students reported

positive learning experiences and derived real benefits or skills from their participation.

Contextual Influences on Responses to Conflict: Peers and Adults

The qualitative data revealed several factors in the environment that influenced

their response to conflict and their decisions to seek help. Peer pressures against

“snitching” and cultural norms about asking for help were significant barriers for many of

the adolescents in this study. In their interviews, students also described various

characteristics of their relationships with adults that influenced how they responded to

conflict or sought help for problems. Interview questions about the differences between

mediators and teachers illustrated these important aspects of the adults’ response to

conflict. Students’ comments suggested that they were more likely to seek out help when

their relationships with adults had high levels of trust and responsiveness. Students’

described these relationships as ones where adults had adequate time available, protected

their confidentiality and provided effective assistance. While seeking help can be daunting

102

for adolescents, it also appears to be a viable way for students and adults to work together

to prevent the escalation of conflict into more severe forms of violence.

Peer norms. During the interviews, students discussed various kinds of barriers or

challenges they encountered if they wanted to resolve conflicts with their classmates.

Many of their comments indicated that peer attitudes about acceptable behavior and

concerns over reputation, which we often associate with adolescence, deter students from

finding constructive solutions for their conflicts. For example, one student at a high school

commented on how the pressure to conform makes it difficult for students to not only

make good decisions during conflicts, but also follow the advice that adults give them

about their problems:

I would still give it [mediation] to anybody and tell them to go, but it’s just how certain

people are. They just don’t want to hear that, or they just don’t want to go with that. They’ll

just go right back and be like, ‘whatever, that didn’t mean anything.’ It really has something

to do with society. You want to act like the next person so you’re going to do what the next

person does. You’re gonna act all hard and be like oh whatever, I don’t care; then go back to

being how you were instead of taking the advice that people are trying to give you and just

use it and be like, ‘oh, I don’t have to act like that. I can act like my own person. I can be

whoever I want to be.’

Many of the students in this study also shared this description of peer culture, norms

about “snitching” and the need to maintain an image of toughness or independence. One

high school student who was harassed with sexually inappropriate comments described

how the pressure to act independently discourages students from seeking help:

Because like I said, they feel like it’s not going to help. A lot of influences from the school

tell them, ‘you know, if you tell someone, you’re a snitch. If you tell someone, it’s because

you can’t do it yourself.’ And I feel like that’s a lot [of pressure]…not because of you adults,

but as in the kids, they feel like they’re either going to be snitching or…since they feel like

my friend can beat that guy up, then I could too. I feel like they don’t come to you guys just

because their friends [would say], “you’re snitching or you’re not going to help”. You have

to be the bigger person and you deal with it yourself, but I feel like if they did know and

talk to you guys, they would think differently. But as long as they have those guys and girls

who influence them, it’s not going to work.

103

In a similar fashion, one high school boy described how maintaining an image or

reputation of toughness may be just as important of a survival strategy as not being labeled

a snitch:

They’ll probably think that they’ll [the mediators] get them in trouble or something like

that. They’ll probably think it won’t help when it really does. The other person that’s

involved will probably think the person would…probably look at them like as a bitch and

shit. He’ll probably think he ratted on him or something like that. They have their own

reasons for getting in the fight. If you snitch on them, you get it twice as worse…you’re

getting the other person in trouble and that’s just another reason to kick their ass [laughs].

In this example, the student described several negative consequences for so^called

snitching such as appearing weak as well as getting beat up. Unfortunately, the pressure to

avoid being a ‘snitch’ is so strong that students may ultimately choose to escalate the

conflict in the absence of a viable alternative. As one high school student who was

harassed in the classroom put it:

I would be scared to tell him that I would tell a teacher because you know how people start

calling you a ‘snitch’, so I would just talk back I guess. Usually that’s what happens.

In all three schools, students often pointed out that serious consequences resulted if peers

discovered their attempts to tell someone about a conflict. One of the mediators described

the difficulty she had with addressing this aspect of peer culture with her students:

They said they’re scared. They’re scared of the consequences. Considering it’s a middle

school, the most important thing to them is people spreading rumors, people telling your

secrets, your reputation at school. The biggest issue for them is if we tell on them, we’re

going to have to face the consequences because we snitched, or they’ll start spreading

rumors and telling everyone we’re tattle^tells…But when we brought up the whole “telling

someone” and they were afraid of the consequences, they brought up, what if they start

spreading rumors about me? And that’s where I got stuck because I didn’t know what to

tell them. That is really big, especially in middle school…

Aggressive behaviors such as spreading rumors, damaging others’ reputations and physical

violence are all ways that students may retaliate when peers violate these cultural norms.

104

Unfortunately, as a result, the consequence of snitching may hamper students’ motivation

to seek help from adults as a possible nonviolent response to conflicts.

While certain aspects of peer culture in adolescence make it difficult to get help,

these barriers are not completely immune to outside influences. Some students, who might

be considered mavericks among their peers, did not seem as concerned about their

reputations. One of the mediators at a high school noted that some of the students in her

mediations learned about how projecting an image of toughness could be a way of

achieving an ulterior motive such hiding feelings of hurt or vulnerability:

And for that person to see them in a different light too because they’re over here thinking

this person has this image and she’s tough. From a lot of the students, it sounds like they

want to have this image with their peers because it gives them some kind of status. The fact

that they were able to see beneath that, I think that was a breakthrough.

Participation in mediation can expose students to alternative perspectives on their

problems, which, as this mediator points out, was a key part of resolving this conflict.

However, not all students respond to peer attitudes or cultural norms in exactly the same

way. In fact, some students reported that they were not very concerned about their image

or reputation amongst their peers. When I asked one high school student about whether

she was concerned about seeking help for a conflict, perhaps as a sign of maturity, she said:

No, there’s the, ‘oh, you’re a snitcher.’ I would rather be called that then to get in a physical

fight or for the problem to get bigger, so I’d rather just settle down the problem.

When I asked another student at the other high school about getting help, she also shared

the previous student’s sentiment and declared similarly:

Like they’ll think I’m a tattle^tell or something? That doesn’t bother me. When I was little, I

would tell on people and they would be like, ‘oh you’re a snitch.’ I really don’t care what

people think about me.

105

Of course, while only a few students made comments like this, it’s important to note that

peer attitudes about these behaviors does not affect all students in the same way; however,

the majority of students’ comments remind us that certain aspects of peer culture present

real barriers to seeking help and the consequences for violating these norms are real.

Adult relationships and response to conflicts. Since the mediators have a much

different role than most of the adult staff on campus, I had a unique opportunity to examine

students’ relationships with these groups of adults influenced their responses to conflicts

and help seeking. For example, I asked students about their perspectives on getting help

from teachers compared to the mediators at school. Their responses revealed several

relationship qualities that seemed to influence the likelihood that students would seek help

for a conflict. Students said they would be more likely to get help from adults who had

enough time available to devote to their problem, who protected their confidentiality, and

who were responsive to their needs. Students trusted so^called responsive adults with their

conflicts because they provided a high level of competent support and did not discipline

them for having problems.

Time available. Several students mentioned that the amount of time available was

an important difference between the teachers and the mediators in their schools. Some

students, perhaps accurately, perceived their teachers and administrators as being too

busy with other things to help them with non^academic problems or, when they did help,

they tried to solve the problem too quickly. One student at the middle school summed it up

succinctly:

Sometimes the teachers have time, but sometimes they don’t. You could talk to the

mediators for a while and they could hear us out and sometimes the teachers don’t have

that much time.

106

Another middle school student made a similar comment, but instead focused on how the

teacher responded to her problem more quickly in comparison to the mediators:

She [the mediator] paid a little more attention. She focused on our problem because you

guys are here for that. She focused on our problems, but other teachers will just try to solve

it fast.

At one of the high schools, a student made an almost identical remark by comparing the

mediator to her school’s principal who, I should acknowledge, often made it a point to be

aware of what her students went through:

Just the fact that she [the mediator] was very focused on the situation and she made sure

she was paying attention. Whenever she would understand exactly what I saying, she made

me stop and asked me and took her time. If I would do that with my principal, she would

always be like “okay, hurry, hurry, hurry. I have to do this and I have to do that.” It wasn’t

the same, but with her it was more like taking the time because we practically stayed there

the whole day.

While this student’s situation may have been unusual because the mediators spent almost

their entire day on the conflict, the additional time that mediators devoted to a problem

seemed to be positively associated with students’ attitudes about seeking help.

Confidentiality. Whether adults would handle students’ conflicts or problems

confidentially was another important factor related to help seeking. The confidential nature

of mediation was arguably one of the most important qualities that set mediators apart

from teachers or other school staff. One high school student commented on how this

mattered for her:

Yes, everything I told them [the mediators], they’re not going to tell nobody. If you tell a

teacher something, they’ll go tell somebody. I know that whatever I say is going to stay

right there in the group versus if I tell a teacher something and I don’t want nobody else to

know, and they’re going to tell somebody else.

In a similar way, one middle school student even remarked how mediators resemble close

friends who don’t share sensitive information with anyone else:

107

I guess she’s [the mediator is] like a friend to me. With my teachers, say Ms. Lisa, I don’t

think I would talk to her about my personal things. Knowing that [the mediator’s name]…

her work is about not saying anything and helping us out. That makes me feel better.

Having the trust that someone will keep your problems confidential is critical when

students decide where they should go for help. One high school student observed that

while the trustworthiness of teachers varies, mediators were consistently more reliable

when it comes to protecting confidentiality:

So you wouldn’t really know if they [the teachers] would go back or tell. Some teachers I do

know won’t do that. They’ll just keep it between you and them. And they’ll be like, ‘if you

need to talk, just come in at any time.’ Like Ms. [teacher’s name], I can tell her anything and

she’ll just be okay. Some teachers, they’ll just be like ‘okay’, and the next thing you know,

you get home, ‘oh what happened at school?’ I want somebody you can talk to and it’s just

between me and them. With the mediators, I feel like it’s just between all of us unless the

other person goes back and tells. I’d rather have conversations like this because I know for

a fact that they won’t tell unless you say something like, ‘oh I want to kill that person,’ or

whatever.

Although I only presented three quotes here, privacy arose repeatedly in student’s

comments about whether they perceived the adults around them as viable sources of help.

While it almost goes without saying, students across all grade levels and schools valued

confidentiality very highly when seeking help for their conflicts. The professional role of

teachers and mediators to some extent determines the information they share with others,

but the way the adults choose to respond to student disclosures also makes a difference.

Responsiveness. Students acknowledged significant differences in the ways

teachers and mediators responded to their problems. In addition to time and

confidentiality, the responsiveness of the adult was an important contributor to students’

willingness to seek help for their conflicts. Students commented on how they disliked or

even feared the way their schools and teachers would punish them if they found out about

their conflicts. On the other hand, students also recognized that differences in the

108

professional capacity or training of the teachers and mediators was somewhat responsible

for the type or quality of the help they received. Responsive adults, or those who were less

likely to punish students and more likely to provide them with effective support for their

conflicts, seemed to be associated with favorable attitudes towards seeking help among the

students in this study.

School discipline. Some students who participated in mediation seemed to fear that

teachers or administrators would discipline them for their involvement in a conflict, which,

as one can imagine, would have a deleterious effect on their motivation to seek help. In the

first of several examples, a high school student who endorsed going to adult for help,

commented on how a school’s disciplinary response can negatively influence decisions to

get help:

I think some kids don’t feel comfortable because they feel like they’re going to get in

trouble for going to them [school staff], ‘oh and what did you do?’ but I think it’s better to

go to at least an adult.

In the next example, a middle student described how her school would respond to student

problems in a different way than the mediators:

It [the mediation] was different because teachers don’t do that. They just tell Ms. [Dean’s

name] that we called each other names and we get in trouble. We didn’t really get in

trouble. We just talked about our problems.

While the student does not directly discuss the helpfulness of either approach, one

implication is that students may be less inclined to seek help from adults under the threat

of punishment. In another example, even after the conflict was brought to the attention of

the school staff, a high school student was still reluctant to talk about her problem because

she wanted to avoid any potential consequences:

My teacher stopped me and he was like, ‘you know you need calm down. You’ve been

doing good. You’re going to let a little incident like this lose your chance here. It’s not even

worth

109

it,’ because this is my last chance here. I was so mad I couldn’t like stay still. I had to move

around so he brought me to office and he was trying to calm me down, but it wasn’t

working out. I was just really mad so I had to talk to the counselor. I didn’t want to tell her

anything because I was trying to avoid any situation so I had decided to let it go…

Similar to the last example, the student then went on to compare how the teachers

responded to her problem in a different way than the mediators:

Yeah, because the teacher, it’s like whatever you can say, you can get in trouble for it even if

you’re just talking to them. Teachers don’t really hear you out. They cut you off. With them

[the mediators], they let me talk until I was done. Then they talked and they gave me their

opinion, and I gave them mine.

Consistent with previous quotes, this student offered several examples of how the

mediators were more responsive than the teachers by not getting them in trouble for

talking and spending more time on the problem. As this case and others illustrate, a

response that penalizes students for talking about or becoming involved in conflicts may

reduce the likelihood that students would turn to adults for help with their problems.

Professional capacity. Variations in the type or quality of help students receive for

conflicts, however, may simply reflect differences in the professional capacities of the

adults at the school. For example, teachers may have more time commitments or less

formal training in conflict resolution than the mediators. In one example, a high school

student described how she considered the responsibilities of the adults at school when

deciding where to get help:

I feel like talking to a teacher here is kind of, not awkward, but kind of uncomfortable

because they’re my teachers and I see them every day. I don’t know who they are, but I see

them, so I’m just with them, but with the mediators, I don’t see them every day, so if I were

to talk to them, I’d pick them over a teacher because I feel like mediators are more helpful.

They’re focused on one thing and teachers are focused on school and subjects and all that.

110

Students like this one here who participated in the program may also discover that the

mediators have specialized training for responding to conflicts. Another high school

student commented on how the mediators’ techniques were helpful to her:

I feel like with the mediation people, they’re more like…they tell us the ground rules and

things like that. Just the fact that they’re saying you can’t do this, you have to be respectful,

and everyone was calm, no one was cussing. If I were to talk with my teachers, and they

would tell the girl, the girl would be cussing.

Sometimes students may perceive the specific strategies that mediators employ during

mediation sessions as more effective than what their teachers provide. In the last example,

a middle school student observed how teachers don’t always respond in the most

constructive way even though they may have sincere intentions to help:

The mediators understood more and the teachers will try to go into a conflict with you.

They’ll fix the problem, but then they’ll say, ‘well why did you do that. You shouldn’t….’

They’ll put more into it, and they’ll fight back what we’re saying, and the mediators don’t

do that.

These examples may help explain some of the differences in student perceptions of

teachers and mediators, which include factors associated with their professional capacities

and training at school. Students may see mediators as more responsive because of their

specialized training or techniques for handling conflicts.

Mediators as Mentors

Another major theme in the qualitative data was how students experienced their

mediators as mentors, or adults who were more than just simply facilitators of their

mediation sessions. Students made ample references to their relationships with the

mediators and reported that they discussed a range of topics beyond the original conflict.

Students also identified with the mediators’ social background and life experiences. I

organized student comments about their relationships with the mediators into three sub^

111

areas including their observations about shared similarities in background, life advice the

mediators gave to the students, and how mediators used disclosures about their own

experiences to build rapport with the students. In these ways, the mediators appeared to

act as mentors to the students, which was a consistent theme in students’ discussions of

their relationships with them.

Shared backgrounds. In the first subtopic of mentoring, students talked about

how the mediators were similar to them in terms of age, gender or social background. Both

middle and high school students found it helpful that the mediators were closer in age to

them than other teachers or adults on campus. One middle school student who had a social

conflict with another girl said she felt:

Comfortable because they [the mediators] didn’t seem like they were that…like if it was

someone older, then I’d feel kind of weird because they’re like older. With them, they went

through it; they’re probably going through it.

Another high school student who also had a social conflict with one of her best friends

at school described the age of the mediators as a kind of “sweet spot” that was

particularly helpful for her:

I liked the fact that these are roughly people who aren’t too old and do have a sense of

understanding. They’re young enough to understand where we’re coming from and have a

point of view where it’s not that you’re not too young and it’s sort of immature, but they

have this logic behind it and they do have this matureness in it. I liked that very much.

Other students also referenced gender as something they appreciated about their

mediators. In fact, gender was significantly associated with program satisfaction as twice as

many female students as males reported high feelings of satisfaction. Male students,

however, made up a much smaller percentage of the student participants and, in addition,

only two out of ten mediators in total were male from all three participating schools. While

students’ comments did not go in depth about why gender was important to them, they

112

briefly mentioned it in their interviews nonetheless. I asked a middle school student if she

was satisfied with the mediators and she responded:

Yes, because she’s a girl. She goes through that, like fights with her friends. She can kind of

relate to it.

A high school student who was involved in a conflict about rumors similarly remarked:

It was good because we were all girls and there were girls too. They said they had

experience. They’re wiser so they know what to do.

As their comments indicate, some students attributed their similarities in age and gender

with the mediators to their feelings of satisfaction with the program. Though speculative,

shared social characteristics in other forms such as ethnicity or cultural background may

also have also been important to the students. While students did not specifically address

these additional aspects, I discuss the potential impact of my own social identify on their

comments in the discussion chapter.

More than just mediation. In the second subtopic of mentoring, some students

observed that their relationships with the mediators were exceptionally supportive as they

received useful advice about their situations and other aspects of their lives. For example,

one high school student described how she appreciated the way the mediators supported

her during the mediation session:

What comes to mind is definitely like a counselor or a therapist. They do give the same

advice and they do have that listening that a counselor or a therapist would have. Just from

personal experience, when I think of these mediators, I don’t think of them just facilitating

or supervising a conflict, I do see them as people who do help others with conflict and they

do engage with wanting to resolve the conflict in a good manner, even if two people aren’t

still friends but they ended on a good note.

This student suggests that the mediators were especially invested in the outcome of her

problems above and beyond their assistance with facilitating the session. Another high

113

school student commented on how the mediator not only talked to her about college, but

also helped her with other relationships or concerns in her life:

That’s what mediation is there for…I would just go to them and I would talk about my

other life problems and my other relationships with people and they would just give me

advice on what to do, what’s the best option for me, and I took it. Even though I’m going

through this stuff, I’m learning to kind of fix my own problems and cope with my situation.

It’s not just about college; it’s life in general.

Indeed the number of stressors often increases as students navigate the various stages of

adolescence and it should come as no surprise that mediators find opportunities to provide

additional support and mentorship. Shared social characteristics like age, gender or culture

may also encourage greater trust between the students and mediators compared to other

school staff, for example. Though these comments were few in number, it seems that some

mediators were able to step outside of their formal roles and tailor their assistance to the

students’ presenting needs and life circumstances.

Mediator self-disclosure. In third and last subtopic of mentoring, one of the most

important ways the mediators engaged with students was through their use of self, which

happened most often in selective, personal disclosures. Some mediators were explicit in

their comments about using this kind of strategy to build rapport with students. For

example, one of the mediators at the small high school described how she used this

approach with students:

What helped to engage with them was the whole talk about school and me being a college

student and me putting myself in their shoes. That’s always how I try to build rapport and

engage with the student…just coming to their level. I always emphasize to them before we

start the mediation what my role is here…that I’m not an authority and I don’t work here.

I’m neutral to the situation. I think right away that let’s them put them guard down and

lets them be honest about what’s really going on because there is stuff that I know that is

shared with me and [the other mediator] that they would not feel comfortable telling [the

principal], or any other authority here on campus.

114

In this example, the mediator highlights her identity in order to draw a distinction between

her self and the rest of the school staff. Another mediator at the middle school commented

on how she disclosed to her group of students an example of a conflict with one of her own

friends from middle school:

I took this mediation close to heart because I had a best friend. I knew her for a while and I

felt forced to have that friendship because of the time that I’ve known her, but it was kind

of similar to like, oh she’s shady, but I’ve known her for this long. I honestly put that on the

table and was like, ‘just because you’ve known a person for this amount of time, that

doesn’t mean you’re forced to be their friend. It’s a matter of how you treat each other and

the relationship you all have created.’ I mean people change and you experience different

things. It’s just a matter of where do you want to take it now. I shared that with them. So in

the mediation sessions, I felt like we did more than mediate. We mentored and gave them

advice about life…because no one is there for you. For myself, no one was there for me

telling me like, ‘oh this happened, and this and this happens.’ So it was kind of like a

mentoring guidance of what happens in life, with friends, with people.

Sharing personal stories like this was an important technique that mediators used to build

rapport with students and provide them with timely advice. These disclosures did not go

unnoticed by the students. In fact, they played a powerful role in building trust and helping

students feel comfortable during the mediation sessions. One of the students recalled how

she identified with her mediator’s story about when she was a high school student:

She kind of opened up first. The mediators did. She said her side of the story when she was

a teenager. She was like, ‘I was not like this [the way I am now]. I did not help people. I was

one angry girl. I was just like you, but worse.’ She started telling her side of the story,

which was like, okay I can kind of relate to this, which kind of made me open up and tell

her, okay, this is my story. I’m not just telling my story to some random person who

doesn’t really care, but they know where I’m coming from because we went through the

same thing just in high school.

Not only can an approach like this result in greater trust for some students, but it also helps

students feel more motivated to participate in the session, which, in turn, can lead to more

effective results overall. One high school student talked about how he was able to reveal

more about himself after listening to the mediator share her own college experiences:

115

I actually felt pretty comfortable talking to one of the mediators because, at the end, she

had some questions about colleges and she’s a UCLA student, so I actually got some

information about college, so I got more open speaking I guess. Either way, she was like,

‘oh it would be great if you tell the truth.’ There’s nothing to hide, especially something this

little…but at one point, I did explain my background and certain things to the mediator…

She told me that she got to UCLA with a 3.4. I was like, wow. She told me to do more clubs

and stuff like that and it’ll look really good on your application. That’s when I started giving

her more about myself.

After the mediator shared her own experience and showed interest in the student’s future

goals, he acknowledged that felt more comfortable participating in the process. In fact,

college was a frequent topic of discussion with high school students especially. While some

mediators may have focused on this area more than others, it was clear that some students

really appreciated these discussions. When I asked one high school student if she learned

anything from the mediators, she described how she gained advice about conflict

resolution as well as motivation for college:

Just like how they said it’s always going to be like that about rumors. It wasn’t going to

stop, but just to focus on college and school. They were giving me school advice, like how

they both started and how they’re ending. One of the girls, she said she didn’t go to college

right away, but she went after and she recommends me to go right away. I really liked that

because sometimes my counselor doesn’t say that to me, so I really liked that from her. She

was motivating me to do it.

Similarly, a high school student’s testimony about his experience with his mediator was

particularly convincing because, according to one of the mediators, other school staff had

described this student as a “troublemaker” with little interest in school. Towards the end of

our interview, I asked this student what he thought about his mediator:

She told me about her past and how I’m going through the same things she went through

when she was my age. I thought it was pretty cool. She didn’t really like high school or

school in general, and she’s going to college now and shit. That’s pretty cool. I don’t know…

[It made me think] that I could go to college one day.

For this student as well as others, the mentoring aspect of their relationships with the

mediators appeared to convey direct benefits including increased trust, participation and

116

feelings of satisfaction with the program. Shared similarities that came from mediators’

personal disclosures of their own life experiences related to conflicts or going to college

helped build rapport with students. It seems likely that mediators who employed this kind

of mentoring based approach were able to not only create more trusting relationships with

students, but also positively influence their motivation, sense of satisfaction and

engagement in the mediation process.

Program Satisfaction

Similar to the other major areas, the qualitative data revealed several themes

related to program satisfaction. I organized student comments into three main areas of

general satisfaction with the program including helping students with self^expression,

seeing mediation as a unique process, and helping students avoid physical fights. In each

of these areas of program satisfaction, students generally described various aspects of the

program that were enjoyable or especially helpful to them. I also view these areas as

important contributions of the mediation program to students’ social development as well

as the prevention of violence and improved safety at school. Finally, I close this section

with several important reasons that students felt dissatisfied with the program.

Considering the reasons that students feel both satisfied and dissatisfied allows for a better

assessment of the strengths and weaknesses of the conflict mediation program and the

extent to which it meets students’ needs for resolving peer conflicts in prosocial ways.

Self-expression. In the first subtopic of general satisfaction, some students

commented on the benefits of being able to express one’s self or emotions. Students found

it difficult to talk about their feelings especially in the physical presence of another student

or if others had responded negatively to them for this in the past. Given these challenges,

117

however, students were still able to comment on how mediation helped them to express

their feelings more openly. In a bullying conflict, one middle school student described how

she felt after expressing herself:

I felt surprised because I had shared my feelings with everyone and they didn’t do

anything. They just went with it. Usually when I say my feelings, everyone’s against me.

They knew my feelings about what happened with the problem and they find out why I

made a joke out of it. It didn’t hurt, it helped.

In a structured setting such as mediation, students may be more likely to have a positive

experience after they share their feelings about a conflict instead of being ridiculed or

harassed by their peers. Another student in this same mediation on bullying also similarly

commented on how it helped her afterwards:

I felt like I was better than I was before because before I wouldn’t talk as much and I would

always keep my problems [to myself], but when I was there, I would always say them out

loud and I started being more out there instead of inside my shell.

Talking to someone in a conflict directly about how you feel can be difficult even for

students who are more confident with expressing themselves in general. One high school

student commented on how the mediators made this easier for her:

The many things I liked about it is that I was able to express everything that I could not

express to other people. I’m really, really good at expressing how I feel talking to my sister

and my siblings, but with them [the mediators] they were exactly like my sisters. It was just

super easy to just say it, but when it comes to the person I have to talk to, it’s hard. I feel

really comfortable with them…it’s not really hard to open up, so that’s why it was easy.

As these students attest, in some situations, the mediators were able to help facilitate

dialogue despite serious challenges that come with having negative experiences or fears

related to expressing emotions to another person in a conflict. While outside the scope of

this research project, it is possible that these kinds of experiences would provide some

degree of positive reinforcement for students’ future attempts at self^expression as well.

118

Mediation as a unique process. The second subtheme in general satisfaction was

that students seemed to appreciate the unique aspects of the conflict mediation process

that they would not otherwise experience in their schools. For example, students

commented on how they appreciated the fairness of the mediators, the controlled

structure of the mediations and having a sense of autonomy over the outcome of their

conflict. One high school student talked about how the mediators’ impartial response to her

situation helped them arrive at a solution:

Since they [the mediators] were the bystanders, they didn’t take one side or the other. They

put it all out there. They said it as it was, what it seemed like. She pointed out some

important points and perspectives that she saw. After that, me and the girl were like we

agree on that, she has a point. It was brought to a further conclusion.

A high school student at another school talked about how the structured mediation process

helped her avoid the way conflicts typically get resolved in her experience:

I think I strongly agreed on a part [of the survey] because they did have control on us with

the talking. I just felt like it was really organized. Where if it had been somebody else, we

would’ve just been yelling at each other. I think it was really good to have it set up like that:

where it’s just the 2 of us and like no other people were able to put their input or say, ‘no,

you’re lying,’ or you’re this that, or whatever. They actually had us take turns and say what

we actually felt about the whole situation…getting your point across, actually having time

to be the only one talking, not talking over each other, to actually say what you have to say,

no argument, or saying you did something wrong, just actually saying how you felt. I liked

that part of it because in most cases, you really don’t get that. You just get, it’s time to

argue, go. That was the best part of it.

The mediation sessions provide students with the opportunity to hear another person’s

perspective, which is crucial for reaching an acceptable resolution. Many students

recognized how difficult it was to accomplish this kind of understanding on their own. For

some, the presence of a neutral, third party facilitator was a real improvement over the

way they normally approached conflicts. Another unique benefit that mediation provided

119

was having some sense of autonomy over the situation. For example, one middle school

student described how the mediators let her decide how to move forward:

I liked the way they let us both talk and got both sides of the story first. I liked the way they

asked us what we wanted. They didn’t just say, “oh you guys just need to get along,” like

most of the teachers and principals here. They let us speak and that’s what I liked: that I

could choose what I wanted to do instead of being forced to talk to her again or something

like that.

The student’s comment also seems to suggest that there are indeed differences in how

schools and mediators respond to the same conflict whereas schools may rely more

on proscriptive solutions instead of allowing students to act more independently.

Avoiding physical fights. In the last subtopic of general satisfaction, students

described how mediation helped them avoid escalating their conflict into something more

serious such as a physical altercation. One high school student talked about how he might

have ended up in a fight without mediation:

I think if the mediation wasn’t here, he probably would’ve told me he wanted to fight me

and we would’ve ended up fighting. I know personally, since he’s my friend, I would’ve told

him to calm down before anything bad happens… It actually solved what me and [student’s

name]…we were close to fighting and mediation actually helped us. We ended up talking as

friends again, so it was pretty cool.

Another high school student also echoed this sentiment about how mediation helped her

avoid a physical confrontation:

I probably still would’ve gone to the Deans, but without mediation, we probably still

would’ve gone to the Deans and the next day we would’ve had another argument. And she

would threaten to jump me or whatever and eventually…she talks like all this mess like, ‘oh

I’m gonna sock you right now. I’m gonnna jump you.’ I said, ‘then do it. If you’re talking all

this, then do it.’ I think probably without mediation, I think we would’ve kinda fought. I

would go to the Deans and suggest mediation again because it helped. I don’t know how it

helped, but we’re not arguing anymore.

Even though this student acknowledged she probably would have asked the Deans for help,

she also remarked that the mediation program helped prevent the conflict from flaring up

120

again. Of course, without long term follow^up, there is no way to actually know whether

students would behave differently from what they report in their interviews. The student’s

comment does not necessarily reveal why they stopped arguing; however, she does say she

would return to mediation, which suggests that something about the process was helpful

for resolving her conflict and, at best, may have prevented future violence.

Dissatisfaction with mediation. Despite the overall reports of high satisfaction,

many students still mentioned that they felt unhappy, or dissatisfied, with their

participation in the conflict mediation program. Some reasons reflected the natural

variation that exists in program quality or facilitation while others were related to the

difficult nature of conflict itself for both students and mediators. I organized student

feelings of dissatisfaction into five subtopic areas including a lack of investment, a lack of

confidentiality, mediator biases, emotionally difficult nature of conflict, and facilitation

problems including a lack of mediator experience. It was also interesting to note that some

of these problem areas were unique to either the middle school or high school level. The

discussion of student dissatisfaction is important not only because it provides a balanced

perspective to the previous sections, but it also suggests additional ways the program could

be improved in the future.

Lack of investment. In the first subtopic area of dissatisfaction, several students

made negative comments about how either the students or the mediators were not very

invested in the mediation process. This appeared as a lack of participation, not taking the

process seriously or making light of the situation. In the first example, two high school

students in the same mediation were particularly frustrated with each other. One student

121

described how she did not believe the other student was talking as much as she could

have, which likely contributed to her feelings of dissatisfaction with the process:

I don’t think anybody can really get you to talk, say what you have to say, but I wish she

would’ve actually talked or just participated a little more. I just felt like she was just sitting

there…“talk! I know we don’t like each other, but just talk. Say something!” I just felt like

me and the mediators were just there. Say something! It’s not just my mediation. It’s yours

too. It’s for us both to say what we have to say. I just felt like I was getting nothing out of

her.

Interestingly, the student on the other side of this mediation was irritated with the

previous commenter as well. From her perspective, the other student was not only being

dishonest, but she also felt like she would be attacked for what she said:

To me, to hear her side of the story, it was just like whatever because it was a lie, so I’m just

like whatever. I know the truth and I was just like whatever about it. If she wants to lie,

that’s on her… I knew what happened. The people that were around us knew what

happened. She brought up all this other stuff that never even happened. And if I said that

she said something, she would be like, ‘oh no, I didn’t say that.’

The example illustrates how students can have negative experiences when they perceive,

either accurately or not, that the other party is not fully engaged in the process. On the

other hand, students also became displeased when they thought the mediators did not

take their concerns seriously. One middle school student reported that the mediator

treated her problem as if it were a “joke”:

She [the mediator] wasn’t helpful. She was just there. And some of the girls are like giggly

about it. They weren’t that serious about it. So she didn’t really help at all. Yeah, like

some people said that she thought it was a joke, but it wasn’t.

In this case, it may be possible that the mediator misunderstood the seriousness of the

conflict, but the student’s detection of a lack of sincerity from the others had a negative

effect on her experience. In fact, this was not an isolated incident. Another student at the

same middle school described how the mediators were dismissive of her problem:

122

What I didn’t like was sometimes they’re [the mediators] just mad and they just don’t fix

the problem. They like just say, ‘why are you guys doing that,’ and all that. They could

do that better I guess.

While we may never know the mediators’ true intentions or conduct here, the student

implies that mediator was not very responsive to her needs. For these students, their

perceptions of a lack of engagement or investment on the part of the other students or the

mediators appeared to contribute to their feelings of program dissatisfaction.

Lack of confidentiality. The second subtopic area of dissatisfaction included breaks

in confidentiality that happened in some of the mediations. Only middle school students

commented on this during my interviews; however, confidentiality is a major feature of

the mediation process and almost all students are concerned if others will respect this

rule. In one example, a middle school student commented quite simply that she did have

much confidence in the ability of her peers to maintain her privacy:

Talking about the problem when they’re there, I’ll feel uncomfortable. I think they’re going

to tell their friends.

Even more troubling, in another example, a student was unfairly accused of breaking

confidentiality when other classmates found out about the mediation from someone else:

My friend that saw her sign up for mediation had told the girls, that’s why…because we

were both emotional so we were both crying. And the girls were like, ‘oh you guys came

out of mediation.’ And then I got in trouble because, as they said, it has to be confidential,

so they thought I told them and I got in trouble.

This student’s experience also highlights the challenges of maintaining confidentiality

throughout the entire process from the time of referral all the way up to returning to class.

However, given the sensitivity of this area, it is also seems significant that complaints about

breaches in confidentiality did not arise more often in the students’ comments.

123

Lack of neutrality. In the third subtopic area of dissatisfaction, some students

perceived a lack of mediator neutrality during their sessions, which only arose in

interviews with high school students. Student perceptions of mediator bias appeared to

have a negative effect on their satisfaction. For example, one student, who thought another

student had intentionally hit her with an orange at lunch, described how the mediator

seemed to take the other student’s side by referring to the incident as an accident:

What I didn’t like was that how in one part she [the mediator] would say that it was an

accident and then it wasn’t. She would say that, basically, it was an accident like if she was

standing there.

Even if the mediator was ambivalent or unclear about what happened, the student was

dissatisfied because she perceived bias in the mediator’s response. This bias may even be

transmitted nonverbally as the next student’s example illustrates. This high school student

described a sensitive situation where he felt like the mediators were judging him:

I think they [the mediators] did pretty good, but the only thing I didn’t like was just how

they were looking at me. Well they were looking at [other student’s name] too but they

weren’t looking at him the same way they were looking at me…I’m going to be really

honest. All 3 of them were looking at me as if I was racist and I felt uncomfortable. I think

they thought I was racist for awhile because of what I told [other student’s name]. It made

me a little bit uncomfortable because they were just looking at me like that.

This student also reported a low satisfaction rating of his mediation session and, taken

together, these two examples suggest how both verbal and nonverbal communication by

the mediators can result in perceptions of impartiality toward one student or another.

Emotional challenges. In the fourth subtopic area of dissatisfaction, students

described the challenging nature of responding to conflict, which often included difficult or

intense emotions. Some students, especially those in high school, commented on how it was

hard to confront or talk directly to someone when they had strong feelings of anger or

frustration. One high school student remarked how she didn’t even want to be in the same

124

physical space as the other student because she harbored strong, negative feelings

towards her:

I didn’t like that, after they asked me questions, they put us in the same room together.

Yeah, I just didn’t want to be in the same room with her. When we have the same class

together, she used to roll her eyes at me and stuff, so I didn’t want to be in the same room

as her. But then I just decided that it would help.

Similarly, another student commented on how her feelings of anger made it difficult for her

to participate in the mediation session:

I didn’t want to meet with the person because I still had feelings. I still was really mad. I

didn’t like the fact that even after I had given my story, she was still looking like the victim

when even the mediation people saw that she did her wrong part too, but I still looked like

the one attacking her.

While this student’s dissatisfaction appeared to result from a combination of factors

including her anger and feeling misunderstood, the process of working through sensitive

issues may be especially challenging for students, or anyone for that matter. One high

school student aptly noted that even though she wasn’t accustomed to the process, she still

recognized it was an important skill to learn for the future:

I think the one thing I didn’t like about it just because…it’s not something that I’m used to…

is to talk about it with a person who you are directly having conflict with because it’s not

something a lot of people do nowadays. It’s definitely something I’m still trying to learn as

I’m growing up. That was something I don’t like, but I know I will like it eventually and I’m

still learning.

These students remind us about the inherently difficult nature of working with someone

directly when negative feelings abound. While the emotional context of conflict resolution

may be unavoidable, it may unintentionally increase students’ feelings of dissatisfaction if

the mediators do not adequately address it during their mediation sessions.

Poor facilitation. The fifth and last subtopic area of dissatisfaction among the

students included poor or inexperienced facilitation of the mediation sessions. Some

125

degree of skill is involved in the successful completion of a conflict mediation session.

Novice mediators may be more likely to skip steps in the mediation process, misuse certain

techniques or neglect important emotional cues. In the first example, one high school

student who participated in the mediation program for two years described a negative

experience from her first year:

The people that came and talked to me, they weren’t the same ones. They kept changing. I

had to explain it a lot of times. I was okay with it. I had to explain. They would do some

things…they would try to make us calm down, but it would only make me more mad…I

guess it was only the first year or something like that…we had mediation and they made us

sign a contract at the end. Three weeks after that, we got in a fight with the girl. When we

had the mediation, they clearly saw how she didn’t agree. She didn’t like it. I knew that she

wasn’t going to respect it.

Not only were the mediators inconsistent in their follow^ups with the students, they also

may have neglected the students’ underlying emotions by prematurely signing a contract

when the students did not fully agree with the resolution. Even worse, this same student

went on to describe a near traumatizing experiment the mediators attempted with her

during her session from the previous school year. Not surprisingly, the result had a kind of

iatrogenic effect on the student:

I actually saw improvements this time from last time. They didn’t make us do some weird,

cheesy stuff. They made me and that girl write a card to each other saying what we liked

and disliked about them, give it to them, and I was just like, ‘oh my god, I can’t wait to read

that note!’ It was all negativity and I just didn’t want to read that. Honestly, that’s what I

thought when I walked in. I was like, oh they’re going to make us write something and I

didn’t want to [laughing].

However, I should note that this student also reported an improvement in the second year

of the mediation program at her school. In the next example, one high school student’s

dissatisfaction revealed the importance of enforcing ground rules in communication such

as taking turns or being respectful. The student described how the mediators did not

adequately address the way the students disrespected each other during the session:

126

I think they handled it well, but I think they should’ve been like…she said that I had said

something about her and I did. They said don’t [interrupt her]…let the person talk, just

listen and then once it’s your turn to talk. But, I did interrupt her because I felt like, ‘don’t

lie because I really didn’t say that.’ We had witnesses there. I can bring your friends and I

can bring my friends and tell you I never said that. The mediators were like, ‘okay, don’t

jump in. Just wait your turn.’ There were a couple of noises, you know, like if somebody

says something and you go, ‘ohh, whatever…’ and that was done a couple of times, and I

would just look at her and then continue talking or whatever. I think there somebody

should’ve said, ‘don’t do that. That’s disrespectful.’ So that made me say, ‘don’t lie,’ because

she did it to me.

New mediators can have difficulty reigning in these micro^aggressions such as offhand

remarks or interruptions especially when they arise frequently in the session. As this

student’s experience suggests, however, a failure to enforce respectful standards of

communication can also result in an unsatisfactory mediation experience for everyone.

While the majority of students reported that they felt satisfied with the program,

dissatisfaction was nonetheless a prominent theme from the students’ comments. The

most likely culprits in the majority of students’ negative experiences included a lack of

investment from students or mediators in the process, a lack of confidentiality, a lack of

mediator neutrality, the difficult nature of conflict and, finally, poor facilitation of

mediation sessions. Even so, most if not all of these problematic areas that students

identified also provide excellent opportunities for program improvement in the future.

Discussion

This study attempted to answer questions about the impact of a conflict mediation

program in several secondary schools of a large, urban school district. The study found that

school violence and conflict are an unfortunate and all too common reality that student

participants must confront on a frequent, and for some, even daily basis. Out of a separate

group of students who participated in the conflict mediation program, the study found that

the mediation program helped students resolve conflicts with their peers, taught them new

127

coping and conflict resolution skills and provided them with mentorship and guidance for

working through problems in nonviolent ways. Students also shared their perspectives

on how adults could best support them in their efforts to seek help for and resolve

conflicts.

One of the major challenges of this study was documenting the impact of the

program across multiple sites with a high degree of variability in the program’s

implementation and the school environment. However, this variation also contributed to a

range of student experiences, both positive and negative, which provided insightful

feedback, criticisms and suggestions for areas of improvement. In this chapter, I will first

discuss the results of the classroom questionnaires on the school context and, second,

highlight the three main findings from the qualitative results. For the first two findings, I

argue that the mediation program has two areas of high impact including helping students

manage their negative emotions and providing students with access to undergraduate

mentors. For the third finding area, I underscore the important ways that contextual

factors including adult response to student conflict and peer culture influence student

decisions about how they respond to conflicts. I then describe the implications for

research as well as the implementation of the mediation program. Lastly, I review the

limitations of the study and close with some final reflections on the process of conducting a

dissertation study under the imperfect and sometimes chaotic conditions in schools.

The School Context: Exposure to Conflict and Help Seeking

Exposure to conflict. Students in the four participating schools who took the

classroom questionnaires reported high levels of conflict including bullying and fighting

that sometimes exceeded rates from previous research. While there were notable

differences among schools, the rates of bullying victimization often exceeded those

128

reported in large^scale samples though the rates of physical fighting were at similar levels.

Students in the present study reported higher rates of any kind of bullying, although

cyber^ bullying and fighting (at the high school level) were nearly identical to previous

research. Joad appeared to be an outlier among the schools in terms of high reports of

conflict, which seemed to surpass the rates found in other middle school survey data.

Across all schools, students consistently reported a high degree of “low level” conflict

including both bullying and harassment, which would make these schools prime

candidates for any interventions that seeks to reduce conflict or violence and improve

school safety. Researchers on bullying in schools have suggested that when prevalence

rates reach such high levels, students may accept these behaviors as normal and resign

themselves to a kind of helplessness where they feel they have nothing to gain by seeking

help for conflict (deLara, 2012). Any intervention that both prevents conflict from

escalating into violence as well as improves students’ willingness to seek help for conflict

would be especially helpful for creating a positive school climate where students feel safe

and supported if they should decide to get help for their problems.

Grade level and gender differences in exposure to conflict were also consistent with

previous research, although race differences were slightly more complicated. Reports of

victimization were highest in 6th grade and decreased linearly to their lowest point in 12th

grade. Male students consistently reported higher involvement in all kinds of conflict than

females for physical fighting as well as witnessing and experiencing bullying or

harassment. Black students also reported the highest levels of conflict compared to other

racial groups while Asians reported the lowest levels. Latinos, who made up the majority of

the sample, were most similar to, though still higher than, Asians in reporting relatively low

129

levels of conflict. White students often reported the second highest rates of exposure to

conflict and were the top reporters of racial harassment, although their numbers

represented a very small proportion of the overall sample. These findings are consistent

with previous research on group differences with the exception of race (Wang, Iannotti, &

Nansel, 2009; Swearer et al., 2012). Racial differences are more complex and may depend

on a number of factors such as how bullying is defined or the racial composition of the

school. In terms of the students in the present study, however, male students in the earlier

grades and those from ethnic minority groups in the sample, such as black and white

students, would be most at risk for exposure to conflict, although this general conclusion

may not be true for each subtype of bullying or harassment. At the very least, the finding

that students most commonly reported verbal bullying (32%) out of all types suggests that

conflict mediation is an appropriate intervention in these schools.

Willingness to seek help. Across the sample, students generally reported favorable

attitudes toward help seeking. The two “small schools”, Casy and Sharon, appeared to have

a better help^seeking atmosphere than the regular, public schools, Milton and Joad. Casy

had the highest percentage (59) of students who said they got help the last time they were

bullied or harassed at school and Sharon had the highest percentage (70) of students who

said they could report a problem without being considered a snitch. While the two small

schools ranked the highest on these two questions, there were no statistically significant

differences among the schools. In two previous studies, 75% of students had told anyone

they were bullied and, more stringently, 40% of students said they notified an adult after

being bullied (Unnever & Cornell, 2004; Robers, Kemp, & Truman, 2013). The question

130

about getting help in the current study falls within this range and seems consistent with

these findings from previous research.

Help seeking scores and group differences were also consistent with findings from

previous research with the exception of a student’s grade level. The average score of 3.1 for

the sample was slightly higher than an average score of 2.8 from a previous study using the

same scale with 9th graders only (Eliot et al., 2010). Help seeking at the “small schools”,

Sharon and Casy, was higher than the regular public schools, Milton and Joad. Indeed, the

average help seeking score of 3.3 at Sharon was significantly higher than the other three

schools. Group differences in help seeking were also similar to previous studies with the

exception of a student’s grade level. For example, girls were more willing than boys to seek

help for conflict or problems, which was a consistent finding in previous research. In the

present study, Asians reported the highest levels of help seeking (3.3) closely followed by

Latinos (3.2), while black (2.9) and white students (3.0) were the least willing to seek help.

This is mostly consistent with previous research, which found that Asian and white

students were most willing to seek help and black and Native American students were the

least likely, while Latino students were in the middle of the high and low groups (Unnerver

& Cornell, 2004; Eliot et al., 2010). Still other researchers found no significant differences

at all by racial group (Williams & Cornell, 2006). These findings of higher help seeking

levels among Latino students are perhaps not completely surprising since the two small

schools, Casy and Sharon, which report the highest help seeking scores, are also almost

entirely Latino. However, Latinos are still among the highest reporters of willingness to

seek help even at Joad and Milton, though they are still the majority ethnic group there as

well.

131

However, the positive association between help seeking and grade level was even

more surprising. Previous researchers found that 8th grade students were less willing to

seek help than 6th graders and speculated that increased independence or ability to solve

problems might explain these differences (Unnerver & Cornell, 2004; Williams & Cornell,

2006). In this study, a positive correlation was found between help seeking and grade,

although the difference between middle school (3.1) and high school students (3.2) was

slight but significant. This suggests that the relationship between help seeking and grade

may not be consistently linear from middle to high school, although school level factors

certainly appear to play a role in students’ willingness to seek help for conflicts.

School context implications. Taken together, these findings suggest there are

significant differences in the overall climate of schools as well as implications for groups of

students would benefit the most from, but may also be the least likely to participate in the

mediation program. Sharon was clearly the leader of the four schools in terms of positive

climate as students there reported the lowest rates of exposure to violence and the most

agreeable attitudes about help seeking. Along these same lines, the students at Joad

unfortunately had to contend with more conflict than any other school. The safety related

aspects of a school’s climate may affect how students interact with the mediation program

and, in fact, students at Sharon reported the highest levels of program satisfaction out of

all three schools. These differences, however, were not statistically significant and could be

attributed to other factors such as the quality of the mediators or any number of

differences in program implementation or student and school characteristics.

From the statistics I presented, the students most likely to report exposure to

conflict, and, by extension, potentially the best candidates for the conflict mediation

132

program, would include black and white, young male students with a low willingness to

seek help for problems. Indeed, this somewhat fits the description of traditionally

estranged youth who are most likely to feel disconnected from school and experience

conflict and least likely to use these types of programs (Theberge & Karan, 2004). Perhaps

ironically and not unexpectedly, the single largest group of students to participate in the

program was Latinas, who are also among the most willing students to seek help. Schools

like Joad would clearly benefit the most from conflict mediation, although certainly all

schools in the study experience enough conflict to take advantage of these kinds of

interventions. Lastly, if conflict mediation programs actually encourage help seeking

among participants, this could be associated with gains in a positive school climate. In

fact, willingness to seek help may also be a good barometer of a school’s climate as

previous research found that high help seeking scores were associated with less school

disorder (Bandyopadhyay et al., 2009).

The data presented here could also serve as a reasonable baseline for any future

assessments of school culture, student exposure to conflict or the impact of interventions

that target any of these areas. In addition, previous research has shown that, when asked,

school faculty does not always agree on the amount or type of conflict that occurs in their

schools (Matloff & Smith, 1999). This type of school data may help staff reach consensus on

safety aspects of their school climate and develop a plan for change if necessary. While

conflict mediation has much to offer students and schools, in order to truly maximize its

potential for reducing conflict and violence, it would be prudent for administrators of these

programs to consider contextual factors that may potentially hamper their effectiveness.

133

High Impact Areas of the Mediation Program

Despite the modest pre^post findings, students’ descriptions of their experiences in

conflict mediation suggest that some of the most important benefits included reductions in

feelings of anger and developing relationships with the undergraduate mediators who

acted as mentors to them. Students reported that participating in conflict mediation helped

them manage their anger, which, in many ways, is an important antecedent to the practice

of nonviolent responses to conflict. In addition, the undergraduate mediators were able to

provide students with safe and confidential spaces for solving conflicts; dedicated time and

support beyond the typical staff member; guidance for students’ personal concerns;

motivation to succeed in school; and, lastly, enhanced feelings of program satisfaction.

Together, these various qualities and benefits comprise the kind of responsive relationships

that may help protect students from the negative effects of high levels of exposure to

conflict in school. In the next section, I first review the results of the pre^post measures

and then I discuss the significance of these two “high impact” areas of the mediation

program.

Summary of pre-post program outcomes. Students who participated in the

conflict mediation programs at the three schools reported a drop in negative affect, or

feelings such as anger or sadness, from an average of 2.1 (SD = 1.0) to 1.4 (SD = 0.5; p <

.001), but did not report any quantitative change in measures of positive affect, nonviolent

response endorsement or help seeking scores. Students, however, did report acceptable

levels of program satisfaction (3.3; SD = 0.7; range 1 ^ 4), which suggests that they

believed the program was fair, helpful and enjoyable. Surprisingly, very few studies have

looked at either negative or positive affect despite early calls in the conflict resolution

literature to

134

assess emotional changes in participants (Johnson & Johnson, 1996). Also, to my

knowledge, no previous studies examined changes in help seeking attitudes.

This study’s absence of significant findings in these outcomes is, however, not

terribly surprising given the number of factors that could potentially limit the detection of

meaningful changes. For example, previous studies found effects for some groups of

students but not others such as high and low risk, voluntary and involuntary, and high and

low pre^test participants (Harris, 2005; Shuval et al., 2010). Researchers also

acknowledged that the short duration of the conflict intervention and powerful school or

neighborhood effects such as climate, size or disorder, could all make detecting changes in

students’ self^ reports even more difficult (Shuval et al., 2010). While all of these

observations may certainly be true, it would also be reasonable to expect larger effects in

this study because, in theory, undergraduate mediators should be able to deliver a higher

quality intervention than peer^aged mediators. However, given the significant differences

in climate and program implementation among the schools, it would also be unrealistic to

expect to find uniform change across these outcome measures. Thus, the positive findings

for negative affect are encouraging and the negative results for the rest of the measures are

understandable, if not slightly premature, considering the positive comments I received

from the student interviews for each of these outcome areas.

The finding of positive program satisfaction, however, was consistent with most

previous research on conflict mediation. In other studies, students cited confidentiality,

lack of teacher involvement, having a structured opportunity to resolve a problem by

talking, and being treated fairly as important reasons for feeling satisfied with the program

(Harris, 2005; Nix & Hale, 2007). Consistent with earlier studies, students described how

135

the mediation program helped them express themselves and avoid fights, and how it also

gave them a unique, structured way to resolve their problems. In order to improve the

visibility and acceptance of the program by students, program administrators and

mediators could also highlight the reasons students gave for their feelings of satisfaction,

which may be especially helpful for drawing in the students least likely to participate.

Reductions in negative affect. The findings in this study strongly support the

ability of conflict mediation to reduce students’ negative feelings of anger, but also of

sadness, and in some cases, even feelings of fear and school avoidance. An analysis of

potential moderators, or additional factors that could affect this change, also revealed that

these reductions in negative feelings were greater as more time passed after the conflict,

for girls, and for high school rather than middle school students. Students also indicated

that mediation alleviated strong feelings by helping them express themselves and giving

them deeper insight into their conflicts. While this may seem intuitive, one previous study

on peer mediation contradicted this finding somewhat as students reported that defusing

anger or other strong emotions was among the least helpful skills of the mediators (Harris,

2005); although, to be fair, a “least helpful” rating didn’t necessary mean unhelpful.

Unfortunately, as noted earlier, there is a paucity of research on the effects of school^based

mediation on student emotions. The finding that girls reported greater reductions in

emotions than boys is problematic, however, especially since boys report more exposure to

and participation in incidents of conflict at school. From the available data in this study, it

is unclear why high school students reported more reductions in anger than middle school

students. Conflict theory might suggest that increases in affective or coping skills

associated with older students could be an obvious explanation in addition to many other

136

factors (Johnson & Johnson, 1996). Also, the significant differences among schools in terms

of student characteristics makes it difficult to determine whether these grade level effects

are not just a reflection of variation among the schools. Unfortunately, the passage of time

is another confounding factor that limits confidence in this finding. Because of the

challenge of being spread across three schools, I was not able to follow^up with all

students within the same, consistent timeframe.

However, students’ comments from their interviews about how mediation helped

them more effectively manage their negative feelings lends additional support to the

finding that negative affect decreased after students participated in mediation. This is

consistent with research on conflict resolution that identifies changes in emotions and

emotional regulation skills as key factors for success in constructively resolving conflict

(Farrell et al., 2008; Sullivan et al., 2010). Indeed, students in the present study reported

that conflict mediation helped with awareness, expression, and management of negative

feelings including anger, which research suggests may ultimately result in fewer incidents

of both verbal and physical conflict at school. In fact, other research also supports findings

from this study that the sole act of getting help for problems can provide assistance to

students for coping with and alleviating negative emotions (Hunter et al., 2004). Students

appear to benefit from just seeking out help from adults or other sources in schools and

may experience relief for difficult or negative emotions independent of the actual type of

help provided. However, when the quality of help provided is poor or incompetent,

adolescents appear to learn from these experiences and may avoid reaching out to adults

and making the same mistake twice (Oliver & Candappa, 2007).

137

Mentoring and responsive relationships. Students’ comments indicated that one

of the most beneficial aspects of the mediation program was the relationship they had with

their mediators. The undergraduate mediators were central to students’ feelings of

satisfaction with the program, but they also played a distinctive role by offering students

an alternative to the more traditional relationships they had with teachers and other

school staff. Mediators were mentors to students in several ways such as being closer in

age than teachers, sharing some similarities in background, providing guidance on

personal concerns, and, at times, enhancing their academic aspirations and motivation to

succeed in school. In addition, the mediator relationships were responsive to students in

ways that were different than other adults at school. The mediators were able to provide

students with dedicated time, competent support, and safe and confidential spaces for

solving conflicts. Mediators did not get students in trouble for their conflicts, which likely

contributed to feelings of greater trust. The qualities found in these responsive

relationships may provide students with several benefits including an increased

willingness to seek help for problems, improved acquisition of new conflict resolution

skills, and, ultimately, some degree of protection from the negative effects associated with

exposure to high levels of conflict or violence in school.

Previous research has documented how positive relationships with adults provide

several benefits for students in schools. The kinds of mentoring relationships reported by

students resemble many of the qualities of positive, student^adult relationships that

researchers report in the literature. After conducting focus groups with adolescents about

reporting bullying at school, researchers concluded that the quality of the adult

relationship was a deciding factor for those students who were willing to seek help

138

compared to those who were not (Oliver & Candappa, 2007). Another study indicated that

students with existing adult relationships with high levels of trust and positive experiences

of seeking help were associated with a greater likelihood of approaching adults for help

with emotional distress (Wilson & Deane, 2001). In addition to trust, other researchers

found that adults with supportive qualities such as appearing interested, caring and

respectful were all associated with students’ willingness to seek help (Elliot et al., 2010).

This is consistent with a long history of previous research on general characteristics of

helpers that included qualities such as being genuine, empathetic and accepting, as well as

specific approach factors that adolescents prefer in adults like making themselves

available, having expertise, and being able to relate to teen issues (Lindsey, 1998).

Teachers as well as mediators bring many of these qualities to their relationships with

students. However, the mediators’ ability to provide confidential help differentiates them

from most other school staff who traditionally respond more harshly with discipline or

consequences, which can undermine feelings of trust between students and adults. A

reoccurring theme in this study was that students felt safe discussing their concerns with

the mediators because they knew they would respect their privacy. Some students in the

study also commented on how previous, negative experiences with getting help from their

schools made them feel less willing to get help in the future. In fact, some students in this

study had voluntarily returned to mediation several times, which suggests that positive

experiences combined with a consistent and accessible mediation program may have a

positive influence on students’ willingness to seek help.

Other studies have examined how mediators model certain kinds of skills in conflict

mediation and how this affects student learning. If students view mediators as mentors, the

139

learning outcomes could be enhanced even further. Other research on mentor^mediation

models with adolescents specifically has recognized the advantages of peer mediators

learning from older, more competent adults, but they were not able to describe how the

disputants themselves benefited from the mentoring with older mediators (Lane^Garon &

Richardson, 2003; Cassinerio & Lane^Garon, 2006). As a theoretical basis for many

mediation programs, researchers claim that students derive significant learning benefits

from those relationships with mediators where students perceive high levels of similarities,

competence and status (Lane^Garon, 2003; Harris, 2005). The expected benefits would be

even higher in cross^age relationships where university students mentor younger,

adolescent students because, presumably, they have greater competence and status as role

models than peer^aged mediators would (Lane^Garon & Richardson, 2003; Cassinerio &

Lane^Garon, 2006). While I did not directly observe any of the mediation sessions because

of confidentiality concerns, Harris (2005) conducted both surveys and observations of the

process and found evidence that modeling occurred during sessions and was significantly

related to the amount of students learning. In their interviews with me, students remarked

they were able to recognize some shared similarities with their mediators, the competence

demonstrated by some of the mediators, and the elevated status of the mediators as college

students, which, taken together, implies that students would be in an ideal position to

maximize their potential for learning new skills from the mediators. While the use of

undergraduate mediators has not been studied extensively, if at all, in the field of conflict

mediation, the findings from this study support a mentor^mediator model, which gives

students access to trusting and responsive relationships with competent adults who can

also model conflict resolution skills in a safe and supportive environment.

140

Environmental Influences on Student Responses to Conflict

Many of the students’ comments from the current study also highlighted several

contextual factors that play a significant role in either impeding or supporting a student’s

choice to respond to conflicts in positive and prosocial ways. The first major theme was

that peer culture represents a significant challenge for students as they attempt to seek

help and use non^aggressive approaches for resolving conflict. In particular, students

often commented on how peer pressures about projecting images of strength and

independence and the culture of snitching were substantial deterrents to their own efforts

at resolving conflicts peacefully. The second major theme was how school culture

influences students’ decisions to seek help for problems, such as volunteering for

mediation for example, or even putting prosocial conflict skills into practice during an

actual conflict. School staff plays a key role in the creation of a safe and supportive climate

that emphasizes trusting relationships with adults and encourages help seeking behaviors

among students. A student^centered climate where adults are responsive to students’

needs and foster trusting relationships with them may help mitigate negative aspects of

peer culture, promote the use of help seeking and conflict skills and, ideally, reduce the

overall levels of conflict and violence in schools.

Challenges in adolescence and peer culture. Students commented on how certain

kinds of peer pressures made it difficult for them to seek help for problems or adopt a

prosocial approach to conflict even if they had intentions of doing so. Students in this study

were under particular pressure to project images of strength and independence, avoid

appearing weak and not “snitch” on their classmates for any peer related conflicts. Students

also explained that serious consequences would follow if they violated these behavioral

141

expectations such as getting beat up, excluded from social groups or having their

reputations damaged. This is consistent with other empirical research on adolescence,

which has noted similar aspects of peer culture including the need to assert toughness and

independence, hide weakness, not tell on peers and enforce consequences when norms are

violated (DeLara, 2012; Oliver & Candappa, 2007; Farrell et al., 2010). High^risk students

may be especially vulnerable to peer influences as well (Wright et al. 2007), which is

worrisome as these students are more likely to experience conflict, but less likely to use

mediation programs. The culture of not snitching also poses challenges to building a

positive school culture based on principles of conflict resolution where programs like

mediation would encourage students to get help for peer problems. Other researchers have

also acknowledged that certain peer factors create barriers for using mediation specifically

such as adolescents’ feelings of distrust of adults, need for independence, concern over

reputations, prevalence of rumor^based conflicts and misconceptions about the program

itself (Theberge & Karn, 2004). In fact, even some mediators expressed difficulties with not

knowing exactly how to respond to these aspects of peer culture when students directly

inquired about them. Some students in the current study also endorsed aggressive

responses under certain circumstances. Students in urban, low^income, communities may

have to contend with even harsher behavioral norms where adolescents must respond to

certain kinds of aggression or provocations with force or violence or risk serious loss of

status or reputation, which may trump the use of prosocial skills learned from programs

like conflict resolution or others (Farrell et al., 2010).

The reality of social dynamics during adolescence reminds us that conflict

resolution programming must consider cultural expectations of acceptable behavior and

142

offset any barriers to accessing resources such as adult help or strategies for resolving

conflict peacefully. Not all adolescent communities follow the same codes of conduct and an

investigation into the surrounding cultural context would be an important planning step

before implementing conflict prevention programs. Researchers have found some ethnic

group differences in help seeking behaviors as well as responses to conflict (Elliot et al.,

2010; Farrell et al., 2010). Though inconclusive, these cultural variations among student

groups should still be considered. Implications of these findings may include incorporating

additional mediator training or other components, which would directly address ways of

responding to difficult peer pressures and cultural messages. Also, a mentor mediator

model of conflict mediation has certain advantages over peer^based mediation in terms of

its confidential nature. Adolescents may perceive compromises in confidentiality for peer^

based programs while those with neutral, third party adults may convey a stronger

message of support for privacy (Oliver & Candappa, 2007). Both the findings from this

study and previous research indicate that developmental features of adolescence like peer

norms represent important challenges in conflict resolution, but if adequately understood,

programs may also leverage these factors to enhance their reach and effectiveness.

School climate and adult response to conflict. Findings from the current study

suggest that schools have a direct impact on how students attempt to resolve their

conflicts including efforts to seek outside help if necessary. The ways that schools respond

to conflict as well as the quality of relationships that adults develop with students appear

to be influential factors in students’ decisions to get help. Similarly, a lack of these positive

qualities, i.e. a lack of trust, such as ineffectual adults, not keeping confidentiality or not

taking the problem seriously all contribute to the underreporting of incidents of bullying

143

harassment (DeLara, 2012). Students in this study also commented that other important

factors included a positive history of getting help, having a conflict of low severity and

developing a personal or trusting relationship with an adult. A school’s disciplinary

response to conflict is also a tricky issue as students report that consequences at school

appear to discourage fighting (Farrell et al., 2010). However, as students in this study

suggest, a one size fits all response that penalizes any involvement in conflict unfortunately

discourages students from trusting adults and seeking help for peer problems. A more

nuanced or tiered approach to school discipline that includes thoughtful, alternative

responses to aggression and conflict would help create greater trust among students and

an overall more positive school climate.

Some researchers conclude that close and supportive relationships with adults may

also help students solve problems nonviolently based on empirical links between feelings

of belonging in school and reductions in risky behaviors (Farrell et al., 2010). In fact, what

may be even more critical is teacher awareness of peer conflict and training in how to

respond to it effectively (Greene, 2003). For example, teachers who feel the most prepared

to handle bullying are also the most likely to intervene above and beyond students

disclosing these incidents (Novick & Isaacs, 2010). In fact, teacher behaviors appear to

influence even the delivery of conflict resolution programs as students who received their

own training in effective listening techniques tended to judge their untrained teachers

more negatively afterwards (Smith et al., 2002). Teacher beliefs about the effectiveness of

conflict resolution are also highly related to the adoption and support of mediation

programs in schools (Matloff & Smith, 1999). Similarly, if school administrators or staff

support are unaware of the importance of developing a positive help seeking culture in

144

their school and alternative approaches to traditional discipline, they may be contributing,

albeit unintentionally, to increased levels of mistrust between students and adults, greater

student exposure to conflict and lower levels of help seeking (Lindsay, 1998; Theberge &

Karan, 2004). My anecdotal experience with the three schools in this study seems to be

consistent with these implications of student^adult relationships and adult responses to

conflict. Milton high school, which had one of the lowest help seeking scores and fewest

percentage of students reporting getting help for problems among the schools, also had an

administrative staff who seemed committed to a punitive approach to discipline. While

awareness of serious consequences may actually contribute to less conflict in a school

overall, my experience with Milton suggested that the staff’s resistance to providing an

alternative response to student conflict may have been related to the difficulties the

mediators had with implementing the program and getting referrals for mediation.

This combination of factors that includes a school’s response to conflict, discipline

policies, the quality of student^teacher relationships, and the prevalence of conflict and

violence are all important indicators of a school’s climate, particularly for student safety. If

this aspect of school climate is overly negative, students may have no reason to believe that

turning to adults would actually help with their problems especially if they experience

conflict as unavoidable or have negative experiences with previous help seeking attempts

(DeLara, 2012). A negative school climate may also compromise learning outcomes for

students. For example, high levels of conflict could undermine students’ sense of self^

efficacy for resolving conflict, or a belief that one’s skills will produce an acceptable

outcome, which would negatively affect the long^term adoption of newly acquired skills

(Farrell et al., 2008). The findings from this study indicate that students who have trusting

145

and responsive relationships with adults are more likely to seek out help for problems,

which is a key component of creating a positive school culture and reducing violence and

aggression (Henry et al., 2011). In addition, schools that take steps to reduce risks

associated with negative peer attitudes about “snitching” or getting help should be more

likely to encourage help seeking behaviors among students and create a safer culture

(Oliver & Candappa, 2007). The findings in this study also suggest that conflict mediation

may reduce the stigma associated with snitching as students have a confidential space to

negotiate their problems, which, in theory, is also free from administrator intrusion.

However, a long history of research on school conflict such as bullying, for example,

suggests that changing the culture of a school requires a comprehensive, collective effort

(Jones, 2004; Olweus & Limber, 2010). Conflict mediation programs are most likely to be

successful when they are incorporated into a larger, school^wide strategy to improve the

safety and climate of the school. Even then, school climate outcomes are especially

resistant to change, although previous research has found some reductions in disciplinary

incidents and improvements in peer mediator perceptions of school climate (Smith et al.,

2002; Cassinerio & Lane^Garon, 2006). Ideally, a conflict mediation program represents

just one piece of a larger strategy to improve the safety climate of the school. The most

successful programs would also include high administrative support, high and low risk

students, ongoing training for mediators, and staff exposure to the principles of conflict

mediation (Lindsay, 1998). In the absence of a rigorous, school^wide approach, however, a

mentor mediation program still represents an appropriate and effective strategy for

reducing peer stigma and helping students manage conflict and develop responsive and

caring relationships with adults.

146

Concluding Remarks

The findings from this study suggest that the conflict mediation program, as

implemented in the three schools, has both strengths and weaknesses in terms of its

potential to help students resolve conflicts and improve the safety climate of schools. In

terms of its strengths, first, conflict mediation is an intervention that meets the

developmental needs of adolescents: it helps students develop social skills, involves the

peer group and encourages autonomy by allowing students to resolve their own conflicts.

Peer relationships take on critical importance in adolescence and learning how to manage

interpersonal conflicts is a key developmental task (Yoon et al., 2004). By discussing

conflicts directly with another disputant, some students in the program learned how to see

the issues from their peer’s perspective. In addition, the mentor model of conflict

mediation, used in this program specifically, provides students with access to positive role

models who assist students with their social and academic development. This mentoring

aspect of the program with college^aged mediators also offsets imbalances of power that

frequently accompanies bullying, harassment and other types of peer conflicts (Theberge

& Karan, 2004). Furthermore, the mediation program is, to some extent, confidential,

which respects the increased privacy and independence needs of adolescents as well as

reduces the risks associated with violating peer norms around reputations and snitching

(Oliver & Candappa, 2007). While students commented on several limitations to the

confidentiality of their mediations, the potential of the program to reduce this cultural

barrier for resolving conflict nonviolently and help seeking remains significant.

Second, mediation offers a unique approach for reducing conflict by creating safe,

confidential spaces for students when schools traditionally rely on top down, disciplinary

147

strategies to deal with aggression. Unfortunately, many schools have been notorious for

enforcing vertical and punitive discipline practices that punish negative behaviors and

reward positive ones, which neglects emerging research on the importance of horizontal

and social^emotional approaches (Osher et al., 2010). In fact, zero tolerance policies may

have the most detrimental effects as they remove far too many students from the

classroom, including many who are most at^risk for academic problems (Cornell & Mayer,

2010). Also, suspensions and other types of traditional discipline methods do not address

the underlying causes of violence in which verbal conflicts often precede physical

aggression (Talbott et al., 2002; Farrell et al., 2010). In contrast, conflict mediation takes a

horizontal approach to conflict as student disputants share a degree of power over the

decision^making process and the outcomes (Theberge & Karan, 2004). Many students who

participated in conflict mediation felt satisfied because they had never before experienced

such a non^punitive approach for dealing with conflicts in school. Students commented on

how the fear of punishment was a substantial deterrent for seeking help for their conflicts

as well. While conflict mediation was never intended to replace school discipline practices

per se, its ability to address underlying causes of physical aggression and violence makes it

an ideal intervention for reducing the harmful effects of exposure to conflict in schools.

The strengths of conflict mediation should also be considered alongside realistic

expectations about what the program can accomplish. Students did not always report

positive experiences, maintain friendships or change their attitudes about nonviolent

responses to conflict. In some ways, these attenuated findings conflict with other research

that found improvements in student relationships because of mediation specifically

(Harris, 2005). In terms of weaknesses, first, mediation is limited in duration, which may

148

reduce the program’s long^term effects. For example, sessions can be short, students

usually attend once, and there is often little or no follow^up or reinforcement. This is

generally consistent with the consensus in intervention research that short^term

programs are the most likely to produce short^term effects (Webster, 1993; Wright et al.,

2007; Shuval et al., 2010). However, other meta^analyses conclude that more program

hours do not always result in better outcomes as long as there is strong program

implementation (Garrard & Lipsey, 2007). Second, the mediation program may fail to

include the most at^ risk students as well as seem disconnected from the rest of the school.

Unfortunately, it is not uncommon to neglect the highest risk students in school^based

intervention research (Shuval et al., 2010). Furthermore, tight integration with and lasting

change to school culture may be limited without significant administrative support or

additional training components for school staff. However, if these schools implement

conflict mediation while taking into account the surrounding cultural context including

barriers and limitations, it should only improve their ability to offer positive programming

that prevents conflict and creates safer schooling experiences for students.

Lastly, this study’s findings suggested that several factors must converge in order

for students to successfully acquire new conflict resolution skills. While students’

comments were mostly consistent with previous research, they also raised concerns about

the difficulty of actually applying their skills in specific moments when emotions run

highest. Adolescents have even reported elsewhere that they believe fighting is necessary

or justified in some situations (Farrell et al., 2008). In fact, adolescents react with a wide

range of responses to conflict, from passivity to aggression, which may be appropriate in

the context of peer culture, but judged more harshly by adults who have different

149

expectations of behavior (Theberge & Karan, 2004). In other words, the social context

plays a role in how youth respond to conflict and what types of behaviors are considered

normative or adaptive to the circumstance. This would seem to suggest that the process of

teaching and learning new conflict resolution skills is not simply about the transmission of

information, modeling behaviors or shoring up “deficits” in students, but that careful

attention must be given to the surrounding context and how the culture of schools,

communities and even families influence students’ confidence and beliefs in prosocial

responses as viable and effective alternatives to violence. This also implies that adopting a

multi^pronged approach to conflict resolution would be most effective for reducing conflict

and violence in schools. Despite this challenging proposition, the advantages would include

not only imparting new skills to students, but also potentially achieving the same kinds of

competencies that researchers have found to make a difference for resilient youth.

Research Implications

This case study on four schools in a large, urban school district contributes

important contextual knowledge about students’ exposure to conflict and the impact of a

mediation program on student adjustment and development in several secondary schools.

The program under investigation was unique in several ways, which offers new insights on

conflict mediation with adolescents in urban settings, but also makes interpreting the

study’s findings in the context of previous research difficult. To my knowledge, there were

no existing studies on mediation programs with undergraduates as mediators in

secondary schools. While there are exceptions, the majority of research on conflict

mediation has not included students of color, low^income students or urban school

settings (Jones, 2004; Garrard & Lipsey, 2007). This study also contributed to the research

base on disputants’

150

perceptions of the mediation process, which has also been neglected to some extent

previously. Findings from this study would have high external validity and may inform the

evaluation or delivery of other mediation programs in urban school districts, which may

have comparable levels of conflict in addition to other cultural similarities.

While many of the results from this study’s questionnaires were consistent with

previous research, some findings were not easily explained, in part, because of

methodological limitations, but also because they were different or not explored by other

researchers. For example, this study was unique because it included both middle and high

school students and found differences in help seeking attitudes by grade level. Within

middle and high schools, help seeking attitudes decreased as grade level increased, but

high schoolers, on average, had higher levels of help seeking than middle schoolers, which

other studies had not reported previously. The relationship between age or grade level to

help seeking still needs further exploration. Also, despite early calls to include emotional

affect in studies of school^based mediation (Johnson & Johnson, 1996), this study was one

of the first to find significant reductions in negative affect after participation in mediation.

While variations in how I delivered the post^questionnaires in combination with the

significant interaction between time and negative affect limits the validity of these findings,

emotional changes in participants continues to be an important area worth exploring in

future research especially given their important role in resolving conflicts (Farrell et al.,

2008). Lastly, this study found some differences in help seeking attitudes and behaviors

among the schools. In particular, students at the small, pilot and charter public schools

appeared to be more willing to seek help for conflicts or, in fact, reported seeking help

more frequently. While no reliable conclusions can be made from this finding because

there

151

were so few schools in this study, it raises the question of whether smaller schools create

more positive school cultures in terms of safety or student^adult relationships. However,

existing evidence on the issue of school size continues to be mixed (Lee & Smith, 1997;

Klein & Cornell, 2010).

Findings from the student interviews contributed to a rich understanding of the

impact of conflict mediation and the contextual factors that influence these effects. Several

reoccurring themes from students’ interviews suggested that multiple factors at both an

individual and environmental level affected how likely they were to learn new skills from

mediation, feel satisfied with the outcomes of mediation and seek help for future peer

conflicts in school. For example, students’ comments implied that factors like the severity

of the conflict, a history of friendship among disputants, prior mediation experience and

the students’ role in the conflict all had an effect on their attitudes or experiences in the

program. Researchers have recommended testing several of these factors in future studies

of program effects including the degree of closeness or friendship among students and the

type of resolution, but, for some reason, subsequent research has not followed this advice

very closely (Johnson & Johnson, 1996; Roberts, Yeomans, & Ferro^Almeida, 2007; Farrell

et al., 2008). In addition, participation in the program appeared to have differential effects

for students depending on their role in the conflict such as being a victim, bystander or an

aggressor. Unfortunately, I did not have a large enough sample to unpack these

relationships or find statistical effects. However, given the strong and consistent findings

of more beneficial effects for peer mediators than disputants, an investigation of a

student’s role in peer conflict appears to be long overdue in studies of conflict

interventions (Yoon et al., 2004). Lastly, findings on the mentorship qualities of students’

relationships with the

152

mediators were unexpected, but highlight this critical component of cross^age mentoring

in mediation. It is also not surprising that this hasn’t been examined in previous research

because most mediation programs include cadres of peer mediators, which reflects a much

different type of relationship. With cross^age mediators, there would also be further

potential to build on previous work that opens the “black box” of mediation and looks at

the underlying process (e.g., how skills are modeled in sessions) in addition to outcomes

(Harris, 2005). The study of mediation process with older mediators may also prove useful

for developing program theories of change since the mediations will arguably be of higher

quality than peer models.

Program Implications

While many of the positive findings in this study support the mentor mediation

program in its current form, student and mediator comments’ also indicate that there is

ample room for improvement. Students mentioned several reasons for feeling dissatisfied

with the program including not taking mediation seriously, a lack of confidentiality and

neutrality, difficulty dealing with strong emotions and poor facilitation of mediation

sessions. This is mostly consistent with previous research on dissatisfaction among

disputants and underscores the importance of mediator training and levels of

preparedness as direct contributors to students’ program satisfaction (Nix & Hale, 2007).

These mediator factors also raise issues related to program implementation quality, or

fidelity, which intervention researchers have studied quite extensively. Historically,

research on the effectiveness of evidence^based practices has had a large influence on

programs in violence prevention (Cornell & Mayer, 2010). While researchers recognize the

importance of intense monitoring of program processes and transactions for maximizing

153

results (Garrard & Lipsey, 2007), this study along with other research suggests that the

process of mediation may be more malleable than other kinds of interventions (Nix & Hale,

2007). This flexibility, or deviations from the “training script”, might actually confer certain

kinds of benefits. Disputants may feel less satisfied with their mediations when mediators

force them back to the original issue after veering “off course” to vent emotions, or, in other

words, when mediators resist departures from the scripted process (Nix & Hale, 2007).

However, mediators may also need to maintain a fine balance here because students also

report high satisfaction when mediation is a controlled process where mediators treat

disputants fairly and enforce ground rules (Harris, 2005). In the present study, many of the

students’ complaints as well as my own observations suggested that the mediators did not

always implement the program faithfully. However, future training of mediators may need

to emphasize how to handle these kinds of disputant departures rather than strict

adherence to a predetermined, or an “evidence based”, set of rules and procedures.

The findings from this study and other research suggest that certain contextual

factors have a significant impact on the effectiveness of the mediation program. This

implies, for example, that the intervention should be well integrated into the culture of the

schools as well as adapted to reflect the heightened importance of peers in adolescence.

One of the most consistent and reoccurring recommendations in school^based research is

the need to adopt whole school approaches for anti^violence interventions, which offer

intentional, coordinated and sustained program elements at every level of the school

environment (Jones, 2004; Yoon et al., 2004; Theberge & Karan, 2004). Examples of

additional school components that could enhance the delivery of the program would

include training and building awareness for school staff, administrative buy^in and

support,

154

classroom integration and inclusion of peers at various points in the process such as

referrals or in the mediations themselves3. Developmental considerations would also

include reducing barriers to help seeking behaviors for adolescents by emphasizing a

referral process through peer networks or self^referral rather than adults (Wilson &

Deane, 2001). In addition, how the referral process is integrated with other school policies

for handling conflict such as discipline is extremely important because associating

mediation with traditional discipline may deter students from seeking out the program.

Mandatory referrals may also discourage motivation and reduce learning benefits, but, on

the other hand, they may increase the involvement of high^risk students who wouldn’t

otherwise participate (Theberge & Karan, 2004). Greater efforts must be made to recruit

“high risk” students into the program as previous research as well as evidence from this

study indicate that students with the highest exposure to conflict are also the least willing

to seek help or participate in mediation. There is a clear message that if these school and

social context factors are ignored, programs that target individual students will not

produce their desired effects (Coyle, 2008; Farrell et al., 2010).

Limitations and Caveats

It is important to delineate limitations and caveats to the conclusions I make about

the relationships among the variables in this study. In order to make claims about this

study’s findings, I must first draw inferences from the data, which can vary in their degree

of validity, or, in other words, the extent to which they are an accurate or truthful

representation of reality. Shadish, Cook and Campell (2002) outline a framework for

examining the truthfulness of a study that breaks validity down into four, main components

including statistical, internal, construct and external types. The authors explain that the

155

statistical subtype includes the validity of inferences about the correlations among the

variables; internal includes inferences about the relationship between the treatment and

outcomes (if A causes B); construct includes inferences about the measurement of

psychological constructs and their reliability; and external includes inferences about

whether the observed relationships would also be true for (or generalize to) different

people, settings, treatments and outcomes. Extensive reviews on conflict resolution

research have also identified a host of methodological problems, which appear in the

present study as well, such as a lack of program specificity (i.e., the independent variable),

no observations of the mediation process, few or no experimental designs, short^term

studies, lack of valid measures and an overreliance on self^report (Johnson & Johnson,

1996). Similar to this study, these so^called “threats to validity” reduce our confidence in

causal relationships, but, on the other hand, these kinds of studies also increase our

confidence that findings would hold across other settings because researchers conducted

all of them in schools. I will consider similar threats to validity according to this

framework and some steps I took to address them for each study component including the

classroom questionnaires, the pre^post tests and the student interviews.

Statistical validity. There were several factors that may have limited my ability to

detect statistical relationships among the variables in the study. For the classroom

questionnaires, many of the questions about exposure to violence were heavily skewed

towards “never” or low frequency, which indicated that students either underreported

their exposure or the measures were not sensitive enough to detect differences on a

weekly, monthly or yearly basis. The reduced range on these measures can weaken their

relationship with other variables in the study. The intervention sample was also small,

156

which may have reduced the likelihood of detecting significant changes from pre^ to post^

test. Also, based on my observations and interviews with the mediators, the

implementation of the intervention seemed to vary widely. These two factors may have

contributed to the lack of significant pre^post findings when, if adequately addressed,

participating in the intervention could have produced measureable changes in the

dependent variables. In order to partially account for these threats to statistical validity, I

took several different actions. I would sometimes collapse the range of measures when

appropriate to report “any” vs. “no” exposure. As the study progressed, I included three

schools instead of one from the original design in order to increase the sample size. Also,

as appropriate, I used non^parametric, rank order tests when ANOVA assumptions about

the data were violated, which most often occurred because the residuals were either not

normal or homogeneous.

Internal validity. For the pre^post tests, one of the largest difficulties in

determining whether the intervention caused a change in the dependent variables was the

“block box” nature of the program. Two factors including the lack of program specificity as

well as an inability to observe mediation sessions (i.e., measure fidelity) obscured the

actual mediation process and made it difficult to attribute outcomes to specific

components of the program. For example, there was no way to know whether a positive

relationship with the mediator or adherence to the conflict mediation steps was

responsible for reductions in negative affect. Another contributing factor was an

inconsistent data collection process. Because I was spread out among three different

schools where mediations occurred simultaneously, it was difficult if not impossible to

collect data within the same time frame for all participants. I also only collected data

during the 3rd quarter of

157

the program’s implementation during the school year. Unexpected changes in the study

design did not allow me to collect data in both the 2nd and 3rd quarters of the program. This

may have biased the sample towards mediators who were more or less motivated, who

had different reasons for participating, or who had more experience than 2nd quarter

mediators. It’s difficult to say whether the 2nd or 3rd quarter implementation of the

program would have been more effective. On one hand, mediators had less experience in

the 2nd quarter; however, on the other hand, there seemed to be a significant number of

mediators in the 3rd quarter who seemed unmotivated to participate, which raised

questions about their intentions in taking the class. This may either over or underestimate

the strength of study’s findings depending on whether the 2nd or 3rd quarter class had more

of a motivated or effective team of mediators.

In addition, significant differences among the schools such as size, type, ethnic

composition and number of years they participated in the program may have added other

unmeasured variables that could have affected the relationship between the program and

the outcomes. Lastly, students’ initial exposure to an outcome measure in the pre^test may

have influenced how they answered the same measure in the post^test, which is always a

concern in pre^post designs. In general, these threats to the internal validity of the study

make it impossible to determine whether the mediation program actually caused changes

in students’ responses. However, in order to account for some of these threats, I employed

a mixed^methods approach to gather additional, qualitative data on the potential,

underlying change processes. I also conducted interviews with mediators in order to

gather additional data (albeit self^report) on what actually occurred during the mediation

sessions. Both student and mediator sources provided rich descriptions of the context and,

158

after data analysis, revealed potential explanatory causes for changes that students

experienced after they participated in the program. I also attempted to see whether

differences in measurement time were related to changes in the outcome measures and, in

fact, I found a significant relationship. In the presentation of the qualitative data, I also

attempted to highlight strengths, weaknesses and exceptions to common themes in order

to present a balanced perspective, i.e., a more valid description of what students said.

Construct validity. There were several threats to the construct validity in the study

that included my selection of measures and how students responded to my questionnaires

and interview questions. First, as noted earlier, all the study data was self^report, or “soft”,

and there were no other “hard” measures of student attitudes or behaviors such as direct

observations or school records of discipline or referrals. Also, as much as I attempted to

make students feel comfortable and assure them their answers were confidential, there

was likely some degree of pressure to provide desirable answers especially as it related to

their learning or satisfaction with the program and the mediators. Furthermore, students,

especially those in middle school, may have felt intimidated during our individual

meetings, which also would have affected their responses. In order to address some of

these issues, I gathered data using multiple methods including questionnaires and

interviews. I used the student questionnaire data to some extent to inform my

interpretation and analysis of both student and mediator interview data. As much as

possible, I also took the measures in this study from examples in previous research in

order to compare findings for consistency. I also only included existing measures that had

either been previously validated or shown to have satisfactory ratings of reliability.

However, the internal consistency of the nonviolent

159

response strategy scale was fairly low, which compromises the integrity of this measure.

Fortunately, this is good news for the study since the measure did not detect any changes.

External validity. There were several threats to external validity in this study as

well as additional factors that may strengthen its ability to generalize findings to other

students or schools with conflict mediation programs. For the classroom questionnaires,

there were differences among the schools in how I sampled classrooms. The two small

schools were small enough to include all the classrooms. At the two large schools, I had to

rely on a staff member to help select the classrooms, which were chosen in a way

convenient to this contact person. While the small schools included complete populations,

the larger school samples may not have been representative of their school populations.

Both the classroom and intervention samples were primarily female and Latino, which

were the groups least likely to report exposure to violence and most likely to report

willingness to seek help for conflicts. The findings of the study may not generalize to

schools that are not majority Latino. While there were no differences in changes in

dependent measures by school, the results also may not apply to large, comprehensive

schools since two of the three schools in the sample were small, pilot or charter schools.

For the intervention results in particular, participant data also suggests that the sample

was predominately “low risk”. Therefore, results may not generalize to higher risk

populations. The conclusions I made about contextual factors from the student interviews

only included students who participated in mediation, which was not representative of the

whole school. For example, these students may have been more likely to seek help or have

better relationships with adults, although they did have to experience at least some

conflict in order to participate in the program.

160

Despite the inclusion of schools that significantly differed from each other in

multiple ways, I tested for school interactions with outcomes but did not find any

differences with the exception of negative affect. I did find significant differences by school

and group for the classroom data analysis and this was reported in the study’s results. In

order to provide a more balanced perspective on conflict and contextual factors for

adolescents, researchers recommend recruiting students who were involved in conflict,

but did not participate in conflict mediation sessions (Smith et al., 2002). As noted earlier,

the absence of these students from the general school population limits the applicability of

this study’s findings to other schools, but I also highly recommended their inclusion in

future research. At the same time, including student perspectives from multiple school

sites may also improve the external validity of the current study. The broader diversity of

student and mediator experiences across multiple schools should strengthen our

confidence that the findings would generalize to other schools within the district.

Reflections

In this concluding section, I discuss several themes about the process of conducting

an investigation into the workings of a school^based program, which in this case,

happened to be a conflict mediation program, but really could have been any number of

similar, social^emotional interventions for students. The singular, overarching theme of

this study was how to continually adapt its design to the rapidly shifting conditions and

relationships in schools while still attempting to preserve some level of validity and an

adequate sample size with meaningful data. Some additional caveats come to mind about

the relationships I had with the UCLA mediators and the staff at the schools and how they

affected my data

161

collection process. Lastly, I will close with reflections on the research process and how I

might do certain things differently after muddling through this experience.

The design of this study went through several dramatic changes from its initial

inception as an action research proposal in only one school. The original intent of the

study was to build upon a longstanding university^community partnership, which, in my

mind, represents the pinnacle of conducting applied research because it melds the

strengths of each institution (Shuval et al., 2010). For example, the university and

partnering school, or organization, collaborate together on identifying problems,

developing research questions and collecting and interpreting data in order to improve a

localized problem, which, in theory, should confer more direct benefits to either students

or schools than research conducted by outsiders. Moreover, I had intended to have more

overlap between the two phases of the study. I wanted to use school context data from

phase 1 to understand the experiences of students in the program reported in phase 2,

which would have been easier with one site. I changed school sites three times and ended

up with a case study and final sample of four schools, over 800 classroom questionnaires

and over 40 students who participated in the mediation program. The sheer size of the

study grew to something that was almost unmanageable for one person and it was a far

cry from my original intention of taking a deep look at the context of school conflict at one

site where I had already become familiar with the school community including some of the

students and staff.

While the changes in this study were necessary and often beyond my control, they

also introduced several unanticipated challenges. The need to monitor the implementation

of the program in three schools was especially difficult. Not only did I have to quickly

introduce myself to the school site coordinators and teams of mediators, but I also had to

162

figure out how the team had implemented the program in each site in order to track

mediation referrals and follow up with students in a timely fashion. Also, I had to spread

myself across three different schools because I was uncertain about how many referrals

and participants I would get for the study until the school year was almost over.

Unfortunately, I believe this resulted in too much data, although this may be preferable to

not having enough data from a research perspective.

These growing pains were not just challenging for me, but they also strained my

relationships with school staff and the undergraduate mediators. Relationships and trust

often take time to build especially in school communities and I only had a few days or

weeks to acquaint myself with the coordinators and the mediators. To this day, I’m still not

sure if this was only in my own head, but many if not most of the mediators and at least

one school site coordinator seemed to view me with suspicion. Despite my best intentions

of clarifying my role and explaining the voluntary nature of the study, my relationships

with the mediators and some school staff members became strained. In fact, at one point, I

was called into a meeting because, apparently, some of the mediators felt intimidated by

my presence at their school site. Unfortunately, although I was somewhat aware of how my

presence might affect the implementation of the program, I underestimated the extent to

which the undergraduate students associated me with an evaluative role of their own

performance especially since they were volunteering in the schools for class credit. And,

understandably, some of the mediators and school staff were protective of their students

and initially did not seem comfortable having an outsider meet with them.

Lastly, trust was also difficult to establish possibly because of differences in identity

or social background. Mediators participated in the program through a Chicano Studies

163

class at UCLA and most were Latina, undergraduate students. There were a few male

students and others from different ethnic backgrounds, though they were a small minority.

The schools predominately served students of color and many if not most of school staff

members I interacted with were also women of color. For all practical purposes, others

tend to view me phenotypically as a white male, which I believe accentuated my status as

an outsider and was partially responsible for difficulties in creating trust in these

relationships especially given the considerable time pressure. Trust is often a sensitive

issue in underserved communities and in no way would I minimize the importance of

establishing long^term relationships before embarking on a research project that has high

potential for either misunderstanding or misuse. My original plan was to stay within one

school community; however, out of practical necessity, I had to spread myself across

several schools, which introduced unexpected challenges such as these. Although, to be

fair, similar issues may have arisen even if I had stayed in only one school site.

As noted earlier, there were many complications in this study and there are several

methodological improvements I might make should I ever have another opportunity to

conduct a school^based research study. For example, the classroom data from phase 1

provided meaningful data about the context of exposure to violence and student

perceptions of the safety climate. This data could be used to inform the data collection in

phase 2 by purposefully recruiting students who may be at elevated “risk” for experiencing

or participating in school conflict. In terms of the study’s measures, I believe I over^relied

on the use of existing measures and did not devote sufficient time to pilot testing them for

use in this study. Even though I adapted my exposure to conflict measures from national

surveys, many of my frequency scales were not sensitive enough to detect infrequent

164

reports and were positively skewed towards the “never” end of the spectrum. This lack of

range likely obscured significant relationships especially in the within^school analyses. If I

had sufficiently pilot tested my questionnaires, hopefully I would have been able to identify

these weaknesses and improve them for the actual study. Also, student learning is a

complex process and measuring learning outcomes with respect to such a short^term

intervention requires very precise measures that are tightly bound to the theoretical model

of change. In hindsight, I probably should have chosen a different indicator of learning

prosocial conflict resolution strategies that would have been more consistent with either

the program theory or the short^term nature of the intervention. In fact, if I had stayed at

one school site, I had planned to include other kinds of “hard” measures such as behavioral

referrals and discipline incidents, which is another way of understanding the impact of the

program on the school community as students often use the program only once.

My own experience with conducting this research study over the past year has also

helped me understand the value of having a flexible research design, using multiple

methods, and distinguishing between “process” and “outcomes” in intervention research.

This study went through numerous changes often in fast succession as the conditions in the

schools were constantly changing, which prevented me from sticking to my original plan. A

case study in some ways resembles an action research approach as it allows the researcher

to make “mid^course corrections” not possible in traditional research designs (Erickson,

2006). In addition, mixed^methods, which is the cornerstone of case studies, provides an

additional “safety” plan when the original design does not work as intended. For example,

shifting to qualitative or descriptive, process data is especially helpful when either the

collection or analysis of quantitative data does not go as planned (Shuval et al., 2010; Smith

165

et al., 2002). I believe these aspects of my research study were extremely helpful for

allowing me to adapt to the challenging conditions of school^based research.

Lastly, the shift in this study’s focus from outcomes to process in some ways

represents a central lesson of effectiveness research with school^based interventions. As

noted earlier in the literature review, whole school, anti^violence interventions have not

necessarily lived up to their promises in part because of shortcomings in program

implementation or fidelity. The difficulties I had with tracking program referrals and

recruiting participants was likely somewhat related to problems with program

coordination and implementation in the school sites. In fact, I am almost sure that my

presence in the schools increased the number of mediations that took place because of how

I had to follow up with both school staff and mediators to track what was happening with

the students. Schools are often communities unto themselves and, in some ways, the

mediation program looked different at each school site. The quantitative outcome findings

in this study were fairly modest, which may reflect some of this variation in program

implementation. The use of qualitative methods in this study to collect data on process was

also especially instructive in this situation. If this contextual data on the mediation

program is included in efforts to improve program fidelity or implementation, then future

studies of the mediation program may be in a better position to utilize more quantitatively

focused designs such as control groups or even treatment comparisons with peer

mediation models, for example. Because of the sometimes disruptive conditions of real

world research, it will always remain challenging to understand the true nature of the

impact of these kinds of programs without sufficient attention to program coordination

and fidelity.

166

Footnotes

1The school names have been changed in this report to protect their privacy.

2The school sites in the final study were not the intended sites from the original

study design. The changes in the study design and school site selection happened for

several different reasons including some factors that were in my control and others that

weren’t. During the 2012^‘13 school year, the UCLA conflict mediation program operated

in at least seven different schools in the district. Originally, I had planned to complete a

different iteration of this study, which included both phases 1 and 2, at only one school,

Joad, where I had established relationships from the previous school year. Joad was also an

ideal site because of the program history at the school and the large number of referrals in

the previous academic year (see Figure 1). I started to collect classroom data for phase 1 at

Joad during Winter Quarter, 2013. However, midway through Winter Quarter, several

administrative problems arose and the program was no longer able to stay at Joad for the

rest of the year. It was obviously not possible to continue with phase 2 of the study, which

is why I have data for all four schools for Phase 1, but not for phase 2. I then switched the

site of the study to SOUL, an alternative, charter high school district for students who were

at risk of dropping out of school. I chose this new setting because I anticipated a high

number of student conflicts and, therefore, a sufficient amount of referrals to the conflict

mediation program. Unfortunately, the mediation program was in its first year at the school

site and difficulties with integrating the program into the school resulted in almost no

referrals during Winter Quarter (see Figure 1).

To minimize the risk of not having enough mediation participants for phase 2, I

made the decision to switch to three other school sites that could produce more referrals.

167

The three schools, Casy, Sharon and Milton, each had teams of mediators continuing into

Spring Quarter, 2013, which was the earliest opportunity I had to resume data collection.

Because I now only had one quarter instead of two for data collection, I chose to stay in

these three schools to increase the odds of recruiting a large enough sample for the second

phase. In Spring Quarter, I completed phases 1 and 2 at each of these three schools. As it

turned out, the number of mediation referrals in the three program schools during Spring

Quarter was roughly equal to Joad’s referrals from Winter Quarter in the previous school

year (see Figure 1). The original intention was never to make any comparisons across

schools, but rather conduct a focused study of just one school on exposure to conflict,

program implementation and effects, and the school climate. However, especially in

schools, even the best of intentions do not always work out as expected, so the current

study reflects a diversity of student experiences across several different kinds schools,

which hopefully will improve the value of the study’s findings.

Figure F1: Mediation Referrals, 2012N2013

SOULJoad '11^'12MiltonSharonCasy

JanFebMarchAprilMayJune

168

3Milton’s HEART program is a good example of incorporating peers into the conflict

mediation program model. At Milton, the UCLA undergraduates train high school students

as mediators and also work with them on other kinds of school issues related to conflict

and social^emotional learning. The HEART students along with the undergraduates make

classroom presentations with a performing arts focus on various adolescent issues as well

as build awareness about the importance of resolving conflict peacefully.

169

Appendix A

Classroom Questionnaire

Thank you for your interest in taking the School Safety Survey. This survey should only

take 5 – 10 minutes of your time. Please answer each question the best you can by circling

the correct answer choice. There are no right or wrong answers. Please DO NOT put your

name on the survey so your answers will be private. You will not get in trouble for how

you respond so please be as honest as possible.

1. What grade are you in? (please circle)

6th 7th 8th 9th 10th 11th 12th

2. What is your gender? (please circle)

Male Female

3. What is your race or ethnic background? (please circle all that apply)

Hispanic/Latino Black Asian White Other

Please circle one answer that best describes what you think for each question.

4. Have you been in a physical fight at school in the last year? (please circle)

a. Yes

b. No

5. Do you see/witness physical fights at school? If so, how often? (please circle)

Please answer some questions about what students can do at school that can make you feel

bad or hurt your feelings. Sometimes we call this bullying or harassment. Has any student

bullied or harassed you? This can include any of these:

6. Are you physically bullied in school? If so, how often? (please circle)

Examples: being hit, pushed, slapped, kicked, spit at, or beaten up, had things stolen or

destroyed, forced to do something you don’t want to do like give money or something

a. b. c. d. e.

Never Once/Year Once/Month Once/Week Every Day

School Conflict and Bullying Questions

170

7. Are you verbally or socially bullied in school? If so, how often? (please circle)

Examples: called names, teased hurtfully, insulted, made fun of, humiliated, threatened,

excluded from a group or activities on purpose, spread rumors or gossiped about you

a. b. c. d. e.

8. Are you bullied online, or electronically, by other students at school or at home?

Examples: using cell phones, texting, Facebook, videos, or other web sites to tease,

threaten or put people down.

a. b. c. d. e.

Never Once/Year Once/Month Once/Week Every Day

9. Are you racially harassed? If so, how often? (please circle)

Examples: called names or treated differently or badly because of your race, culture,

ethnic background

a. b. c. d. e.

Only answer this next question if you said you were bullied or harassed at least once in

questions 6 – 8. If you’ve never been bullied or harassed, please skip this question.

10. Try to think of the last time you were bullied or harassed in school. Did you get

help from another student or an adult at school for what happened?

a. Yes

b. No

i. If yes, who did you get help from?

i. If no, why didn’t you get help?

11. How often do you see any kind of bullying or harassment at school that happens to

other people (not including yourself)? (please circle)

a. b. c. d. e.

12. If you see or witness conflict or bullying at school, does it make it hard for you to

get your schoolwork done or concentrate in class? (please circle)

a. b. d. e.

Not at all A little bit Sometimes All the time

Never Once/Year Once/Month Once/Week Every Day

171

If you hear about a threat to someone’s safety or a conflict at school, can report it to an

adult without “snitching”?

a. Yes

b. No

For the next set of questions, please circle the answer that best describes how much you

agree or disagree with the following statements:

Statement Strongly

agree

Somewhat

agree

Somewhat

disagree

Strongly

disagree

(For practice): I like to run

during PE.

1. If I tell a teacher that someone is

bullying or harassing me, the teacher

will do something about it.

2. It’s okay for students to talk to

someone when other students are

being bullied or have conflicts.

3. Teachers make it clear that bullying

and harassment are not acceptable.

4. There are adults here I can talk to if I

have a personal problem.

5. If another student were bullying or

harassing me, I would tell someone

about it.

6. Students here try to stop bullying and

fighting when they see it.

7. Teachers at this school are genuinely

concerned about me.

8. If another student said they were

going to hurt someone, I would tell

another teacher or adult at school.

9. If another student brought a weapon

to school, like a knife or gun, I would

tell someone like another teacher or

adult.

172

Appendix B

Program Post (Pre) Questionnaire

Thank you for your interest in taking the Conflict Mediation Survey. This survey should

take less than 10 minutes of your time. There are no right or wrong answers. Please be as

honest as possible. Your answers are completely private and you will not get in trouble for

any of your answers.

You will see a list of words below that describe different feelings and emotions. Read each

word and then circle the appropriate answer next to that word.

Please think of the problem that brought you to the conflict mediation. Tell us how much

you have felt this way the last time you saw the person you had a conflict with:

Feeling or Emotion: Not at all A little Moderately Quite a bit Extremely

Upset 1 2 3 4 5

Mad 1 2 3 4 5

Alert 1 2 3 4 5

Ashamed 1 2 3 4 5

Excited 1 2 3 4 5

Nervous 1 2 3 4 5

Strong 1 2 3 4 5

Interested 1 2 3 4 5

Afraid 1 2 3 4 5

Energetic 1 2 3 4 5

The next time you find yourself really angry with someone or something, how likely is it

that you would…?

Very unlikely

Unlikely

Likely Very likely

1. Ignore the situation? 1 2 3 4

2. Ignore it, but get the person back later? 1 2 3 4

3. Try to talk it out? 1 2 3 4

4. Suggest going to mediation? 1 2 3 4

5. Channel your anger into something constructive? 1 2 3 4

6. Laugh it off? 1 2 3 4

7. Try to calm down? 1 2 3 4

Questions About How You Feel

Questions about Anger

173

8. Try to see the other person’s point of view? 1 2 3 4

Please circle your answer based on how much you agree or disagree with these statements.

Strongly Agree

Agree

Disagree

Strongly Disagree

The mediators clearly explained the rules. 1 2 3 4

The mediators really listened to me. 1 2 3 4

The mediators understood my point of view. 1 2 3 4

The mediators knew what they were doing. 1 2 3 4

One thing I liked about this mediation was:

One thing I would change about the mediation or do differently for next time is:

Please circle a number for how much you agree or disagree with each of the following

statements.

Statement: Strongly Agree Agree Disagree Strongly Disagree

I liked going to the mediation.

1 2 3 4

The mediators treated me fairly and listened to me.

1 2 3 4

I’m satisfied with the agreement from the mediation. I

would use mediation again.

I would recommend mediation to my friends or classmates.

1 2 3 4

Since the mediation, I get along better with the other person.

1 2 3 4

For the next set of questions, please circle the number that best describes how much you

agree or disagree with the following statements.

Statement: Strongly Agree Agree Disagree Strongly Disagree

If I tell a teacher that someone is bullying or harassing me, the teacher will do something

about it.

Mediator Feedback Questions

1 2 3 4

174

1 2 3 4

It’s okay for students to talk to someone when they have problems or conflicts with other

students.

1 2 3 4

There are adults here I can talk to if I have a personal problem.

1 2 3 4

If I had a problem or a conflict with another student,

I would tell someone about it.

1 2 3 4

Adults at this school are genuinely concerned about me.

1 2 3 4

If another student said they were going to hurt someone,

1 2 3 4

I would tell someone like another teacher or adult at school.

1 2 3 4

If another student brought a weapon to school, like a knife or gun,

1 2 3 4

I would tell someone like another teacher or adult.

1 2 3 4

175

Appendix C

Student Interview Protocol

Q1: Previous Mediation History

1) Was this your first time in mediation? How many times have you been before?

a. What did you think of your other mediation(s)? Was it helpful?

Q2: Current Mediation Context

2) How were you sent to the mediation program this time? Did you have a choice?

a. How many other students were involved? Individual or group meetings?

b. What was your relationship with the other person like before the problem?

c. How did you feel about the person when the conflict/problem happened?

Q3: Mediation Process and Relationship to Mediators

3) What did you think about the way the mediators handled your conflict?

a. Did they do or say anything that you remember? That was helpful?

b. Do you think you learned anything from them? What?

c. How did you feel talking about your problem with the mediator? Why?

d. How is talking to a mediator different than a teacher or staff at school?

e. How did it feel to share your side of the story? How was it to hear the

other person’s side? What was your reaction?

Q4: Mediation/Program Outcomes

4) Were you able to reach an agreement at the end of the mediation?

a. What was it? And how did you come up with that?

b. Did the agreement work? Were you satisfied with it?

c. If it hasn’t been working, why do you think that’s the case?

d. Did how you feel about the person change after the mediation? How?

Q5: Conflict Resolution Skills/Learning

5) Looking back, how would you normally have tried to solve this conflict/problem if your

school didn’t have a mediation program?

a. What if you had a similar problem in the future with another person and

how would you try to solve or fix it then?

Q6: Mediation/Program Satisfaction

6) What did you like about going to mediation? What didn’t you like about it?

a. Would you try mediation again if you had another problem with someone?

b. Would you recommend mediation to your friends? Why/why not?

Q7: Program Improvement/Barriers

7) Can you think of reasons why students wouldn’t want to use mediation?

a. What would you say to these students to encourage them to use it?

b. Is there anything you would change about the mediation program to make it

better or to have more students use it at school?

176

Appendix D

Mediator Interview Protocol

Who was involved in the mediation?

How many students? Were they participating voluntarily? How

many times did you meet with them? Follow^ups?

What was their relationship like before/after?

What was your relationship like with the students? Before/after?

How did the mediation session/process go for you?

Did you use the steps (ITUNA)?

How did the students engage in the mediation?

Did they take it seriously? Share their perspectives? Hear the other person’s side? Were

there any “breakthrough” moments during the mediation?

How did the mediation end?

What was the agreement and how did you get there? Were you satisfied with it? If not,

what could you have done differently? Did they stick to the agreement?

What parts of the mediation were the most challenging? What

parts of the mediation were the most rewarding?

Learning:

What do you think the students learned from you? How do you know? Was

there anything non^mediation related you talked about?

How did this mediation affect you? Do you think you learned anything? What?

Summative:

What would you like to improve on for your next mediation? Do differently? What

would you like to repeat/keep the same in your next mediation?

What would you change about the how the program is currently working?

Q1: Mediation Context

Q2: Mediation Process

Q3: Mediation Outcomes/Program Satisfaction

Q4: Learning Outcomes / CR Skills

Q5: Program Improvement

177

Table A1

Appendix E

Primary Mediation Outcomes by School

Primary outcome scores from pre to post intervention at Casy (n = 18)

Outcome Measure (DV)

Pre^test

M (SD)

Post^test

M (SD) n

Negative affect 1.9 (1.0) 1.7 (0.7) 18

Positive affect 2.1 (1.0) 2.2 (1.0) 16

Nonviolent response 2.2 (0.6) 2.4 (0.5) 17

Help seeking 3.1 (0.5) 3.3 (0.6) 18

Satisfaction

3.3 (0.7) 18

a Included in post^test only

Table A2

Primary outcome scores from pre to post intervention at Sharon (n = 10)

Outcome Measure (DV)

Pre^test

M (SD)

Post^test

M (SD) n

Negative affect * 2.2 (1.0) 1.4 (0.7) 10

Positive affect 2.4 (1.1) 2.4 (0.9) 10

Nonviolent response 2.6 (0.5) 2.6 (0.5) 10

Help seeking 3.2 (0.5) 3.2 (0.5) 9

Satisfaction

3.2 (0.8) 10

* p < .001

a Included in post^test only

Table A3

Primary outcome scores from pre to post intervention at Milton (n = 12)

Outcome Measure (DV)

Pre^test

M (SD)

Post^test

M (SD) n

Negative affect * 2.0 (1.1) 1.3 (0.4) 12

Positive affect 2.3 (1.2) 1.6 (0.8) 12

Nonviolent response 2.4 (0.4) 2.6 (0.4) 12

Help seeking 2.9 (0.6) 3.1 (0.5) 10

Satisfaction

3.2 (0.6) 11

* p < .001

a Included in post^test only

178

References

Anderson, E. (1999). Code of the street: Decency, violence, and the moral life of the inner city.

New York: Norton & Company, Inc.

Bandyopadhyay, S., Cornell, D. G., & Konold, T. R. (2009). Validity of three school climate

scales to assess bullying, aggressive attitudes, and help seeking. School Psychology

Review, 38(3), 338^355.

Bauer, N. S., Lozano, P., & Rivara, F. P. (2007). The effectiveness of the Olweus Bullying

Prevention Program in public middle schools: A controlled trial. The Journal of

Adolescent Health, 40(3), 266–74.

Brand, S., Felner, R., Shim, M., Seitsinger, A., & Dumas, T. (2003). Middle school

improvement and reform: Development and validation of a school^level

assessment of climate, cultural pluralism, and school safety. Journal of Educational

Psychology, 95(3), 570^88.

Cassinerio, C., & Lane^Garon, P. (2006). Changing school climate one mediator at a time:

Year one analysis of a school based mediation program. Conflict Resolution

Quarterly, 23(4), 447–460.

Centers for Disease Control and Prevention. (2012a). Youth Risk Behavior Surveillance —

United States, 2011. Atlanta, GA: U.S. Department of Health and Human Services.

Centers for Disease Control and Prevention. (2012b). Middle School Youth Risk Behavior

Survey: 2011. Atlanta, GA: U.S. Department of Health and Human Services.

Chambless, D. L., & Hollon, S. D. (1998). Defining empirically supported therapies. Journal of

Consulting and Clinical Psychology, 66(1), 7^18.

179

Cohen, J., Mccabe, E. M., Michelli, N. M., & Pickeral, T. (2009). School climate: Research,

policy, practice, and teacher education. Teachers College Record, 111(1), 180–213.

Cornell, D.G. (2012). School Climate Bullying Survey. Charlottesville VA: University of

Virginia, Virginia Youth Violence Project. Retrieved from:

http://curry.virginia.edu/resource ^library/school^climate^bullying^survey

Cornell, D.G., & Mayer, M. (2010). Why do school order and safety matter? Educational

Researcher, 39(1), 7^15.

Coyle, H. E. (2012). School culture benchmarks school culture benchmarks: Bridges and

barriers to successful bullying prevention program implementation. Journal of

School Violence, 7(2), 105–122.

Crawford, D. K., & Bodine, R. J. (1996). Conflict resolution education: A guide to implementing

programs in schools, youthNserving organizations, and community and juvenile justice

settings: Program report. Washington, DC: Office of Juvenile Justice and Delinquency

Prevention, U.S. Dept. of Justice.

Dahlberg L. L., Toal, S. B., Swahn, M., & Behrens, C. B. (2005). Measuring violenceNrelated

attitudes, behaviors, and influences among youths: A compendium of assessment tools,

2nd ed., Atlanta, GA: Centers for Disease Control and Prevention, National Center for

Injury Prevention and Control.

Daunic, A. P., Smith, S. W., Robinson, T. R., Miller, M. D., & Landry, K. L. (2000). School^wide

conflict resolution and peer mediation programs: Experiences in three middle

schools. Intervention in School and Clinic, 36(2), 94^100.

deLara, E. W. (2012). Why adolescents don't disclose incidents of bullying and

harassment. Journal of School Violence, 11(4), 288^305.

180

DeVoe, J. F., & Bauer, L. (2011). Student victimization in U.S. schools: Results from the 2009

School Crime Supplement to the National Crime Victimization Survey (NCES 2012N

314). Washington, DC: U.S. Department of Education, National Center for Education

Statistics, U.S. Government Printing Office.

Dinkes, R., Cataldi, E. F., & Lin^Kelly, W. (2008). Indicators of school crime and safety: 2007.

Washington, DC: National Center for Education Statistics, Institute of Education

Sciences, U.S. Department of Education, and Bureau of Justice Statistics, Office of

Justice Programs, U.S. Department of Justice.

Eccles, J., & Roeser, R. (2009). Schools, academic motivation, and stage^environment fit. In

R. Lerner & L. Steinberg (2009), Handbook of adolescent psychology (3rd Ed., Vol. 1,

pp. 404^434). Hoboken, N.J.: John Wiley & Sons.

Eliot, M., Cornell, D., Gregory, A., & Fan, X. (2010). Supportive school climate and student

willingness to seek help for bullying and threats of violence. Journal of School

Psychology, 48(6), 533–53.

Erickson, F. (1986). Qualitative methods in research on teaching. East Lansing, Mich:

Institute for Research on Teaching.

Erickson, F. (2006). Studying side by side: Collaborative action ethnography in educational

research. In G. Spindler & L. A. Hammond (Eds.), Innovations in educational

ethnography (pp. 235^258). London: Lawrence Erlbaum Associates.

Espelage, D. L. (2002). Bullying in early adolescence: The role of the peer group. Champaign,

IL: ERIC Clearinghouse on Elementary and Early Childhood Education (ERIC

Document Reproduction Service No. ED471912).

181

Espelage, D. L., & Swearer, S. M. (2003). Research on school bullying and victimization:

What have we learned and where do we go from here? School Psychology Review,

32(3), 365–383.

Fabelo, A. (2011). Breaking schools’ rules: a statewide study of how school discipline relates

to students' success and juvenile justice involvement. College Station, TX: Justice

Center, Council of State Governments, Public Policy Research Institute.

Farrell, A. D., Erwin, E. H., Bettencourt, A., Mays, S., Vulin^Reynolds, M., Sullivan, T., … Meyer,

A. (2008). Individual factors influencing effective nonviolent behavior and fighting

in peer situations: A qualitative study with urban African American adolescents.

Journal of Clinical Child and Adolescent Psychology, 37(2), 397–411.

Farrell, A. D., Mays, S., Bettencourt, A., Erwin, E. H., Vulin^Reynolds, M., & Allison, K. W.

(2010). Environmental influences on fighting versus nonviolent behavior in peer

situations: A qualitative study with urban African American adolescents. American

Journal of Community Psychology, 46(1^2), 19^35.

Farrington, D. P., & Ttofi, M. M. (2009). SchoolNbased programs to reduce bullying and

victimization. Washington, DC: United States Department of Justice.

Flannery, D. J., Vazsonyi, A. T., Liau, A. K., Guo, S., Powell, K. E., Atha, H., … Embry, D. (2003).

Initial behavior outcomes for the PeaceBuilders universal school^based

violence prevention program. Developmental Psychology, 39(2), 292–308.

Flaspohler, P. D., Elfstrom, J. L., Vanderzee, K. L., Sink, H. E., & Birchmeier, Z. (2009).

Stand by me: The effects of peer and teacher support in mitigating the impact of

bullying on quality of life. Psychology in the Schools, 46(7), 636–649.

182

Garrard, W. M., & Lipsey, M. W. (2007). Conflict resolution education and antisocial

behavior in U.S. schools: A meta^analysis. Conflict Resolution Quarterly, 25,

Goldstein, S. E., Young, A., & Boyd, C. (2007). Relational aggression at school: Associations

with school safety and social climate. Journal of Youth and Adolescence, 37(6), 641–

654.

Greene, M. B. (2003). Counseling and climate change as treatment modalities for bullying in

school. International Journal for the Advancement of Counselling, 25(4), 293–302.

Harris, R. D. (2005). Unlocking the learning potential in peer mediation: An evaluation of

peer mediator modeling and disputant learning. Conflict Resolution Quarterly, 23, 2.

Henry, D. B., Farrell, A. D., Schoeny, M. E., Tolan, P. H., & Dymnicki, A. B. (2011). Influence of

school^level variables on aggression and associated attitudes of middle school

students. Journal of School Psychology, 49(5), 481–503.

Hertz, M. F., Donato, I., & Wright, J. (2013). Bullying and suicide: A public health

approach. The Journal of Adolescent Health, 53(1), 1^3.

Heydenberk, R. A., Heydenberk, W. R., & Tzenova, V. (2006). Conflict resolution and bully

prevention: Skills for school success. Conflict Resolution Quarterly, 24(1), 55–69.

Hunter, S. C., Boyle, J. M. E., & Warden, D. (2004). Help seeking amongst child and

adolescent victims of peer^aggression and bullying: The influence of

school^stage, gender, victimization, appraisal, and emotion. British Journal of

Educational Psychology, 74(3), 375^390.

Johnson, D. W., & Johnson, R. T. (1996). Conflict resolution and peer mediation programs in

elementary and secondary schools: A review of the research. Review of Educational

Research, 66(4), 459^506.

183

Johnson, D. W., & Johnson, R. T. (2007). Preventing bullying: Developing and maintaining

positive relationships among schoolmates. In National Coalition Against Bullying

Conference (pp. 1–59). Minneapolis, MN.

Johnson, S. L., Cheng, T., Finigan, N., Bradshaw, C., & Haynie, D. (2012). Urban African

American parents' messages about violence. Journal of Adolescent Research, 28(5),

511^534.

Jones, T. S. (2004). Conflict resolution education: The field, the findings, and the future.

Conflict Resolution Quarterly, 22(1^2), 233–267.

Juvonen, J., & Graham, S. (2014). Bullying in schools: The power of bullies and the plight of

victims. Annual Review of Psychology, 65, 1.

Juvonen, J., Yueyan, W., & Espinoza, G. (2011). Bullying experiences and compromised

academic performance across middle school grades. Journal of Early

Adolescence, 31(1), 152^173.

Kasen, S., Johnson, J., & Cohen, P. (1990). The impact of school emotional climate on student

psychopathology. Journal of Abnormal Child Psychology, 18(2), 165^77.

Kiyama. M.J. (2010). College aspirations and limitations: The role of educational ideologies

and funds of knowledge in Mexican American families. American Educational

Research Journal, 47(2), 330^356.

Klein, J., & Cornell, D. (2010). Is the link between large high schools and student

victimization an illusion? Journal of Educational Psychology, 102(4), 933^946.

Klein, J., Cornell, D., & Konold, T. (2012). Relationships between bullying, school climate,

and student risk behaviors. School Psychology Quarterly, 27, 3, 154^69.

184

Kuperminc, G. P., Leadbeater, B. J., Emmons, C., & Blatt, S. J. (1997). Perceived school

climate and difficulties in the social adjustment of middle school students. Applied

Developmental Science, 1(2), 76–88.

Lane^Garon, P. (1998). Developmental considerations: Encouraging perspective Taking in

student mediators. Mediation Quarterly, 16, 2, 201^218.

Lane^Garon, P. S., & Richardson, T. (2003). Mediator mentors: Improving school climate,

nurturing student disposition. Conflict Resolution Quarterly, 21(1), 47–67.

LaRusso, M., & Selman, R. (2011). Early adolescent health risk behaviors, conflict resolution

strategies, and school climate. Journal of Applied Developmental Psychology, 32(6), 354–

362.

Laurent, J., & Catanzaro, S. J. (1999). A measure of positive and negative affect for children:

Scale development and preliminary validation. Psychological Assessment, 11(3),

326^ 338.

Lee, V. E., & Smith, J. B. (1997). High school size: Which works best, and for whom?

Educational Evaluation and Policy Analysis, 19(3), 205^227.

Lindsay, P. (1998). Conflict resolution and peer mediation in public schools: What

works? Mediation Quarterly, 16, (1), 85^100.

Long, J., Fabricius, W., Musheno, M., & Palumbo, D. (1998). Exploring the cognitive and

affective capacities of child mediators in a "successful" inner^city peer mediation

program. Mediation Quarterly, 15(4), 289^302.

Matloff, G., & Smith, S. W. (1999). Responding to a school^wide conflict resolution^peer

mediation program: Case study of a middle school faculty. Mediation

Quarterly, 17(2), 125^141.

185

Mayer, M. J., & Furlong, M. J. (2010). How safe are our schools? Educational

Researcher, 39(1), 16–26.

Merrell, K. W., Gueldner, B. A., Ross, S. W., & Isava, D. M. (2008). How effective are school

bullying intervention programs? A meta^analysis of intervention research.

School Psychology Quarterly, 23(1), 26–42.

Merriam, S. B. (2002). Qualitative research in practice: Examples for discussion and analysis.

San Francisco: Jossey^Bass.

Meyer^Adams, N., & Conner, B. T. (2008). School violence: Bullying behaviors and the

psychosocial school environment in middle schools. Children & Schools, 30(4), 211–

221.

Nansel, T. R., Haynie, D. L., & Simons^Morton, B. G. (2003). The association of bullying

and victimization with middle school adjustment. Journal of Applied School

Psychology, 19(2), 45–61.

Nansel, T. R., Overpeck, M. D., Haynie, D. L., Ruan, W. J., & Scheidt, P. C. (2003). Relationships

between bullying and violence among US youth. Archives of Pediatrics & Adolescent

Medicine, 157(4), 348–53.

Nansel, T. R., Overpeck, M., Pilla, R. S., Ruan, W. J., Simons^Morton, B., & Scheidt, P.

(2001). Bullying behaviors among us youth: Prevalence and association with

psychosocial adjustment. JAMA, 285(16), 2094–2100.

Newman, R. S. (2008). Adaptive and non^adaptive help seeking with peer harassment:

An integrative perspective of coping and self^regulation. Educational

Psychologist, 43(1), 1^15.

186

Nix, C., & Hale, C. (2007). Conflict within the structure of peer mediation: An examination of

controlled confrontations in an at^risk school. Conflict Resolution Quarterly, 24(3),

327^348.

Novick, R. M., & Isaacs, J. (2010). Telling is compelling: The impact of student reports of

bullying on teacher intervention. Educational Psychology, 30(3), 283^296.

Oliver, C. & Candappa, M. (2007) Bullying and the politics of ‘telling’. Oxford Review of

Education, 33(1), 71^86.

Osher, D., Bear, G., Sprague, J., & Doyle, W. (2010). How can we improve school

discipline? Educational Researcher, 39(1), 48^58.

Olweus, D. (1996). Bullying at school: Knowledge base and an effective intervention

program. Annals New York Academy of Sciences, 794, 265–276.

Olweus, D., & Limber, S. P. (2010). Bullying in school: Evaluation and dissemination of the

Olweus Bullying Prevention Program. The American Journal of Orthopsychiatry,

80(1), 124–34.

Ringeisen, H., Henderson, K., & Hoagwood, K. (2003). Context matters: Schools and the

“research to practice gap” in children’s mental health. School Psychology Review, 32,

153–168.

Ripski, M. B., & Gregory, A. (2009). Unfair, unsafe, and unwelcome: Do high school students’

perceptions of unfairness, hostility, and victimization in school predict engagement

and achievement? Journal of School Violence, 8(4), 355–375.

Robers, S., Zhang, J., & Truman, J. (2012). Indicators of School Crime and Safety: 2011.

Washington, DC: National Center for Education Statistics, U.S. Department of

187

Education, and Bureau of Justice Statistics, Office of Justice Programs, U.S.

Department of Justice.

Robers, S., Kemp, J., and Truman, J. (2013). Indicators of School Crime and Safety: 2012

(NCES 2013N036/ NCJ 241446). Washington, DC: National Center for Education

Statistics, U.S. Department of Education, and Bureau of Justice Statistics, Office of

Justice Programs, U.S. Department of Justice.

Roberts, L., Yeomans, P., & Ferro^Almeida, S. (2007). Project WIN evaluation

shows decreased violence and improved conflict resolution skills for middle

school students. Rmle Online: Research in Middle Level Education, 30(8),

1^14.

Schensul, S. L., Schensul, J. J., & LeCompte, M. D. (2013a). Initiating ethnographic research: A

mixed methods approach. Ethnographer's toolkit (Book 1). Lanham, Md: AltaMira

Press.

Schensul, S. L., Schensul, J. J., & LeCompte, M. D. (2013b). Essential ethnographic methods: A

mixed methods approach. Ethnographer's toolkit (Book 3). Lanham, Md: AltaMira

Press.

Schensul, S. L., Schensul, J. J., & LeCompte, M. D. (2013c). Analysis and interpretation of

ethnographic data: A mixed methods approach. Ethnographer's toolkit (Book 5)

Lanham, Md: AltaMira Press.

Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasiNexperimental

designs for generalized causal inference. Boston, MA: Houghton Mifflin.

Shuval, K., Pillsbury, C. A., Cavanaugh, B., McGruder, L., McKinney, C. M., Massey, Z., & Groce,

N. E. (2010). Evaluating the impact of conflict resolution on urban children’s

188

violence^related attitudes and behaviors in New Haven, Connecticut, through a

community^academic partnership. Health Education Research, 25(5), 757–68.

Sipe, L. R., & Ghiso, M. P. (2004). Developing conceptual categories in classroom descriptive

research: Some problems and possibilities. Anthropology & Education

Quarterly, 35(4), 472^485.

Smith, S. W., Daunic, A. P., Miller, M. D., & Robinson, T. R. (2002). Conflict resolution and

peer mediation in middle schools: extending the process and outcome knowledge

base. The Journal of Social Psychology, 142(5), 567^86.

Smith, J. D., Schneider, B. H., Smith, P. K., & Ananiadou, K. (2004). The Effectiveness of

whole^school anti^bullying programs: A synthesis of evaluation research. School

Psychology Review, 33(4), 547–560.

Sullivan, T. N., Helms, S. W., Kliewer, W., & Goodman, K. L. (2010). Associations between

sadness and anger regulation coping, emotional expression, and physical and

relational aggression among urban adolescents. Social Development, 19(1), 30^51.

Swearer, S. M., Espelage, D. L., Vaillancourt, T., & Hymel, S. (2010). What can be done about

school bullying? Linking research to educational practice. Educational Researcher,

39(1), 38–47.

Swearer, S. M., Wang, C., Maag, J. W., Siebecker, A. B., & Frerichs, L. J. (2012). Understanding

the bullying dynamic among students in special and general education. Journal of

School Psychology, 50(4), 503–20.

Talbott, E., Celinska, D., Simpson, J., & Coe, M. G. (2002). "Somebody else making somebody

else fight": Aggression and the social context among urban adolescent

girls. Exceptionality, 10(3), 203^220.

189

Thapa, A., Cohen, J., Guffey, S., & Higgins^D'Alessandro, A. (2013). A review of school climate

research. Review of Educational Research, 83(3), 357^385.

Theberge, S. & Karan, O.C. (2004). Six factors inhibiting the use of peer mediation in a

junior high school. Professional School Counseling, 7(4), 283–290.

Thompson, E. (2007). Development and validation of an internationally reliable

short^form of the Positive and Negative Affect Schedule (PANAS). Journal of

CrossNCultural Psychology, 38(2), 227^242.

Unnever, J. D., & Cornell, D. G. (2004). Middle school victims of bullying: Who reports being

bullied? Aggressive Behavior, 30, 373^388.

Wang, J., Iannotti, R. J., & Nansel, T. R. (2009). School bullying among adolescents in the

United States: Physical, verbal, relational, and cyber. The Journal of Adolescent

Health, 45(4), 368^75.

Way, N., Reddy, R., & Rhodes, J. (2007). Students’ perceptions of school climate during the

middle school years: Associations with trajectories of psychological and

behavioral adjustment. American Journal of Community Psychology, 40(3^4), 194–

213.

Webster, D. W. (1993). The unconvincing case for school^based conflict resolution

programs for adolescents. Health Affairs, 12(4), 126–141.

Williams, F., & Cornell, D. (2006). Student willingness to seek help for threats of violence in

middle school. Journal of School Violence, 5(4), 35^49.

Wilson, C. J., & Deane, F. P. (2001). Adolescent opinions about reducing help^seeking

barriers and increasing appropriate help engagement. Journal of Educational and

Psychological Consultation, 12(4), 345^364.

190

Wright, R., John, L., Livingstone, A., Shepherd, N., & Duku, E. (2007). Effects of school^based

interventions on secondary school students with high and low risks for antisocial

behavior. Canadian Journal of School Psychology, 22(1), 32–49.

Yablon, Y. B. (2010). Student^teacher relationships and student's willingness to seek help

for school violence. Journal of Social & Personal Relationships, 27, 8.

Yin, R. K. (1992). The case study method as a tool for doing evaluation. Current

Sociology, 40, 121^137.

Yoon, J. S., Barton, E., & Taiariol, J. (2004). Relational aggression in middle school:

Educational implications of developmental research. Journal of Early Adolescence,

24(3), 303–318.

Students also viewed