|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
The Relationship Between School Connecte dness and Bullying Vicitmization in Secondary Students by Janet Urbanski A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Education Department of Educational Leadership College of Education University of South Florida Co-Major Professor: Arthur Shapiro, Ph.D. Co-Major Professor: Steve Permuth, Ed.D. William Benjamin, Ph.D. Sue Street, Ph.D. Steve Lang, Ph.D. Date of Approval: October 22, 2007 Keywords: school violence, resiliency, school climate, harassment, social emotional learning, risk and protective factors Copyright 2007 Janet Urbanski
Dedication To my husband, Robert, for believing in me and helping me find the time to finish this journey. To my children, Kimberly, Danielle and Michael who are always a source of pride and inspiration. Words of thanks seem such a small toke n for the treasures you have provided me.
Acknowledgement A journey of this magnitude is not ta ken alone. Along the way there have been many people who offered assistance and inspirati on. It is with treme ndous gratitude that I offer thanks first to my family. Hours spen t writing a dissertati on are hours spent away from your family. Thank you for your patience and understanding. I also want to thank the many colleagues who offered encouragement. Each time someone asked how the dissertation was progressing, I moved a bit clos er to completion. I offer a special thank you to Jim Dean for asking every time I saw him and for talking statistics with me. Another special thank you is given to Jo an Reubens who is as passionate about preventing bullying in schools as I am. Your ent husiasm helped fuel the fire to finish; I could do no less for the children who are bullied. I am also grateful to Dr. Philip Fitch Vincent who took an interest in helping a docto ral student just because it is important to give back. You truly are a man of character. And finally, I offer si ncere appreciation to my committee members, Dr. Shapiro, Dr. Pe rmuth, Dr. Benjamin, Dr. Street, and Dr. Lang who challenged me, believed in me, and would never let me settle for less.
i Table of Contents List of Tables List of Figures Abstract Chapter One: Introduction Problem Statement Purpose Rationale of the Study Research Questions Definitions Bullying Victimization School Climate School Connectedness Social Emotional Learning Risk and Protective Factors Academic Achievement Delimitations and Limitations Significance of the Study Summary vi vii viii 1 4 7 7 8 9 9 10 11 11 12 13 13 14 15 17
ii Chapter Two: Literature Review Social Emotional Learning Theory Risk and Protective Factors Bullying Behavior Impact and Consequences The Student Who Bullies The Victim of Bullying The Bystander Prevalence of Bullying Behavior Bullying and the Special Needs Child Positive Interpersonal Relationships Positive Adult Student Relationships Peer Relationships The Role of the School Environment School Culture School Climate School Climate and Bullying School Connectedness Summary Chapter Three: Methodology Problem Purpose Research Questions 19 20 23 30 36 37 39 42 44 46 50 52 56 58 58 60 65 67 77 79 79 80 81
iii Data Source Sampling Procedures Definitions School Bullying Academic Achievement School Connectedness Key Variables Data Analysis Research Question 1 Research Questions 2, 3, and 4 Limitations Summary Chapter Four: Results Overview of Results Summary of Results Preliminary Analysis Data Management Factor Analysis Descriptive Statistics Sample Dependent Variables 82 86 87 87 88 88 89 89 100 103 103 104 105 106 106 108 109 109 110 112 112 116
iv Independent Variables Moderating Factors Multivariate Analysis Assumptions Research Question 1 Research Question 2 Research Question 3 Research Question 4 Follow Up Analysis Summary Chapter Five: Discussion Overview of the Study Review of the Method Discussion of Findings School Connectedness and Bullying Victimization Relationships Additional Analysis Implications Study Limitations Recommendations for Future Research Summary References 118 121 123 123 123 128 129 130 131 134 136 136 138 139 139 145 146 147 149 151 153 155
v Appendices Appendix A: School Crime Supplement to the National Crime Victimization Survey Appendix B: Pearson Correlation Coefficients for Variables with Bullied About the Author 181 182 191 End Page
vi List of Tables Table 1. Definitions of Bullying Victimization Table 2. School Level Indicators of Risk and Protective Factors Table 3. Social Emotional Learning Competencies Table 4. Risk and Protective Factors Table 5. Types and Definitions of Cyberbullying Table 6. Components of School C onnectedness Construct Variables Table 7. Study Variables a nd Operational Definitions Table 8. Analysis to Answer Research Questions Table 9. Factor Analysis of Variab les Related to School Connectedness Table 10. Demographics of Study Sample Table 11. Frequency of Dependent Variables Table 12. Frequency of School Connectedness Components Table 13. Frequency of Moderating Factors Table 14. t-values of Predictor Variables Table 15. t-values of Predicto r and Demographic Variables Table 16. t-values of Moderating Variables Table 17. Revised Factor Analysis including Discipline Variables 10 14 21 25 32 93 96 101 113 114 117 119 122 126 127 130 132
vii List of Figures Figure 1. Distribution of Grade Level Figure 2. Distribution of Self-Reported Grades Figure 3. Distribution of Race Figure 4. Distribution of Freque ncy of Bullying Victimization Figure 5. Distribution of Emotional Safety Scores Figure 6. Distribution of Relationship Scores 115 116 116 117 118 119
viii The Relationship between School Connecte dness and Bullying Victimization in Secondary School Students Janet Urbanski ABSTRACT Bullying is a complex behavior that can cause academic and social problems for students and can contribute to a negative school climate. Students who feel isolated or do not feel connected to their school may e xperience similar risks to those who are victimized by peers. Recent school violence inci dents have led to an increase in bullying behavior research. The importance of the school climate is also emerging in educational discourse prompting a growth of research in school connectedness and positive relationships. However, research on the impact that relationships and school connectedness may have on bullying vi ctimization at school is limited. This is a secondary analysis of a national data set from the 2005 administration of the National Crime Victimization Survey School Crime Supplement. The study focused on the relationship between school connectedne ss and bullying victimization and whether gender, race, grade level, and academic achie vement moderate the relationship. The role of relationships in bullying victimization was considered. Weighted regression analys es were conducted to determine the relationship between bullying behaviors and school connect edness and to identify the combination of factors that may influen ce the relationship. Component s of school connectedness identified through factor analysis were stat istically significant predictors of occurrence
ix and frequency of bullying victimization, but accounted for a very small amount of variance in the outcome. Adding demographic va riables of race, gende r, grade level, and academic achievement produced a slight incr ease in the proportion of variance accounted for. Race did not have a statistically significant impact on occurrence of bullying victimization; neither race nor gender was statistically significant in variance of frequency of bullying victimization. Peer relationships proved to be statistically significant in bullying victimization freque ncy but neither adult-student nor peer relationships were statistically related to bullying victimization occurrence. Overall, school connectedness predicted a very sma ll proportion of variance in occurrence and frequency of bullying victimization, sugges ting that bullying prevention efforts should include strategies beyond thos e to improve a students sens e of connectedness to school. A comprehensive approach is needed to addr ess bullying in schools effectively.
1 Chapter One Introduction Imagine that you are in school and very much alone. Every day you dread tomorrow and you have to drag yourself to class where children tease and taunt and point at you. The more you s quirm, the sweeter the chase. There is no escape and hardly anyone ever comes to your assistance. Sometimes a child does not have to have a partic ular physical trait for other children to hone in on, sometimes they just pick someone for the heck of it and hate becomes infectious with group leaders and their followers within a school (Dellasega, 2001, p. 85). Many school environments are confronted with bullying, sexual harassment, and mean spirited teasing that have become a normative process poisoning the climates of schools (Sprague & Walker, 2005) depriving child ren of the right to be educated in an environment that is both physically and emotiona lly safe. If students are afraid to attend school or spend time at school worrying about sa fety rather than academics, they can not learn. A climate of safety, respect, and em otional support in schools not only helps to diminish the possibility of targeted violence in schools, it impacts academic achievement as well (Fein, Vossekuil, Pollack, Borum, Modzeleski, & Reddy, 2002). Therefore, it can no longer be viewed as an either or proposit ion; academics and safety are both essential components of educational discourse. Regr ettably, since academic standards,
2 accountability, and high stakes testing have become co mmon dialogue in American schools, educators can be so keyed into curric ulum that they forget the importance of the classroom and school climate even though real learning can only take place when the climate is positive and the children feel secu re, respected, confident, and safe (Abourjilie, 2000). Fear of harm or embarrassment creates a threat which shuts down the learning process (Mendler, 2001). In contrast, achieveme nt is increased when the culture of a school supports learning for both students a nd adults (Walker & Lambert in Lambert, et al., 1995). Additionally, educators who have a positive relationship with students have better discipline and more time for instruction (Mendler) resulting in higher achievement. Knowing this, the traditional focus of sc hool discipline is cha nging from a focus on student behavior to a concentration on em otional and physical safety of the school community (Calabrese, 2000). In order to ensure that no child is left behind, we have to first ensure that each child is safe at school. It is critical that schools are places where student feel safe, respected and able to share their concerns openly without fear of shame or pun ishment. It is essential that students connect positively with at least one caring adult, and also that they get the emotional support they need to break the pervasive and dangerous code of silence that sways todays youth (Paige, 2002). Connection through human relationships is a vital component of a safe school environment that works to bond students with ea ch other and with the adults charged with meeting their educational needs (Fein, et al., 2002). In addition, childrens peer
3 relationships are an essential contributo r to their social-emotional and cognitive development; maladaptive peer relations in childhood strongly predic t negative outcomes including school drop-out, delinquency, a nd mental health problems (Shea, 2003). Educators can develop this connection with st udents and enhance peer relationships by creating personal connections characterized by trust, academic conne ctions consisting of strategies to encourage succe ss in content areas, and soci al connections among students (Mendler, 2001). Bullying in schools works in opposition to effo rts to create this type of safe and healthy learning environment. Research s hows a relationship betw een student bullying and school issues such as academic ach ievement, school bonding, and absenteeism (Telljohann, 2003). Anti-social behavior s including bullying, harassment, and victimization compete directly with the instructional mission of schools resulting in decreased achievement (Sprague & Walker 2005). The bullying cycle becomes an obstacle to learning, self-development, and effective citizenship (Morrison, 2002) and can contribute to a climate of fear and intimidation in schools. This can be compounded when the bullying is encouraged and s upported by the presence and attention of bystanders. When teachers and students participate in bullying and harassment or witness the actions and do nothing, they are sending th e message that it is acceptable (Olweus, Limber, Mullin-Rindler, Riese, & Snyder, 2004). This perpetuates an unhealthy and unsafe environment for all students. When a chil d feels unsafe, that ch ild is vulnerable to anxiety and a diminished capacity to discover, to remember, and to find joy in the process of learning (Cohen, 1999).
4 Problem Statement Students must feel safe in order to learn (Abourjilie, 2000; Fein, et al., 2002). Unfortunately, reality for many students is a sc hool day filled with fear, intimidation and the ridicule of bullying. Comments from seve ral middle school students draw attention to what is real life for many students (J. Reubens, persona l communication, February 8, 2007). There are some kids that always make fun of me. Calling me names and saying that I make everyones life miserable. They talk behind my back and exclude me from everything. I just dont know what to do. He smashed a package of crackers in my face and then crushed them on me and dumped the crumbs over my head. They tell me that no one likes me and they tell everyone to ha te me. If they see Im happy they make me sad on purpose. I need help! I have been bullied in school. No t any violence, just a shove now and then. And name calling, mostly because of my last name. Some people think it is fine and th at no harm is done, but it does, deep down, hurt my feelings. A final example highlights the anxiety a nd long lasting consequences suffered by students who experience bullying victimization. Ive gotten bullied a lot 4th grade through 8th grade. Of course it doesnt happen any more but all the damage is done. I just cant get rid of all the stress that has built up inside of me. I just need some help before its too late and before I lose it all Researchers in the area of school safety ha ve consistently show n that the lack of physical and/or emotional safety is likely to result in negative educational outcomes
5 including violence, truancy, and poor academ ic performance (McEvoy & Welker, 2001; Zins, Bloodworth, Weissberg, & Walberg, 2004) whereas a sense of belonging to a school community has a positive impact on academic performance as well as psychological adjustment (Kent, 2003). More sp ecifically, research in the area of bullying prevention highlights the negative outcomes of bullying and victimi zation including an increased risk of mental health disorder s, antisocial behavior, and poor academic achievement (Hawker & Boulton, 2000; Mishna, 2003; Olweus, 1993; Telljohann, 2003). Despite this knowledge, there is an increas ed emphasis on academic achievement that ignores the importance of school climate and th e negative impact of bullying behaviors. Much of the formal research on bullying has occurred in Scandinavia, Great Britain, and Japan. However, within the la st decade, there has been an increasing awareness in the United States that bullying is a serious form of peer violence that plagues the school system. Following the schoo l shootings such as those at Columbine and Santana high schools and the Se cret Service findings that tw o-thirds of all the school shooters since 1974 had been vi ctims of bullying prior to the shootings (Brady, 2001), bullying has jumped to the forefront as an issue that schools must deal with and work to prevent. Bullying can no l onger be viewed as a right of passage because teasing, name-calling, and harassment that is not stopped often escalates to threats and physical violence. Furthermore, frustrated pa rents are suing school districts for failing to protect their children from hara ssment and abuse (Shariff, 2005). Concurrently with the issue of bullyi ng in schools, academic standards and high stakes testing have become a primary focus in American schools. A driving force behind
6 this accountability movement is the Federal No Child Left Behind Act of 2001 that requires all states to measure each public sc hool's and district's achievement and establish annual achievement targets for the state with the overarching goal of all students meeting or exceeding standards in reading and mathematics by 2014 (U.S. Department of Education, Office of the Under Secretary, 2003). However, education driven by high stakes testing ignores the intr icacy of human beings and dism isses the complexity of the teaching and learning process. The test scores at Columbine High School were among the highest in Colorado, highlighting the fact th at a narrow focus on improving test scores makes it more difficult for teachers to get to know their students well (Kohn, 2004) and offers little consideration for the heteroge neous population of students that must be educated in our public schools. Social emotional learning theory suggests that learning is relationship centered, compelling teachers to know their students, show concern for their academic progress and create a caring clas sroom environment (Ragozzino, Resnik, Utene-Obrien, & Weissberg, 2003). A safe school environment is one where students are able to know and to trust and to be known and trusted by adults (Kohn). With the recent emphasis on accountability and high stakes testing, society seems to be indifferent to the basic foundations of childrens well being and the role the school plays in educating the whole child. Even though learning can only take place when children feel physically and emotionally safe, these areas have traditionally been treated as separate conditions. It is essential that educators fi nd ways to support students emotional well being as well as finding ways to support their academic achievement.
7 Purpose The study focused on the relationship betw een bullying victimization and school connectedness. Considering previous bullying research as well as risk and protective factor research, the study examined whethe r the presence of school connectedness serves as a protective factor diminishing bullying vi ctimization. Gender, ra ce, grade level, and academic achievement were considered as moderating factors. The role that adult to student and peer to peer relationships play in the bullying phenomenon was also investigated. The study looked at relationshi p differences by investigating whether a students level of school connectedness pr edicted bullying victimization. Finally, it identified risk and protective factors that may a llow educators to target students at risk of victimization for proactive intervention as well as indicating prevention efforts at the school level. The earlier the bully/victim pattern can be broken, the less negative the effects will be. The results of this study may provide inform ation to guide educators as they develop and modify bullying prevention pr ograms to meet the needs of all children. Rationale of the Study The environment in a school impacts how students learn and teachers teach. A positive school climate has been shown to have an influence on student behavior, including achievement, with both student -peer and student-teacher relationships positively correlated to student academic achievement (Niebuhr, 1999). According to Social Emotional Learning theory, when school s attend systematically to the students social and emotional needs, the academic achie vement of children increases, incidents of problem behavior decrease, and quality of relationships surrounding the child improves
8 (Elias, et al., 1997). Therefore it is imperative that school di stricts pay attention to the social and emotional needs of students as well as their academic achievement. The prevention field lacks a sufficient research base to characterize the effectiveness of most types of activities in schools intended to reduce or prevent delinquency or problem be havior (Gottfredson & Go ttfredson, 2001). Although a significant amount of research exists regarding academic ac hievement and an increasing amount of research is emerging that c oncludes that school climate matters, few comprehensive studies exist that have inves tigated the reciprocal relationship between school climate and academic achievement. N one investigate bullying as a mitigating factor in the relationship. Research Questions Based on social emotional learning theory and risk and protective factor research, this study analyzed data from the 2005 Na tional Crime Victimization Survey School Crime Supplement (see Appendix A) to examin e whether a relationship exists between bullying victimization and school connectedness. It investigated whether factors such as grade level, race, gender, or academic ach ievement level have an effect on the relationship. The role positive adult student re lationships and peer relationships have in victimization in school were also examined. The research questions addressed in this study were: 1. What is the relationship between bullying victimization and school connectedness? a. Is the frequency of bullying victim ization related to the level of
9 connectedness a student has with school? b. Does the level of connectedness of students who report no experience of bullying victimization differ from the level of connectedness of students who report bullying victimization? c. Is the relationship between occurren ces of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? d. Is the relationship between freque ncy of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? 2. Is bullying victimization moderated by th e strength of adult-student relationships that a student develops? 3. Is bullying victimization moderated by th e strength of peer relationships that a student develops? 4. Does the impact of adult-student rela tionships on the frequency of bullying victimization differ from the impact of peer relationships on the frequency of bullying victimization? Definitions Bullying For the purpose of this study, bullying is defined as repeated behavior that is intended to harm or disturb a nother person. It is proactive ag gression that usually occurs without provocation or threat on the part of the victim (Olweus, 2003) and involves
10 an imbalance of power with a more powerfu l person physically or emotionally attacking a less powerful one (Nansel, Overpeck, Ra mani, Ruan, Simons-Morton, & Scheidt, 2001). Bullying encompasses a spectrum of aggr essive behaviors ranging from overt acts of physical violence to more subtle patterns of verbal or relational cruelty (Feinberg, 2003) and can be categorized as four types of victimization as defined in Table 1. Table 1 Definitions of Bullying Victimization Type Definition Physical Verbal Relational Aggression Cyberbullying Harm to anothers person or property Taunting, teasing, name calling, extortion, or threats Harm to anothers self esteem or group acceptance Using technology to harass or intimidate another person A key difference between behavior defined as antisocial and that considered to be bullying is the persistent repeated nature of the behavior that occurs between the perpetrator and the victim. Youth identified as an tisocial tend to direct their aggression in a random fashion and towards large pools of pot ential targets whereas bullying is directed towards a specific individual (Sprague & Walker, 2005). Victimization Bullying victimization is th e experience among children of being a target of the aggressive behavior of other children (Hawker & Boulton, 2000) A student who is
11 bullied, the victim, is defined as a person w ho is repeatedly exposed to negative actions from peers in the form of physical attack s, verbal assaults, or psychological abuse (Olweus, 1993). School Climate School climate is the shared perceptions of a school and consists of the attitudes, beliefs, values, and norms that underlie the in structional practices and operations of a school (Welker, 2000). It embodies the physic al and psychological environment of a school with a specific link to student academ ic achievement (Niebuhr, 1999) and can be defined as the pervasive qua lity of a school environmen t experienced by students and staff which affects their behaviors (Roach & Kratochwill, 2004). Elements of the school environment that contribute to the school cl imate include continuous academic and social growth, levels of respect, trust and integrity, morale and cohesiveness, caring, opportunities for input (Rinehart, 1993), leve l of orderliness, degree of satisfaction experienced, amount of productivity possible, sense of belonging (Flo rida Department of Education, 2002), and degree of connecte dness (Doan, Roggenbaum, & Lazear, 2003). This study focused on one aspect of sc hool climate, schoo l connectedness. School Connectedness In this study, school connectedness refers to a students relationship to school and is defined as a component of school climate that creates a feeli ng of belonging to the school and being accepted by others (Blu m, 2005; Blum & Libbey, 2004; Nakammura, 2000; Perlstein, 2004). It includes students experi ences of caring at school and a sense of
12 closeness to school personnel and environmen t (Smith, 2004). It involves the students comfort level at school and come s from the feeling that adults there care about them. In alignment with the Natio nal Longitudinal Study of Adolescent Health, school connectedness includes the degree to which st udents feel close to people at school, are happy to be at school, and feel like a part of the school (Lib bey, 2004). It is a function of attachment, interpersonal suppor t, and experiences of belongi ng and includes the acts of giving back to, being involved with, and being affectively inve sted in other people, places and activities (Gerler, 2004). Although connectedness to friends is a component of school connectedness, it can have a different impact on studen t behavior. Gerler reports that a strong connectedness to friends pa ired with a low connectedness to school increases the students risk for engaging in vi olence. Therefore, peer relationships were also considered separately in this study. Social Emotional Learning Social emotional learning is rooted in the fields of medicine and psychology and supports a school-based emphasis on emotiona l development (Halford, 1996). It is the process through which children develop the sk ills necessary to acquire social and emotional competence and addresses the stud ents ability to understand, manage and express the emotional aspects of life in ways that enable successful management of life tasks (Elias, et al., 1997).
13 Risk and Protective Factors Risk factors are any cond ition that increases a childs likelihood of engaging in unsafe behaviors. Risk factors function in a cu mulative manner with the number of risk factors positively correlated to the likelihood that a child will engage in delinquent behaviors (Introduction to Risk Factors and Protective Factors, n.d.). Protective factors are any condition that promotes healthy beha viors and decreases the chance that a child will engage in unsafe behavior. Protective f actors are generally the opposite of the risk factors that make it likely for a child to e ngage in risky behavior Risk and protective factors are typically organized into five ca tegories: individual, fa mily, school, peer group, and community. Given the scope of this study, only school level fact ors were considered. School level indicators are list ed in Table 2 (Introduction to risk factors and protective factors, n.d.). Academic Achievement Academic achievement is defined as adequate progress towards meeting state content and performance standards. For the purpose of this study, academic achievement level is defined by the students self-report of their letter grade average over the past school year. Previous research has shown th at although self reported grades tend to be slightly inflated, they highly correlate ( r = .78) with official transcripts (Dornbusch, Ritter, Leiderman, Mont, Reynaud, & Chen, 1990).
14 Table 2 School Level Indicators of Risk and Protective Factors Risk Factors Protective Factors Low academic achievement Negative attitude toward school Low bonding, low school attachment Truancy, frequent absences Suspensions Inadequate school climate Identified as learning disabled Frequent school transitions Academic achievement Student bonding and connectedness Opportunities and reward for prosocial involvement Clear standards and rules High expectations for students Presence and involvement of caring supportive adults Delimitations and Limitations Conducting secondary data analysis offers the benefits of large, complex surveys including a large sample size, increase d power, and generaliz ability due to the representation of the populati on (McCall & Applebaum, 1991). In contrast, a commonly noted concern of secondary data analysis is the verity that the data were collected for purposes other than the purpose of the seconda ry analysis. Given that the data typically has been collected for a different purpose, an existing data source is rarely a perfect match necessitating a balance between comp romise and gains (McCall & Appelbaum). Although the advantages and disa dvantages inherent to second ary analysis also apply in this study, the use of existing data is a nonintrusive means of an alysis (Yegidis &
15 Weinbach, 1991) and can be valuable in term s of the number of subjects and variables assessed (McCall & Appelbaum). With this in mind, severa l limitations to this study must be noted. Although the sample is natio nally representative, it is not a random sample. The School Crime Supplement is a sample survey so non response bias can affect the strength and application of data. As a self report survey, unit non response may have also biased results. The survey is read to respondents so students do not have the opportunity to respond anonymously. Responde nts may have reported on victimization that occurred outside the six month reference period, artificially infla ting the report. The survey lists bullying incidents as single points in time rather than a state of victimization which may have resulted in the re port being artific ially deflated. Respondents recall of bullying episodes may have been inaccurate leading to an und erestimate of victimization. Finally, the survey did not include cyberbully ing as an extension of bullying behavior which may have artificially deflated the results of the survey. To the extent that parents refused their childs participation, students chos e not to participate, failed to answer some questions, or provided false or misleading re sponses, a less reliable basis to form generalizations may have resulted. Thus, conclusions are delimited; if a different population was included, results may differ. Significance of the Study A major mental health concern facing our nation is the early identification and prevention of antisocial beha vior in youth (Miller, Breh m, & Whitehouse, 1998). Social withdrawal, excessive feelings of isolation and rejection, vict im of violence, feelings of being picked on and persecuted, low interest in school, and pattern s of impulsive and
16 chronic bullying behavior ar e included as early warning signs for violence in Early Warning, Timely Response: A Guide for Safe Schools (Dwyer, Osher, & Warger, 1998). In comparison, emphasis on positive relationships among students and staff and open discussion of safety issues are listed as characteristics of safe schools. Supportive and caring relationships promote academic motivat ion in schools but less is known about the influence of school connectedness on adolescent health risk behavior s (McNeely & Falci, 2004). Participants in a learning community are connected in their learning and their work. Understanding this community and its c ontext is at the core of building a safe school (Calabrese, 2000). In public school classr ooms there are a representative mix of values and cultures; educators are respons ible for building learning communities that consist of students who may have nothing more in common than attending the same school. Teachers and administrators must ensure that each student feels a sense of safety and well-being to participate fully and e qually in the educatio nal process (McEwan, 2000). When problems stay unresolved, the whole community suffers (Lambert, Collay, Dietz, Kent, & Richert, 1996). A safe school environment results from the collaboration among administrators, teachers, parents, and students (Calabrese 2000). If schools promote a sense of caring and fairness for all, the members of the schoo l will have a greater opportunity to meet the standards of a civil and caring commun ity (Vincent, Wangard, & Weimer, 2004). Successful interventions focus on a reduction of stressful events, a change in normative belief about approval of aggr ession, and teaching coping and social skills (Batsche, 2000). Understanding and evaluation of character istics of the larger school context allow
17 educators to become aware of the school wide risk or protec tive factors that may influence intervention outcomes and trends in student or staff behavi or and attitudes that call for systemic intervention efforts (Roach & Kratochwill, 2004). Considering the consequences of bully ing behavior and the impact the school environment has on its members, an increas ed knowledge of school connectedness as a factor associated with bullying victimization is needed. Results of this study enhance the knowledge of school connectedness by providing information that can be used in the development of successful bullying preventi on and intervention programs designed to address the needs of all th e students in the school. Summary The emotional well being and physical safety of students are an integral part of a successful learning environment. A students sense of belonging to school along with the development of positive adult to student and peer to peer relationships can impact educational outcomes. Bullying is a phenomenon that po tentially interferes with the healthy social emotional development of students thus impacting academic success. Bullying behavior can have negative conseque nces for the individual student as well as the school climate creating an unhealthy learning environment for students who experience bullying victimization. This s econdary analysis of data from the 2005 National Crime Victimization Survey Sc hool Crime Supplement investigated the relationship between school connectedness and bullying victimization and the impact of gender, race, grade, and academic achievem ent on the relationship. Chapter 1 provided the purpose and rationale of the study al ong with an introduction to the problem
18 addressed in the study. Research questions, de finitions of terms, and delimitations and limitations of the study were presented. The chap ter concluded with an explanation of the significance of the study.
19 Chapter Two Literature Review Traditionally, bullying behavior in school s has been ignored or viewed as a normal rite of passage that all children mu st go through; however there is evidence to show that bullying can have se rious consequences for all invo lved (Florida Office of Safe and Healthy Schools, 2005). Research over th e past decade shows that low level, underlying violence in schools may not be as overtly threatening as weapons but occurs with greater frequency and has a profound im pact on a students emotional health and school performance (Dupper & Adams, 2002). Bully ing behavior is an example of this type of violence that influences a schools climate, thereby impacti ng all students at the school. In addition, there is an increasing am ount of research that shows a correlation between students engaging in bullying behavior and engaging in s ubsequent violence. Much of the early research in bullying fo cused on identifying characteristics that differentiate bullies and vict ims from one another and from those not involved in the behavior (Greene, 2003). Mo re recently, research has examined individual level predictors related to bullyi ng but school level predicto rs have not been thoroughly investigated (Sanders & Phye, 2004). Research in the area of school climate and school connectedness is also emerging. Typically it is related to school violence. The role each plays in bullying and its subsequent impact on academics is minimal.
20 Although educational literature indicates a correlation between antisocial behavior and academic failure, they are commonly treated as separate conditions (Welker, 2000). Social emotional learning theory takes ex ception to this disconnect by stressing the importance of the affective aspect of edu cation as well as the academic (Weissberg & OBrien, 2004). Similarly, risk and protectiv e factor research pr ovides insight on the factors contributing to a positive school clim ate, highlighting the importance of positive interpersonal relationships with both peer s and adults (Resnick, Ireland, & Borowsky, 2004). Numerous researchers ha ve supported the premise of the protective impact of adult relationships for young people (Commission on Children at Risk, 2003) suggesting that adult relationships are in tegral to school connectedness. However, the impact adult relationships and connectedness may have on bullying victimization at school is not addressed. There is a gap in the literature as to the interrelationship of these factors. This chapter is divided into six secti ons. The first two sections provide an overview of social emotional learning theory and risk and protective factor research. The next section is a review of the literatur e on bullying behavior. The fourth section addresses the significance of positive interper sonal relationships in the school setting. The fifth section focuses on the role of th e school environment. The final section is dedicated to one specific area of sc hool climate, school connectedness. Social Emotional Learning Theory Social emotional learning is the proc ess through which children learn to recognize and manage emotions and is based on the a ssumption that optimal learning emerges from supportive and challenging rela tionships (Weissberg & OBrien, 2004). It focuses on the
21 ability to understand, manage, and express th e emotional aspects of ones life in ways that enable successful management of life tasks including learning (Elias, et al., 1997). Social emotional learning activities allow students to develop social and emotional competence, defined as a students ability to handle emotions, problem solve and maintain positive relationships e ffectively (Ragozzino, et al., 2003). The Collaborative to Advance Social and Emo tional Learning identified competencies essential to the successful social and emoti onal development of youth: awareness of self and others; positive attitudes and values; re sponsible decision making; and social interaction skills (Payton, Wardlaw, Graczyk, Bloodworth, Tompsett, & Weissberg, 2000). The competencies are fu rther defined in Table 3. Table 3 Social Emotional Learning Competencies Competency Definition Self awareness Social awareness Self management Relationship skills Decision making Recognition of ones emotions, strengths, self efficacy and self confidence Empathy, respecting others, abilit y to see others perceptions Impulse control, motivation, and goal setting Cooperation and communication Evaluation, reflection, and responsibility Principles of social emotional learning theory provide a conceptual framework to address a schools academic activities as well as prevention initiatives (Weissberg &
22 OBrien, 2004). Addressing social emotiona l learning in schools develops caring classroom environments, contributes to the development of positive relationships, and provides students with personal skills su ch as managing emotions and developing motivation, working cooperatively, and setting academic goals (Ragozzino et. al., 2003). Dating back to 1918, the National Education Asso ciation listed seven aims of education: health, command of fundamental proce sses, worthy home membership, vocation, citizenship, worthy use of leis ure time, and ethical characte r suggesting that promoting academic competence should be accompanied by development of the other aims (Noddings, 2005). More recently, research on school culture and school effectiveness indicates that providing for th e whole student contributes to the success of effective schools and school improvement (Sergiovann i, 1992). Effective schools focus on academics. However, recognizing that lear ning problems are systemic, the school does everything possible to attend to the developm ental, physical, and social needs of its students. Caring is viewed as a ke y to academic success (Sergiovanni). An expanding body of research demonstrates that social emotional learning is integral to academic learning. Social emo tional learning programs provide instruction that enhances students ability to recognize and to manage emoti ons, appreciate others perspectives, establish goals, problem solve, and develop su ccessful interpersonal skills (Payton, et al., 2000). A meta-analysis of 165 studies examining the effectiveness of prevention activities found that social emo tional learning programs increased school attendance and decreased dropout rates (Wilson, Gottfre dson, & Najaka, 2001). Confirming these results, a quantitative anal ysis of more than 300 studies on social emotional learning found that ch ildren who are given clear be havioral standards and are
23 taught social skills feel safe, valued, and challenged resulting in better academic performance, better attendance, and more constructive behavior (Zins, et al., 2004). Research in the field of ne uropsychology supports the idea th at learning is relational and social emotional skills are necessary fo r development of cognitive activities. Brain research shows that optimal learning takes place in an emotional and behavioral context and that memory is linked to social and em otional situations (E lias, et al., 1997). Highlighting the importance of the soci al emotional aspect of learning, the Association for Supervision and Curric ulum Development position paper on School Safety and Violence states that educators s hould create safe school s that are emotional and intellectual centers of their comm unity and develop positive and trusting relationships with students (Halford, 1996). When students view the environment as hostile, they are likely to dislike the cla ss, teacher, and classmates (Ecstrom, Goertz, Pollack, & Rock, 1986). An environment of fear creates an atmosphere of distrust and promotes anxiety, stress, and depression. Under conditions of real or imagined threat or high anxiety, there is a loss of focus on lear ning and a reduction in th e ability to problem solve (Elias, et al., 1997). W ithout a feeling of physical and emotional safety and security, students will find it difficult to move beyond fear and a nxiety to explore new challenges willingly (Ecstrom, et. al). Addressi ng social emotional learning contributes to a safe learning environment where children can learn. Risk and Protective Factors Risk and protective factors are factor s in a young persons life that contribute, either positively or negatively, to the extent of his or her involvem ent in health risk
24 behaviors. Whether biological or environmental, risk and protective factors transcend socioeconomic status and ethnicity and repr esent continuing inter actions between the child and the environment from birth thr ough adolescence (Robins & Rutter, 1990). Risk factors are not causal factors; rather they are conditions that increase the li kelihood of an individual engaging in risk behaviors (Florida Department of Education, 1998). In comparison, protective factors are the even ts, opportunities, and experiences that diminish or buffer against the likelihood of involvement in risky behaviors (Resnick, Ireland, & Borowsky, 2004). According to noted resilience researcher, Werner (1994), protective factors have a stronger influen ce on individuals who grow up and overcome adversity than do the risk factors that are pr esent in their lives. Protective factors are called upon when necessary but have to be in place prior to being needed. A significant difference between risk and protec tive factors is that risk f actors may lead directly to a disorder but protective factors only operat e when a risk is pr esent (Clark, 1995). However, it is important to note that prot ective factors are central to understanding how to reduce the impact of risk factors and how to encourage positive behavior and social development (Hawkins, Ca talano, & Miller, 1992). Risk and protective factors can be categor ized into several spheres of influence: individual, family, peer, school, and community (Minnesota Department of Health, 2002; Florida Department of Education; 1998; Hawk ins, et al., 1992). Individual factors that impact a childs health and well being can be determined biologically or socially and include a view of self, attitudes and beliefs, sens e of future, and ability to interact socially with others. Hawkins, et al. also list gende r, personality, and intelligence as innate characteristics that may help protect a child exposed to risk fact ors. Their research
25 indicates that outgoing children and bright children are more protected than peers who do not have these characteristics. Additionally, ma les were more likely to engage in health risk behaviors in adolescence than females. Fa mily factors incorporate the level of family conflict, stability, and supervisi on. Peer influence is either dire ct or indirect and includes the attitudes and behaviors of peers or the perception of their att itude and behaviors. School factors center around a feeling of conn ectedness to school and academic success. Finally, the community influence involves a sense of belonging to the community as well as the levels of poverty and vi olence within the community. The five domains are further described in Table 4 (Florida De partment of Education, 1998). Table 4 Risk and Protective Factors Domain Risk factor Protective factor Individual Early aggressive behavior Low behavioral inhibition Poor cognitive development Victimization, exposure to violence Self control Religiosity Problem solving skills Resilient temperament Family Lack of supervision Family violence Familial antisocial behaviors Poor family bonding Monitoring Stability Clear expectations Family bonding Peers Peer rejection Deviant peers Academic competence Positive peer group
26 Table 4 (continued). Risk and Protective Factors Domain Risk factor Protective factor Substance abuse Parental approval School Failure to bond with school Low expectations/achievement Truancy Suspensions Frequent transitions Bonding with school High expectations Opportunities for Involvement Clear standards/rules Supportive adults Community Poverty Concentration of deviant youth Access to weapons Economically stable Neighborhood cohesion Visible law enforcement Risk and protective factors can affect ch ildren at different stages of development (NIDA, n.d.). Early risks, such as aggressive behavior, can be changed or prevented with family, school, and community interventions that focus on helping children develop appropriate, positive behaviors. If not addres sed, negative behaviors can lead to more risks, such as academic failure and social difficulties (NIDA). Research on risk factors reveals severa l common risk elements that increase the likelihood that a youth will engage in risk be haviors including violence. According to a Search Institute (1991) survey of 47,000 youth in grades 6 through 12, common behaviors that poten tially limit successful development during adolescence include: involvement with alcohol, tobacco, illic it drugs; early onset sexual activity;
27 depression/suicide; anti-social behaviors; poor school attendan ce; and the desire to drop out of school. Expanding on the idea of common elements related to risk factors by investigating the role of gender, the Minnesota Adolescen t Health Survey (Mi nnesota Womens Fund, 1992) of 36,000 youth in grades 7 through 12 fou nd more girls than boys exhibit covert behaviors such as emotional stress, poor body image and self esteem, eating disorders, and attempted suicide. More boys than girl s act out behaviors by committing delinquent acts, taking physical risks, engaging in unprotected sex, and abusing substances (Minnesota Womens Fund). This research also revealed th at patterns of co-occurrence exist among risk factors. Prevention research indicates that expos ure to risk factors in the absence of protective factors dramatically increases th e likelihood that a young person will engage in problem behaviors. Therefore, reducing risk factors and increasing protective factors is an effective way to improve the lives of young people (Introduction to Risk Factor and Protective Factors, n.d.). In an analysis of data from the Nationa l Longitudinal Study of Adolescent Health designed to identify indi vidual, family, and comm unity level risks and protective factors, Resnick, et al. (2004) found substantial redu ctions in the percentage of youth involved in violence when protective fact ors were present, regardless of the level of risk factors. The researchers defined prot ective factors as factors that, if present, diminish the likelihood of negative health a nd social outcomes. Results of the study showed that school and peer rela ted factors are stronge r predictors of part icipation in risk behaviors than demographic variables such as family structure, social class, race, or ethnicity. Results also suggest ed that a sense of connecte dness to adults outside the
28 family was a significant protectiv e factor for both males and females, especially for those without a strong connection to family, T = 2.02, p < .043. The researchers concluded that a perceived connectedness with school provide d a buffer against adolescents emotional distress, suicidal thoughts and behaviors, enga ging in violent behaviors, use of tobacco, alcohol and marijuana, and age of first sex. Findings of the National Longitudina l Study of Adolescent Health support conclusions reached in earlier research. Participants in a study conducted by Hunt, Meyers, Davies, Meyers, Grogg, and Neel (2002) perceived five items as significant factors contributing to violent beha vior: disrespect for authority ( M = 4.60, SD = .72); lack of parental support ( M = 4.38, SD = .93); poor anger management skills ( M = 4.31, SD = .85); disrespect for peers ( M = 4.23, SD = .85); lack of academic interest ( M = 4.18, SD = .89). Racial and ethnic differences ( M = 3.36, SD = .1.23) and socioeconomic status ( M = 3.12, SD = .1.07) were rated as less signific ant. Miller, Brehm, and Whitehouse (1998) also found variables that distinguish youth who engage in antisocial behavior from those who do not. Their research followed boys who were identified as aggressive in elementary school until they reached age 14. Find ings showed that the presence of higher degrees of prosocial skills, school bonding, academic achievement, and avoidance of peers involved in antisocial behavior were significant deterr ents to increased involvement in delinquent behaviors. As la ter research supports, these fa ctors were more predictive than sociodemographic characteristics. Although fewer studies have been done in the area of prot ective factors, researchers believe protective factors serve as a buffer to risk factors, interrupt the process through which risk fact ors operate, and may prevent the initial occurrence of a
29 risk factor (Florida Department of Edu cation, 1998; Introduction to Risk Factor and Protective Factors, n.d.). In fact, research s uggests that protective factors can mitigate the effects of a risk environment (Reid, Rei d, & Peterson, 2005; Resnick, et al., 2004) and schools can help buffer the effects of risk factors on adolescent development (Florida Department of Education, 1998). According to Benard (1992), 50% to 80% of students with multiple risk factors in their lives do succeed, especially when they have the experience of a caring school environment. Pr otective factors are important even in the absence of risk factors. In a study of more than 13,000 adolescents, among students without any identified risk factors, the presence of protective factors decreased the proportion of both boys and girls involved in violence (Resnick, Ireland, & Borowsky, 2004). According to Olweus (1993), some conditions tend to create or enhance bully/victim problems while other factors have mitigating effects. Olweus also concluded that the attitudes, routines, and behaviors of school personn el can be a contributing or countervailing force in bullying behavior Risk factors for bullying include poor childhood conditions and child rearing and family problems. This includes an inadequate amount of care or supervision and a lack of clear behavioral limits (Olweus). Schools can also function as a risk or protective factor for antiso cial behavior with school connectedness as the most salient protective factor against acting out behavior (Clark, 1995). Therefore, while reducing risk factors, schools should also work to foster the development of protective f actors that create an environm ent of caring and connection (Clark).
30 Bullying Behavior Bullying is unprovoked and intentiona lly aggressive physical action or psychological control exercised from a positi on of power by one individual or group over another person or group (Florida Office of Safe and Healthy Schools, 2005). According to Olweus (1993), a student is being bullied when he or she is exposed, repeatedly and over time, to negative actions on the part of one or more students. Bullying behavior can include physical aggression, verbal harassment, psychological intimidation and threat, or harassment through electronic communications. Bullying encompasses a spectrum of aggres sive behaviors rangi ng from overt acts of physical violence to more subtle patterns of verbal or relational cruelty (Feinberg, 2003). Physical bullying is a direct form of bullying that involves causing harm to a person or to someones property. Hawkers and Boulton (2000) descri be physical bullying as behavior in which the victims physical in tegrity is attacked and verbal bullying as when the victims status is threatened with words or vocalizations. Verbal bullying is the more common means of direct bullying on a sc hool campus (Olweus, 1993). This type of bullying includes taunting, teasing, name calli ng, extortion, or threats and can be as devastating as physical bullyi ng. A study conducted for the U. S. Department of Health and Human Services reported that students between the ages of 9 and 13 consider namecalling the worst kind of verbal bullying and that threats or taunts based on race or appearance have as much a negative im pact as physical bullying (Windemeyer Communications, 2003). If only overt acts of bullying are incl uded when considering the problem of bullying and victimization, a large percentage of victims will not be identified (Young,
31 Boye, & Nelson, 2006). This is because bullying th at occurs in more subtle ways is more challenging for adults to recognize and unde rstand (Packman, 2005). In fact, some students may not exhibit behavi oral problems in class under th e watch of the teacher, but will harass and bully others when unsupervise d. This type of bullying behavior is known as covert or indirect bullying and in cludes relational aggression and cyberbullying. Relational aggression, defined as harm that occurs through manipulation of a relationship (Young, Boye, & Nelson, 2006), has recently received increased attention in the literature. This type of indirect bullyi ng focuses on social manipulation and includes gossiping, spreading rumors (Young, et al.), exclusion, alliance bu ilding, and ignoring (Nixon, 2005). According to Nixon, gender does not play a role in relational aggression occurring but does differ in how it is mani fested. Typically girls exhibit relational aggression within a circle of friends while boys exhibit relational aggression outside their circle of friends. This pattern begins to change in eighth gr ade when gender differences in relational aggression occurrenc es emerge (Nixon). Young, et al. offer a somewhat different opinion, suggesting that although research does not yiel d a consistent pattern in gender differences, proportionall y, girls engage in more re lational aggression than boys because they are more likely to use this form of bullying over physical aggression. As with more direct forms of bullying, victimization through relational aggression has both short term and long term conseque nces. Ophelia Project research found that students who experience high le vels of relational aggression are less connected to their school and participate in fewer activities (O phelia Project, n.d.). According to Young, et al. (2006), being a victim of relational aggr ession is significantly associated with concurrent social-psychological maladjustment, peer rejection, internalizing problems,
32 and externalizing difficulties. A study conducte d at Ohio State University showed that girls who initiate or engage in relational aggression exhibited adjustment difficulties and had higher self reports of depression, loneli ness, and social isolat ion than their peers (Mounts, 1997). This emotional maladjustment ca n have a long lasting impact similar to that of more direct bullying. A study of 205 fifth and sixth grade urban students to determine the frequency child ren experience overt and rela tional victimization found a significant and positive relationship of medium size effect among overt, r =.37, p < .001 and relational victimization, r = .33, p < .001 and posttraumatic stress (Esposito, 2003). A relatively new and less research ed area of the bullying phenomenon is cyberbullying, defined as online harassing, in timidating, or threatening others by sending or posting harmful or cruel text or images using the internet and other electronic communication devices (Willard, 2005). This type of bullying is furthered defined in Table 5. Table 5 Types and Definitions of Cyberbullying Type Definition Flaming Harassment Cyberstalking Denigration Sending angry, rude, or obscene messages Repeatedly sending offensive messages Repeatedly sending messages that include threats Sending or posting harmful, untrue statements about a person
33 Table 5 (continued): Types and Definitions of Cyberbullying Type Definition Impersonation Outing and trickery Exclusion Pretending to be someone else and sending or posting material Sending or posting sensitive, priv ate, or embarrassing material about a person Actions that intentionally exclude a person from an online group Although a new form of bullying, with estima tes that 99% of teens regularly use the internet (Shariff, 2005), cyberbullying is a rapidly increasing problem. A recent study conducted through the University of New Hamp shire Crimes against Childrens Research Center found that 1 in 17 children aged 10 to 17 have been threatened or harassed online (Florida Office of Safe and Healthy Sc hools, 2005). Similarly, a study of 177 middle school students conducted in Canada reveal ed that 23% of responding students were bullied by email, 35% in chat rooms, and 41% by text messaging (Li, 2005). Although cyberbullying begins anonymously, it can impact learning in the school environment by creating a hostile environm ent where students feel unwelcome and unsafe (Shariff, 2005). Cyberbullying is unique in that the student who is bullying is removed from the immediate feedback of th e victim because there is no face to face confrontation. However, the impact on the victim is not unique. A survey of 5,500 students revealed that 72% of students reporte d online bullying as distressing as face to face bullying (Coady, 2005). Additionally, students who would not engage in face to face
34 bullying may engage in cyberbullying belie ving that technology will allow them to remain anonymous. A common myth about bullying is that children are victimized because of outward appearance. However, the most frequent reason cited by youth for persons being bullied is that they didnt fit in (Nansel, et. al 2001). Olweus (2003) stated that accumulated research indicates personality characterist ics or typical reacti on patterns, along with physical weakness in the case of boys, are si gnificant contributors to the development of bullying problems. Children who bully tend to focus on peers who seem vulnerable. Longitudinal studies show that bullies te nd to gravitate towards children who are physically weak, exhibit internal izing behaviors, lack prosocia l skills, and have low self worth and perceptions of social competence (Rodkin & Hodges, 2003). Schools themselves can also serve as a risk factor for bullying to occur, with teachers attitudes of major significance for the extent of bully/victim problems at school (Olweus, 1993). Teachers may contribute to bullying behavior through a lack of awareness, a nonchalant attit ude, or using inappropriate in terventions (Rodkin & Hodges, 2003). Additionally, teachers typically underestim ate the prevalence of bullying and often fail to stop the behavior when they see it (Olweus; Rodkin, & H odges). The rules and procedures in a school may also indirectly encourage bullying behavior. When students are given no genuine authority over their da ily circumstances in school, they may seize power out of sheer frustration; such drasti c measures are often the only means to power that the educational system allows studen ts (McEwan, 2000). Without the opportunity to exercise a voice in matters that concern them teens may be driven in a nonproductive or self injurious directi on taking the form of high risk e xperimentation, defiance of adult
35 authority and values, or development of a peer culture that ignores rules and expectations (Lickona & Davidson, 2005). In cont rast, when they are given a voice, students will have a stronger commitment to the mission of th e school and develop the skills needed to problem solve and become active citizens (Lickona & Davidson). One high school students comment gathered in qualitativ e research conducted by Lickona and Davidson highlights the importance of empowering students. For students, it is very important that their voice be heard. It gives them a chance to tell the school wh at they think. There would be a major difference in a schools moral character if the students were just given a chance to express themselves. This would show students that administrators and te achers respected them, and then students would be more likely to show respect in turn. (p. 42) Increasing student power requires oppor tunities for meaningful participation. Lickona and Davidson (2005) suggest cla ss meetings, student surveys on school improvement issues, and student led discussions on school concerns as means to increase students voices in decisions affecting the school. Based on resear ch that students in democratic schools develop greater concer n for the welfare of the group, authentic student government where students seek input and report back to the student body is another strategy Lickona and Davidson suggest to positively influence the peer culture. In order for faculty and staff to successfully impl ement strategies to give greater voice and responsibility to students, it is essential for ad ministration to treat fa culty and staff in the same way (Lickona & Davidson).
36 Similar approaches to empower students have also been effective in bullying prevention initiatives. According to Olweus (1993), an important aid in counteracting bullying problems and creating a better social climate in school is for the teachers and students to agree on rules about bullying and engage in discussion about the rules in forums such as class meetings. When student s have the opportunity to get involved, they are likely to experience greater responsibility for conformity to the rules (Olweus). The researcher found that the classes that showed a larger reduction in bully/victim problems as a result of the intervention program had implemented class meetings to a greater extent than those with smaller or no changes in bu llying behavior. Olweus also suggests class PTA meetings as opportunities for students, teachers, and parent s to problem solve bullying issues. Another illustration is th e workshops and summits conducted by the American Association for Univ ersity Women that engaged st udents in dialogue on sexual harassment and bullying allowing for better understanding of per ceptions and conflict resolution with hopes of transf orming a culture of fear and harassment to a culture of camaraderie and trust (American Associ ation of University Women, 1993). Impact and Consequences The American Medical Association (2002) recognizes bullying as a complex and abusive behavior with potentially serious so cial and mental hea lth consequences for children and adolescents. The United States Su rgeon General also declared that bullying and peer harassment are public health probl ems that require federal attention and intervention in order for them to be solv ed (Sprague & Walker, 2005). Bullying occurs within a group context with different stude nts taking on different roles (Packman, 2005)
37 making it a complex problem for schools to de al with. A study cont rolled for background characteristics showed that students in sc hools with high levels of violence had lower math scores by 0.20 of a standard deviation a nd were 5.7 percentage points less likely to graduate (Marzano, 2003). Similarly, Dake, Price, Telljohann, and Funk (2004) report that academic achievement was found to be lower for students involved in all forms of bullying behavior. School adjustment and school bonding were also found to be less likely to occur in students who were engage d in bullying behaviors. In addition to the overall impact on the school, there are shor t term and long term consequences for the students who bully, the victim s of bullying, and the bystande rs who see the bullying or know it is occurring. The student who bullies. A student who bullies is someone who repeatedly hurts another person on purpose. Bullying threatens the social emo tional development of students because it allows children to achieve immediate goals wi thout learning socially acceptable ways to deal with others, resulting in persistent maladaptive patterns (Haynie, et. al, 2001). Bullying behavior has both short term a nd long term consequences for the child who bullies others. Without intervention, children who bully are more likely to develop a criminal record and engage in antisocial behaviors (Olweus, 1993). Olweus found that those who bully have more cases of alcohol ism and substance abuse, more antisocial personality disorders, and are more likely to drop out of school. More recently, the Indiana White Paper on Bullying concluded th at children who bully are more likely to become violent adults while victims of bu llying often suffer from anxiety, low self-
38 esteem, and depression into adulthood (Indiana Department of Education, 2003). Bullying can also be a risk factor for more serious violent behaviors. According to Olweus, by age 23, approximately 60% of boys identified as bullies in middle school had at least one conviction of a crime and 35% to 40% had three or more convictions; 50% of all identified bullies became criminals as adults. Research also supports the hypothesis that children who bully also engage in othe r violent behaviors at school. While both the student who bullies and the victims report a higher likelihood of carrying a weapon to school, the chances of this behavior are higher for students who bully (Viadero, 2003). Self reported bullies were 3.2 times more likel y to carry a weapon to school and 3.1 times more likely to fight often. The researchers also found the correlation increased as the frequency of bullying experi ences increased. These findings are supported in a study by Nansel, Overpeck, Haynie, Ruan, and Sche idt (2003) that showed a consistent relationship between bullying and interperso nal violence with both bullying and being bullied related to higher freque ncies of violence. Regression analyses results indicated a consistent pattern of results with involvem ent in bullying, both for bullies and targets, associated with greater odds of weapon carrying, fighting, and injury from fighting. These relationships were strongest for wea pon carrying but were notable for fighting and fighting injuries as well. Th e highest risk for weapon ca rrying was associated with bullying others in or away from school and being bullied away from school, with 70% of boys and 30% to 40% of girls involved in bullying reporting carrying a weapon in the past month.
39 The victim of bullying. The victim of bullying is defined as a student who is repe atedly exposed to negative actions from peers (Olweus, 1993) in the form of physical attacks, verbal assaults, or psychological abuse. Someone who is bullied is less powerful than the person who is bullying and may be physically smaller than the child who bullied. Victims are typically unable to defend themselves give n that they may be outnumbered, have less physical strength, or be less psychologically resilient (Smith & Ananiadou, 2003). Both boys and girls are at equal risk of bei ng victimized (Siris & Osterman, 2004). Research indicates that children who ar e bullied have lower self esteem and higher levels of stress, anxiety, depression, i llness, and suicidal ideation (Olweus, 1993). The Hawker and Boulton (2000) meta-analysis also established depr ession, anxiety, and low self esteem as consistent correlates of victim experience. Similarly, bullying has a negative impact on students social-emotional and educational lives. A research study involving 300 nine to eleven year olds indi cated that victims have more problems with social skills than non victims (Fox & Boult on, 2005). In this study, the researchers found that internalizing problems such as w ithdrawal, anxiety, de pression, and physical weakness were independently predictive of increase in victimization while having a friend was associated with decreased victimi zation. Six social skills were identified as effective predictors of vi ctimization: looks scared, r = .68, p < .05, gives in to the bully too easily, r = .56, p < .05, cries when picked on, r = .50, p < .05, stands in a way that appears weak, r = .49, p < .05, talks very quietly, r = .34, p < .05, looks unhappy, r = .33, p < .05. In addition, the absen ce of helplessness in girls and the absence of counter aggression in boys were found to be factor s that make bullying diminish or stop.
40 Research conducted by Veenstra, Lindenber, Oldehinkel, DeWinter, Verhulst, and Ormel, (2005) investigating individual char acteristics that predict bullying behavior reached similar conclusions. Using peer nomination, a multivariate analysis distinguished aggressiveness, isolation, dislik eabilty, and gender as strong predictors of bullying while socioeconomic status, parenting, and academic performance were found to be weak predictors. Victims may suffer greater psycholo gical maladjustment than nonvictims (Hawker & Boulton, 2000; Mis hna, 2003). Telljohann (2003) st ated that victimization correlates positively with loneliness and ne gatively with self-esteem and found that students who were bullied were 3.2 to 4.2 times more likely to report anxiety symptoms compared to noninvolved children. Rigby (2003) found that frequently victimized students may have mixed emotions, show va rious symptoms of di stress, and showed significantly more depression th an non-victimized peers. Ex periencing bullying is also associated with poorer psychosocial adjust ment (Malecki, 2003; Nansel, et al., 2001; Rigby, 2003). Victims of bullying suffer from anxiety, depression, impaired concentration, poor self esteem, and avoi dant behavior (Fei nberg, 2003; Hawker & Boulton; Storch, Brassard, Masia-Warner 2003; Unnever & Corn ell, 2003b); they experience acute feelings of re jection and loneliness and in ex treme cases are at risk for suicide (Kumpulainen, Rasanen, & Puura, 2001; OMoore, 1998; Rigby, 2003). Children who are seen as victims by their peers tend to report greater distress than children who are not seen as victims (Haw ker & Boulton). The impact of bullying can also have long term effects. For example, adolescents abus ed by peers report elevated depression and low self-esteem 10 years later (Olweus, 1993).
41 The problems associated with victimiza tion have an impact on school success. According to the National School Safety Center, an estimated 160,000 children miss school every day due to fear of attack or intimidation by other students and 10% of students who drop out of school do so because of repeated bullying (Weinhold, 1999). In addition, the Secret Service found two-thirds of all the school shooters since 1974 had been victims of bullying pr ior to the shootings (Brady, 2001). Thousands more attend school every day filled with fear spending a significant amount of time and emotional strength thinking about ways to avoid teasing and taunti ng leaving little energy for learning. Victims of bullying can experience withdrawal, aggression, and feelings of rejection resulting in both social and academic consequences (Siris & Osterman, 2004). These students may become detached from a dults and peers, have poor attitudes about themselves and others, have difficulty deve loping positive relationships, and begin to reject classroom norms (Siris & Osterman). Loneliness and insecurity are common for victims but responses to bullying vary. Some victims withdraw while others react more aggressive ly. Students who react with aggression, known as provocative victims, fre quently tease and a nnoy the person who is bullying further alienating them from their peers (Siris & Osterman, 2004). Students who bully and victims are not always mutually excl usive with nearly half of bullies reporting being victims as well (Veenstra, et al., 2005). Bullies who are also vi ctims are a distinct group from those who bully or those who are victimized, hence they face unique challenges. These bully/victims primarily a ggress in reaction to aggression and are generally among the most disliked members of the class. They face segregation and rejection from their peers (Rodkin & Hodge s, 2003) and are at risk of both the
42 consequences related to bullying behavior as well as those of the victim. They demonstrate higher levels of aggression a nd depression as well as lower scores on measures of academic achievement, prosoc ial behavior, social acceptance, and selfesteem (Veenstra, et al.). Despite the speci al challenges of this group, there is less research addressing this element of the bu llying phenomenon. Haynie, et al. (2001) did investigate the prevalence of bully/victims in a study of 4,263 middl e school students and found that of 53% of the 301 students who repor ted bullying three or more times over the past year also reported being victimized three or more times. Of the 1,257 frequently victimized students, 64% reported never bul lying. The researchers also found that the bully/victims showed the leas t optimal psychosocial function ing in comparison to those who bully or bullying victims. The bystander. Bystanders are also affected by the chroni c presence of bullying in schools. These students are onlookers to bullying situations and can stand by and do nothing, encourage the bullying behavior, or inte rvene by helping the victim. According to Olweus (1993), there are several different bystander roles. The followers, also called henchmen, stand back and wait for an opportunity to take an active part in the bullying activity; never instigating, always following a nothers lead. The supporter or passive bully stands by to watch but shows no direct support of the bullying behavior. The disengaged onlooker ignores the behavior while th e possible defender wants to help but does not because of fear of the bully. Finally, the defender of the victim stands up for the person who is being bullied and attempts to stop the behavior.
43 Self respect and self conf idence can be eroded when a child witnesses bullying behaviors and is unable or unwilling to re spond effectively. Students who continually observe bullying behavior without interv ention may develop a decreased sense of individual responsibility (Olweus, 1993). Th ese students commonly experience fear and worry that there will be retaliation if they get involved (U.S. Department of Education, 1998). The impact of bullying on the bysta nder includes a sense of anger and helplessness, guilt for not taking action or for enjoying the role of witness, an avoidance of areas in the school where bullying occu rs, and an underlying feeling of safety concerns. One study conducted for the U.S. De partment of Health and Human Services reports that bystanders suffer from feelings of helplessness, powerlessness, and develop poor coping and problem solving skills (Windemeyer Communications, 2003). The majority of students in a school are bys tanders rather than students who bully or victims of bullying. In a study of 108 urba n schools in 13 states, 50.2% of students reported seeing others bullied at least on ce a month (Perkins, 2006). Although bystanders comprise the largest percentage of students in a school, they seldom intervene on behalf of the victim. A study of bullying on pla ygrounds conducted by Hawkins, Pepler, and Craig (2001) found that observers were pres ent in 88% of bully ing situations but intervened in only 19%. An earlier study by Peplar (1998) had similar results. An examination of the roles of peers in bu llying situations observed on urban school playgrounds revealed that peer s were involved in 85% of bul lying incidents. Peers were active participants in 48% of the episodes and reinforced the bullying in 81% of the episodes but intervened in only 13% of the ep isodes they observed. Peers were also more respectful and friendly toward the bullies than the victims.
44 Although bystanders are not directly i nvolved in bullying, these students are affected because they suffer from a less secure learning environment, fear that they may be the next target, and have knowledge th at teachers and other adults are unable or unwilling to control bullying behavior (Flori da Office of Safe and Healthy Schools, 2005). Peer involvement may also be a factor that perpetuates bu llying interactions whether peers are active participants or passive bystanders (Peplar, 1998). Prevalence of Bullying Behavior Every day in our Nations schools, childre n are threatened, te ased, taunted, and tormented by bullies (U.S. Department of E ducation, n.d.). Survey results consistently report the existence of bullying in schools. Scientific studies show that bullying is an international problem with remarkable sim ilarity in the incidence of bullying from country to country (Cleary, 2000). Although resear chers agree that bully ing is a pervasive problem, estimates vary. The prevalence of bullying in elementary schools worldwide varies from 11.3% to 49.8% with estimates in the United States around 19% (Dake, et al., 2004). Storch, et al. (2003) estimate that as many as 20% of children and adolescents are exposed to negative peer interactions on a fre quent basis. According to Feinberg (2003), an estimated 15% to 30% of students nationwid e are either bullies or victims of bullying. The National Institute of Child Health a nd Human Development found that almost onethird of U.S. students in grades 6 to 10 were directly or indirec tly involved in serious, frequent bullying (Nansel, et al., 2001). The U.S. Department of Education (1998) reports that 77% of middle and high sc hool students surveyed had been bullied at some point in their school career. A 2003 Gallup Youth Survey of teens aged 13 to 17 indicated that
45 37% of teens admitted to being teased or picked on at school (Mason, 2003). A 2001 survey conducted by the U.S. Center for Di sease Control and Prevention found that 12% of students said someone at school had calle d them a degrading word having to do with race, religion, ethnicity, disability, gende r, or sexual orientation (Donald, 2002). The American Association of University Woman (1993) reported that 81% of public school students in grades 8 through 11 experience d some form of sexual harassment. A Nickelodeon survey conducted in conjuncti on with the Kaiser Family Foundation and International Communications Research found th at 55% of 8 to 11 year olds and 68% of 12 to 15 year olds reported bullying as a big problem for people their age (Kaiser Family Foundation, 2001). In addition, 74% of 8 to 11 y ear olds and 86% of 12 to 15 year olds indicated that kids get teased or bullied at their school. According to the U.S. Department of Justice (2004), in 2003 7% of students aged 12 through 18 reported being bullied within the last six months. The U.S. Depart ment of Justice report also indicated grade level was inversely related to students likelihood of being bullied. As grade level increased, students likelihood of being bulli ed decreased. In this 2003 School Crime Supplement study 14% of sixth grade students, 13% of seventh grade students, 7% of ninth grade students, and 2% of twelfth grade students had been bullied at school. Trends in current research also indicate that the prevalence of bullying in schools may be increasing. The 2005 administration of the School Crime Supplement survey showed an increase in bullying behavior from 7% in 2003 to nearly 29% of students reporting having been bullied at school during the previous 6 months (U.S. Department of Justice, 2006). Although grade level continued to be inve rsely related to students likelihood of being bullied, there was an increas e in the percentage of students reporting
46 victimization with 37% of 6th grade students, 28% of 9th grade students, and 20% of 12thgrade students reporting that they had been bullied at school. Of these students, 52% reported being bullied once or twice, 25% re ported once or twice a month, 11% reported being bullied once or twice a week and 8% reported daily bullying. This trend is also seen in the results of a 2001 survey done with 11,000 elementary and middle school students when compared to 1983 results from the same survey instrument. The percentage of victim ized students had increased by approximately 50% and the percentage of st udents involved in frequent, serious bullying had increased 65% (Olweus, 2003). According to the 2002 Indicators of School Crime and Safety Report, in 2001 8% of students reported that they had been bullied at sc hool in the last six months, up from 5% in 1999. The percentage of students who reported that they had been bullied increased between 1999 and 2001 with bot h males and females more likely to be bullied in 2001 than in 1999. Although ther e was an increase from 1999 to 2001, there was no significant change between 2001 and 2003. Bullying and the Special Needs Child Disability harassment, a form of bullyi ng specifically based on or because of a disability, is another possible fo rm of bullying that relates to children with special needs. This type of harassment creates a hostile e nvironment by denying access to, participation in, or receipt of benefits, services, or opportunities at school (Hoover & Stenhjem, 2003). Adjustment problems and difficulty with so cial functioning occu r among all children; however, these risks increase when a child ha s a disability or ch ronic illness (Moore, 2002; Yude & Goodman, 1999). According to Barabarin, Whitten, and Bonds (1994),
47 chronically ill children experience these probl ems at rates twice as high as those for healthy children. The Office of Special Educa tion and Rehabilitative Services issued an official statement in July of 2000 stating th at the increasing number of complaints and consultation calls regarding di sability harassment demonstrates the steadily increasing allegations and proven situations of disa bility harassment (Hoover & Stenhjem, 2003). Equal access to educational benefits for these youth can be eroded through bullying, including denial of ri ghts under the Indi viduals with Disabilities Act, Section 504 of the Rehabilitation Act, Title II, and provisions of a Free Appropriate Public Education. According to Roberts (2003), estimates of the number of youths under age 18 who experience a chronic health condition range fr om 10% to 30% and are rising, due in part to significant advances in medical care that reduce mortality. Children and adolescents with chronic illness experience more academic difficulty than their healthy peers. As many as 45% of students with chronic illness report falling behind in their school work, leading them to dislike school (Lynch, Lewis, & Murphy, 1992). In addition, Lynch, et al. estimate that 58% of students with chr onic conditions routinely miss school and 10% miss more than 25% of the year. No consistent association between vict imization and physical characteristics has been found (Olweus, 1994). Other than the tend ency for male bullies to be physically stronger than their victims, being weak and having weak friends significantly enhances a students likelihood of becoming a victim of peer abuse (Hodge, Malone, & Perry, 1997). Highly disliked or peer rejected children are thought to be at greater risk of victimization than non-rejected peers because of their devalu ed status in the peer group (Shea, 2003). In fact, Shea reports that victimized ch ildren generally have fewer friends than non-
48 victimized children. Research has shown that issues facing children with special needs have the potential to impact peer rejection resulting in in creased victimization. Children with chronic illness may be at risk for experi encing social difficulties due to the physical effects of their disease, its treatment, or the disruptions to daily lif e that they experience (Reiter-Purtill, Gerhardt, Va nnatta, Passo, & Noll, 2003). Ch ildren with special needs may have cognitive and/or physical limitations that can contribute to ineffectiveness in play activities, sports, and self-defense (Heinrichs, 2003). According to Heinrichs, children diagnosed with attention deficit di sorder, oppositional-defiant disorder, bipolar disorder, obsessive-compulsive disorder, Aspe rger Syndrome, and learning disorders are typically more rejected by their peers. In addition, Mishna (20 03) reported that children with learning disabilities had fewer friends and were teased significantly more than children without learning disabilities. Bara barin, et al. (1994) f ound that approximately one in three children with Sickle Cell Disease was teased because of the illness or some visible sequelae of the illness. In addition, t easing was more likely to be experienced by those with pain than those without pain lead ing to their conclusion that illness severity places children at greater risk of teasing. In support of the hypothesis that severe fo rms of chronic illness seem to present an increased risk of social pr oblems than milder forms, Reite r-Purtill, et al (2003) report that multiple studies suggest that children with severe forms of Juvenile Rheumatoid Arthritis experience more social difficulties than patients with mild or inactive forms. Additionally, Noll (2003) reported that chil dren with severe hemophilia are more adversely affected by the disease than children with milder forms. However, after controlling for issues of pain, Reiter-Purtill, et al. found no differences between children
49 with Juvenile Rheumatoid Arthritis and cont rols in difficulties with social functioning, measures of social reputati on and social acceptance or th e number of times they were chosen by peers as a best friend or re ciprocated friend. Similarly, Noll found no significant differences in social functioning or measures of social acceptance between children with hemophilia and the control group. Research in the area of bullying has s hown that friendships can serve as a deterrent to victimization. This protective fa ctor, or lack of it can have significant implications for children with special need s. Mishna (2003) found that approximately 25% to 30% of students with a le arning disability are socially rejected compared to 8% to 16% of peers without a lear ning disability. In addition, a study conducted by Barabarin, et al. (1994) found that more th an one in five children with Sickle Cell Disease had no close friends. Children with AD/HD are consistently found to be less popular and more rejected by their peers than children without AD/HD pu tting them at increased risk of peer victimization (Shea, 2003). A study of middle school stude nts conducted by Unnever and Cornell (2003a) found that studen ts with AD/HD are at increased risk of being victimized by bullies. Hodge (2003) found th at hyperactivity is likely to annoy peers and provoke potential aggressors, thus lead ing to increases in victimiza tion. Hodges data showed that 34% of students who reported taking medication for AD/HD al so reported being bullied at least two or three times a month compared to 22% of the students in the control group. In a study of 60,000 children in Finland, Kumpulainen, et al. (2001) found that psychiatric disorders were common among child ren impacted by bullying behaviors with 29% of children who bullied and 14% of vic tims having attention deficit disorder.
50 Children with hemiplegia in mainstr eam schools were compared with their classmates and found to have two to three times the rate of peer problems. Twenty seven percent of the children with hemiplegia had two or more peer problems compared with only 9% of controls. They were twice as like ly to be rejected, tw ice as likely to lack friends, and three times as likely to be victimized (Yude & Goodman, 1999). Hugh-Jones and Smith (1999) surveyed dysflue nt adults about the quality of their life in school with a focus on e xperiences of being bullied. When asked if they were ever teased or bullied at school, 83% of thos e surveyed responded yes with 18% reporting bullying every day and 41% replying a few times per week. The methods of bullying included name-calling, threatening, rumors, physi cal bullying, and racial insults. Results of the study indicated a signifi cant relationship between the severity of dysfluency and the reported likelihood of being bullied. Ho wever, they found that the only successful predictor for severity of bullying was diffi culty in making friends indicating dysfluent children may be bullied more because of thei r difficulties in friendship-making than from the stammering itself. Positive Interpersonal Relationships Meeting human needs is a prerequis ite to a healthy classroom (Benard, 1991). The affective dimension of the school da ythat is how students feel about their experience at schoolsis as important as the academic dimension. Without trust and respect, without a p hysically and psychologically safe
51 environment, teaching and learning cannot reach their maximum potential (Perkins, 2006, p. 6). Based on Maslows hierarchy of needs, the need to belong and feel love must be met prior to moving to a highe r level of growth (Calabrese 2000) with psychological and emotional safety necessary prior to lear ning and intellectual development (Halford, 1996). According to Maslows continuum of needs, every individua l requires security and freedom from fear, anxiety, and chaos along with structure, orde r, established limits and protection from harm (Drapela, 1987). Safety needs promote the physiological survival of the individual and are esp ecially strong during infancy and childhood (Drapela). Once the lower level safety needs ar e satisfied, the individuals need for love, affection and belonging emerge and loneliness and isolation become painful experiences (Drapela). Glassers (1995) c ontrol theory also in cludes the need for love and belonging, defined as a need to feel c onnected to others, as one of five basic needs of survival. Positive relationships help to balance em otions and develop a sense of security, resilience to stress, and an ability to make sense of life (Pierson, 2005). In fact, the desire to belong to a community is a part of human nature (Lambert, et al., 1996) with patterns of relationships serving as the system synapses through which meaning and knowledge are constructed and the basi s through which humans integr ate emotion, identity, and cognition (Covey, 1991). According to the Commission on Children at Risk (2003), humans are born to form attachments because the brain is physically wired to develop in tandem with anothers through emotional communication. The Commissions report argues that humans are biologically primed to find meaning through attachments to others. The report also expresses concern that some youth experience a deficit of
52 connectedness due to a difference between what is biologically required and what their social situation provides. Not only can th is lack of social connection influence development, there is a significant impact on education and lear ning. Noddings (1995) contends that there is more to learning than academic proficiency and test scores; schools will not achieve adequate academic achievement unless the students believe they are cared for and learn to care for others. When humans care, they want to do their best for the object of their caring (N oddings) resulting in increased performance. The role of positive relationships is also noteworthy in a ddressing the problem of bullying in school in that research in the area of bullying preven tion suggests that risk of victimization by peers is related to the quantity and qual ity of interpersonal relationships (Rodkin & Hodges, 2003). It is important to belong to a group and to feel a sense that one is capable of receiving and giving love (Calabrese, 2000). Gi ven that education is a social endeavor (Lambert, et al., 1996; Shapiro, 2000), this sense of belonging can emerge from an accepting school environment (Calabrese). The work of Dewey, Piaget, Vygotsky, and Feurerstein, suggests that learni ng is a social activity in which knowledge is constructed as a result of interaction and shared efforts to make sense of new information (Walker & Lambert, 1995). Positive Adult Student Relationships Social control theory states that the st rength and quality of relationships with significant others is crucial to an individuals tendency to engage in deviant behavior (Hirsch, 1969). An expansion of this theory to the classroom environment shows that the
53 interpersonal relationship between teacher a nd students is an important element that contributes to the learning process of students (Brekelman, Wubbles, & denBrok, 2002; Elias, et al., 1997). There is an extensive know ledge base on the behavi oral correlates of effective teacher-related and peer-related so cial emotional adjust ments that children negotiate within the context of schooling; st udents who fail to make these adjustments are behaviorally and academically at risk (Spr ague & Walker, 2005). In fact, a report on school shootings from the U.S. Department of Education (2002) states that an important effort in prevention is to ensure that yout h have opportunities to ta lk and connect with caring adults. Research show s that students with caring and supportive interpersonal relationships in school report more positive academic attitudes and values, satisfaction with school, and are more engaged in the le arning process (Klum & Connell, 2004). An investigation focusing on nurturing relationshi ps and preventing exclusion to increase students sense of belonging to school resu lted in improvement in the social, academic, and emotional behavior of victimized stude nts (Siris & Osterman, 2004). In this study teachers attempted to connect with victim s by spending more time with the students, creating more opportunities for students to get to know their classmates, and asking questions to learn more about the students. Th ese changes in teacher behavior resulted in an observed increase in students self c onfidence and improved social skills. Teachers also observed an improved classroom climate noting that the more positive the teacher, the more supportive and accepting the students were. In a similar study, Murray and Malmgren (2005) used a randomized control group design to examine the effects of a program designed to improve urban adolescen ts relationships with one teacher and found that supportive ad ult-child relationships can pr omote social, emotional, and
54 academic adjustment among children exposed to multiple risks. The interventions were designed to establish communi cation and warmth in teache r-student rela tionships and involved teachers meeting weekly with a ssigned students, developing goal sheets, increasing teacher praise and encouragement, and calling the students at home to discuss progress at school. After the five month progr am, students in the experimental group had higher grade point averages following the intervention than did the control group, F (1, 47) = 4.36, p < .05. Multiple research studies link student learning achievement and engagement in school to meaningful supportive relationships and membership in a community (Lambert, et al., 1996). Fewer studies a ddress the level of support need ed to impact learning. Klum and Connell (2004) investigated the threshold level on teacher support and engagement as well as the level of impact that achievi ng the threshold limit provided. Elementary students experiencing high levels of support were 89% more likely to feel engaged whereas unsupported students were 93% less likel y to feel engaged in school. Elementary students reporting high levels of engagement were 44% more likely to do well on performance and attendance measures. At th e middle school level, students with high levels of teacher support were three times more likely to have high levels of engagement and those reporting low levels of teacher s upport were 68% more likely to be disengaged from school. Middle school students with hi gh levels of engagement were 75% more likely to do well on achievement and attendan ce indices. The researchers concluded that their findings provide support for the existe nce of an indirect link between teacher support, student engagement, and academic achievement for both elementary and middle school students.
55 Supporting these findings, results from a study testing a path model for explaining school engagement among Lati no middle school students within the context of risk and protective factors showed that parent support ( = .14), teacher support ( = .32) and friend support ( = .17) were all directly related to school engagement (Reid, et al., 2005). Teacher support had the strongest correlation with school engagement, R = .35, p < .01. Isernhagen and Harris (2002) concluded that faculty must be encouraged to build relationships with students in order to address the problem of bullying on campuses. In a study examining the relationship of school conne ctedness and adolescent risk behaviors, McNeely and Falci (2004) f ound that violence was the only outcome for which teacher support was both a protective factor (risk ratio = .90, p < .001) and was also associated with cessation of the beha vior (risk ratio = 1.07, p < .01). Adults in the school setting play a prominent role in determining the ex tent to which bullying problems will arise and grow into problem behaviors (Florida Office of Safe and Healthy Schools, 2005). Research by Peplar (1998) showed that childr en bully when they are not with the teacher. A chi square analysis revealed a significant association with classroom activity (2, N = 60 = 199.37, p < .001). Sixty-five percent of bullying incidents occurred when children were involved in solitary tasks, 23% occurred in group tasks, and 12% occurred when students were engaged in teacher led activities. Positive interpersonal relationships with school staff can function as a protective factor against violence in general as we ll as bullying in pa rticular. Developing relationships and trust in adults is critical in dealing with bullying because students often feel that school faculty do not interv ene in bullying incidents (Packman, 2005).
56 Perception of the frequency of intervention al so varies between st udents and adults in school. A study by Garrett (2003) stated that 71% of teachers say they intervene in bullying situations whereas only 25% of st udents reported teacher intervention when bullying occurs. However, Peplar (1998) found that teachers intervened significantly more than peers when proxima l to the bullying situation, z = 2.7, p < .05. Peer Relationships The ability to interact cooperatively with peers, inhibit anti social behavior, and form close relationships such as friendsh ips are important developmental tasks for children as they enter school and they pr ovide the foundation for subsequent skill development (Pellegrini & Bohn, 2005). Developi ng the skills necessary for membership in peer groups provides a basis for successful interaction with peers and teachers and is necessary for adjustment to school. Having at l east one friend also serves as a protective factor for the negative effects of peer rejection (Mounts, 1997). Children who fail to establish a reciprocated best friend are more victimized than those who have a reciprocated best friend (Rodki n & Hodges, 2003). This also ha s the potential to impact academic achievement. In a longitudinal study conducted by Flook and Repetti (2005), peer acceptance was significantly associat ed with academic performance. Less peer acceptance was consistently associated with poorer academic achievement, r = -.49, p < .01. A lack of peer acceptance also predicted academic performance in sixth grade, F (1, 151) = 12.37, p < .01. Student who bully and their victims are em bedded within a social system made up of other students, teachers, and their interr elationships (Rodkin & Hodges, 2003). In view
57 of the fact that childrens social competence develops in the contex t of interacting with their peers, bullying is a phenomenon that ca n interfere with the normal development of peer relationships. Research conducted with 4,263 middle school students showed that adolescent and peer relations hips influence bullying and vi ctimization (Haynie, et al., 2001). The researchers found that deviant peer influences, F = .630, p < .001 were stronger predictors of bullying behavior than gender, F = -.160, p < .001, grade, F = .207, p < .001, school bonding, F = -.510, p < .001 or parental support, F = -.433, p < .001. Peer groups where norms favor bullying influe nce the level of bullying for both boys and girls. When students with high status engage in or endorse bullying, they send a message to other peers and contribute to an emerging norm accepting of bullying (Rodkin & Hodges). Another distressing as pect of this social phenomenon is that some socially savvy students may use bullying, particularly relational aggression, to maintain or improve their social status (Young, et al., 2006). Peer relationships can also have a positive impact by serving as a protective factor against bullying. According to research, best friend support and classmate support scores were significantly related to experiencing bul lying for both boys and girls (Rigby, 2000) whereas the number of friends students have is negatively as sociated with victimization (Hodge, 2003). According to Sanders and Phye (2004), having at least one friend at school is a fundamental resource for the prevention of bullying. Longitudinal studies show that children who have at least one friend that they can count on to stick up for them are relatively unlikely to become vic tims of peer harassment (Shea, 2003) whereas rejection by peers leaves students unprotected and susceptible to further victimization (Mishna, 2003). Moore (2002) also found that support from close friends can buffer the
58 impact of stressors such as adjusting to a chronic illness, coping with medical treatments, restrictions on activitie s, and teasing from peers. Havi ng a close friend may also supply the social support to provide a buffering eff ect when a student is victimized (Young, et al., 2006). Schwartz, Dodge, Petit, and Bates (2000) studied students raised in homes characterized by high levels of marital conflict, stress, abus e, and harsh discipline and found that victimization by peers was non exis tent for children with many friends and intensified for children with few friends. The Role of the School Environment School Culture As people develop relationships within thei r social system, they develop a culture of shared patterns and expecta tions that all members within the culture learn (Shapiro, 2000). Schools are no exception. In fact, the nature of th e relationship among staff members can set the tone for the school. Students are more likely to be motivated and engaged in schools where the staff is energetic and positive and openly demonstrates their care for others (Lambert, et al., 1996). In contrast, schools may provide greater or lesser opportunities for bullying a nd violence to take place in terms of the nature of the school environment (Smith & Ananiadou, 2003). Bu llying has the potential to impact all students at a school by negatively influencing a schools climate. The culture of the school is shaped by daily experiences and created by the complex pattern of norms, attitudes, beliefs, behaviors, values, ceremonies, traditions, and myths that are deeply ingrained in the core of the organization and dictate the way things are done in a school (Barth, 2001) From a psychological perspective, it is
59 important for teachers to es tablish an atmosphere of mutual support, caring, and community (UCLA School Mental Health Project, 1998). This environment exists when a critical mass of stakeholders are committed to each other and to the goals and values of the school and extend effort towards meeting the goals and maintaining the relationships (UCLA School Mental Health Project). Much of the literature on school culture addresses the importance of a caring culture which includes caring for and about ot hers. However, community refers not just to the sense of cohesion among students and t eachers, but also to the notion that the educational environment plays an important ro le in how students learn and teachers teach (Walker & Lambert, 1995). Students work hard er, have better attendance, and higher academic achievement in schools with str ong communities (Stolp, 1995). Teachers also work harder and enjoy their work more in this type of environment (Stolp). According to the UCLA School Mental Health Project ( 1998), learning and teaching are experienced most positively when the learner cares a bout learning and the teacher cares about teaching. This caring environment is characte rized by an atmosphere where students feel welcome, respected and comfortable; opportunities exist for developing caring relationships with peers and adults; inform ation and expectations enable students to determine what it means to care for themse lves and the group; and opportunities, training and expectations encourage students to c ontribute to the greater good. A caring school culture also attends to stude nts who have difficulty making friends. When all facets of caring are present and balanced, they can nurtu re individuals and f acilitate the learning process.
60 There is a tendency for schools to focus on structure rather th an culture partly because the culture of the school is resist ant to change (Abourjilie, 2006). Change in structure is tangible whereas culture changes are less visible and more difficult to make due to the fact that one can pronounce a ch ange in policy or procedure but cannot proclaim change in attitudes, beliefs, or behaviors (DuFours & Eaker, 1998). Unless teachers and administrators act to change th e culture of the school, all innovations will have to fit in and around existing elements of the culture (Barth, 2001). However, adults cannot act in a vacuum. In addition to a dult intervention, McQuill an (2005) found that when student participation is encouraged and nurtured, classroom and school changes are likely to be deepened and sustained. School Climate The culture of the school creates the c limate of the school (Abourjilie, 2006). School climate is the learning environment created through the in teraction of human relationships, physical setting, and psychologica l atmosphere (Perkins, 2006). It is the shared perceptions of a school and consists of the attitudes, beliefs, values, and norms that underlie the instructional practices a nd operations of a school (Welker, 2000). It embodies the physical and psychological enviro nment of a school with a specific link to student academic achievement (Niebuhr, 1999). El ements of the school environment that contribute to the school climate include conti nuous academic and social growth, levels of respect, trust and integrity, morale and cohesiveness, caring, opportunities for input (Rinehart, 1993), level of or derliness, degree of satisfa ction experienced, amount of productivity possible, sense of belonging (F lorida Department of Education, 2002), and
61 degree of connectedness (Doan, et al., 2003). According to Sprague and Walker (2005), school-based protective factors that contribute to a safe, healthy school are a positive school climate and atmosphere; clear and high performance expectations for all students; inclusionary values and pr actices throughout the school; st rong student bonding to the school environment; high levels of student participation; and parent involvement in schooling. The researchers also included the provision of opportunities for skill acquisition and social development and school -wide conflict resolution strategies as protective factors. Comparatively, Fein, et al. ( 2002), listed six factors that characterize a climate of safety and respect within an educational setting. These factors in cluded having positive role models within the faculty; the pr esence of a positive connection between each student and at least one adu lt in authority; an openness for discussion where diversity and differences are respected; an atmosphere where communication between adults and students is encouraged and supported; and an environment in which conflict is managed and mediated constructively. Wilson (2004) adds an emphasis on academic achievement; respect for all in the school community; fair and consistent discipli ne policies; attention to safety issues; and family and community i nvolvement to the list of characteristics of a positive school climate. The National Association of Attorneys General (1999) proclaims that a supportive school climate is the most important step in ensuring that schools provide a safe and welcoming environment for all students. This is supported by increasing research indicating that students w ho feel connected and safe in school perform better academically (Zins, et al., 2004). According to Deborah Price, Deputy Undersecretary of
62 the Office of Safe and Drug Free Schools, kid s must feel safe and have a sense of wellbeing. If there is bullying, drug use, and an absence of commitment to character, kids dont learn (U.S. Department of Education, n.d., p. 10). Educators have the ability and responsibi lity to create a clim ate that addresses harassment and cultivates the courage and l eadership of students In an educational setting where there is a climat e of safety, adults and student s respect each other (Fein, et al., 2002). Research suggests that schools can strengthen the school climate and a students sense of belonging by adopting co mmunity building stra tegies including actively cultivating respectful, supportive re lationships among students, teachers, and parents (Schaps, 2003). Achievement is more likely to occur in a fr iendly classroom envir onment with a teacher who connects with students and encourages them to create, take risks, and share ideas (Mendler, 2001). Research done with a sample of schools in Michigan that controlled for the effects of race and socioeconomic factor s found that school climate factors accounted for 63% of the variation in mean school achievement between low and high achieving schools (Welker, 2000). Fraser (1999) found simila r results in a review of 40 studies of the effects of classroom environment on student outcomes and found that learning environment was consistently and strongly a ssociated with achievement and affective outcomes. Higher achievement occurred in classes perceived as having greater cohesiveness, satisfaction, and goal directio n; and less disorgan ization and friction. Results from the Gay, Lesbian, and Straig ht Education Network 2003 National School Climate Survey of 887 lesbian, gay, bisexual, and transgender (LGBT) youth also found a direct relationship between in school vict imization, grade point average, and college
63 aspirations of LGBT youth (US Newswire, 2003). Key findings of the survey included the finding that unchecked harassment correlat es with poor performance and diminished aspirations and that supportive teachers can make a difference in that relationship. Results also showed that LGBT students who di d not have or were unaware of a policy to protect them were 40% more likely to skip school. Additionally, 84% of LGBT student report being harassed and 82.9% of LGBT students report that faculty never or rarely intervenes when present. A study conducted by Hoy and Sabo (1998) investigated the hypothesis that teachers perceptions of the openness of the school climate affect student achievement. The researchers defined an open climate as one with a high degree of trust and low disengagement, principal and faculty who are genuine and open in their interactions, a principal who leads by example providing a blen d of direction and support, and teachers who work well together and are committed to the task at hand. Results of the study demonstrated a significant and positive rela tionship between school climate and student achievement; the more open and healthy the envi ronment, the greater the levels of student achievement in reading, writing and math. The health of the school climate had a significant relationship with math, reading, and writing achievement as measured by the New Jersey Eighth Grade Ea rly Warning Test, EWT, r = .61, .58 and .54; p < .01. In addition, a multiple regression of six climate dimensions found that a lack of principal restrictiveness combined with collegial and committed teacher behavior are the best predictors of achievement. When contro lling for socioeconomic status, the six dimensions had multiple R s of .69, .68 and .61 with math, reading, and writing achievement scores respectively and explai ned 44%, 42%, and 33% of the variance for
64 respective tests. The elements of climate ope nness considered were: supportive, directive, restrictive, collegial, committed, and disengage d. Disengagement was the only factor that made no independent contribu tion to the explanation of achievement variance. Fraser and Fisher (1983) considered bo th student and teacher perceptions when they investigated the differences between st udents and teachers in their perception of the classroom environment and of differences between the actual environment and that preferred by students or teachers. They used a person-environment framework to explore whether student outcomes depend on the natu re of the classroom environment and the match between the students preferences and th e actual environment. Results showed that students prefer a more positive classroom environment than was perceived to be present and teachers perceived a more positive actual environment than students perceived to be present. The researchers conc luded that class achievement can be enhanced by changing the actual classroom environment making it more congruent with the environment preferred by the class. Hoy and Feldman (1999) state that healthy schools are better places to work than unhealthy ones. Teachers are more productive, administrators are more reflective, and students achieve at higher levels. Accordi ng to Brooks (1999), the underpinnings of a healthy school climate include empathy, the ability to see the world through anothers eyes, and self-esteem: the feelings and t houghts students have about their competence, ability to make a difference, ability to lear n from success and failures, to have control over their lives and to treat th emselves and others with respect. Academic emphasis is also an integral part of a healthy school; when there are high expectations, the learning environment is orderly and serious, teacher s believe students can achieve, students are
65 committed to doing well, schools are successful and students achieve at high levels. A students sense of security a nd self-worth in a classroom provides the scaffolding that supports increased learning, motivation, self-d iscipline, realistic risk-taking, and the ability to deal effectively with mistakes (B rooks). Students will learn most effectively in an atmosphere in which they feel safe and do not fear being ridiculed or humiliated, in which they are challenged and assisted to meet realistic goals, in which they feel teachers genuinely care about them and respect their individuality, in which learning is seen as exciting rather than drudgery (Brooks). An example of this can be found in a qualitative study conducted by Haynes and Marans (1999) to assess school climate in an urban elementary school. The research revealed a connection between the climate and high rates of absenteeism and low levels of academic achievement. The students achievement, behavior, and attitude towards school were all conn ected to students perceptions of being isolate d, disregarded, and treated disr espectfully by their peers and teachers. School Climate and Bullying Bullying and harassment pose serious ps ychological and behavioral risk for victims and students who bully; these events can also have a serious, negative impact on the climate and social ecology of schools (Limber & Small, 2003). Students depend on adults to provide an environment that is sa fe and free from fear. Changing the climate of a school is imperative if it is embedded w ith beliefs and conduct that support bullying behaviors. Bullies do not stop their behavior for no reason; they persist until confronted by adults who change the environment in whic h the behavior occurs or change the mind
66 set of the perpetrator (Florida Office of Safe and Healthy Sch ools, 2005). Attending to the school climate decreases th e likelihood of students interact ing in an aggressive or threatening manner. Schools that implemen t strategies to create a safe learning environment benefit from fewer office re ferrals, less physical bullying, and more appropriate social intera ction (Young, et al., 2006). According to Elliot (2003), stopping bu llying among students is difficult when adults in the community are actively demons trating this same behavior. For example, teachers can be overheard gossiping about st udents in the hallway. Students who hear malicious gossip, rumors, and ridicule among adults in their lives take that as a signal this it is acceptable behavior. Smith and Brain (2000 ) define this culture of bullying as a multidimensional phenomenon characterized by a normative set of shared beliefs that support and even encourage bullying. These beliefs result in behaviors that support bullying behavior by rewarding, enabling, and empowering the bullies. Key elements in this culture of bullying are th e levels of adult intervention and peer intervention. Results from a study of six middle schools in Virg inia conducted by Unnever and Cornell (2003b) indicated that fewer than half of the students felt that teachers intervened to stop bullying behavior and two-thirds felt that their teachers did li ttle to counteract bullying. Another key finding was that 10% of the studen ts not identified as bullies reported that they would join in bullying another student leading to their conclu sion that even though bullies may more strongly identify with the culture of bullying, students who do not bully also adhere to its beliefs. Bullying in schools contributes to a climat e of fear and intimidation where some students feel unhappy and unwelcome (Batsche & Knoff, 1994). A school climate that
67 reinforces or ignores bullying can contribute to a victims sense of helplessness and reluctance to report bullying incidents (Gr eene, 2003). Although vict ims are universally unhappy about their status, the victimization is unlikely to stop unless there is a shift in the social climate of the school (Gottheil & Dubow, 2001). Humans have a basic need for emotional and physical safety, for close supportive relationships and connectedness (Schaps, 2003). An effective way to reduce lo w level forms of violence in schools is to create a school culture characterized by warmth, tolerance, sensitivity, diversity cooperation, and expectations of appr opriate behavior (Dupper & Adams, 2002). Creating a positive school climate rooted in shared values and responsible student participation can help student feel safe, s upported, and engaged in school. In contrast, students who feel isolated or lonely expe nd more energy seeking friends than learning (Mendler, 2001). In a healthy en vironment students are free from harassment and know that adults care for them whereas sc hool environments char acterized by bullying and meanness can lead to student isolation a nd fear inhibiting learning and growth (Fein, et al., 2002). School Connectedness Successful schools are communities of learne rs that are not based on contracts or commitments but on shared values and relations hip (Horn & Kincheloe, 2001) that create a caring environment. School climate is defined and fostered by students having a positive connection to at least one adult in aut hority (Fein, et al., 2002). This aspect of the school climate known as school connectedness, commonly referred to as school bonding, can be defined as students experience of car ing at school and a se nse of closeness to
68 school personnel and environment (Smith, 2004). School connectedness is a component of school climate that creates a feeling of belonging to the school and being accepted by others and is an outcome of an environmen t where students are able to know and trust and be known and trusted by adults (K ohn, 2004). School connectedness includes the sense of attachment and commitment a student feels as a result of perceiving that students and peers care about them (Wilson, 2004). The importance of connecting with a car ing environment is evident in the Wingspread Declaration: A National St rategy for Improving School Connectedness document created by the Adolescent Health and Developm ent at the University of Minnesota in collaboration with the U.S. Ce nters for Disease Control and Prevention and the Johnson Foundation (University of Minneso ta, 2003). Based on empirical evidence, it declares that two c ritical requirements for feeling c onnected to school include students experiencing positive adult/student relationshi ps and physical and emotional safety. In agreement with this Declaration, Blum (2005) suggests several factors that influence school connectedness: students like school a nd feel they belong, believe teachers care about them and their learning and that educa tion matters, have frie nds at school, believe discipline is fair, and have opportunities to participate in extracurricula r activities. For comparison, Blum also lists three threats to connectedness: social isolation, lack of safety in schools, and po or classroom management. Studies have shown that connecting to sc hool is important be ginning as early as preschool and continuing through high school. At the preschool level, children with more secure attachments to their teachers and more positive interactions were more engaged in complex social play, demonstrated more a dvanced cognitive abilit y, and showed more
69 resiliency (Howes & Ritchie, 2002). Eval uation of the elementary level Child Development Project revealed increased e ffectiveness of the program when students perceived their school as a caring community (Solomon, Battistich, Watson, Schaps, & Lewis, 2000). At the middle school level, We ntzel (2002) reported that when students perceive their teachers as supporting, fair, respectful, and having high expectations, the students show an increased self-efficacy, se lf-regulation and academic achievement. A quasi-experimental study of the Seattle Soci al Development Project involving 808 youth showed that bonding during the middle and hi gh school years was ne gatively associated with substance abuse, delinquency, gang i nvolvement, violence, and academic problems (Catalano, Haggerty, Oesterle, Fleming, & Hawkins, 2004). School bonding was also related to lower rates of alcohol consum ption and smoking initiation. This study also revealed that students who were bonded to schoo l by fifth and sixth grade were less likely to become offenders in seventh grade and t hose attached to school in grade seven were more likely to refrain from delinquent be havior through ninth gr ade. This negative association held true for student s with elevated risk factors as well. Another analysis of this sample of students found th at adolescents bonded to school in ninth grade engaged in less violent behavior through ag e 18 while students less bonde d in eighth and tenth grade were twice as likely to engage in violent be havior in twelfth grad e (Catalano, et al.). A final finding of this evaluation was that an increase in schoo l bonding between seventh and twelfth grade correlated positively with grade point average and negatively with school misbehavior. The Raising Healthy Ch ildren Project also investigated the relationship of school bonding with problem behaviors and academic achievement (Catalano, et al.). Findings showed that school bonding in grades three and four was
70 negatively associated with problem behavior The level of bonding in third grade was also positively associated with academic test scores in seventh grade. In support of the theory that school bonding serves as a protectiv e factor, this same evaluation showed that school bonding had a stronger protective eff ect for children whose parents reported involvement in antisocial behaviors. Although connecting students to school is important at all grade levels, its especially crucial during the a dolescent years. Over the past decade, educators and school health professionals have increasingly point ed to school connecte dness as an important factor in reducing the likelihood that adoles cents will engage in health compromising behaviors (Blum, 2005). Bullyi ng and other violent acts ar e less likely to occur in a school that feels like a caring community, a pl ace where children experience a sense of connection to one another and to adults, a place where they feel a sense of belonging (Kohn, 2004). Students in this type of environm ent are less likely to skip school, or be involved in fighting, bullying or vandalism (Blum). Students who feel connected to school are also less likely to use substances exhibit emotional di stress, demonstrate violent or deviant behavior, experience suic idal thoughts, or attempt suicide (Blum). Confirming the importance of belonging, a na tional study of school dropouts found that a consistent characteristic of high school stude nts who drop out of school is alienation from school life (Ecstrom, et al., 1986). In addition, the National Longitudinal Study on Adolescent Health survey of 90,000 middle and high school students found that a feeling of connectedness was the number one protective factor against suicid al behavior (Doan, et al., 2003). Survey results also support previous research fi ndings that students who feel
71 a sense of connectedness with school are less likely to dri nk alcohol, carry weapons, or engage in other delinquent behaviors. According to the Wingspread Declarati on, by increasing the number of students connected to school, educator s will see a positive change in academic performance, incidents of fighting, bullying and vandalism, absenteeism, and school completion rates (University of Minnesota, 2003). Research indicates students with a high degree of school connectedness exhibit a consistent positive developmenta l pattern including improved academic achievement, reduced delinquency rates and decreased rates of health-risk behaviors (Wilson, 2004). After the school shooting at Columbine High School, the Colorado professional education asso ciations created a safe school model that focused on creating an overall school climat e where students feel safe, valued, and supported in their learning and subsequen tly implemented the model in 32 schools. Evaluation of the initiative showed that schoo l connectedness was negatively related to physical aggression and demonstrated the strongest pr edictive ability, b = -.344, p =.000 for aggression (Wilson). The analysis also show ed that school climate is inversely related to relational aggression, b = -.181, p = .003 and that as connectedness increased, relational aggression decreased, b = -.600, p = .000. Evaluation of this model also provided evidence that regardless of climat e, strong connectedness to school provides protective effects. In school s with a negative climate, 39% of low connected students demonstrated high levels of physical aggres sion and 56% demonstrated high levels of relational aggression whereas only 17% of their highl y connected classmates demonstrated physical aggression and 46% demonstrated relational aggression. Highly connected students were also less like ly to experience victimization.
72 A students sense of belonging to school is highly associated with student outcomes including academic achievement and i nvolvement with a range of health risk behaviors. Youth do better when they feel conne cted to school, feel that they belong, and that teachers are supportive (Li bbey, 2004). In contrast, years of research consistently list no connection and no support as reasons student s drop out of school (Abourjilie, 2006). As early as 1969 Glasser emphasized the role of warm teacher involvement and its relationship to school success (Niebuhr, 1999). Mo re recently this has been aligned with the ideas of Demings concept of quality management explaining that part of the necessary quality conditions in the classroom is that as teachers allow students to know them, and hopefully like them, the students will work harder, t hus increasing their opportunities for success (Niebuhr). To further support the significance of relationships, science is increasingly demonstrating that human beings are hardwired for close attachments beginning with parents, families, and extending out to the community. These nurturing environment or lack of them affect gene transcription and the development of brain circuitry (Commission on Children at Risk, 2003). All adolescen ts need to achieve a minimum amount of connectedness but not all are able to establish sufficient connectedness within conventiona l contexts such as family and school. Youth who have an imbalance, more unconventional than conve ntional, in connecte dness are at risk for engaging in violent beha viors (Gerler, 2004). According to data from the National L ongitudinal Study of Adolescent Health, junior and senior hi gh school students found that school connectedness, defined as a feeling of being a part of and cared for at sc hool, is a key to reducing risk for engaging in violent behavior, substance abuse, and suicide (Blum, McNeely, & Rinehart, 2002).
73 Students who feel close to others, fairly treat ed and vested in school are less likely to engage in risky behaviors than those who do not (Resnick, et al., 1997). School connectedness also emerged as an important f actor in research c onducted with 304 school employees (Hunt, et al., 2002). Educators were as ked to identify and rate ways to help students feel connected to school. Resu lts showed that teacher attitude ( M = 4.80, SD = .50), students feel someone cares for them ( M = 4.79, SD = .50), students feel safe at school ( C = 4.69, SD = .57), student has an adult at sc hool the can go to with a problem ( M = 4.66, SD = .61) and a friendly, positive climate ( M = 4.60, SD = .64) were perceived as important factors in cr eating a students sense of connectedness to school. In comparison, researchers have found a direct relationship between school disconnectedness and outcomes such as deli nquency, truancy, drug use, and a number of physical and mental health indicators (Smith, 2004). As self worth within a community incr eases, social coopera tion within that community also increases (M orrison, 2002).When a student id entifies with the school community he or she develops a sense of pr ide and will behave cooperatively to uphold the school rules and values. Comparatively, a lack of cooperation has been correlated with high involvement in school bullying (Morrison, 2001). Supporting the link with antisocial behaviors, a retr ospective study involving youth ag es 15 to 18 incarcerated for murder, found the existence of several comm on elements among the youth: inability to talk with anyone about feelings, loss of c onnection to school and the feeling that no one wanted them in the school anyway (Pharris, 2002). In addition to behavioral outcomes, c onnectedness appears essential to many educational processes and schooling outcome s. Multiple regression analysis of 612
74 students revealed that sense of belonging in a class was relate d to students expectations of academic success, intrinsic interest in academic work, course grades, and teachers rating of students academic effort (Ma, 2003) Multiple regression analysis of a survey of 301 multiethnic students designed to ex amine the correlation between sense of belonging and measures of motiv ation and achievement showed that a students sense of belonging has a statistically significant impact on motiva tion as well as engaged and persistent effort in difficult academic work (Ma). Research conducted with 2,169 Mexican American students identified as eith er high achievers or low achievers showed that a sense of belonging to school was the only statistically si gnificant predictor of student academic grades (Ma). An exploratory study conducted by Thorpe (2004) examined the hierarchical nature of school connectedness and mathematic s proficiency. Using gender and ethnicity as covariates, the researcher hypothesized that connectedness to school would be a predictor of achievement. In support of the hypothesis, the re sults showed that gender, ethnicity, and school connectedness significan tly predicted math proficiency with no variability between schools for gender and et hnicity. The effects of connectedness varied across schools with 21% of the variance in math scores explained by between school variability. Percentage of minority students, = -.11, p < .001 and mobility rate, = -.76, p < .0001 predicted math proficiency across schools. Percent of minority students, = -.01, p < .01 and mobility rate, = -.45, p < .05 also predicted school connectedness. Schools with high mobility rates and a higher pe rcentage of minority students had lower scores on of math proficiency and fewer st udents who felt connected to the school.
75 Friendship and a sense of belonging to a school community are important in that there is a positive relationship between academic performance and psychological adjustment (Kent, 2003). Research indicates that there is a significant relationship between overall support and being victimi zed; social support is related to positive outcomes for students whereas the lack of soci al support is related to negative outcomes (Malecki, 2003). Positive peer relationships can be associated with reductions in rates of peer victimization and may even serve a pr otective function agai nst future negative outcomes (Hodge, 2003; Storch, et al., 2003). When students feel a strong sense of connectedness to school they are less likely to engage in violent behaviors or tolerate them among peers (Halford, 205). Thorpe (2003) investigated school-initiated connectedness and peer-initiated connectedne ss as they relate to student-initiated connectedness and academic achievement in 1,758 seventh grade students and found school initiated connectedness ha s an indirect effect on stud ent achievement through peer desire to be connected to one another and individual choice to be connected to school. A sense of school connectedness allows a dolescents to identify with those who are different from them. In contrast, student s with a weaker sense of connectedness begin to feel distant from others leading to a se nse of social isolati on (Thorpe, 2003). Social isolation is also a risk fact or and a consequence of vic timization. Students involved in school bullying were significantly less likely to reflect high levels of school adjustment or bonding and victimized children tend to beco me more school avoidant after they are victimized by peers (Rigby, 2003; Telljohann, 2003). Similar to the consequences suffered by students who are bullied: lower se lf esteem and higher levels of stress, anxiety, depression, illness, and suicidal ideation (Olweus, 1993), students who cannot
76 easily participate in social activities with peers due to restriction on activity or communication difficulties may begin to iden tify themselves as helpless individuals (Kent, 2003). For example, the pe rception of being left out or undesirable is characteristic of the social relationships of hard of heari ng students putting them at risk of alienation leading to adverse outcomes including social cognitive processing and social maladaption (Kent). Kent found that students who self id entified as having a hearing disability reported statistically significan t levels of feeling lonely and a correlation between self identification and reports of being bullied, r = .273, p = .050. In order for students to feel connected to school, they must feel accepted. This can be a unique challenge for minority student s. Minority students may be connected differently in a school with a majority of wh ite students than they would be in a school with a majority of minority students (Thorpe, 2003). This may be intensified for black male students. Researchers inve stigating the su ccess of black males in schools cite a lack of commitment to create a culture of care a nd nurturance for black boys as a systematic problem in schools (Varlas, 2005). Research in this area also states that the perceived lack of caring is the most devastating factor for black youth and that the single most important thing in turning lives around is the ongoing presence of a caring adult (Varlas). Developing friendships and relationships with adults appears to be a key component to creating a sense of school connectedness for all students. Trusting relationships between adults and students are the pr oduct of quality connection, interaction, and communications. This sense of connectedness to a dults has protective effects for both boys and girls across ethnic, racial, and so cial class groups (Commission on Children at Risk, 2003). Students are more likel y to feel connected if they believe they
77 are being treated fairly, feel safe, and be lieve that teachers are supportive (Blum, 2005; Doan, et al., 2003). Developing a classroom characterized by a general physical and emotional well being, based on mutual resp ect and trust is a continuous process. Relationships evolve and do not develop simply because an adult and student have been assigned to interact with one another (Fein, et al., 2002). Despite this knowledge, with the current emphasis in education on standards and accountability, time dedicated to building rela tionships and a sense of connectedness may seem like a luxury. Educators are held accountab le for student success without regard for the personal and social conditi ons that affect students (M endler, 2001). However, school connectedness can have a substantial impact on the measures of student achievement for which schools are currently being held acc ountable (Blum, 2005). A review of the research shows that a connected school e nvironment increases th e likelihood of academic success. Summary Bullying is a complex phenomenon with long and short term consequences for both the child who bullies and the victim of bullying behavior. Students who do not feel connected to their school experi ence similar risks to those wh o are victimized by peers. These difficulties can be compounded for stude nts with special need s. Bullying effects the climate of a school by inte rfering with student learning and creating an environment of disrespect and fear. An investigation of th e factors related to bully ing reveals that peer support and a positive school climate are vita l for reducing victimization. School culture research supports the fact that if students ar e well connected to thei r schools and teachers
78 in a supportive environment, they're more likely to do well academically and stay in school (McGlynn, 2004). Research also suggests that adult relationships are integral to a positive school climate but does not address the impact that connections with adults may have in preventing bullying behaviors and vic timization at school. In addition, research has investigated the impact of school connectedness on achievement separately from the impact it has on bullying behavior. There is a gap in the literature as to the interrelationship of these factors.
79 Chapter Three Methodology This chapter describes the method used to address the research questions developed to examine the relationship between school connect edness and bullying victimization. The chapter is presented in eigh t sections. The first two sections provide an overview of the problem and purpose of th e study. The next section describes the National Crime Victimization Survey School Crime Supplement where the research data was drawn from. The fourth section outlines the sampling procedures used for the School Crime Supplement. This is followed by key definitions used in the survey and a description of the variables th at were used in the study. Th e next section describes the analytic procedures that were used to answ er each research question and the final section details the limitations of the study. The ch apter concludes with a summary of the methodology. Problem Bullying and victimization in school can create an environment that interferes with student learning (Fein, et al., 2002); whereas a school climate characterized by cohesiveness, satisfaction, openness, and a high degree of trust allows for higher student achievement (Fraser, 1999; Hoy & Sabo, 1998) Bullying prevention research highlights
80 other negative outcomes of bullying and vict imization including an increased risk of mental health disorders and antisocial behavior that can also negatively impact academic achievement (Hawker & Boulton, 2000; Mi shna, 2003; Olweus, 1993; Telljohann, 2003). In contrast, a students sense of belongi ng to school can have a positive impact on academic performance and social emotional growth (Kent, 2003). Although educational researchers concur that learning can only ta ke place when a student feels physically and emotional safe, emphasis continues to be on accountability and high stakes testing ignoring the role the school plays in educating the whole child. It is essential that educators consider both academic and social emotional aspects of learning to create environments conducive to learning. Purpose The study focused on the relationship betw een bullying victimization and school connectedness. Considering previous bullying research as well as risk and protective factor research, the study examined whethe r the presence of school connectedness serves as a protective factor diminishing bullying vi ctimization. Gender, ra ce, grade level, and academic achievement were considered as moderating factors. The role that adult to student and peer to peer relationships play in the bullying phenomenon was also investigated. The study looked at relationshi p differences by investigating whether a students level of school connectedness pr edicted bullying victimization. Finally, it identified risk and protective factors that may a llow educators to target students at risk of victimization for proactive intervention as well as indicating prevention efforts at the school level. The earlier the bully/victim pattern can be broken, the less negative the
81 effects will be. The results of this study can provide inform ation to guide educators as they develop and modify bullying prevention pr ograms to meet the needs of all children. Research Questions The research questions addr essed in this study were: 1. What is the relationship between bullying victimization and school connectedness? a. Is the frequency of bullying victim ization related to the level of connectedness a student has with school? b. Does the level of connectedness of students who report no experience of bullying victimization differ from the level of connectedness of students who report bullying victimization? c. Is the relationship between occurren ces of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? d. Is the relationship between freque ncy of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? 2. Is bullying victimization moderated by th e strength of adult-student relationships that a student develops? 3. Is bullying victimization moderated by th e strength of peer relationships that a student develops? 4. Does the impact of adult-student rela tionships on the frequency of bullying
82 victimization differ from the impact of peer relationships on the frequency of bullying victimization? Data Source The research is a secondary analysis of selected data from the 2005 National Crime Victimization Survey School Crime S upplement. Nationally representative data sets have become increasingly available, making secondary data analysis of large, complex sample surveys more common among so cial and behavioral scientists (BellEllison & Kromrey, 2007). According to Jacobson, Hamilton, and Galloway (1993), secondary analysis of existing data from larg e-scale studies can be reanalyzed from a different perspective, thus enhancing the original studys contribution to scientific knowledge. Secondary data analysis also has limita tions in that the data source is rarely a perfect match for the secondary analysis due to the fact that it was or iginally collected for a different purpose. However, the Cens us Bureau conducted the National Crime Victimization Survey for the purpose of providi ng statistical information about the nature and extent of crime throughout the United States (U.S. Department of Justice, 2006), thus allowing for a more direct matching of data in the study. The National Crime Victimization Survey is the nations primary source of information on crime and victimization (Di nkes, Cataldi, Kena, & Baum, 2006). It is administered annually for the U.S. Bureau of Justice Statistics by the U.S. Census Bureau to collect information on the frequency and nature of crimes experienced throughout the United States. In addition to the information collected on the regular survey, additional surveys are periodically administered in or der to obtain specific information. The School
83 Crime Supplement was created as a supplement to the National Crime Victimization Survey to collect additional information a bout school related vict imization on a national level. The survey was designed by the National Center for Educational Statistics and the Bureau of Justice Statistics to assist po licymakers and academic researchers in making informed decisions concerning crime at schoo l (Dinkes, et al.) and is considered a primary source of national level data on cr ime victimization in schools throughout the United States (U.S. Department. of Justice, 2006). In addition to e xploring the frequency and nature of crime at school, questions are directed at preventive measures employed by schools; students' participati on in extracurricular activitie s; perception and enforcement of school rules; the presence of weapons; dr ugs, alcohol, and gangs in schools; student bullying; hate-related incidents; and attit udinal questions relati ng to the fear of victimization at school. Demographic characte ristics are also provi ded (U.S. Department. of Justice). The School Crime Supplement asks students to self-report incidents of crime and victimization at school as we ll as perceptions and attitude s about school. Self-reports can be an efficient means of getting informati on about behavior, but efficiency must be balanced with accuracy (Young, et al., 2006). Su rvey research literature addresses the extent that respondents answer honestly. Panel analysis of Monitoring the Future data showed a high degree of stability in the self report of drug use (Johnston, OMalley, Bachman, & Shulenberg, 2006). Additionally, infe rential evidence from the survey data suggested there was little under reporting by skipping que stions, over reporting was minimal, and anonymity made little difference in student self-reports of substance use. In a research study investigating whether adolescen ts told the truth on a survey about sexual
84 behavior, 83% of respondents reported they were honest in reporting, leadi ng researchers to the conclusion that the data added to th e predictive capacity of models of behavior (Newcomer & Udry, 1988). Gold (1977) revi ewed the literature on self reported delinquent behavior of adolescents and found th at self report was the best single measure of delinquent behavior and concluded it is accurate enough for use in rigorous research designs. Although non response and misreporting can be a problem with sensitive questions, a strategy to overcome this tendenc y is to assure conf identiality of answers (Rasinski, Visserm Zagatsky, & Rickett, 2004 ). Additionally, if the focus is on large groups, the effect of inaccurate responses is usually very sm all compared with the total sample (Fan, et al., 2006). The School Crime Supplement is a confiden tial survey administered in person and via phone by an interviewer who reads the questions and records the participants responses. The administration guidelines al so allow for proxy responses in which one household member answers questions for anothe r. However, bullying research shows that adult perception of the frequency of bullying behavior varies from student perception. Pepler and Craig (2000) reporte d that 71% of teachers say th ey almost always intervene in bullying situations whereas only 25% of students report that teachers almost always intervene. Additionally, their observations s howed that teachers intervened in 14% of bullying incidents in the classroom and onl y 4% of episodes on the playground. Parents are also often unaware of bullying problems. Pepler and Craig repor ted that only 48% of students who bully and 62% of victims indicated that they talked to their parents about the bullying problem. Due to this inconsistenc y between adult and st udent perceptions of bullying, proxy interviews were excluded from this study sample.
85 The School Crime Supplement survey was conducted in 1989, 1995, 1999, 2001, 2003 and 2005. In 2005, the question related to bully ing was modified to include a series of questions rather than the single questi on asked on previous versions. Due to this substantive change in wording, only the 2005 data were used in this study. The 2005 School Crime Supplement consisted of eight se ctions: screening; environmental; fighting, bullying and hate behaviors; av oidance; fear; weapons; gangs; a nd student characteristics. 1. Screening questions were asked to determ ine if the respondent was eligible to participate in the survey. 2. Environmental questions gathered inform ation on the type of school the student attended, measures taken by the school to ensure student safety, and the availability of drugs and alcohol in school. 3. Fighting, bullying, and hate behavior questi ons identified the nature and extent of these crimes at school. 4. Avoidance questions identified the eff ects of fear of crime on behavior. 5. Fear questions pertained to how often a st udent feared an attack or being harmed at school. 6. Weapons questions were designed to dete rmine if students brought weapons to school for protection. 7. Gang questions related to th e presence of organized ga ngs at school and whether students came into contact with gangs or gang members at school. 8. Student characteristic questions asked a bout the respondent's attendance, grades, and plans regarding college.
86 Sampling Procedures Households were randomly selected to participate in the National Crime Victimization Survey with all age-eligible individuals becoming part of the sample. Respondents were interviewed every six months for a total of seve n interviews over a three-year period. The first interview was f ace-to-face; the rest were by telephone (U.S. Department of Justice, 2006). The School Crime Supplement was administered at all National Crime Victimization Survey househol ds interviewed from January through June 2005. For both surveys, the methods of data coll ection included interviewing in person or by telephone with the responses being entered on a paper instrument and computer assisted telephone interviewing using an auto mated version of the paper instrument to administer the questions and record the re spondent's answers. Each interview took an average of 10 minutes to complete. The National Crime Victimization Survey School Crime Supplement employed a stratified multistage cluster sample of households with children between the ages of 12 and 18 who had attended school at any time during the six months prior to the month of the interview, and who were enrolled in a school that would adva nce them toward the eventual receipt of a high school diploma (U.S. Department of Justice, 2006). Students who were home schooled were not included in the sample. The survey was conducted during a six month period from January to June 2005 in all households selected for participation in the National Crime Victim ization Survey. Eligible respondents were asked the supplemental questions on the Sc hool Crime Supplement only after completing the National Crime Victimization Survey.
87 Of the 11,525 National Crime Victimizat ion Survey responde nts eligible, 7,112 (61.7%) students participated in the 2005 administration of the School Crime Supplement. The remaining 38.3% were non-interviews (U.S. Department of Justice, 2006) The overall unweighted School Crime Supplement completion rate was 56% (Dinkes, et al., 2006) and the item res ponse rate for most items was 95% (U.S. Department of Justice). Because intervie ws with students were only completed after households responded to the National Crim e Victimization Survey, the overall completion rate reflects both the household a nd student interview co mpletion rates. The data from the School Crime Supplement is weighted to represent the population from which the sample was drawn. Although the or iginal School Crime Supplement survey sample size was 7,112, after removing proxy interviews and limiting the sample to those with complete data on all variables of inte rest, the sample for this study included 5,780 respondents. Definitions School For the purposes of the National Crim e Victimization Survey School Crime Supplement, school was defined as any instit ution designed to advance a person toward a high school diploma (U.S. Departments. of Justice, 2006). The definition of school included in the school building, on school pr operty, or on the way to or from school. Students schooled at home were not consider ed as attending school in administration of this survey.
88 Bullying Bullying is a form of aggression in which there is an imbalance of power between the bully and victim (Pepler & Craig, 2000). A st udent is being bullied when he or she is exposed, repeatedly and over time to negative actions on the part of one or more students (Olweus, 1993). Bullying can be verbal, physical or psychological in nature and can be done directly or indirectly. Direct bullying involves face to face aggression while indirect bullying traditionally involves more covert behaviors such as exclusion and gossip (Pepler & Craig). With the explosion of t echnology, indirect bully ing has expanded to cyber space. Cyberbullying is a psychological form of soci al cruelty conveyed through electronic media (Shariff, 2005) such as cel l phones, web sites, chat rooms, text messages, e-mails, and web logs. During administration of the National Cr ime Victimization Survey School Crime Supplement, respondents were provided an expl anation of what being bullied meant prior to being asked questions about bullying vict imization. For the purpose of the study it was defined as what students do at school that make you feel bad or are hurtful to you and included questions about teasi ng, rumors, threats, physical attacks, coercion, exclusion, and destruction of property. Cyberbullyi ng was not included in the 2005 survey. Academic Achievement Academic achievement is defined as adequate progress towards meeting state content and performance standards. In th is study, academic achievement was measured by students self-reported grades.According to Dornbusch, et al. (1990), educators consensus is that grades are the most a ppropriate measure of a students current
89 performance. They found self-reported gr ades were a close approximation to school transcripts ( r = .78) with only a slight inflation of gr ades at the lower end of the spectrum. School Connectedness School connectedness refers to a students relationship to school and is defined as a component of school climate that creates a feeling of belonging to the school, and being accepted by others (Blum, 2005; Blum & Libbey, 2005; Nakammura, 2000; Perlstein, 2004). It represents the sense of attachment and commitme nt a student feels towards school and includes students experiences of car ing at school and a se nse of closeness to school personnel and environment (Smit h, 2004). School connectedness includes the degree to which students feel close to people at school, are happy to be at school, and feel like a part of the school (Libbey, 2004). E ducational research offers several common components of school connectedness incl uding academic engagement and high expectations; a physically and emotionally sa fe environment; feeli ng respected; feelings that adults care about students; having a s upportive relationship with at least one person; opportunities to participate in extracurricular activities; presence of friends; discipline; fairness; and liking school (Blum; Blum & Libbey; Nakammura; Perlstein). Key Variables The dependent variables in this study we re the occurrence and the frequency of bullying victimization as measured by the School Crime Supplemen t self-report survey. The independent variables in this stu dy were components of school connectedness determined through factor analysis including se nse of emotional safety, relationships, and
90 involvement in extracurricular activities. Addi tional independent variables were strength of adult-student relationships, strength of p eer relationships, gender, race, grade level, and self-reported academic achievement. Schoo l safety indicators were included in the school connectedness factor analys is but not in the final regr ession analysis. In order to answer the research questions examining th e relationship between bullying victimization and school connectedness, selected survey questions representing school connectedness and bullying victimization were used. Data fr om these questions were coded numerically for ease of analysis. Due to their comparability to school connectedness components outlined in educational research, variables from several survey questions were selected to include in factor analysis of school connectedness vari ables. The first variable corresponds to participation in extracurricular activities a nd was measured by the question During the last six months, have you participated in any of the following extra-curricular activities at school? a. Athletic teams at school? b. Spirit groups, for example, cheerleading or Pep Club? c. Performing arts, for example, band, orchestra, or drama? d. Academic clubs, for example, debate team, honor society, Spanish club or math club? e. School government? f. Service clubs, for example, Key Club or other service oriented groups? g. Other school clubs or school activities?
91 The second and third variables represent th e students sense of physical safety and were measured by two questions: Does your sc hool take any measures to make sure students are safe? For exam ple, does the school have: a. Security guards or assigned police officers? b. Other school staff or other adults supervising the hallway? i. A code of student conduct, that is, a set of written rules or guidelines that the school provides you? During the last six months, did you stay at home from school because you thought someone might attack or harm you at school, or going to or from school? The fourth variable measured the factors related to the students emotional safety and bonding with the school and were measured by the questions: During the last four weeks, did you skip any classes? Thinking about your school over the last six months, would you strongly agree, agree, disagree, or str ongly disagree with the following: b. The school rules are fair. c. The punishment for breaking school ru les is the same no matter who you are Thinking about the teachers at your school, during the last six months, would you strongly agree, agree, disagree, or strongly disagree with the following: a. Teachers treat students with respect b. Teachers care about students Thinking about the teachers at your school, during the last six months, would you strongly agree, agree, disagree, or strongly disagree with the following:
92 a. At school, there is an adult I can talk to who cares a bout my feelings and what happens to me. b. At school, there is an adult who helps me with practical problems, who gives good suggestions and advi ce about my problems Thinking about friends at your school, during the last si x months, would you strongly agree, agree, disagree, or str ongly disagree with the following: a. At school, I have a friend I can talk to who cares about my feelings and what happens to me. b. At school, I have a friend who helps me with practical pr oblems, who gives good suggestions and advice about my problems Factor analysis is a statistical approach us ed to analyze the inte rrelationships of a large number of variables for th e purpose of condensing them in to smaller sets of factors with a minimum loss of information (Heir, et al., 1992). According to Heir, et al., factor analysis can be used to verify a conceptualiz ation of a construct of interest. In order to find the least number of fact ors that account for the comm on variance in the variables identified to measure school connectedne ss and to validate the school connectedness model, a principal factor analysis was done to determine the school connectedness construct variables. Analysis resulted in the creation of two construct variables: emotional safety and relationships. The com ponents included in the construct variables are outlined in Table 6. As a result of the fact or analysis, extra curr icular activities were recoded to be measured as participation or nonparticipation and questions related to physical safety were not considered as part of school connectedness. The question related to skipping classes was also eliminated as a result of the factor analysis.
93 Table 6 Components of School Connectedness Construct Variables Emotional safety Teachers care about students Teachers treat students with respect School rules are fair Punishment for breaking the rules is the same Relationships Have friend at school w ho helps with problem Have friend at school to talk to There is adult at school who cares about me Have adult who helps with problems Although questions relating to adult-student relationships and peer relationships were included in the school connectedness factor analysis, the questions were also analyzed separately to answer research questions 2.) Is bul lying victimization mediated by the strength of adult-student relationships that a student deve lops? 3.) Is bullying victimization mediated by the stre ngth of peer relationships th at a student develops? and 4.) Does the impact of adult-student re lationships on the fr equency of bullying victimization differ from the impact of peer relationships on the fre quency victimization? Bullying victimization encompasses the st udents experience of being bullied and has a twofold purpose: determining if bullyi ng victimization occurr ed and if so, the frequency of the behavior. The bullying vi ctimization variable was derived from questions regarding whether bullying occurr ed and how frequently it happened. More
94 specifically, the first question addressed th e occurrence by asking D uring the last six months, has any other student bullied you? That is, has another student Made fun of you, called you names or insulted you? Spread rumors about you? Threatened you with harm? Pushed you, shoved you, tripped you or spit on you? Tried to make you do things you did not want to do, for example, give them money or other things? Excluded you from activities on purpose? Destroyed your property on purpose? An affirmative response to any segment of this question put the respondent in the category of having experienced bullying victim ization while a negative response to all segments placed the respondent in the category of no experience of bullying victimization. Students who responded affirmatively to occurrence question were then questioned about the frequency of the behavi or During the last six months, how often did this happen to you? once or twice in the last six months once or twice a month once or twice a week almost every day Responses to this question were numerically coded to obtain the variable for the frequency of bullying victimization. Responde nts who skipped this question due to a
95 negative response on question 19a were code d (0). Students res ponding dont know to this question were excluded from frequency calculations. In addition to the bullying victimizat ion variables and school connectedness construct variable, individual responses to questions were included in the analysis to investigate moderating factors. Grade level was determined using data from the question During the last 6 months what grade were you in school? Possibl e responses included fifth or under, sixth, seventh, eighth, ninth, tenth, elevent h, twelfth, or college/GED. Responses within the range of sixth through twelfth were used in the study. The academic achievement variable was taken from the question: Duri ng this school year, across all subjects, ha ve you gotten mostly As Bs Cs Ds Fs The race and gender variables were taken from the race/ethnicity and sex variables collected in the National Crime Victimiza tion Survey and appended into the School Crime Supplement data file. Operational definitions and coding for each variable included in the study are provided in Table 7.
96 Table 7 Study Variables and Operational Definitions Variable Operational definition Dependent variables: bullying victimization Occurrence of bullying victimization Experienced any of the following within the previous six months: made fun of, calle d names, or insulted; spread rumors about student; threatened with harm; pushed, shoved, tripped or spit on; tried to make do things that student did not want to do; excluded from activities on purpose; had prope rty destroyed on purpose No (0), Yes (1) Frequency of bullying victimization How often bullying happened with in the last six months: once or twice (1), once or twice a month (2), once or twice a week (3), al most every day (4), never (0) Independent variables: school connectedness Emotional safety Level of agreement with: The school rules are fair The punishment for breaking school rules is same Teachers treat students with respect Teachers care about students Mean of likert scale respons es: Strongly agree (4), agree (3), disagree (2), strongly disagree (1)
97 Table 7 (continued). Study Variable s and Operational Definitions Variable Operational definition Relationships Extracurricular Involvement Level of agreement with: At school, there is an adult I can talk to who cares my feelings and what happens to me. At school, there is an adult who helps me w ith practical problems, who gives good suggestions and advice about my problems. At school, I have a friend I can talk to who cares about my feelings and what happens to me At school, I have a friend who helps me with practical problems, who gives good suggestions and advice about my problems. Mean of likert scale respons es: Strongly agree (4), agree (3), disagree (2), strongly disagree (1) Participation in: Athletic teams Spirit groups Performing arts Academic clubs School government
98 Table 7 (continued). Study Variab les and Operational Definitions Variable Operational definition Service clubs Other school cl ubs or school activities Participation in any: No (0), Yes (1) Moderating Factors Race/Ethnicity Gender Grade White: No (0), Yes (1) Black: No (0), Yes (1) American Indian, Alaskan Native: No (0), Yes (1) Asian: No (0), Yes (1) Hawaiian/Pacific Islander: No (0), Yes (1) Hispanic: No (0), Yes (1) Multi: No (0), Yes (1) Male (1), Female (2) Numeric grade level Six: No (0), Yes (1) Seven: No (0), Yes (1) Eight: No (0), Yes (1) Nine: No (0), Yes (1) Ten: No (0), Yes (1)
99 Table 7 (continued). Study Variab les and Operational Definitions Variable Operational definition Achievement Level Strength of Adult Student Relationships Strength of peer relationships Eleven No (0), Yes (1) Twelve: No (0), Yes (1) Most frequent grade over the past year A (5), B (4), C (3), D (2), F (1), no grades given (0) Presence of trusted adult at school There is an adult I can talk to who cares my feelings and what happens to me There is an adult who helps me with practical problems, gives good suggestions, and advice about my problems Presence of trusted friend at school I have a friend I can talk to who cares about my feelings and what happens to me I have a friend who helps me with practical problems, gives good suggestions and advice about my problems Mean of Likert scale res ponses: strongly agree (4), agree (3), disagree (2) or strongly disagree (1)
100 Data Analysis This study was a secondary analysis of data collected in the 2005 administration of the School Crime Supplement of the Nati onal Crime Victimization Survey. Measures of school connectedness, bullying victimizati on, and student characte ristics from selfreported cases were used. Quantitative data was analyzed using SAS software. Unit and item response rates in surveys are rarely close to 100%, and non response rates are commonly assumed to be ignorable (Copas & Farewell, 1998). However when the topic of interest is sensitive, informants propensity to res pond may result in non-ignorable non-response (Copas & Farewell). Although bully ing victimization can be a sensitive topic for those enduring bullying behavior, standard techniques for regression models require full covariate information, with the most commonly used technique for handling missing data being complete case analysis (Ibrahim, Chen, Lipsitz, Stuart, & Herring, 2005). Therefore, only respondents with complete data for all variables of interest were included in the study and a discussion of non response is included in chapter five. Principal factor analysis was done to confirm the commonalities among the variables included in the school connectedness constructs and reliability estimates were determined. Weighted descriptive statistics including means and st andard errors were computed for each individual variable and cons truct. In order to answer the research questions, a weighted multivariate analysis was done to determine the relationship between bullying behaviors and school connect edness and to identify the combination of factors that influence the relationship. Anal ysis of data from complex sample surveys requires the use of software that incorporat es sample weights and accurate estimates of variance (Bell-Ellison & Kromrey, 2007). Therefore, to a ccount for the complex nature
101 of the sampling procedures employed, SAS SURV EY procedures were used in analysis. Table 8 lists the research questions, variables, and method of analysis. Table 8 Analysis to Answer Research Questions Research question Variable Analysis 1.What is the relationship between b ullying victimization and school connectedness? 1a. Is the frequency of bullying victimization related to the level of connectedness a student has with school? 1b. Does the level of connectedness of students who report no experience of bullying victimization differ from the level of connectedness of students who report bullying victimization? 1c. Is the relationship between occurrences of bullying victimization and Level of connectedness Occurrence of victimization Level of connectedness Frequency of victimization Level of connectedness for those reporting victimization Level of connectedness for those reporting no victimization Occurrence of victimization Level of connectedness Occurrence of victimization Logistic regression Multiple regression Logistic regression Logistic regression
102 Table 8 (continued). Analysis to Answer Research Questions Research question Variable Analysis connectedness impacted by gender, race, grade level, or academic achievement levels? 1d. Is the relationship between frequency of victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? 2. Is bullying victim ization moderated by the strength of adultstudent relationships that a student develops? 3. Is bullying victim ization moderated by the strength of peer relationships that a student develops? 4. Does the impact of adult-student relationships differ from the impact of peer relationships on the frequency bullying victimization Gender, race, grade level Academic achievement Frequency of connectedness Frequency of victimization Gender, race, grade level Academic achievement Frequency of victimization Adult relationship strength Frequency of victimization Peer relationship strength Adult relationship strength Peer relationship strength Multiple regression Multiple regression Multiple regression Multiple regression
103 Research Question 1 The initial research question explored th e extent to which school connectedness relates to bullying victimization. A principal factor analysis examined the relationship between the individual variab les included in the school conne ctedness constructs prior to regression analysis. A weighted logistic regression analysis was conducted to determine the degree of the relationship between the tw o variables and to evaluate how well the level of school connectedness predicts bully ing victimization in general. A regression analysis was also conducted to examine th e level of connectedness of non victims in comparison to those experiencing bullying victimization. The effects of specific components of school connectedness were consid ered in the regression analyses as well. The first research question also looks at mode rating factors in the relationship. To answer this segment of the question, logistic regres sion equations were deve loped that included interaction terms for race, gender, academic achievement, and grade level. These regression equations were used to identify if the relationship between a students level of connectedness with school and bullying vi ctimization can be accounted for by the students race, gender, academic achievement, or grade level. Research Questions 2, 3, and 4 The remaining three research questions ex amined the impact of relationships on bullying victimization. Multiple regression analyses were done to examine the relationship between adult-student relati onships, peer relationships, and bullying victimization while controlli ng for race, gender, academic achievement, and grade level. The assumptions required to do multiple regression analysis were assessed.
104 Limitations There are several limitations inherent in the research design. Although the sample is nationally representative, it is not a random sample. The sample selected is just one of many possible samples that could have been selected resulting in sampling error variability (DeVoe & Kaffenberger, 2005). B ecause the School Crime Supplement is a sample survey, non-response bias can affect the strength and app lication of data. The sampling frame has four student or school ch aracteristic variables for which data is known for respondents and non-respondents: sex, race, household income, and urbanicity. To the extent that there were differential responses w ithin these groups, nonresponse bias would be a con cern (Dinkes, et al., 2006). Another potential limitation is instrumentation. The School Crime Supplement is a self-report survey used to gather data on crime victimization at school allowing participants to express bias in their responses. Additiona lly, the survey was read to respondents so students did not have the opportunity to respo nd anonymously. Several other limitations of victimization surveys ma y also impact the estimates of bullying reported in the School Crime Supplement. The survey asked about incidents of bullying that occurred during the last six months. Respondents may have reported on victimization that occurred outside that refe rence period artificially inflatin g the report. The survey also listed bullying incidents as single points in ti me. However, because victims often live in a state of victimization where they are regular ly threatened (DeVoe & Kaffenberger, 2005), the report may be artificially deflated. Additionally, respondents recall of bullying episodes may have been inaccurate leading to an underestimate of victimization.
105 Conclusions reached in this study are delim ited to the extent that parents refused their childs participation, st udents chose not to participat e, failed to answer some questions, or provided false or misleading res ponses resulting in a less reliable basis to form generalizations. If a different popul ation was included, results may differ. Summary This chapter outlined the design of the investigation to answer the research questions and fulfill the purpose of the study. The study examined the relationship between a students level of connectedness with school and his/ her experience with bullying victimization utilizing data from the 2005 National Crime Victimization Survey School Crime Supplement. The impact of race, gender, academic achievement, and grade level were considered. A comparison of the ro le of adult-student a nd peer relationships was also conducted. Empirical evidence gath ered from this study adds to the bullying prevention knowledge base.
106 Chapter Four Results The purpose of this secondary anal ysis of data co llected in the 2005 administration of the School Crime Supplem ent of the National Crime Victimization Survey was to examine whether the presen ce of school connectedness serves as a protective factor diminishing bullying victimization and to investigate the impact that adult and peer relationships, gender, race, grade level, and academic achievement have on the relationship. This chapter re ports the results of the SAS statistical analysis done to answer the four research questions. The first section provides an overview of the results. The second section presents results of the preliminary analysis including data management and factor analysis. The next se ction reports descript ive statistics and the final two sections present results of the multivariate analysis completed to answer each research question and follow up analysis. The chapter concludes with a summary of findings. Overview of Results The study sample included only observations with complete data for the variables of interest resulting in a sample size of 5,780 students. Weighted multivariate analysis was done to answer the research que stions addressed in the study:
107 1. What is the relationship between bullying victimization and school connectedness? a. Is the frequency of bullying victim ization related to the level of connectedness a student has with school? b. Does the level of connectedness of students who report no experience of bullying victimization differ from the level of connectedness of students who report bullying victimization? c. Is the relationship between occurren ces of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? d. Is the relationship between freque ncy of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? 2. Is bullying victimization moderated by th e strength of adult-student relationships that a student develops? 3. Is bullying victimization moderated by th e strength of peer relationships that a student develops? 4. Does the impact of adult-student rela tionships on the frequency of bullying victimization differ from the impact of peer relationships on the frequency of bullying victimization?
108 Summary of Results The analysis showed a very weak rela tionship between school connectedness as measured by emotional safety, relationships, a nd participation in extracurricular activities and the occurrence and frequency of bullyi ng victimization. The school connectedness variables were statistically si gnificant predictors of bullying but accounted for just 2% of the variance in occurrence of victimization and 1% of variance in the frequency of victimization. Adding gender, race, grade level, and academic achievement to the regression analysis increased the proportion of variance accounted for by the model to 5% for occurrence and 3% for frequency of victimization. Race was the only independent variable that was not a statis tically significant predictor of occurrence of victimization. Neither race nor gender was found to be a sta tistically significant pr edictor of frequency of victimization. Similarly, school connecte dness factors accounted for a very small amount of variance for students who reported that they had not experienced bullying victimization. Correlation analys is also showed a weak re lationship between level of connectedness and both students wh o experienced victimization and those that did not. An examination of the moderating effect of adult-student relationships and peer relationships showed that neither type of relationship was a statistically significant predictor of occurrence of bullying victimization. Peer re lationships were a statistically significant predictor of frequency of vi ctimization. Independently, adult-student relationships were not a signifi cant predictor of frequency but did prove to be statistically significant when combined w ith peer relationships. Based on the literature review, the resu lts were somewhat unexpected. Since the results of the analysis showed a weak rela tionship between the variables selected to
109 represent school connectedness and the occurrence and frequency of bullying victimization, an additional correlation an alysis was done to examine the relationship between occurrence of bullyi ng and all variables of inte rest. Once again, only weak relationships were found. Alt hough weak predictor variables, the strongest relationship was between occurrence of bullying victimi zation and how frequently students were distracted from doing schoolwork because ot her students were misbehaving followed by the variable measuring how often teachers punish students during class. Adding these variables to the constructs developed for measuring school connectedness decreased the efficiency of the regression model. However, a regression analysis using just the four variables with the highest correlations showed that the strongest predictor of occurrence of bullying victimization was a model that included school rules are fair, same punishment for breaking rules, how often teach ers punish other students in class, and how often student is distracted by another stude nts behavior. These variables still only accounted for 9% of the variance in occurrence of victimization. Preliminary Analysis Data Management The original data set was retrieved from the Inter-University Consortium for Political and Social Research website. The data file in cluded 11,525 observations and 4,173 variables from both the National Crim e Victimization Survey and the School Crime Supplement. Prior to data analysis, SAS set up for the ASCII file was done using the SAS set up file provided with the Nationa l Crime Victimization Survey School Crime Supplement data files. Coding to normalize weights was done.
110 Recoding was done to change all No responses from (2) to (0). Reverse coding was done to all Likert scale responses changing strongly agre e to (4), agree to (3), disagree to (2) and strongly disagree to (1). Reverse coding was also done to academic grades coding A equal to (5) and F equal to (1). School does not give grades responses were coded as (0). Items with a dont know response, residuals, and out of universe responses were recoded as missing data. A data set was created that included only the variables of interest in this study and observations with missing responses for the variables of interest we re eliminated. The resulting sample included 5,780 observations and 45 variables. Factor Analysis Based on educational research, 20 variab les related to school connectedness were identified and a principal factor analysis was done to examine commonalities in order to create construct variables for school connect edness. The data included in the analysis were the responses to questions measuring part icipation in extracurricular activities such as athletic teams, spirit groups, performi ng arts, academic clubs, school government, service clubs, and other school activities; physical safety meas ures such as presence of security guards, staff supervision, student code of conduct and missing school due to safety concerns; and emotional safety measur es such as fairness of rules and punishment, respect and caring, having adults and peers to talk to and go to for help, and skipping classes. Three factors were retained by the propor tion criterion. A factor loading cut point of 0.30 was used to identify the components fo r each factor. The firs t factor represented
111 emotional safety and included teachers care a bout students, teachers treat students with respect, school rules are fair and all receive the same punish ment for breaking the rules. The second factor represented relationships and consisted of have friend at school who helps with problems, have adult at school who helps with problems, have a friend at school to talk to, and have an adult at school who cares. Items indi cating involvement in extracurricular activities comprised the third f actor. Variables related to school safety and skipping class did not identify with any of the factors. The factor loading based on the rotated factor pattern is found in Table 9. A reliability test was run in order to measure the internal consistency among the individual items for each factor. Cronbachs alpha increases with the average correlation between items with .7 considered to be an acceptable reliability coefficient (Nunnaly, 1978). Both the emotional safety ( = 0.75) and relationship ( = 0.81) factors had sufficient reliability and each were recoded as a construct variable using the mean of the individual variables with a scale of 1 ( low ) to 4 ( high ). Cronbachs alpha for the extracurricular factor ( = .47) indicated insufficient reliability. Deleting individual extracurricular variables did not increase reliability. Theref ore, rather than creating a construct variable, extracurricu lar variables were recoded to indicate participation in any of the activities as involvement (1) a nd nonparticipation in any activities as noninvolvement (0). Since school safety factors did not fall with any of the factors in the school connectedness factor analysis, a separate factor analysis wa s done on these variables. One factor was retained by the proportion cr iterion. Although securi ty guards (0.32824) and staff in hallways (0.32502) fit together, Cronbach s alpha indicated that they were not a
112 reliable construct for school safety ( = .29). Therefore, school safety was not included in the regression analysis. Descriptive Statistics Sample The National Crime Victimization Survey School Crime Supplement was administered to 7,112 students in grades 6 through 12. This study included only observations with complete data for all variables of interest resulting in a final sample size of 5,780 ( n = 5,780). The weighted sample included slightly more males ( n = 2,941) than females ( n = 2,839) and more ninth grade students ( n = 957) than any other grade level (mode = 9). The survey categorized race as white, black, American Indian, Asian, Hawaiian/Pacific Islander, or multiracial with Hispanic origin considered separately. Most students in the weighted sample were identified as white ( n = 4,426) and 1,006 students were identified as being of Hispanic origin ( n = 1,006). The most frequently selfreported school grade was mostly Bs ( n = 2476). Complete frequency information is reported in Table 10. Despite removal of outliers, demographic variables were not normally distributed. The distri bution of grade leve l was slightly plat ykurtic (skewness = 0.006, kurtosis = -0.79). The dist ribution of selfreported grades was negatively skewed and leptokurtic (skewness = -1.28, kurtosis = 3.44) and the distribution of race was positively skewed and leptokurtic (skewness = 2.79, kurtosis = 8.56). The distributions of demographic variables are s hown in figures 1, 2, and 3
113 Table 9 Factor Analysis of Variables Re lated to School Connectedness Variable Factor 1Factor 2Factor 3 0.70067 0.67978 0.54204 0.501.86 0.09572 0.02460 -0.05733 0.15346 0.27185 0.21026 0.14888 0.13252 0.00792 0.00945 0.01194 0.78727 Teachers care about students Teachers treat students w/respect School rules are fair Same punishment break rules School safety; security guards Skip class Stay home fear of attack Friend help w/problem Friend to talk to Adult who cares about me Adult help w/problem Academic club participation Service club participation School government participation Other school activities Performing arts participation Spirit group participation Athletics participation School safety: Code of Conduct School safety: staff in hallways 0.13886 0.43902 0.45636 -0.02810 -0.01876 -0.03564 -0.00398 -0.06836 0.00560 0.08601 -0.00961 0.01813 0.77768 0.53182 0.50147 -0.05627 -0.05505 -0.04615 -0.04614 -0.08373 -0.03373 -0.06633 -0.00345 -0.02183 -0.08610 -0.07328 -0.03061 0.06870 -0.00786 -0.01898 0.01050 -0.15758 -0.16400 -0.08022 -0.05444 0.48868 0.42206 0.37870 0.29707 0.23969 0.02333 0.18722 -0.01181 -0.05150
114 Table 10 Demographics of Study Sample Variable Wgt. f SD of Wgt f % SE of % Gender Male Female 2941 2839 82.68 78.40 50.89 49.11 0.70 0.70 Race White Black Am Indian Asian 4426 929 71 289 135.96 55.96 15.18 21.51 76.57 16.08 1.24 5.00 1.09 0.97 0.26 0.36 Hawaiian/Pacific Multi Hispanic origin 26 39 1083 6.39 7.10 61.61 0.44 0.67 17.41 0.11 0.12 0.90 Academic grade Mostly As Mostly Bs Mostly Cs Mostly Ds Mostly Fs No grades given 1985 2476 1064 157 55 43 7.88 9.18 13.32 43.38 69.85 70.50 34.33 42.84 18.40 2.70 0.97 .75 0.80 0.76 0.60 0.23 0.15 0.14
115 Table 10 (continued). Demogr aphics of Study Sample Variable Wgt. f SD of Wgt f % SE of % Grade level 6th 7th 8th 9th 10th 11th 12th 466 887 930 957 948 854 738 22.50 32.01 42.62 36.16 39.16 37.26 37.52 8.06 15.35 16.10 16.55 16.40 14.78 12.77 0.33 0.48 0.57 0.55 0.51 0.52 0.57 Figure 1 : Distribution of Grade Level 0 200 400 600 800 1000 1200 6th7th8th9th10th11th12th grade levelnumbe r
116 Figure 2 : Distribution of Self-Reported Grades Figure 3 : Distribution of Race Dependent Variables The dependent variables in this st udy were the occurrence of bullying victimization and the frequency of bullying victimization. There we re 1,671 students who reported experiencing bullying victimizati on and 4,109 who reported not experiencing any bullying victimization duri ng the past six months. Over 87% of students reported that 0 1000 2000 3000 4000 5000 WhiteBlackAm IndAsianHaw/PacMulti racenumbe r 0 500 1000 1500 2000 2500 3000 ABCDFNo Gr self reported gradenumbe r
117 bullying only happened once or twice over the past six months The distribution of scores was positively skewed and leptokurtic (skewness = 3.48, kurtosis = 12.72). Table 11 contains complete frequency information and distribution is shown in Figure 4. Table 11 Frequency of Dependent Variables Variable Wgt. f SD of Wgt f % SE of % Bullying victimization Experienced Not experienced 1671 4109 61.61 117.09 28.90 71.10 0.74 0.74 Freq of victimization 1 or 2 x in 6 months 1 or 2 x in month 1 or 2 times in week Almost every day 5078 399 178 125 122.41 23.71 13.68 14.02 87.85 6.91 3.09 2.15 0.48 0.37 0.22 0.23 Figure 4 Distribution of Frequency of Bullying Victimization 0 1000 2000 3000 4000 5000 6000 1 or 2 x in 6 months1 or 2 x in month1 or 2 x in weekAlmost every day Frequency of VictimizationNumber Reporting
118 Independent Variables The independent variables in the study were school connectedness, grade level, academic achievement, race, gender, peer re lationships, and adult-student relationships. Based on previous research, school connecte dness was measured by three variables: extracurricular involvement and two construc t variables created th rough factor analysis named emotional safety and relationships Sixty six percent of students reported participating in extracurricular activities ( n = 3829) and 24% reported no involvement ( n = 1951). The mean score for emotional safety was 3.1 on a scale of 1 to 4 ( n = 2222). Mode and median scores were also 3 (mode = 3, Mdn = 3). Figure 5 shows a slightly leptokurtic distribution (skewness = 0.13, kurtosis = 1.34). The mean score for relationships was 3.2 with nearly 41% of students reporting level 3 ( n = 2362), 18% reporting level 4 ( n = 1047), and 1% reporting level 1 (n = 82). Scores were normally distributed (skewness = 0.47, kurtosis = 0.35). See Figure 6. Complete frequency information for the school connectedness components is contained in Table 12. Figure 5 Distribution of Emotional Safety Scores 0 500 1000 1500 2000 2500 22.214.171.124.126.96.36.199.188.8.131.52.184.108.40.206.220.127.116.11.75.4. Score# reportin g
119 Figure 6 Distribution of Relationship Scores Table 12 Frequency of School Connectedness Components Variable Wgt. f SD of Wgt f % SE of % Extracurricular Involvement Non involvement 3829.00 1951.00 98.07 69.05 66.23 33.76 0.78 0.78 Emotional safety 1 1 .25 1.5 1.75 2 2.25 2.5 3.43 3.32 17.94 23.68 76.27 174.48 336.87 2.09 1.67 5.17 5.73 9.01 15.16 22.33 0.06 0.06 0.31 0.40 1.32 3.02 5.82 0.04 0.03 0.09 0.10 0.16 0.25 0.35 0 500 1000 1500 2000 2500 18.104.22.168.22.214.171.124.126.96.36.199.188.8.131.52.184.108.40.206.75.4. Score# reportin g
120 Table 12 (continued). Frequency of School Connectedness Components Variable Wgt. f SD of Wgt f % SE of % 2.75 3 3.25 3.5 3.75 4 757.31 2222.00 658.90 639.97 409.06 456.40 40.37 70.32 26.48 32.18 24.40 22.02 13.10 38.45 11.40 11.07 7.08 7.90 0.60 0.84 0.37 0.44 0.39 0.36 Relationships 1 1 .25 1.5 1.75 2 2.25 2.5 2.75 3 3.25 3.5 3.75 4 0.82 3.06 8.90 3.96 26.95 58.23 244.14 309.45 2362.00 460.55 863.77 391.46 1047.00 0.82 1.84 3.87 2.07 6.35 8.34 18.48 20.23 80.44 23.57 34.02 24.11 42.74 0.01 0.50 0.16 0.07 0.47 1.01 4.22 5.35 40.86 7.97 14.94 6.77 18.11 0.01 0.03 0.07 0.04 0.11 0.14 0.31 0.33 0.93 0.36 0.51 0.37 0.61
121 Moderating Factors Peer relationships and adult-student rela tionships were consid ered as moderating factors in the occurrence and frequency of bu llying victimization. Peer relationships were measured using the mean of two questions asking about friends at schools and adultstudent relationships were meas ured using the mean of two questions asking about adults at schools. Both were scored on a scale of 1 ( low ) to 4 ( high ). Strong peer relationships were reported by 58% of respondents ( n = 3361) and very strong p eer relationships were reported by 36% of the participants ( n = 2101). There was an approximate normal distribution of scores (skewness = -0.35, kurtosis = 0.66). Slightly fewer students reported having relationships with adults with 23% reporting very strong adult-peer relationships ( n = 1315) and 65% reporting strong adult-student relationships ( n = 3780). Distribution of scores was normal (ske wness = -0.12, kurtosis = 0.60). Complete frequency information is reported in Table 13. Grade level, academic achievement level, race, and gender were also considered as moderating factors. The frequency and dist ribution of these variab les are described in the section entitled Sample and can be seen in Table 10 and Figures 1, 2, and 3.
122 Table 13 Frequency of Moderating Factors Variable Wgt. f SD of Wgt f % SE of % Peer Relationships 1 1.5 2 2.5 3 3.5 4 14.11 8.10 127.50 168.55 2843.00 517.97 2101.00 4.54 3.67 12.95 15.30 92.78 29.40 68.25 0.24 0.14 2.21 2.91 49.18 8.96 36.35 0.08 0.06 0.22 0.25 0.99 0.50 0.78 Adult-Student Relationships 1 1.5 2 2.5 3 3.5 4 15.74 18.42 262.61 388.09 3227.00 552.59 1315.00 3.84 4.42 19.35 24.47 91.22 28.11 47.32 0.27 0.32 4.54 6.71 55.84 9.56 22.76 0.08 0.07 0.32 0.38 0.81 0.43 0.64
123 Multivariate Analysis Assumptions Prior to conducting the regression anal ysis, assumptions were considered. Although the predictor variables are random rath er than fixed, regression is relatively robust to violations of this assumption. The c onstructs created as pr edictor variables each had acceptable reliability with Cronbachs alphas of .75 and .81. Although the skewness and kurtosis statistics of the residuals (ske wness = 0.94; kurtosis = 1.00) vary from what is generally considered acceptable values fo r normal distribution, re gression is robust to this violation due to the size of the sample ( n = 5,780) and removal of outliers can reduce the probability of Type I and Type II errors and improve accuracy of estimates (Osborne 2001). Examination of the residua l scatterplot revealed no viol ations of homoscedasticity or linearity assumptions. Independence al so did not appear to be violated. Multicollinearity was also investigated by examining correlation among predictor variables and was determined not to be problematic with R values ranging from 0.10 to 0.48. Based on this screening, it seemed appr opriate to proceed with the regression analysis. Research Question 1 : The first research question a ddressed in this study was: 1. What is the relationship between bullying victimization and school connectedness? a. Is the frequency of bullying victim ization related to the level of connectedness a student has with school?
124 b. Does the level of connectedness of students who report no experience of bullying victimization differ from the level of connectedness of students who report bullying victimization? c. Is the relationship between occurren ces of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? d. Is the relationship between freque ncy of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? A regression analysis was done to inve stigate the relationship between bullying victimization and the three co mponents of school connectedness determined by the factor analysis: emotional safety, relationships, and extracurricular activities. The initial analysis examined if the occurrence of vic timization could be predicted by the school connectedness components. The adjusted R2 value was 0.020 suggesting that only 2% of the variance in the occurrence of bullying vi ctimization is account ed for by the set of predictors. Cohens effect size was computed to be .02 which can be interpreted as a small effect size. An examination of the in dividual predictors s howed that all were statistically significant: emotional safety, t (5777) = -8.74; p < .0001, relationships, t (5777) = 3.22; p = .0016, and extracu rricular involvement, t (5777) =3.97; p = 0.0001. Further examination using the regression analys is looked at the differences in level of connectedness between students who experienced bullying victimization and those that did not experience victimization. Results showed that although fewer students experienced victimization, there was no di fference in the proportion of variance
125 accounted for by the school connectedness components for students who were not victimized ( R2 = .02). A second regression analysis was done to examine if the frequency of victimization could be predicted by the sc hool connectedness components. The adjusted R2 value was 0.011 indicating that only 1% of the variance in the fr equency of bullying was accounted for by the set of school connected ness predictor variables. A small effect size was computed ( f 2 = .01). Examination of individual predictors showed that two predictors were statistically significant: emotional safety, t (5777) = -5.95; p <.0001 and extracurricular involvement, t (5777) = 2.48; p = .01. Values for each predictor variable are shown in Table 14. Further regression analysis was done to investigate the moderating effect of gender, race, grade level, and academic achie vement on the occurrence and frequency of bullying victimization. There was a slight in crease in the proportion of variance when these variables were added. In both the occurrence of victimization and the non occurrence of victimization models the adjusted R2 values were .05. The only variable that was not significant in either model wa s race. In the model that looked at the frequency of victimization the R2 value was .03 indicating that 3% of the variance in frequency of victimization was accounted for by the predictor variables. Neither race nor gender was a significant predictor. Values fo r each predictor and demographic variable are shown in Table 15.
126 Table 14 t-values of Predictor Variables Variable t valuep value Unstandardized estimate Occurrence of Victimization Emotional safety Relationships Extracurricular involvement -8.74 3.22 3.97 <.0001 0.0016 0.0001 -0.15218 0.05840 0.05322 Non Occurrence of Victimization Emotional safety Relationships Extracurricular involvement 8.85 -3.46 -3.66 <.0001 0.0007 0.0004 0.01738 0.01832 0.01352 Frequency of Victimization Emotional safety Relationships Extracurricular involvement -5.94 1.41 2.48 <.0001 0.1615 0.0143 -0.14720 0.03459 0.04467
127 Table 15 t-values of Predictor and Demographic Variables Variable t valuep value Unstandardized estimate Occurrence of Victimization Emotional safety Relationships Extracurricular involvement Grade level Academic achievement Race Gender -9.18 3.93 4.34 -10.16 -6.01 -1.50 2.43 <.0001 0.0001 <.0001 <.0001 <.0001 0.1348 0.0163 -0.15777 0.06928 0.06267 -0.03589 -0.04229 0.07982 0.02686 Non Occurrence of Victimization Emotional safety Relationships Extracurricular involvement Grade level Academic achievement Race Gender 9.29 -4.19 -4.02 10.11 5.76 0.63 -2.45 <.0001 <.0001 <.0001 <.0001 <.0001 0.5302 0.0155 0.15928 -0.07468 -0.05890 0.03553 0.04133 0.00514 -0.02721 Frequency of Victimization Emotional safety -6.55 <.0001 -0.15608
128 Table 15 (continued). t-values of Pr edictor and Demographic Variables Variable t valuep value Unstandardized estimate Relationships Extracurricular involvement 2.09 3.14 0.0386 0.0020 0.05025 0.05736 Grade level Academic achievement Race Gender -7.48 -3.14 -1.16 0.16 <.0001 0.0021 0.2496 0.8712 -0.03728 -0.04258 -0.01044 0.00253 Research Question 2 A regression analysis was done to exam ine the second research question in the study that asked if bullying victimization is moderated by the strength of adult-student relationships that a student de velops. Prior to proceeding with the regression analysis, a correlation analysis was done to determine the reliability of the two questions being used to measure adult-student relationships. Th e questions proved to be correlated ( r = .78, p < .0001) and had a Cronbachs alpha indicating sufficient reliability ( = .87). However, the adult-student relationship variab le was not statistically significant and the obtained adjusted R2 value was 0.00016 indicating that adult-student relationships did not account for the variance in bullying victimization. Results from a regression analysis conducted to predict non-victim ization from the strength of adult-student relationships were similar ( R2 = -0.00012). The strength of adult-st udent relationships were also not significant in the frequency of victimization, t (5779) = 0.99, p = 0.3225 and accounted
129 for almost none of the variance in frequency of victimization ( R2 = -0.000069). Table 16 shows the values of the variables for the mode rating factor of adultstudent relationships. Research Question 3 The third research question in the stu dy asked if bullying victimization was moderated by the strength of peer relationships that a student develops. Once again, prior to proceeding with the regression analysis, a correlation analysis was done to determine the reliability of the two questions being used to measure peer relationships. The questions proved to be correlated ( r = .71, p < .0001) and had a Cronbachs alpha indicating reliability ( = .83). A regression analysis wa s done to determine if the strength of peer relationshi ps predicted bullying victimiz ation. The strength of peer relationships accounted for a minute amount of variance in both the occurrence of bullying victimization ( R2 = -0.00003) and the non occurren ce of bullying victimization ( R2 = -0.00009). The peer relationship variab le was not signific ant for either the occurrence of bully ing victimization, t (5779) = -0.71; p = 0.4762 or non occurrence of bullying victimization, t (5779) = 0.54; p = 0.59. However peer relationships did make a significant contribution in accounting for the variance of frequency of bullying victimization, t (5779) = -3.19; p = 0.0017 but only accounted for a very small amount of the variance, ( R2 = .003). The size of the typical pred iction error was also somewhat high considering the range of possi ble scores was 1 to 4; root mean square error = .59. The values of the peer relationship va riable can be found in Table 16.
130 Table 16 t-values of Moderating Variables Variable t valuep value Unstandardized Estimate Occurrence of Victimization Adult-student relationships Peer relationships 0.28 -0.71 0.7824 0.4762 0.003205 -0.01003 Non Occurrence of Victimization Adultstudent relationships Peer relationships -0.52 0.54 0.6067 0.5900 -0.00605 0.00762 Frequency of Victimization Adultstudent relationships Peer relationships 0.99 -3.19 0.3225 0.0017 0.01683 -0.05758 Research Question 4 The final research question addressed in th e study was: Does the impact of adultstudent relationships on the fre quency of bullying victimizati on differ from the impact of peer relationships on the frequency of bullyi ng victimization? As seen in Table 16, peer relationships and adult-student relationships were similar in that neither significantly impacted the occurrence or non occurrence of bullying, but they differed in that peer relationships did significantly impact th e frequency of bullying. Another regression analysis was done that entered both types of relationships as pr edictor variables for bullying victimization. Once again, ne ither adult-student relationships, t (5778) = 0.81; p = 0.4188 nor peer relationships, t (5778) = -0.96; p = 0.3370 was a statistically significant
131 predictor of bullying victimiza tion and accounted for a minimal amount of variance ( R2 = 0.00009). Neither variable wa s a significant predictor of non-occurrence of bullying victimization: adultstudent relationships, t (5778) = -0.98; p = 0.3268; peer relationships, t (5778) = 0.87; p = 0.3840 and had a R2 value of 0.00009 indicating that the variables accounted for almost none of the variance in non occurrence of vi ctimization. However when including both types of relationships as predictors for frequency of bullying victimization, both were significan t: adult-student relationships, t (5778) = 2.48; p = 0.0145; peer relationships, t (5778) = -3.57; p = 0.0005. The effect of the variance was quite small with an R2 value of 0.005. The size of the t ypical prediction error remained somewhat high with a root mean square error = .59. Follow Up Analysis Considering the weak relationships f ound in the original analysis, additional analysis was done to determine if the effect of school connectedness on bullying victimization was washed out in the factor analysis. A correlation analysis was done to examine the relationship between all variables of interest and the occurren ce of bullying victimization. Additional variab les related to discipline a nd school safety were also included in the correlation matr ix. No strong relationships we re discovered. The variable measuring how often students are distracted by the behavior of other students had the strongest correlation ( r = .27, p < .0001). This was followed by how often teachers punish students in class ( r = .14, p < .0001), school rules are fair ( r = 14, p < .0001), and the punishment for breaking rules is the same ( r = .10, p < .0001). The complete correlation matrix can be found in Appendix B. Since th e two classroom discip line variables had not
132 been considered previously, another principal factor analysis was r un, again resulting in three factors being retained. The variables that comprised the extracurricular factor remained the same and the variables represen ting emotional safety and relationships fell together in this factor analysis. Factor anal ysis results are shown in Table 17. A reliability test done on the new factor showed that the new emotional safety/relationship factor had sufficient reliability ( = .78). However, a regression an alysis including the new factor showed that the new emotional safety fact or accounted for even less variance in the occurrence of bullyi ng victimization ( R2 = .009) than the original model. Table 17 R evised Factor Analysis incl uding Discipline Variables Variable Factor 1Factor 2 Factor 3 Adult who cares about me Adult help w/problem Friend help w/problem Teachers care about students Friend to talk to Teachers treat students w/respect School rules are fair Same punishment break rules Skip class School safety; security guards 0.69472 0.68425 0.67462 0.66588 0.65768 0.60238 0.46588 0.43426 -0.02346 -0.06784 0.09075 0.05938 0.27770 0.04403 0.28399 0.03450 0.00007 -0.10221 -0.01560 -0.06426 -0.04620 -0.17470 -0.29465 0.32380 -0.29926 0.38105 0.32436 0.26867 -0.00941 -0.06713
133 Table 17 (continued). Revised Factor An alysis including Discipline Variables Variable Factor 1Factor 2 Factor 3 Academic club participation Service club participation School government participation Other school activities Spirit group participation Performing arts participation Athletics participation School safety: staff in hallways School safety: Code of Conduct Stay home fear of attack How often teachers punish How often distracted by misbehavior 0.04227 0.03726 0.04560 0.02652 0.01463 0.10404 0.09961 0.00421 0.01287 -0.02553 0.01581 -0.06809 0.48187 0.36848 0.36848 0.29495 0.22910 0.88971 0.18301 -0.04043 -0.01925 -0.00875 -0.07157 -0 0.05388 0.04359 0.04184 0.01320 -0.00545 -0.01126 0.05168 -0.01759 -0.02322 -0/07778 -0.30506 One final regression analysis was done to determine if the four factors with the highest correlation predicte d the occurrence of bullying victimization. All were statistically significant predictors: distracted, t (5776) = 14.88; p <.0001, teachers punish, t (5776) = 3.34; p = .001, fair rules, t (5776) = -4.78, p <.0001, and same punishment, t (5776) = -1.96, p = .05. Although still weak, this m odel proved to be the strongest predictor ( R2 = .09). Adding the moderating variable s increased the amount of variance accounted for ( R2 = .10). Once again, neither race nor gender was a significant predictor in the model.
134 Summary Chapter four detailed the results of the st atistical analysis done to answer the four research questions addressed in this study. Resu lts of factor analysis descriptive statistics and results of regression analyses were pr esented. Nearly one-t hird of the students participating in the study had experienced bullying victimizati on with 88% reporting frequency levels of just once or twice w ithin the past six months. Daily bullying victimization was reported by 2% of the stud ents while 3% reported weekly experiences and 7% reported victimization once or twi ce a month. Two thirds of students reported participation in extracurricular activities. Most students reported that they had someone at school to talk with and help with problems with 94% reporting rela tionships with peers and 88% reporting relationships with adu lts. Less than 3% of respondents reported having no relationship with either adults or peers. The majority of students indicated having a sense of emotional safety with 76% scoring three or above on the emotional safety construct. Regression analysis showed that em otional safety, relationships, and extracurricular participation were signifi cantly related to the occurrence of bullying victimization but only emotional safety and extracurricular participation were significantly related to the frequency of victimization. Ho wever, the model only predicted a small proportion of variance in the outcome 2% for occurrence and 1% for frequency. The addition of gender, race, grade level, and academic achievement as moderating factors slightly increased the proportion of variance, 5% for occurrence and 3% for frequency. Race was the only variable that di d not account for a statistically significant amount of the variance in occu rrence of victimization. Race and gender were the only
135 variables that did account fo r a statistically si gnificant amount of the variance in the frequency of victimization. The analysis exam ining adult-student relationships and peer relationships as moderating f actors showed that neither factor was a statistically significant predictor of bullying victimizati on. Peer relationships were a statistically significant predictor of frequency of victimi zation but accounted for li ttle variance in the outcome. Overall, the model of school conne ctedness as measured by emotional safety, relationships, and participation in extracurricular activities predicted a very small amount of variance in the occurren ce of bullying victimization and frequency of bullying victimization. Further analysis including additional variables from the School Crime Supplement did not increase the proporti on of variance accounted for by the school connectedness constructs. The strongest pred ictors were factors related to school discipline.
136 Chapter Five Discussion This study is a secondary an alysis of data collected in the 2005 administration of the School Crime Supplement of the National Crime Victimization Survey. Chapter five reviews the purpose and methodology of the stud y. Results and limitations are discussed and implications of the findings as well as recommendations for future research are presented. The chapter concludes with a summary of findings. Overview of the Study School bullying has gained in creased attention in the Un ited States over the past decade, partially in response to resulting violence and more recently because of its potential effect on student performance. As a result, an increasing number of empirical studies examining bullying behaviors have emer ged. Literature on bully ing is replete with studies highlighting the social emotional consequences of engaging in bullying behavior or being the victim of such behavior. However, most studi es focus on individual rather than school level predictors of bullying. A common theme throughout research in the area of school climate is that students learn better in an environment that is characterized by support, respect, and a sense of belonging at the school. School climate research as serts that the essential
137 components for feeling connected to school incl ude students experiencing positive adult/student relationships and physical and emotional safety. Empirical evidence indicates that these elements of school connectedness also se rve as a protective factor for antisocial behaviors at school (Blum, 2005; Catalano, et al., 2004; Clark, 1995). The influence of school connectedness on the pot ential of becoming a victim of these antisocial behaviors has been studied less extensively. The primary purpose of the study was to examine whether the presence of school connectedness serves as a protective factor diminishing bullying victimization. Gender, race, grade level, and academic achievement we re considered as moderating factors. The role that adult to student and peer to peer relationships play in the bullying phenomenon was also investigated. The study looked at relationship differen ces by investigating whether a students level of school connect edness predicted bullying victimization. Finally, it identified risk and prot ective factors that may allow educators to target students at risk of victimization for proactive intervention as well as indicating prevention efforts at the school level. The four research questions addressed in the study were: 1. What is the relationship between bullying victimization and school connectedness? a. Is the frequency of bullying victim ization related to the level of connectedness a student has with school? b. Does the level of connectedness of students who report no experience of bullying victimization differ from the level of connectedness of students who report bullying victimization? c. Is the relationship between occurren ces of bullying victimization and
138 connectedness impacted by gender, race, grade level, or academic achievement levels? d. Is the relationship between freque ncy of bullying victimization and connectedness impacted by gender, race, grade level, or academic achievement levels? 2. Is bullying victimization moderated by th e strength of adult-student relationships that a student develops? 3. Is bullying victimization moderated by th e strength of peer relationships that a student develops? 4. Does the impact of adult-student rela tionships on the frequency of bullying victimization differ from the impact of peer relationships on the frequency of bullying victimization? Based on a review of literature, a rela tionship between school connectedness and bullying victimization was expected. Previous bullying research concluded that having friends decreased the risk of bullying victimization (Fox & Boulton, 2005; Rodkin & Hodges, 2003; Shea, 2003); therefore it was e xpected that the relationship would be evident in this study as well. Review of the Method This secondary analysis used data co llected in the 2005 administration of the National Crime Victimization Survey School Crime Supplement. The survey asked students to self-report inci dents of crime and victimization at school as well as perceptions and att itudes about school. The 2005 survey was administered to 7,112
139 students in grades 6 through 12. Only observati ons with complete data for the variables of interest were included in the study resul ting in a sample size of 5,780. The variables of interest in the study were: o ccurrence of bullying victimi zation; frequency of bullying victimization; physical and emotional safety; involvement in extracurricular activities; relationships at school; gender; race; gr ade level; and academic achievement. Principal factor analysis was done to e xplore commonalities and create constructs to measure school connectedness. A series of weighted regression analyses were conducted to determine if the variation in the occurrence and fr equency of bullying victimization could be accounted for by the le vel of school connectedness as measured by emotional safety, relationships, and involve ment in extracurricular activities. Discussion of Findings School Connectedness and Bullying Victimization A review of the literature on school connectedness reve als a list of standard components that create a sense of belonging to school. A sense of physical safety is typically included. However, in the factor analysis done to create the school connectedness construct, school safety variable s did not emerge as a factor. Subsequent correlation analysis supported the lack of commonality and resulted in school safety measures not being included in the school connectedness constructs. Although physical safety is considered a component of school connectedness, the school safety measures addressed in this particular survey may actually create a negative school climate and promote alienation.
140 Measures such as video cameras, lock er searches, and metal detectors are designed to make a school more secure but may also affect school climate. These measures may improve the safety element of school climate for some students while for others, creating an atmosphere of fear or intimidation resulting in a negative climate (Peterson & Skiba, 2000). Hyman and Snook (2000) found that in schools that have adopted law enforcement rather than educati onal models, there has been little evidence for the efficacy of prevention approaches such as the use of metal detectors, increased police presence, searches of students and lo ckers, or the use of student and staff identification cards. In fact, overdepende nce on these tactics can have negative consequences. In contrast, schools characte rized by a respectful environment that provides students with a sense of shared responsibility are bett er able to assure safety for all. The same researchers later concluded that a schools rush to implement unproven police tactics and punitive procedures based on exaggerated reports of school violence can intensify school alienation when solutions to school violence require creating positive school climates (Hyman & Snook, 2001). Noddings (2005) agreed that solutions to the problems of violence, alienation, ignorance, and unhappiness can not be addressed by increasing security apparatus. Instead, she beli eves teachers must be allowed to interact as whole persons and schools must develop policies that treat the school as a whole community. The factors that were used to measure school connectedness in this study fit into the model suggested by Noddings: the m ean likert scale scores of variables measuring emotional safety and relationshi ps and whether a student was involved in extracurricular activities.
141 The first construct for measuring scho ol connectedness was labeled emotional safety and was comprised of questions re lating to teachers caring about students and treating them respectfully as well as questi ons pertaining to school rules and punishment for breaking the rules. On a scale of 1 ( low ) to 4 ( high ), 76% of students placed at a level three or above in sense of emotional safety. Le ss than 1% of student s placed at level one which indicated that most students felt some level of emotional safety at school. The second school connectedness constr uct was labeled relationships and included questions related to having both friends and adults at school to talk to and help with problems. It also included having an a dult at school that cares about the student. Based on the same scale as the emotional safe ty construct, over 88% of participants placed at a level three or above. Less than one half of 1% placed at a level 1 indicating nearly all students had some t ype of relationship with either peers or adults. The final measure of school connectedness was involve ment in extracurricu lar activities. Two thirds of students reported participating in some type of extracurricular activity. Considering response to the three measures, most participants in the study felt a sense of connection to school. Nearly 29% of students in this study re ported experiencing bullying victimization ( n = 1671). This is comparable to previous re search that reports rates of victimization ranging from as low as 7% to as high as 77% with most stud ies reporting around onethird of students being the target of bullying. A comparison of the level of school connectedness of students who reported experi encing bullying victim ization with those who did not experience victimization reveal ed similar correlations leading to the conclusion that the level of connectedness of students who did not experience bullying
142 victimization did not differ from the le vel of connectedness of students who did experience bullying victimization. Regression analysis showed that alt hough the measures of school connectedness were statistically significant pr edictors of bullying victimi zation, there was a very weak relationship, accounting for only 2% of variance in occurrence in bullying victimization. A sense of emotional safety and involvement in extracurricular activities were slightly stronger predictors than having relations hips with adults and peers. Although relationships affected the occurrence of bully ing victimization, relationships were not a significant predictor of frequency of victimiza tion. This is contrary to previous research that lists having friends as a protective factor for victimi zation and may be caused by the combination of adult-student relationship with peer relationships in this study. Previous research has shown that st udents with a high degree of school connectedness are at less risk for engaging in delinquent behaviors. It is possible that the relationship between school connectedness and being the victim of such delinquent behaviors is not evident in the literature because a relationship has not been found. Researchers in the areas of social emotional learning and risk and pr otective factors agree that a safe learning environment is essent ial to academic success. Additionally, prior research in school connectedne ss shows that a students sens e of belonging to school and feeling cared about at school is essential for reduci ng the risk of engaging in aggressive behaviors. Results of this study suggest that the reduction of risk does not apply to the occurrence or frequency of a student expe riencing bullying victimization. Unlike the relationship between connectedness and engagi ng in aggressive behaviors, the weak relationship between the measures of school connectedness and bullying victimization
143 indicates that a low level of connectedness is not a risk factor fo r experiencing bullying victimization. Nor does school connectedne ss serve as a protective factor for experiencing bullying victimization. Results of the current study are comparable to those recently reported by researchers at Johns Hopkins Universit y. A survey of 11,000 mi ddle school students showed that despite evidence th at school climate was improvi ng, the self reported rate of being bullied remained unchanged (Bradshaw, Debman, Martin, & Gill, 2006). Outcomes of the current study also concur w ith findings reported in a recently published study on adolescent bullying. Spriggs, Iannotti, Nansel, & Haynie (2007) found that school attachment was inconsistently related to bullying behavior a nd that school factors did not differentiate between victims and noni nvolved peers. However, the researchers found that both students who bullied and vi ctims of bullying reported lower school attachment which is in contrast to findings of this study. The first research question also considered the impact of gender, race, grade level, and academic achievement on the relationship between school connectedness and bullying victimization. Nearly even numbers of males ( n = 2941) and females ( n = 2839) were included in the study sample. Although fewer 6th grade students were included, other grade were represented fairly evenly. Academic achievement was measured by self reported grades. As expected, the distribution was skewed with 95% of students reporting grades of C or better. The distribution of race was also skewed with 76% of the study respondents being white students which likely had an impact on the study outcomes. Regression analysis taking into account the effect of gender, race, grade level, and academic achievement showed a slightly larger proportion of variance accounted for by
144 the variables of interest but still only accounted for 5% of the vari ance in occurrence of bullying victimization and 3% of variance in frequency of bullying victimization. Grade level appeared as the strongest predictor of both occurrence and frequency of bullying victimization. This supports previous research that shows grade leve l is inversely related to experience of bullying. Race was the only factor that did not make a statistically significant contribution to the variance for either occurrence or fr equency of bullying victimization. Since the majority of participants in the sample were white students, this result needs to be regarded with caution. Recently published research investigating bullying across race/ethnicity yielded results that show r acial/ethnicity differences in prevalence of bullying can be observed (Spr iggs, et al., 2007; Nansel et al, 2001). A national study involving more than 11,000 adolescents found th at school satisfaction was relevant for black and Hispanic students only (Spriggs, et al.). Additionally, feeli ng unsafe at school was positively associated with victimization for white students only. Another recent study of the Olweus Bullying Prevention Program found decreases in bullying for white students only (Bauer, Lozano, & Rivara, 2007). These findings indicate that additional research with a more diverse population is needed. Bullying prevention researchers disagree as to a pattern in gender differences in engaging in or experiencing bullying behavior. This study found that although gender was a statistically significant predictor of the occurrence of bullying victimization, it was not significant for the frequency of bullyi ng victimization. These results suggest that while gender impacts whether or not a stude nt experiences bullying victimization, it makes no difference in the frequency of th e victimization. This both supports and
145 contradicts previous research that maintains boys and girls engage in different types of bullying behavior. Boys typically engage in more direct bullying such as physical aggression which is easily observable and reco gnized as bullying but may be limited in frequency whereas girls are more likely to enga ge in indirect bullyi ng such as relational aggression which is more difficult to observe and not always rec ognized as bullying but generally has a more ongoing pattern. Relationships The last three research questions dealt w ith the moderating eff ect of adult-student relationships and peer relations hips. Research on peer relatio nships as a determinant of bullying involvement abounds while studies i nvestigating the role of adult-student relationships is scarce. This study consid ered both types of relationships. Peer relationships were measured using the mean of two likert scale questions that asked if students had a friend at school th ey could talk to and go to fo r help with problems. Adult to student relationships were measured using two similar questions related to adults at school. On a scale of 1 ( low ) to 4 ( high ), 94% of students placed at a level three or above in peer relationships and 88% placed at a level three or above in adult student relationships indicating that most students had develope d relationships at school. The current study produced resu lts that are inconsistent with previous research that concludes bullying behavior s are consistently related to peer relationships. In this study, peer relationships ha d a weak correlation with the occurrence of bullying victimization. They were not a statistica lly significant predictor of occurrence of victimization but were statistically signifi cant in predicting the frequency of bullying
146 victimization. However, peer relationships accounted for a very small amount of variance in frequency of victimization. In comparison, adult-student relationshi ps were similar in that they were not correlate d to occurrence of bullying a nd were not a statistically significant predictor of occurr ence of bullying victimization. Dissimilarly, adult-student relationships were not a statistically signi ficant predictor of frequency of bullying victimization either. Considering these result, the impact of adult-student relationships on the frequency of bullying victimization does differ from the impact of peer relationships on the frequency of bullying victimization. It appears that bullyi ng victimization is moderated by peer relationships but not by adult-student relations hips. Prior research underscores the importance of adult-student re lationships in creati ng a sense of school connectedness. The results of this study sugge st that the importance does not carry over to bullying victimization. Cont rary to gender, peer relations hips do not make a difference in whether or not a student experiences bullyi ng victimization but can have an impact in the frequency of the victimization. This finding concurs with other rese arch that indicates the ineffectiveness of approaches focused so lely on peer relationshi ps (Spriggs et al., 2007) and leads to a conclusion that successful bullying prevention initiatives must have a comprehensive approach. Additional Analysis Since none of the variables of interest had a strong relationship with bullying victimization, a correlation analysis was done to discer n any possible relationship between the individual variable s of interest and the occurr ence of bullying victimization.
147 Additional variables measuring school discipline and school safety were added to the analysis. Consistent with the previous an alysis, none of the variables had a strong relationship with occurrence of bullying vict imization. The strongest correlations were with two new variables related to school discipline: how ofte n a student is distracted by the misbehavior of other students and how of ten teachers punish students in class. Having fair rules and equal punishment for breaking the rules showed a hi gher correlation than the remaining variables. Addi ng the new variables to the sc hool connectedness variables in the regression analysis act ually decreased the proportion of variance accounted for by the model. However, a regression analysis usin g just the four highest correlated variables and moderating variables proved to be the stro ngest model for predicting the occurrence of bullying victimization. Alt hough still a weak relationship, it is relevant for planning bullying prevention initiatives since disciplin e factors are generall y not considered as contributors to bullying problem s. While adult-student relati onships were not found to be statistically related to bullying victimiza tion, discipline factors were related. This supports the conclusion reached by Olweus (1993) that the attitudes, routines, and behaviors of school personnel can be a c ontributing force in bullying behavior. Additional research investigating the impact of educators discipline style and strength would be beneficial. Implications The present study contributes to the pr evention education re search base and enhances knowledge of what works to addr ess bullying in schools. If the information discovered in this study is correct, the imp lications of the findings are that bullying
148 problems in school cannot simply be addressed by working to create a positive school climate. Increasing a students sense of connectedness to school and building relationships with students are not suffi cient components for a bullying prevention program given that they do not have a strong impact on bul lying victimization. Although school connectedness is considered a key to re ducing the risk of e ngaging in aggressive behaviors, it does not app ear to reduce the risk of being a victim of bullying victimization. Considering prior research, the weak relationship between school connectedness and bullying victimization found in this study may have resulted from the focus on occurrence and frequency rather than the victims response to bullying victimization. A students decision to report bu llying, assertively address the problem, or respond with violence may be related to hi s/her level of school connectedness. This possible relationship wa rrants further study. Even though bullying may have a negativ e impact on the clim ate of a school (Limber & Small, 2003; Batsche & Knoff, 1994), there is not a clear inverse relationship. School climate does not have the same eff ect on bullying victimization. To create a school climate that discourages bullying, school staff must be aware of the extent of bully-victim problems in their school and ma ke a commitment to reduce or eliminate bullying (Peterson & Skiba, 2000). Simply deve loping a relationship w ith students is not enough. Results of this study are relevant to planning successful bullying prevention initiatives. Previous research shows up to a 50% reduction in bully ing behavior when a school implements a school wide bullying prevention program that includes school, classroom, individual, and community com ponents (Olweus, 1993). The current study
149 supports the finding that a comprehensive appr oach is needed to address bullying in school effectively. Specific stra tegies that increase awaren ess of bullying, teach students how to respond to bullying, and address indi viduals involved in bullying behaviors appear to be essential in a su ccessful bullying prevention initiativ e. Study Limitations The current study has the strength of a large national sample but there are limitations that emerged as analysis was co mpleted limiting the ability to generalize the findings. The lack of diversit y in the study sample caused co ncern in that it may have skewed the results restricting the ability to generalize findings to sc hools serving students from varied racial or ethnic backgrounds. Another possible limitation is non res ponse. A potential 11,525 students were eligible to take part in th e survey but only 7,112 actually pa rticipated. Analysis of non response found evidence that respondents from households with an income of $35,000 or more had higher response rates than those from households with lower incomes and respondents living in urban areas had lower resp onse rates than those living in rural or suburban areas (Dinkes, et al ., 2006). Weighting adjustment s were made to reduce the problem of non response bias and the weight ed data was used in this study. Unit non response poses another potenti al limitation. The survey was confidential but was not an anonymous survey. Therefore, students who did not feel connected to school or who were victims of bullying may have chosen not to provide accurate responses. Students often do not report being the victim of bullying out of embarrassment or fear of retaliation
150 (Olweus, 1993). These same fee lings may have caused particip ants to skip the questions related to bullying causing them to be removed from the study sample. The study design and survey instrument also posed several limitations. Since the study was a secondary analysis, follow up with respondents was not possible. A qualitative component to the study would have clarified and possibly strengthened the findings. Many of the questions used a Likert scale ranking so comparisons of the degree of differences in rating the strength of connectedness comp onents and relationships could not be made. For example, one students pe rception of the punishment for breaking rules being fair for all students may be very different than anot hers perception of the same question. The School Crime Supplement su rvey asked questions about bullying victimization but did not address students e ngaging in bullying behaviors or witnessing bullying happening to others. Data from st udents who bully and bys tanders would have provided a more complete understanding of the role that school connectedness and relationships play in scho ol bullying. Another missing el ement in the School Crime Supplement survey was cyberbullying. A lthough a relatively new form of bullying, research shows that it is an increasing probl em. By overlooking this form of bullying in the survey, the rate of bullying victimiza tion may be artificia lly deflated. Since cyberbullying is typically begun in a mo re anonymous manner with no face to face contact or immediate feedback, there may be a unique correlation with school connectedness or relationships with adults and peers at school. Th is merits further study. Risk and protective factor research indicates that school climate and connectedness influence aggressive behaviors. There has been minimal research examining the influence of school climate on being the recipient of the aggressive
151 behaviors. This may be due to the lack of a relationship between the two. Based on the results of this data analysis, a relations hip between school connectedness and bullying victimization is very weak and may not exist at all. It is also possible that the survey instrument and study design could not find the relationship due to lack of power. Although the sample size was large, sampling stratification may have contributed to a lack of power. It is also reas onable to consider that the ques tions on the survey instrument could not find the nuances of school conn ectedness that are related to bullying victimization and that results from another survey designed specifically for this purpose would find a stronger relationship. Recommendations for Future Research The results of this study contribute to th e empirical research base exploring school level factors and bullying in schools. A lthough a strong relationship between school connectedness and bullying victimization was not found in this study, previous evidence of the consequences of bullying victimizati on make further research necessary. Despite the limitations noted in the previous secti on, findings can provide direction for future studies. In fact, certain limitations pr esent opportunity for additional research. The National Crime Victimization Survey School Crime Supplement was designed to gather informati on about school related victimi zation on a national level. The survey covered a variety of potential school crime scenarios; bullying was only a small piece of the information collected. Additional research using alternate instrumentation designed specifically to inve stigate the relationship betw een school connectedness and bullying behaviors is reco mmended. Considering the st ronger relationship between
152 school discipline factors and the occurrence of bullying victimization found in this study, new instrumentation should include variables to measure a teachers discipline style and the effectiveness of classroom management strategies. A mixed methods research design that allows for follow up with the responde nts would strengthen the study. Any further research in this area needs to include a diverse population to determine effective strategies for various populations. The current study focused only on bu llying victimization. Although a strong relationship between school connectedness and bullying victimization was not found, prior research asserts that there is a re lationship between school connectedness and aggressive behaviors. Additional research is needed to determine if this relationship exists for students engaging in bullying beha vior. Studies to inves tigate the impact of school connectedness on the re sponse of bystanders would also provide valuable information for determining strategies to in clude in a comprehensive bullying prevention program. Cyberbullying is a form of bullying that has not been studied extensively until recently. Although cyberbullying often occu rs from outside the school campus, it can carry over to the school envi ronment. It is recommended that research examining the relationship between school connectedness a nd cyberbullying be conducted to provide a more complete picture of this emerging problem. Finally, recent school violence, such as that at Virginia Polytechnic Institute and State University, continues to increase the desi re to create school climates that reduce the risk for victimization and violence. In light of this ubiquitous con cern, further research on school safety factors is needed. As schools ar e looking for ways to create safe campuses,
153 overlooking emotional safety could have devast ating consequences. Referring back to the Secret Service findings that two thirds of school shooters had been victims of bullying (Brady, 2001), school safety implies more than hardening the target with locks, fences and armed security. Although this particular study did not find a correlation, a study designed specifically to look at school safety factors and school connectedness as it relates to bullying victimization may find a re lationship that provides valuable insight to school safety efforts. Summary Empirical research suggests that crea ting schools that are both physically and emotionally safe is an integral componen t of student success. Findings from this secondary analysis of data from the 2005 National Crime Victimization Survey School Crime Supplement add to the research base by providing information about the relationship between school connectedness a nd bullying victimization. The information can be used to help schools create effective bullying pr evention programs. While bullying prevention programs can contribu te to a positive school climate, results of this study indicate that strategies to improve climate cannot be th e sole component of bullying prevention initiatives. A very weak relationship between school connectedness and bullying victimization was found sugges ting that more than emotional safety, involvement in school activities, and positive relationships with adults and peers are needed to effectively addr ess bullying victimization in schools. A comprehensive approach that teaches staff and students that bullying is unacceptable and empowers them to address bullying behaviors is required. Additional research to determine the most
154 effective strategies for a ddressing students who bully, victims of bullying, and the bystanders who witness bullying is needed to support development of successful bullying prevention and intervention programs designed to address the needs of all students in the school.
155 References Abourjilie, C. (2000). Developing character for classroom success: Strategies to increase responsibility, achievement, and motivation in secondary students Chapel Hill, NC: Character Development Publishing. Abourjilie, C. (2006, February 4). Developing character for classroom success and beyond: Building character, achievement and motivation in all your students Paper presented February 4, 2006 at Pinellas County Schools, Largo, FL. American Association of University Women (1993). Hostile hallways: The AAUW survey on sexual harassment in Americas schools Washington D.C.: Author. American Medical Association (2002). Report 1 of the Council on Scientific Affairs: Bullying behaviors among children and adolescents Retrieved August 11, 2004 from http://www.ama-assn.org Barabarin, O.A., Whitten, C.F., & Bonds S.M. (1994). Estimating rates of psychosocial problems in urban and poor children with sickle cell anemia. Health and Social Work, 19 112-119. Barth R.S. (2001). Learning by heart San Francisco, CA: Jossey-Bass. Batsche, G.M. (2000, August 31). Creating safe, healthy envi ronments for children and youth: Implementing the blueprint. Paper presented August 31, 2000 to Pinellas County Schools, Largo, FL.
156 Batsche, G.M. & Knoff, H.M. (1994). Bu llies and their victims: Understanding a pervasive problem in schools. School Psychology Review, 23 (2), 165-175. Brady, J. (2001, September 18). An interview w ith Dr. Russell Skiba of the Safe and Responsive School Project. Retrieved June 21, 2003 from www.guidancechannel.com Bauer, N.S., Lozano, P. & Rivara, F.P. (2007) The effectiveness of the Olweus bullying prevention program in public middl e schools; A controlled trial. Journal of Adolescent Health, 40, 266-274. Bell-Ellison, B.A. & Kromrey, J.D. (2007). Alternatives for analysis of complex sample surveys: A comparison of SAS, SUDAAN, and AM software. Unpublished manuscript, University of South Florida at Tampa. Benard, B. (1992, June). How schools c onvey high expectations for kids. Northwest Regional Educational Laboratory, 5 (3), 17-19. Benard, B. (1991, September). Prevention should emphasize protective factors. Western Center News, 4 (4), 1-4. Blum, R.W. (2005, April). A cas e for school connectedness. Educational Leadership, 62 (7). 16-20. Blum, R.W. & Libbey, H.P. (2004) School connectedness: Strengthening health and education outcomes for teenagers. Journal of School Health, 74 (7). 229-299. Blum, R. W., McNeely, C., & Rinehart, P. M. (2002). Improving the odds: The untapped power of schools to improve health of teens. [Electronic version] Minneapolis, MN: Center for Adolescent Health and De velopment, University of Minnesota.
157 Bowen, N.K. & Bowen, G.L. (1999). Effects of crime and violence in neighborhoods and schools on the school performance of adolescents. Journal of Adolescent Research, 14 319-342. Bradshaw, C.P., Debnam, K.J., Martin, L. & Gill, R. (2006, September 20). Assessing rates and characteristics of bullying th rough an internet-based survey system. Paper presented at the Persistently Safe Schools Conference. Washington, D.C. Brady, J. (2001, September 18). An interview with Dr. Russell Skiba of the Safe and Responsive School Project Retrieved June 21, 2003 from www.guidancechannel.com Brekelman, M, Wubbels, T., & den Brok, P. (2002).Teacher experience and the teacher student relationship in the classroom. In S.C. Goh & M.S. Kline (Eds.) Studies in educational learning environmen ts: an international perspective (pp. 73-99) New Jersey: World Scientific. Brooks, R.B. (1999). Creating a positive school climate: Strategies for fostering self esteem, motivation and resilience. In J. Cohen (Ed.) Educating minds and hearts: Social emotional learning and the passage into adolescence. (pp. 158-170). Calabrese, R.L. (2000). Leadership for safe schools: A community based approach. Lanham, MD.: Scarecrow Press, Inc. Catalano, R.F., Haggerty, K.P., Oesterle, S. Fleming, C.B., & Hawkins, J.D. (2004, September). The importance of bonding to school for healthy development: Findings from the social development research group Journal of School Health, 74 (7), 252-261.
158 Clark, P. (1995, winter). Risk and resilien cy in adolescence: Th e current status of research on gender differences. Equity Issues, 1 (1). Cleary, M. (2000, December 21). Bullying information for schools New Zealand: Telecom. Coady, E. (2005, April 15). The bully blight. Time, 144-145. Cohen J. (Ed.) (1999). Educating minds and hearts: Social emotional learning and the passage into adolescence New York: Teachers College Press. Commission on Children at Risk (2003). Hardwired to connect: Th e new scientific case for authoritative communities a report to the nation from the Commission on Children at Risk New York: Institute for American Values. Copas, A.J. & Farewll, V.T. (1998). Dea ling with non-ignorable non-repsonse by using an enthusiasm to respond variable. Journal of Royal Statis tical Society, Series A (Statistics in Society), 161 (3). 385-396. Covey, S. R. (1991) Principle centered leadership. New York: Simon & Schuster. Dake, J.A., Price, J.H., Telljohann, S.K ., & Funk, J.B. (2004, June). Principals perceptions and practices of school bullying prevention activities. Health Education and Behavior. 31 (3), 372-387. Dellasega, C. (2001). Surviving Ophelia: Mothers share th eir wisdom in navigating the tumultuous teenage years Cambridge: MA.: Perseus Publishing. DeVoe, J.F. & Kaffenberger, S. (2005). Stude nt reports of bullying: Results from the 2001 School Crime Supplement to the Na tional Crime Victimization Survey (NCES 2005-310). U.S. Department of E ducation, national Center for Education Statistics. Washington, DC: U.S. Government Printing Office.
159 Dinkes, R., Cataldi, E.F., Kena, G., and Baum, K. (2006). Indicators of School Crime and Safety: 2006 (NCES 2007003/NCJ 214262). U.S. Departments of Education and Justice. Washington, DC: U.S. Government Printing Office. Doan, J., Roggenbaum, S., & Lazear, K. (2003). Youth suicide prevention school-based guide Issue brief 2: School Climate Tampa, FL: Department of Child and Family Studies, Division of State and Lo cal Support, Louis de la Parte Florida Mental Health Institute, Univ ersity of South Florida. Dornbusch, S., M., Ritter, P.L., Leiderman, P.H., Mont-Reynaud, R., & Chen, Z. (1990). Family decision making and academic performance in a diverse high school population. Journal of Adolescent Research, 5, 143-160. Donald, B. (2002, December 10). Students still afraid de spite fewer school weapons, crime. Why? Bullies Retrieved September 20, 2004 from http://www.bridges4kids.org/articles Drapela, V.J. (1987). A review of personality theories Springfiled, IL: Charles C. Thomas Publisher. DuFours, R. & Eaker, R. (1998). Professional learning communities at work: Best practices for enhancing student achievement. Bloomington, IN.: National Education Service. Dupper, D.R. & Adams, N.M. (2002, May). Lo w-level violence: A neglected aspect of school culture. Urban Education, 37 (3), 350-364. Dwyer K. Osher, D., & Warger, C. (1998, August). Early warning timely response: A guide to safe schools. Washington DC: U.S. Department of Education.
160 Ecstrom, J., Goertz, M. Pollack, J., & Roc k, D. (1986). Who drops out of high school and why? Findings from a national study. In G. Natriello (Ed.), School dropouts: Patterns and policies New York: Teachers College Press. Elias, M.J., Zins, J.E., Weissberg, R.P., Fre y, K.S., Greenberg, M.T., Haynes, N.M. et al. (1997). Promoting social and emotional lear ning: Guideline for educators. Alexandria VA: Association for Supe rvision and Curriculum Development. Elliot, G.P. (2003). School mobbing and emotional abuse New York: Brunner Routledge. Esposito, L.E. (2003, June 1). Peer victimization and posttraumatic stress among children Child Study Journal. Retrieved April 10, 2004 from http://www.highbeam.com Fan, X., Miller, b.B., park, K., Winward, B.W. Christensen, M. Grotevant, H.D., et al. (2006, August). Anexploratyory study a bout inaccuracy and invalidity in adolescent self-report surveys. Field Methods, 18 (3), 223-244. Fein, R. A., Vossekuil, B., Pollack, W.S., Borum, R., Modzeleski, W. And Reddy, M. (2002, May). Threat assessment in schools: A gu ide to managing situations and to creating safe school climates Washington D.C.: U. S. Secret Service and U.S. Department of Education. Feinberg, T. (2003). Bullying prevention and intervention. Principal Leadership (Middle School Edition).4 (1), 10-14. Flook, L. & Repetti, R. L. (2005). Classroom social experiences as predictors of academic performance. Developmental Psychology, 42 (2), 319-327.
161 Florida Department of Education, (1998). School staff guide to risk and resiliency Tallahassee, FL: Author. Florida Department of Education (2002, February). What is school climate? SDDFS Notes, 5 (2), 1-2. Florida Department of Education (2004). Assessment and Accountability Briefing Book. Tallahassee, FL: Author. Florida Office of Safe and H ealthy Schools. (2005, April). SDDFS Notes, 8(2) Tallahassee, Florida: Department of Education. Fox, C.L. & Boulton, M.J. (2005). The social ski lls problems of victims of bullying: Self, peer and teacher perceptions. British Journal of E ducational Psychology, 75 33328. Fraser, B.J. (1999). Using l earning environment assessments to improve classroom and school climates. In H.J. Freiberg (Ed.), School climate: Measuring, improving and sustaining healthy learning environments (pp. 65-83). Philadelphia, PA: Palmer Press. Fraser, B. J. & Fisher, D.L. (1983) Use of actual and preferred classroom environment scales in person-environment fit research. Journal of Educational Psychology, 75 303-313. Gall, M. D., Gall, J. P., & Borg, W. R. (2003). Educational research: An introduction. New York: Pearson Education, Inc. Garrett, A.G. (2005). Bullying in American schools Jefferson, N.C.: McFarland & Company, Inc.
162 Gerler Jr., E.R. editor (2004). Handbook of School Violence New York: Haworth Reference Press. Glasser, W. (1995) Staying together New York: Harper-Collins. Gold, M. (1977). The validity of self-reports of delinquent behavior. Unpublished manuscript, Institute for Social Research. Gottfredson G.D. & Gottfredson, D.C. ( 2001) What schools do to prevent problem behavior and promote safe environments. Journal of Educational and Psychological Consultation, 12 (4), 313-344. Gottheil, N.F. & Dubow, E.F. (2001) Tripartite beliefs models of bully and victim behavior In R.S. Geffner, M. Loring, & C. Young (Eds) Bullying Behavior: Current issues, resear ch and interventions (pp. 7-23). New York: Haworth. Greene, M.B. (2003). Counseling and climate change as treatment modalities for bullying in school. International Journal for the Advancement of Counseling, 25 (4), 293302. Halford, J.M. (1996, fall). Policies fo r safety, caring and achievement. ASCD Infobrief, (6). Retrieved October 20, 2005 from http://www.ascd.org Hawker, D.S.J. & Boulton, M.J. (2000) Twenty years research on peer victimization and psychosocial maladjustment: A meta-a nalytic review of cross-sectional studies. Journal of Child Psychiatry and Psychiatry, 41 441-455. Hawkins, D.L., Pepler, D.J., & Craig, W.M. (2001). Naturalistic obs ervations of peer interventions in bullying. Social Development, 10 (4), 512-527.
163 Hawkins J. D., Catalano R. F., & Miller J. Y. (1992). Risk and protective factors for alcohol and other drug problems in adoles cence and early adult hood: Implications for substance abuse prevention. Psychological Bulletin, 112 (1), 64-105. Haynes, N.M. & Marans, S. (1999 ).The cogni tive, emotional and behavioral framework for promoting acceptance of diversity. In J. Cohen, (Ed.) Educating minds and hearts: Social emotional learni ng and the passage into adolescence. (pp. 158170). New York: Teachers College Press. Haynie, D.L., Nansel, T., Eitel, P., Crum p, A.D., Saylor, K., Yu, K. et al. (2001, February). Bullies, victims, and bully/vic tims: Distinct group s of at-risk youth. Journal of Early Adolescence, 21 (1), 29-49. Heinrichs, R.R. (2003, March 1). A wholeschool approach to bullying: Special considerations for children with exceptionalities. Intervention in School & Clinic, 38 (4), 195-205. Hirsch, T. (1969). Cases of delinquency Berkeley: University of California Press. Hodge, E.V.E. (2003, June 22). Bullies a nd victims in the peer ecology: Four questions for psychologists and school professionals. School Psychology Review Retrieved March 1, 2004 from http://www.highbeamresearch.com Hodge, E.V.E., Malone, M.J., & Perry, D.G. ( 1997). Individual risk and social risk as interaction determinants of vi ctimization in the peer group. Developmental Psychology 33, 1032-1039.
164 Hoover, J. & Stenhjem, P. (2003, Decembe r). Bullying and teasing of youth with disabilities: Creating positive school e nvironments for effective inclusion. National Center on Secondary Education and Transition Issue Brief, 2 (3). Retrieved January 29, 2004 from http://www.ncset.org/publications Horn, R.A. & Kincheloe, J.L. (2001). American standards: Quality education in a complex world, the Texas case. New York: Peter Lang Publishing Inc. Howes, C. & Ritchie, S. (2002). A matter of trust: Connecti ng teachers and learners in the early childhood classroom New York: Teachers College Press. Hoy, W.K. & Feldman J.A. (1999). Organiza tional health profiles for high schools. In H.J. Freiberg (Ed.), School climate: Measuring, im proving and sustaining healthy learning environments (pp. 84-102). Philadelphia, PA: Palmer Press. Hoy, W. K. & Sabo, D. J. (1998). Quality middle schools: Open and healthy. Thousand Oaks CA: Corwin Press, Inc. Hugh-Jones, S. & Smith, P.K. (1999). Self-re ports of short and l ong term effects of bullying on children who stammer. British Journal of Educational Psychology, 69 (2), 141-158. Hunt, M.H., Meyers, J., Davies, G., Me yers, B., Grogg, K., & Neel, J. (2002). A comprehensive needs assessment to facilitate prevention of school drop out and violence. Psychology in the Schools, 39 (4), 399-416. Hyman, I.A. & Snook, P.A. (2000, March 1) Dangerous schools and what you can do about them. Phi Delta Kappan Retrieved September 7, 2007 from http://www.highbeam.com
165 Hyman, I.A. & Snook, P.A. (2001, Septembe r 22). Dangerous schools, alienated students. Reclaiming Children and Youth Retrieved September 7, 2007 from http://www.highbeam.com Ibrahim, J.G., Chen, M.H., Lipsitz, S.R. & Herring, A.H. (2005, March 1). Missing data methods for generalized linear models: a comparative review. Journal of the American Statistical Associat ion. Retrieved June 30, 2007 from http://www.highbeam.com Indiana Department of Education (2003, December 31). White paper on bullying prevention and education Retrieved June 30, 2006 from http://www.doe.state.in.us/legwat ch/docs/Bullyingpaper2004session.doc Introduction to risk factors and protective factors. (n.d.) Retrieved February 14, 2006 from http://www.helpingamericasyouth.gov Isernhagen, J. & Harris, S. (2002, April). A comparison of 9th and 10th grade boys and girls bullying behavior in two states. Paper presented at the American Education Research Association. New Orleans, LA. Jacobson, A.F., Hamilton, P., & Galloway, J. (1993, August 15). Obtaining and evaluating data sets for secondary analysis in nursing research. Western Journal of Nursing Research, 4 483-494. Retrieved February 2, 2007 from http://www.ncbi.nlm.nig.gov Johnston, L. D., OMalley, P. M., Bachma n, J. G., & Schulenberg, J. E. (2006). Demographic subgroup trends for various licit and illicit drugs, 1975-2005 (Monitoring the Future Occasional Pa per No. 63) [On-line]. Ann Arbor, MI: Institute for Social Research. Avai lable: http://monitoringthefuture.org
166 Kaiser Family Foundation (2001, March 8). Ta lking with kids about tough issues: A national survey of parents and kids. Talking with Kids. Retrieved August 11, 2004 from http://www.kff.org Kent, B.A. (2003). Identity issues for ha rd-of-hearing adolescents aged 11, 13, and 15, in mainstream settings. Journal of Deaf Studies and Deaf Education, 8 (3), 315-324. Klum, A.M. & Connell, J. P. (2004, Septembe r). Relationships matter: Linking teacher support to student engagement and achievement. Journal of School Health, 74 (7), 262-272. Kohn, A. (2004). Rebuilding school cu lture to make schools safer. The Education Digest, (70) 3, 23-30. Kumpulainen, K., Rasanen, E., & Puura, K. (2001). Psychiatric disorders an the use of mental health services among students involved in bullying. Aggressive Behavior, 27 102-110. Lambert, L., Collay, M., Dietz, M.E. Kent, K., & Richert, A.E. (1996). Who will save our schools? Teachers as c onstructivist leaders. Thousand Oaks, CA: Corwin Press, Inc. Li, Qing (2005). Cyber-bullying in schools: The nature extent of adoles cents experience. Paper presented at the American Edu cation Research Association (AERA) Conference in Montreal, Quebec, Canada, April, 2005. Libbey, H.P. (2004, September). Measuring stude nt relationships to school: Attachment, bonding connectedness, and engagement. Journal of School Health, 74 (7), 274-283.
167 Lickona, T. & Davidson, M. (2005). Smart & good high schools: Integrating excellence and ethics for success in school, work and beyond Cortland, NY: Center for the 4th and 5th Rs. Washington, DCL Charac ter Education Partnership. Limber, S. P. & Small, M. S. (2003). Mini-s eries: Bullying prevention and intervention: Integrating research and evaluation findi ngs. State laws and policies to address bullying in schools. School Psychology Review, 32 (3), 445-456. Lynch, E.W., Lewis, R.B., & Murphy, D.S. (1992) Educational services for children with chronic illnesses: perspectives of educators and families. Exceptional Child, 59 210-220. Ma, X. (2003, July/August). Sense of belonging to school: Can schools make a difference? Journal of Educational Research, 96 340-349. Malecki, C.K. (2003). Perceptions of the fr equency and importance of social support by students classified as victims, bullies, and bully/victims in an urban middle school. School Psychology Review, 32 (3), 471-489. Marzano, R.J. (2003). What works in schools: Trans lating research into action Alexandria, VA: Association for Superv ision and Curriculum Development. Mason, H. (2003, December 30). Teens and bullying: Whos taking abuse? The Gallup Organization. Retrieved September 13, 2004 from http://web.lexisnexis.comproxy.usf.edu/uniiverse MCall, R.B. & Appelbaum, M. I. (1991). So me issues of conducting secondary analyses. Developmental Psychology 27 (6), 911-917.
168 McEvoy, A. & Welker, R. (2001). Antisocial behavior, acad emic failure and school climate In H.M. Walker & M.H. Epstein (Eds) Making schools safer and violence free: Critical issues solutions and recommended practices (pp. 28-38). Austin, TX: PRO-ED Inc. McEwan, B. (2000). The art of classroom management: Effective practices for building equitable learning communities. Upper Saddle, NJ: Prentice-Hall. McGlynn, C. (2004, September 22). Healthy sc hools healthy students: Positive school culture promotes student success and well-being. The Journal of Addiction and Mental Health Retrieved March 5, 2005 from www.highbeam.com McNeely, C. & Falci, C. (2004, September). School connectedness and the transition into and out of health risk behavi ors among adolescents: A comparison of social.belonging and teacher support. Journal of School Health, (74) 7, 284-292. McQuillan, P.J. (2005, winter). Possibilities an d pitfalls: Comparative analysis of student empowerment. American Educational Research Journal, 42 (4), 639-670. Mendler, A.N. (2001). Connecting with students. Alexandria, VA.: Association for Supervision and Curriculum Development. Miller, G.E., Brehm, K., & Whitehouse, S. (1998). Reconceptualizing school-based prevention for antisocial behavior within a resiliency framework. The School Psychology Review, 27 (3), 364-379. Minnesota Department of Health (2002). Being, belonging, becoming: Minnesotas adolescent health action plan: 2002 executive summary St. Paul, MN: Author. Minnesota Womens Fund (1992). Reflection of risk: Growing up female in Minnesota Minneapolis, MN.: Amhe rst Wilder Foundation.
169 Mishna, F. (2003, July 1). Learning disa bilities and bullyin g: Double jeopardy. Journal of Learning Disabilities 36 (4), 336-347. Moore, H. (2002, August 1). Peer relations of youth with pediat ric conditions and health risks: Promoting social support and hea lthy lifestyles. Journal of Developmental & Behavioral Pediatrics, 23, 271-280. Morrison, B. (2002). Bullying and victimization in sc hools: A restorative justice approach. Canberra Australia: Australian Inst itute of Criminology Trends and Issues in Crime and Criminal Justice. Morrison, B. (2001). Restorative justice and school violence: Building theory and Practice [Electronic version]. International Institute for Restorative Practices. Paper presented March 2001 at Internati onal Conference on Violence in Schools and Public Policies, Paris France. Mounts, N.S. (1997, summer). What about gi rls? Are they rea lly not aggressive? Human Development and Family Life Bulletin: A Review of Research and Practice, 3 (2). Murray, C. & Malmgren, K. (2005). Implem enting a teacher-student relationship program in a high poverty urban schoo l: Effects on social, emotional, and academic adjustment and lessons learned. Journal of School Psychology 43 137-152. Nakuramura, R.M. (2000). Healthy classroom management: Motivation, communication and discipline. Belmont, CA: Wadsworth Thomson Learning. Nansel, T. R., Overpeck, M., Haynie, D. L ., Ruan, J., & Scheidt, P. C. (2003). Relationships between bullying and violence among U.S. youth. Archives of Pediatrics and Adol escent Medicine, 157 (4), 348-355.
170 Nansel, T.R., Overpeck, M., Ramani, S., Ru an, W.J., Simons-Morton, B., & Scheidt, P. (2001, April 25). Bullying behavi ors among US youth: Prevalence and association with psychosocial adjustment. Journal of the American Medical Association, 285 (16), 2094-2100. National Association of A ttorneys General. (1999). Protecting students from harassment and hate crime Washington D.C.: U.S. Department of Education, Office for Civil Rights. Newcomer, S. & Udry, J.R. (1988). Adolescent s honesty in a survey of sexual behavior. Journal of Adolescent Research, 3 (3-4), 419-423. NIDA (n.d.). Preventing drug abuse among children and adolescents. Retrieved February 23, 2007 from http://www.nida.ni h.gov/Preventio n/risk.html Niebuhr, R. E. (1999, June 22). An empirical st udy of student relationships and academic achievement. Educational Research, 87, 5-17 Nixon, C. (2005, October 28). Peering into the world of covert aggression: Relational aggression Paper presented at Na tional Bullying Prevention Conference, Atlanta, Georgia. Noddings, N. (1995, May). Teaching themes of care. Phi Delta Kappan, 76 675-679. Noddings, N. (2005, September). What does it mean to educate the whole child? Educational Leadership, 6 (1), 8-13. Noll, R.B. (2003, August 1). Social, emoti onal, and behavioral functioning of children with Hemophilia. Journal of Developmental & Behavioral Pediatrics, 24 225-232. Nunnaly, J. (1987). Psychometric theory New York:McGraw Hill.
171 Olweus, D. (1993). Bullying at school Cambridge, MA: Blackwell Publishers. Olweus, D. (2003, March). A prof ile of bullying in schools. Educational Leadership, 60 (6), 12 17. Olweus, D., Limber S. P., Mullin-Rindler, N ., Riese, J., Flerx, V., & Snyder, M. (2004). The Olweus bullying prevention prog ram training of trainers notebook Clemson, SC: The Olweus Group. OMoore, M. (1998, May15). Critical issues fo r teacher training to counter bullying and victimization. Paper presented at the Eu ropean Conference on Initiatives to Combat School Bullying, London, England. Ophelia Project. (n.d.). Relational aggression the facts Retrieved August 5, 2004 from http://www.opheliaproject.org Osborne, J. W. (2001). A new look at outliers and fringeless: Their effects on statistic accuracy and Type I and Type II error rates. Unpublished manuscript, Department of Educational Research and Leader ship and Counselor Education, North Carolina State University. Packman, J. (2005). Education: Were not going to take it: A student driven anti-bullying approach Retrieved August 28, 2005 from http://www.highbeam.com Paige, R. ( 2002, June 2). Education Depart ment and Secret Service Release Threat Assessment Guide for Schools [Press releas e]. Washington, DC: U.S. Department of Education.
172 Payton, J.W., Wardlaw, D.M., Graczyk, P.A .., Bloodworth, M.R., Tompsett, C.J., & Weissberg, R.P. (2000, May). Social and emotional learning: A framework for promoting mental health and reducing risk behavior in children and youth. Journal of School Health, 70 (5), 179-185. Pellegrini, A.D. & Bohn, C.M. (2005, January/February). The role of recess in childrens cognitive performance and school adjustment. Educational Researcher, 34 (1), 13-17. Pepler, D.J. (1998, November 1). Observat ions of bullying in the classroom. The Journal of Educational Research Retrieved June 12, 2006 from www.highbeam.com Pepler, D.J. & Craig, W. (2000, April). Making a difference in bullying LaMarsh Centre for Research on Violence and Conf lict Resolution, York University. Perkins, B. K. (2006). Where we learn: The CUBE survey of urban school climate. Alexandria, VA: National Sc hool Board Association. Perlstein, L. (2004). Allies can reach middle schoolers. The Education Digest 70 (1). 56-58. Peterson, R.L. & Skiba, R. (2000, March 22) Creating school clim ates that prevent school violence. Preventing School Failure Retrieved September 7, 2007 from http://www.highbeam.com Pharris, M. D. (2002). Coming to know our selves as community through a nursing partnership with adolesce nts convicted of murder. Advances in Nursing Science, 24 (3), 21-42. Pierson, R. (2005, February 27). Giving the best that youve got. Paper presented at the National Students Assistance Conference, Atlanta GA.
173 Ragozzino, K., Resnik, H., Utene-Obrie n, M., & Weissberg, R.P. (2003, summer). Promoting academic achievement through social emotional learning. Educational Horizons 169-171. Rasinski, K.., Visser, P.S., Zagatsky, M., & Richert, E.M. (2005, May). Using implicit goal priming to improve the quality of self-report data. Journal of Experimental Social Psychology, 41 (3), 321-327. Reid, P.G., Reid, R., & Peterson, N.A. (2005, May). School engagement among latino youth in an urban middle school context: Valuing the role of social support. Education and Urban Society, 37 (3), 257-275. Reiter-Purtill, J., Gerhardt, C.A., Vannatta, K., Passo, M.H., & Noll, R.B. (2003). A controlled longitudinal study of the social functioning of children with juvenile rheumatoid arthritis. Journal of Pediatric Psychology, 28 (1), 17-28. Resnick, M., Bearman, P., Blum, R., Bauman, K., Harris, J., Tabor, J., et al. (1997). Protecting adolescents from harm: Findi ngs from the National Longitudinal Study on Adolescent Health, JAMA, 278 (10), 823-832. Resnick, M.D., Ireland, M., & Borowsky, I. (20 04). Youth violence perpetration: What protects? What predicts? Findings fro m the national longitudinal study of adolescent health. Journal of Adolescent Health, 35 (5), 423e1-424e10. Rigby, K. (2003, October). Consequences of bullying in schools. Canadian Journal of Psychiatry, 48, 583-590. Rigby, K. (2000). Effects of peer victimi zation in schools and perceived social support in adolescent well being. Journal of Adolescence, 23 57-68.
174 Rinehart, J. S. (1993, June 22). Teac her empowerment and school climate. Education Retrieved January 27, 2004 from http://www.highbeam.com. Roach, A.T. & Kratochwill, T.R. (2004). Eval uating school climate and school culture. Teaching Exceptional Children, 37 (1), 10-17. Roberts, M.C. (2003). Handbook of pediatric psychology New York: Guilford Press. Robbins, L. & Rutter, M. eds. (1990). Straight and devious pathways from childhood to adulthood. NY: Cambridge University Press. Rodkins, P.C. & Hodges, E.V.E. (2003). Bullies and victims in the peer ecolgy: Four questions for psychologist and school professionals. School Psychology Review, 32 (3), 384-400. Sanders, C.E. & Phye, G.D. (2004). Bullying: Implications for the classroom New York: Academic Press. Schaps, E. (2003, March). Creating a school community. Educational Leadership, 60 (6), 31-33. Schwartz, D., Dodge, K.A., Pettit, G.S., & Bate s, J.E. (2000). Friendship as a moderating factor in the pathway between early hrsh home environment and later victimization in the peer group. Developmental Psychology, 36 646-662. Search Institute (1991). The troubled journey: A profile of American youth Minneapolis, MN: Lutheran Brotherhood. Sergiovanni, T. J. (1992). Moral leadership: Getting to th e heart of school improvement. San Francisco, CA: Jossey-Bass Publishers.
175 Shariff, S. (2005, winter). Cyber-dilemmas in the new millennium: School obligations to provide student safety in a virtual school environment. McGill Journal of Education, 40 (3), 457-477. Shapiro, A. (2000). Leadership for constructivist schools Lanham, MD: Scarecro Press, Inc. Shea, B.M. (2003). Social exile: The cycle of peer vi ctimization for children with ADHD Unpublished masters thesis, Univer sity of Toronto, Toronto, Canada. Siris, K. & Osterman, K. 2004, December). Interrupting the cycle of bullying and victimization in the elementary classroom. Phi Delta Kappan, 288-291. Sprague, J. R. & Walker, H. M. (2005). Safe and healthy schools: Practical prevention strategies. New York: Guilford Press. Smith, H. (2004, October 1). School connectedne ss, anger behaviors and relationships of violent and nonviolent American Youth Perspectives in Psychiatric Care. Retrieved March 16, 2005 from www.highbeam.com Smith, P.K. & Ananiadou, K. (2003, April). The nature of school bullying and the effectiveness of school -based interventions. Journal of Applied Psychoanalytic Studies, 5 (2), 189-209. Smith, P. K. & Brain, P. (2000). Bullying in schools: Lessons from two decades of research. Aggressive Behavior, 26, 1-9. Spriggs, A.L., Iannotti, R.J., Nansel, T.R. & Haynie, D.L. (2007). Adolescent bullying involvement and perceived family, peer and school relations: Commonalities and differences across race/ethnicity. Journal of Adolescent Health, 41 283-293.
176 Soloman, D., Battistich, V., Watson, M., Schaps, E., & Lewis, C. (2000). A six district study of educational change: Direct and me diated effects of the child development project. Social Psychology of Education, 4 3-51. Sprague, J. R. & Walker, H. M. (2005). Safe and healthy schools: Practical prevention strategies. New York: Guilford Press. Stolp, S. (1995). Every school a community: The academi c value of strong social bonds among staff and students Eugene, OR: Oregon School Study Council. Storch, E.A., Brassard, M.R.. & Masia-Warn er, C.L. (2003). The relationship of peer victimization to social anxiety and loneliness in adolescence. Child Study Journal, 33, 1 18. Telljohann, S.K. (2003, May 1). The nature and extent of bullying at school. Journal of School Health, 73 (5), 173-181. Thorpe, P.K. (2003, September). School context, student c onnectedness and mathematics classroom performance. Wichita KS: Wichita Public Schools. (ERIC Document Reproduction Service No. Ed480670). Thorpe, P.K. (2004). A mediation model relating teacher ratings of student achievement to student connectedness at school Paper presented at American Education Research Association Conference, Chicago. IL. UCLA School Mental Health Project (1998, spri ng). Enabling learning in the classroom: A primary mental health concern. Addressing Barriers to Learning, 3 (2), 1-6, 10-11.
177 University of Minnesota, 2003. Wingspread declaration: A national strategy for improving school connectedness Retrieved 2/15/06 from http://www.allaboutkids.umn.edu/Wi ngfortheWeb/schooldeclaration.pdf Unnever, J.D. & Cornell, D.G. (2003a). The culture of bullying in middle school. Journal of School Violence, 2 (2), 5-27. Unnever, J.D. & Cornell, D.G. (2003b, Febr uary). Bullying, self-control, and ADHD. Journal of Interpe rsonal Violence, 18 (2), 129-147. U.S. Department of Educa tion (n.d.). Character and citi zenship: Are schools safe without them? The Challenge. 12 (4). Washington, DC: Author. U.S. Department of Education (1998). Preventing bullying: a manual for schools and communities Washington DC: Department of Education. U.S. Department of Education (2002, Marc h). Preventing school shootings: A summary of a U.S. Secret Service Sa fe School Initiative Report. National Institute of Justice Journal (248), 11-15. U.S. Department of Education, Offi ce of the Under Secretary, (2003). No Child Left Behind: A Toolkit for Teachers Washington, D.C.: Author. U.S. Department of Justice, Bureau of Justice Statistic s (2004, November). School Crime Supplement to the National Crim e, Victimization Survey, 2003 retrieved July 21, 2006 from http://nces.ed.gov/pubs2005/2005002.pdf
178 U.S. Department of Justice, Bureau of Justice Statistics (2006, December 4). National Crime Victimization Survey: School Crime Supplement, 2005 [Computer file]. Conducted by U.S. Dept. of Co mmerce, Bureau of the Census. ICPSR04429-v1. Ann Arbor, MI: Inter-unive rsity Consortium of Political and Social Research [produc er and distributor]. U.S. Newswire (2003, December 2). Release of 2003 national school climate survey sheds light on experiences of lesbian, gay, bisexual, transgender (LGBT) students New York: Author. Varlas, L. (2005, August). Bridging the widest gap: Raising the achievement of black boys. Education Update, 47 (8). 1-3, 8. Veenstra, R., Lindenber, S., Oldehinkel, A.J ., DeWinter, A.F., Verhulst, F.C., & Ormel, J. (2005). Bullying and victimization in elementary schools: A comparison of bullies, victims, bully/victims, and uninvolved preadolescents. Developmental Psychology, 41( 4), 672-682. Viadero, D. (2003, May 14). Two studies high light links between violence, bullying by students. Education Week Retrieved November 14, 2005 form http://www.wdweek.org Vincent, P., Wangaard, D. & Weimer, P. (2004). Restoring school civility. Chapel Hill, N.C.: Character Development Group. Walker, D. & Lambert L., (1995). Learning an d Leading Theory in Lambert, L., Walker, D., Zimmerman,D.P., Cooper, J.E., Lamb ert. M. D., Gardner, M.E. & Slack, P.J.F. (1995) The constructivist leader New York: Teachers College Press.
179 Weinhold, B.K. (2000). Bullying and school violence: The tip of the iceberg. Teacher Educator, 35 (3), 28-33. Weinhold, B.K. (1999, September). Bullying and school violence: The tip of the iceberg. Counseling Today. Alexandria, VA: American Counseling Association. Weissberg, R.P. & OBrien, M.U. (2004, January ). What works in school based social and emotional learning programs for positive youth development. Annals, AAPSS, 591 86-97. Welker, R. (2000, September 22). Antisocial behavior, academic failure, and school climate: A critical review. Journal of Emotional and Behavioral Disorders Retrieved January 27, 2004 from http://www.highbeam.com Wentzel, K.R. (2002). Are effective teacher s like good parents? Teaching styles and student adjustment to early adolescence. Child Development, 73 287-301. Werner, E. E. (1994). Overcoming the odds. Developmental and Behavioral Pediatrics, 15 (2), 131-136. Willard, N. (2005). A parents; guide to cyberbullying and cyberthreats. Center for Safe and Responsible Internet use Wilson, D. (2004, September). The interface of school climate and school connectedness and relationships with aggressions and victimization. Journal of School Health, 74 (7), 293-299. Wilson, D.B., Gottfredson, G.D., & Najaka, S.S. (2001). School-based prevention of problem behaviors: A meta-analysis. Journal of Quantitative Criminology, 17 247-272.
180 Windemeyer Communications. (2003). National bullying prevention campaign formative research report. Washington D.C.: U.S. Department of Health and Human Services. Yegidis, B.L., & Weinbach, R.W. (1991). Research methods for social workers White Plains, NY: Longman. Young, E.L., Boye, A.E., & Nelson, D.A. (2006) Relational aggressi on: Understanding, identifying, and responding in schools. Psychology in the Schools, 43 (3), 297312. Yude, C. & Goodman R. (1999). Peer probl ems of 9 to 11 year old children with hemiplegia in mainstream schools. Can these be predicted? Developmental Medicine and Child Neurology, 41 4-8. Zins, J.E., Bloodworth, M.R., Weissberg, R.P ., & Walberg, H.J. (2004). The scientific base linking social and emotional learning to school success. In Building academic success on social and emotional le arning: What does the research say? (pp. 3-22). Danvers, MA: Teachers Co llege, Columbia University.
182 Appendix A: School Crime Supplement to th e National Crime Victimization Survey
183 Appendix A: (Continued)
184 Appendix A: (Continued)
185 Appendix A: (Continued)
186 Appendix A: (Continued)
187 Appendix A: (Continued)
188 Appendix A: (Continued)
189 Appendix A: (Continued)
190 Appendix A: (Continued)
191 Appendix B: Pearson Correlation Coeffi cients for Variables with Bullied N = 5780 Prob > |r| under H0: Rho=0 emotsafe relation extracurr race VS014 VS028 VS029 VS030 VS031 -0.12215 -0.00271 0.05198 -0.01387 -0.12317 -0.01989 0.02327 0.07905 0.01949 <.0001 0.8367 <.0001 0.2919 <.0001 0.1305 0.0769 <.0001 0.1384 VS032 VS033 VS034 VS035 VS036 VS037 VS038 VS039 VS040 0.03263 0.01495 0.03264 -0.02946 -0.02393 0.01273 0.03523 -0.00600 0.00166 0.0131 0.2557 0.0131 0.0251 0.0689 0.0.3333 0.0074 0.6484 0.8997 VS041 VS042 VS043 VS045 VS046 VS047 VS048 VS049 VS050 0.04891 -0.02798 -0.00648 -0.13732 -0.10277 0.07433 0.08325 0.08583 -0.03913 0.0002 0.0334 0.6221 <.0001 0.0334 <.0001 <.0001 <.0001 <.0001 VS051 VS052 VS053 VS054 VS055 VS116 VS123 VS126 VS129 -0.07304 0.00246 0.00747 -0.00439 -0.01465 0.09150 -0.06078 -0.10001 0.27340 0.0029 0.8515 0.5701 0.7384 0.2654 <.0001 <.0001 <.0001 <.0001 VS130 VS131 VS133 V3018 V3024 0.13912 -0.02449 -0.08154 0.02978 0.06159 <.0001 0.0627 <.0001 0.0236 <.0001 Variable Label bully emotsafe relation extracurr race VS014 SCHOOL GRADE LAST 6 MONTHS VS028 EXTRA-CURR: ATHLETICS VS029 EXTRAC-CURR: SPIRIT GROUPS, PEP VS030 EXTRA-CURR: ARTS VS031 EXTRA-CURR: ACADEMICS VS032 EXTRA CURR: SCHOOL GOVERNMENT VS033 EXTRA-CURR: SERVICE CLUBS VS034 EXTRA-CURR: OTHER SCHOOL ACTIVITIES VS035 SCHOOL SAFETY: SECURITY GUARDS VS036 SCHOOL SAFETY: STAFF/ADULTS IN HALLWAY VS037 SCHOOL SAFETY: METAL DETECTORS VS038 SCHOOL SAFETY: LOCKED DOORS VS039 SCHOOL SAFETY: VISITORS SIGN IN VS040 SCHOOL SAFETY: LOCKER CHECKS VS041 SCHOOL SAFETY: BADGES
192 Appendix B: (Continued) VS042 SCHOOL SAFETY: SECURITY CAMERAS VS043 SCHOOL SAFETY: CODE OF CONDUCT VS045 SCHOOL RULES ARE FAIR VS046 SAME PUNISHMENT FOR BREAKING RULES VS047 SCHOOL RULES STRICTLY ENFORCED VS048 STUDENTS KNOW PUNISHMENTS VS049 TEACHERS TREAT STUDENTS WITH RESPECT VS050 TEACHERS CARE ABOUT STUDENTS VS051 TEACHERS MAKE STUDENTS FEEL BAD VS052 ADULT AT SCHOOL WHO CARES ABOUT ME VS053 SCHOOL HAS ADULT WHO HELPS W PROBLEMS VS054 HAVE FRIEND AT SCHOOL TALK TO VS055 FRIEND AT SCHOOL WHO HELPS W PROBLEMS VS116 STAY HOME: THOUGHT SOMEONE ATTACK OR HARM YOU VS123 KNOW STUDENTS BROUGHT GUN TO SCHOOL VS126 GANGS AT SCHOOL VS129 HOW OFTEN DIST RACTED BY STUDENTS MISBEHAVING VS130 HOW OFTEN TEACHERS PUNISH STUDENTS VS131 SKIP CLASSES VS133 GRADES V3024 HISPANIC ORIGIN
About the Author Janet Urbanski received a Bachelor of Science degree in Elementary Education from Eastern Connecticut State College in 1982, a Master of Ar ts degree in Guidance and Counseling from the University of South Fl orida in 1991, and an Educational Specialist degree in Educational Leadership from the Un iversity of South Florida in 2003. She is currently the supervisor of the Pinellas County Schools Safe and Drug Free Schools Office. She has experience as a Prevention Spec ialist, an elementary teacher, and school counselor. She is a national trainer for the Olweus Bullying Preven tion Program and has conducted workshops on bullying prevention, co nflict resolution and character education at the local, state, and national levels.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001935519
007 cr mnu|||uuuuu
008 080424s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002236
The relationship between school connectedness and bullying victimization in secondary students
h [electronic resource] /
by Janet Urbanski.
[Tampa, Fla.] :
b University of South Florida,
ABSTRACT: Bullying is a complex behavior that can cause academic and social problems for students and can contribute to a negative school climate. Students who feel isolated or do not feel connected to their school may experience similar risks to those who are victimized by peers. Recent school violence incidents have led to an increase in bullying behavior research. The importance of the school climate is also emerging in educational discourse prompting a growth of research in school connectedness and positive relationships. However, research on the impact that relationships and school connectedness may have on bullying victimization at school is limited. This is a secondary analysis of a national data set from the 2005 administration of the National Crime Victimization Survey School Crime Supplement.The study focused on the relationship between school connectedness and bullying victimization and whether gender, race, grade level, and academic achievement moderate the relationship. The role of relationships in bullying victimization was considered. Weighted regression analyses were conducted to determine the relationship between bullying behaviors and school connectedness and to identify the combination of factors that may influence the relationship. Components of school connectedness identified through factor analysis were statistically significant predictors of occurrence and frequency of bullying victimization, but accounted for a very small amount of variance in the outcome. Adding demographic variables of race, gender, grade level, and academic achievement produced a slight increase in the proportion of variance accounted for.Race did not have a statistically significant impact on occurrence of bullying victimization; neither race nor gender was statistically significant in variance of frequency of bullying victimization. Peer relationships proved to be statistically significant in bullying victimization frequency but neither adult-student nor peer relationships were statistically related to bullying victimization occurrence. Overall, school connectedness predicted a very small proportion of variance in occurrence and frequency of bullying victimization, suggesting that bullying prevention efforts should include strategies beyond those to improve a student's sense of connectedness to school. A comprehensive approach is needed to address bullying in schools effectively.
Dissertation (Ed.D.)--University of South Florida, 2007.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 192 pages.
Advisors: Arthur Shapiro, Ph.D. and Steve Permuth, Ed.D.
Social emotional learning.
Risk and protective factors.
x Educational Leadership
t USF Electronic Theses and Dissertations.