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 001937895
007 cr mnu|||uuuuu
008 080505s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002177
Mihalas, Stephanie Tamara.
Positive protective factors as moderators in the relationship between relational victimization and depression in minority adolescents
h [electronic resource] /
by Stephanie Tamara Mihalas.
[Tampa, Fla] :
b University of South Florida,
ABSTRACT: This study explored the relationship between relational victimization and depression among middle school students (n = 153) in an urban school in Florida. The majority of participants were African-American and Hispanic at-risk youth. This study is one of the first to study how positive protective factors (i.e., hope, spirituality, perceived social support) moderate the relationship between victim status and depression. A mixed methods design was used to gain further insight into the survey data collected. Findings from the study indicated that hope and perceived social support were statistically significant moderator variables. Additionally, results from the qualitative interviews suggested that teachers, parents, and siblings play an important role in supporting victimized students. Implications for gender and culturally sensitive interventions are discussed. Possible avenues for future research are also outlined.
Dissertation (Ph.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 203 pages.
Adviser: Linda Raffaele Mendez, Ph.D.
Perceived social support.
x School psychology
t USF Electronic Theses and Dissertations.
Positive Protective Factors As Moderators In The Relationship Between Relational Victimization And Depression In Minority Adolescents by Stephanie Tamara Mihalas A dissertation submitted in partial fulfillment of the r equirements for the degree of Doctor of Philosophy Department of P sychological and Social Foundations College of Education University of South Florida Major Professor: Linda Raffaele Mendez Ph.D. Robert Dedrick, Ph.D. Shannon Suldo, Ph.D. Richard Weinberg, Ph.D. Date of Approval: July 20, 2007 Keywords: hope middle school, perceived social support, spirituality, urban Copyright 2007 Stephanie Mihalas
Acknowledgements While I tend to reflect about my life and all the actions that I take on a continual basis, I find now is the time to take a few moments to openly reflect and thank those persons who have inspired me and worked closely with me to complete this major feat: my dissertation. I have to admit that the strength I felt on a day to day basis to persevere through graduate school was largely in part due to the love and guidance of my parents, Dr. and Mrs. Lawrence and Sharon Mihalas. Second, I would like to thank Keith for making me laugh when I wanted to cry. His encouragement and undying love will never be forgotten. Third, my puppy Yang always knew when it was time for me to take a break from the computer. I thank him for the play or feeding break s which helped me to re center and focus better Finally, the support of my committee was overwhelmingly positive. My committee worked with me to make important deadlines and always provided valuable insight and care, both personally and professionally.
i Table of Contents List of Tables v List of Figures vii Abstract v iii Chapter I: Introduction 1 Relational Aggression and Victimization 2 Significance of R esearch on Relational Victimization 3 Rationale for Current Study 3 Purpose of the Current Study 4 Summary and Description of the Research 5 Research Questions 6 Apriori Hypotheses 7 Definition of Terms 8 Chapter II: Review of the Literature 10 Introduction 10 Relational Aggression 11 Victims of Aggression 13 Victim Classification 14 Victim Profile 16 Ri sk Factors for Victimization 20 Contextual Risk Factors 20 Interpersonal Risk Factors 26 Intrapersonal Risk Factors 27 Protective Factors 29 Social Support 29 Connection to School 31 Intrapersonal Factors 32 Relational Victimization 33 Psychopathology 33 Coping Style 35 Summary and Future Research 36 Current Study 38 Risk versus Protective Factors 4 1 Positive Psychology 42
ii Perceived Social Support 43 Intrapersonal Protective Factors 44 Outcomes of Victimization 46 Chapter III: Method 48 Overview 48 Setting 48 Participants 50 Procedures 51 Pre Data Collection 51 Data Collection 55 Post Data Collection 56 Interviews 57 Measures 59 The Childrens Depression Inventory Short Form (CDI S) 5 9 The Child and Adolescent Social Support Scale (CASSS) 60 The Spirituality Assessment Instrument (SSA) 62 The Childrens Hope Scale (CHS) 63 Social Experience s Questionnaire (SEQ) 65 Data Analysis 67 Analysis Software 67 Descriptive Statistics 67 Relationships among Predictor and Outcome Variables 68 Comparison of Groups 68 Multiple Regression 68 Rationale for Model Testing 69 Ch apter IV: Results 71 Overview 71 Treatment of the Data 71 Reliability of Measures 72 CDI S 73 SEQ 73 CHS 74 CASSS 74 SSA 74 Confirmatory Factor Analysis 75 Demogr aphics 76 Defining Charact eristics of Each Survey 79 Pe rceived Social Support 79 Spirituality 85 Depression 88 Relational Victimization 91 Hope 9 2 Research Question 1 Results 96
iii Research Question 2 Results 98 Research Question 3 Results 101 Research Question 4 Results 104 Spirituality as a Moderator Variable 105 Perceived Social Support a s a Moderator Variable 105 Hope as a Moderator Variable 106 Research Question 5 Results 108 Student Demographics 109 Perceptions of Victimization 110 Location and Frequency of Victimization 111 Safety Issues 112 Coping Strategies 113 Persons who Support Victims 116 Perceptions of Teachers 116 The Role of Victimization on Well Being 118 Cha pter V: Discussion 120 Summary of Study 120 Incid ence of Relational Victimization 120 Rationale for Lying 122 Relationship between Victimization and Depression 123 Perpetrators of Victims 125 Coping Mechanisms 125 Perceived Social Support 125 Hope 128 Spirituality 130 Important Gender Implications 131 L imitations 135 Sample Size 135 Survey Instruments 135 Member Checking 137 Future Research 137 Final Thoughts 140 References 142 Appendices 169 Appendix A Active Parent Consent Form 170 Appendix B Child Assent Form 172 Appendix C Informational Letter in Spanish 174 Appendix D The Childrens Depression Inventory Short Form 175 Appendix E The Child and Adolescent Social Support Scale 176 Appendix F The Spirituality Assessment Instrument 178 Appendix G The Childrens Hope Scale 179 Appendix H Childrens Social Experiences Questionnaire Self Report 18 0
iv Appendix I Semi structured Interview Questions 181 Appendix J Student Demographics Form 182 Appendix K Effect Sizes for Survey Instruments 184 Appendix L Reliability Estimates for All Survey Instruments 189 Appendix M Moderator Model Summaries 191 About the Author End Page
v List of Tables Table 1 Characteristics of Participating Schools 48 Table 2 Effect Sizes for Race and Gender for Spirituality on the SSA 185 Table 3 A nalysis of Categorical Variables for Differences in Participant Sample 79 Table 4 Characteristics of the Social Support Measure (CASSS) by Gender 80 Table 5 Characteristics of the Social Support Measure (CASSS) by Ethnicity 82 Table 6 Characteristic s of the Social Support Measure (CASSS) by Ethnicity and Gender 84 Table 7 Correlation Matrix of Subscales on Perceived Social Support Measure (CASSS) (N = 153) 85 Table 8 Characteristics of the Spirituality Measure (SSA) by Ethnicity 86 Table 9 C haracteristics of the Spirituality Measure (SSA) by Gender 87 Table 10 Characteristics of the Spirituality Measure (SSA) by Gender and Ethnicity 88 Table 11 Characteristics of the Depression Measure (CDI S) by Gender 8 9 Table12 Characteristics of the Depression Measure (CDI S) by Ethnicity 90 Table 13 Descriptive Statistics of the Depression Measure (CDI S) by Ethnicity and Gender 91 Table 14 Characteristics of the Victimization Measure (SEQ) by Gender 92 Table 15 Characteristics of the Hope Measure (CHS) by Gender 93 Table 16 Characteristics of the Hope Measure (CHS) by Ethnicity 94 Table 17 Characteristics of the Hope Measure (CHS) by Ethnicity and Gender 95 Table 18 Sum Scores and Standard Deviations Obtained in Current Study Compare d to Referred Sample 95 Table 19 Percentage of Sample Reporting Various Forms of Relational Victimization 97 Table 20 Percentage of Males and Females Who Reported Varying Degrees of Victimization (N = 152) 98 Table 21 Categorical Representation of Co ping Strategies for All Participants (N = 152) 99
vi Table 22 Comparisons of Mean Scores for Coping Strategies by Gender and Ethnicity 100 Table 23 MANOVA Summary Table for Coping Strategies (N = 153 ) 101 Table 24 Correlations Among Predictor, Moderat or, and Criterion Variables (N = 152) 102 Table 25 Emotional Responses to Bullying Behavior 119
vii List of Figures Figure 1 Conceptual Moderator Model for Depression 69 Figure 2 Perceived Social Support as a Moderat or b etween Relational Victimization and Depression 106 Figure 3 The Contribution of Hope as a Moderator for Depression in Females 107 Figure 4 The Combination of Hope as a Moderator for Depression in Males 108 Figure 5 Coping Strategies Reported by P articipant 115
viii Positive Protective Factors as Moderators in the Relationship Between Relational Victimization and Depression in Minority Adolescents Stephanie Tamara Mihalas ABSTRACT This study explored the relationship between relational vict imization and depression among middle school students ( n = 153) in an urban school in Florida. The majority of participants were African American and Hispanic at risk youth. This study is one of the first to study how positive protective factors (i.e., h ope, spirituality, perceived social support) moderate the relationship between victim status and depression. A mixed methods design was used to gain further insight into the survey data collected. Findings from the study indicated that hope and perceived social support were statistically significant moderator variables. Additionally, results from the qualitative interviews suggested that teachers, parents, and siblings play an important role in supporting victimized students. Implications for gender and culturally sensitive interventions are discussed. Possible avenues for future research are also outlined.
1 Chapter 1 Introduction Gaining a better understanding of the phenomenon of aggression has been a popular focus of research in the social sciences over the past several decades. Aggression is known to have a substantial negative impact on all individuals involved. Although long recognized that aggression may take various forms, the majority of research on aggression has focused on its physical and verbal forms (e.g., Connor, Steingard, Anderson, & Melloni, 2003). As a result, most aggression research has focused on boys based on the fact that boys engage in more physical and verbal aggression than girls (Rauste von Wright, 1989). However, research suggests that when the definition of aggression is expanded to include social ostracism and rumors, girls ar e equally as aggressive as boys (Tapper & Boulton, 2000; Young, Boye, & Nelson, 1996). As research on aggression has evolved to include its more social forms, researchers have identified a specific type of aggression designed to damage social relationship s that has come to be known as relational aggression (Crick & Nelson, 2002). Relational aggression is indirect and covert (sometimes overt) and usually verbal in nature. It occurs in a variety of settings including schools, community activities, and other social venues where people have the opportunity to engage in discussion. Research on relational aggression to date has focused on a variety of issues, including risk and resiliency factors (e.g., Campbell & Frabutt, 1999; Yoon, H ughes, Cavell, & Thompson, 2000 ), assessment tools (e.g., Crick, 1996), adjustment trajectories (e.g., Prinstein, Boergers, & Vernberg, 2001; Tomada & Schneider, 1997;), and effective interventions (e.g., Leff, Power, Manz, Costigan, & Nabors, 2001; Taub, 2002). Perpetrators of
2 rela tional aggression and those who experience it have been shown to experience short and long term consequences, including academic decline, peer rejection ( Prinstein et al. 2001 ) and maladaptive personality features ( Werner & Crick, 1999 ). While researcher s now recognize relational aggression as a specific form of aggression, there has not been much research investigating the victims of relational aggression. As a result, little is known about the prevalence of victims in school settings, the pathology that victims present, and the kinds of services they receive or seek. The purpose of this study is to expand the literature in this area by focusing on youth who experience relational aggression. The term victim will be used throughout this manuscript to des cribe youth who are targets of relational aggression. This term is not meant to personify youth experiencing relational aggression as disempowered individuals; rather, the term victim provides a degree of consistency across the manuscript and allows for common nomenclature when describing and discussing other studies in this particular area of research. Relational Aggression and Victimization The few studies that have been conducted with victims of relational aggression have found that these individual s experience depression, social isolation, anxiety (Walker & Cillessen, 2006), and Post Traumatic Stress Disorder (PTSD) (Gumpel & Kliewer, 2006). Recently, somatization (specifically abdominal pain) and school absenteeism have been linked to relational vi ctimization (Greco, Freeman, & Dufton, 2007). Moreover, victims of relational aggression have been identified as shy individuals who maintain negative attributions about themselves and the world (Dill, Vernberg, Fonagy, Twemlow, & Gamm, 2004). A study con ducted in Finland found that the kinds of coping strategies
3 that relational victims used were maladaptive, including aggression and self destruction (Olafsen & Viemero, 2000). Significance of Research on Relational Victimization Some may question the si gnificance of studying a construct like relational victimization because of long standing societal beliefs about female behavior suggesting that gossip, misuse of confidential information, breaking trust, and exclusionary group tactics are fairly normative behaviors among many middle and high school adolescent females. The reality, however, is that the available research indicates that relational victimization is associated with maladaptive outcomes for both perpetrators and victims. Since relational aggr ession negatively affects both boys and girls (although possibly to differing degrees) (Crick, 1997), it is critical to develop a greater understanding of victimization including how to appropriately identify victimized students, the pathology they presen t, potential coping mechanisms that may serve as protective factors and/or other long term consequences of victimization. Rationale for Current Study In order for psychologists to build culturally sensitive and individualized interventions for victims of relational aggression, they first must understand the underlying issues that place students at risk for victimization (Roosa & Gonzales, 2000), as well as strength based protective factors that moderate the relationship between victimization and pathology Once protective factors are identified, interventions may be tailored in a more positive manner, focusing less on ameliorating problems in youth and instead focusing on enhancing strengths already inherent within them. Notably, the relational victimiza tion research that has been conducted focuses almost completely on
4 Caucasian participants or persons living outside of the United States (e.g., Osterman, Bjorkqvist, Lagerspetz, Kaukiainen, & Landau, et al., 1998; Pakaslahti, Spoof, Peltola Asplu d, & Jarvi nen Keltikangas, 1998 ). Thus, another focus of this study was to evaluate the experience of relational victimization in minority youth (i.e., African American, Hispanic, and mixed race youth). To date, only one other study has addressed middle school mino rity youth, victimization (not relational victimization), and internalizing disorders (c.f., Peskin, Tortolero, Markham, Addy, & Baumler, 2007). This type of research is consistent with the current focus on culturally competent practice and evidence based interventions promoted by national psychological associations and training programs (Ingraham & Oka, 2006). Purpose of the Current Study The short term goals of the current study were to (a) assess the relationship between depression and relational victi mization; (b) determine variables from the positive psychology literature that serve as protective factors against victimization; and (c) gain a richer understanding of individuals experience of victimization. The long term goal of this study was to eluc i date findings that will aid in the development of interventions for minority students who are victims of relational aggression (Miranda, Bernal, Lau, Kohn, Hwang, & LaFromboise, 2005), ultimately expanding options for treatment (Snowden & Yamada, 2005). Finally, the study was intended to empower victimized youth by giving them a voice to express their thoughts and feelings surrounding the experience of being a victim through qualitative investigation.
5 Summary and Description of the Research The curr ent study took place in two middle schools in a county located in Florida. One school was used primarily as a pilot school to determine any possible limitations that needed to be addressed prior to data collection. Both research sites were characterized by a high percentage of minority students enrolled at each school, a large proportion of the students receiving free and reduced lunch, and a low school performance grade. Participants included any student who received parent consent to participate; however data analysis only included African American, Hispanic, and mixed race students. Data from Caucasian students and students from other minority groups that were obtained during data collection will be used in future studies. All participants completed a battery of instruments to assess a variety of constructs including relational victimization, depression, hope, spirituality, and perceived social support. The specific measures that were used in this study included: The Childrens Depression Inventory S hort F orm (CDI S; Kovacs, 1985), The Childrens Hope Scale (CHS; Snyder, Hoza, Pelham, Rapoff, Ware, et al., 1997), The Social Experiences Questionnaire (SEQ; Crick & Grotpeter, 1996), The Child and Adolescent Social Support Scale (CASSS; Malecki, Demaray & Elliott, 2000), and a scale constructed by the primary investigator adapted from The Spiritual Involvement and Beliefs Scale (SIBS; Hatch, Burg, Naberhaus, & Hellmich, 1998) (i.e., SSA). Additionally, in order to complement the quantitative portion of this study, qualitative methods were utilized (Creswell, 1994) as part of an overall mixed methods approach (see Tashakkori & Teddle, 2003). Specifically, interviews with select victims were completed. In order to gain multiple perspectives on the experi ence of
6 victimization, purposive sampling was used to choose victims who (a) presented with high versus low victimization status, (b) obtained high versus low depression scores, and (c) used a variety of coping mechanisms and strategies. Research Questio ns The specific research questions addressed were as follows: 1. What percentage of minority youth in high risk middle schools have experienced relational victimization? a. Do more females or males report relational victimization? b. What levels of (i.e., how muc h) relational victimization do minority youth in high risk middle schools report experiencing? 2. Which coping strategies/mechanisms (i.e., spirituality, hope, perceived social support) are used and/or cited most frequently by minority youth in high risk middle schools? 3. How is victim status related to rates of depression among minority youth in high risk middle schools? 4. Which protective factor(s) (i.e., spirituality, hope, perceived social support) serve as a moderator(s) between victimization and dep ression among minority youth in high risk middle schools? a. Does this relationship differ based on gender, ethnic group (e.g., Hispanic versus African American), grade, and/or school? 5. How do students experience relational victimization? a. What are students perceptions as to why they are victimized? b. Are students able to verbally define how they cope with relational aggression? c. Who do victims specifically feel supported by, if anyone?
7 d. How does the victimization impact their well being? Apriori Hypotheses T his study is both confirmatory and exploratory in nature and as such, research questions 1, 2, and 5 do not lend themselves to apriori hypothesis development. However, based on previous research findings and clinical judgment gleaned from working with vict ims in the field, hypotheses were created for research questions 3 and 4. The data were expected to support the following hypotheses: 1. There will be a significant positive correlation between relational victimization and depression. This hypothesis is cons istent with research conducted by Walker and Cillessen (2006). 2. Perceived social support will moderate the relationship between relational victimization and depression such that victimized youth with higher levels of social support will report lower levels of depression compared to their victimized peers with lower levels of social support. While no studies have directly assessed this relationship, research conducted by Demaray and Malecki (2002a) have found that perceived social support serves as a protecti ve factor among youth 3. Hope will moderate the relationship between relational victimization and depression such that victimized youth with higher levels of hope will report lower levels of depression than their victimized peers with lower levels of hope. While this direct relationship has not been tested previously, studies conducted by Snyder and colleague s (2002) support the notion that hope is protective 4. Spirituality will moderate the relationship between relational victimization and depression suc h that victimized youth with higher levels of spirituality will report
8 lower levels of depression than their victimized peers with lower levels of spirituality. While this direct relationship has not been tested previously, studies conducted by Mofidi, De Vellis, Blazer, DeVellis, and Porter et al. (2006) have suggested that spirituality has been linked to a decrease in depressive symptoms in adults. Definition of Terms Two particular terms require explanations so that readers share a common definition whi le reading this manuscript: at risk and high crime. The term at risk is defined in this study by the following criteria: (1) participant attendance at a Title I school, (2) participant attendance at a school with a high teacher turnover rate, (3) partic ipant attendance at a school has not met goals for annual yearly progress (AYP) in recent years (i.e., AYP), and (4) participant place of residence, such that neighborhood has a high crime rate. A high crime rate in this study was defined by a crime rate of at least double the modal number of crimes reported in the surrounding counties. According to the local police department, between January and December of 2006, a total of 88 crimes were reported for the grid location in which School B is located and the grid location next to the school location (less than three miles away) had a total of 70 reported cri mes Crimes were defined as murder, rape, sodomy, aggravated assault, larceny, burglary, and vehicle theft. To further understand the crime reports in comparison with the other 230 grids in the county where data were collected, the range of crimes reported ranged from 1 230, with a mode of 4 0 crimes. The other data that are collected by the city are entitled Mandatory Primary Offense, which include ( under this title ) acts of fraud, kidnapping,
9 and drug related charges. The range of offenses in all 230 grids was from 1 512; however, the mode was 25. The grid location for the primary data collection school reported 107 offenses.
10 Chapter 2 Revie w of the Literature Introduction Acts of aggression and violence impact many youth today around the world. Aggression, defined as an act that is intended to injure another either physically or emotionally, is commonplace in the media, in many neighborh oods, and in a majority of our schools. Over 5.7 million youth in the United States reported involvement in aggression and violence on school campuses when asked to fill out a nationwide questionnaire on bullying behavior (Nansel, Overpeck, Pilla, Ruan, S imons Morton, et al., 2001). Specifically, students in grades 6 10 identified themselves as perpetrators (13%), victims (11%), or as perpetrator victims (6%). Victimization is experienced by youth around the world (with research documenting its occurrenc e in Greece, China, The Netherlands, England, and Turkey), making victimization a topic that is at the forefront of developmental and applied research (Paul & Cillessen, 2003). Given that a significant number of students experience aggression to some degr ee in schools, the study of this phenomenon continues to develop and evolve. The approach that educators and psychologists use to define and assess aggression is gradually shifting away from a focus on physical aggression to a focus that encompasses indir ect and social forms of aggression (Cullerton Sen & Crick, 2005). The research base on social forms of aggression, however, remains limited. Hawker and Boulton (2000) found only five studies that assessed relational victimization over a 20 year period.
11 Given that the trajectories of psychosocial adjustment may be poor for many students who engage in, or are victims of, aggressive behavior, more information is needed how educators and mental health practitioners may impact the relationship between victim ization and maladjustment. This chapter provides a summary of the deleterious effects victimization has on children and adolescents. It a lso conveys a number of gaps and limitations in the current research literature, as well as a rationale for the curre nt study. The goals of this literature review will be accomplished by (a) briefly explaining relationally aggressive behavior, (b) explaining characteristics and behaviors of victims, (c) clarifying how aggression impacts victim well being, (d) discussing victim risk and protective factors, and finally (e) highlighting the necessity of the current study and the implications of the study for at risk minority youth populations. Relational Aggression Relational aggression (RA) is an indirect and manipulative form of aggression that intends to harm others through damage to peer relationships in a way that blocks the social goals of the target peer. RA and social aggression differ because RA may be covert or overt whereas social aggression is almost always indir ect. There are no other major differences that researchers agree upon, hence the problem with misidentification of students who are either socially or relationally aggressive. Some authors actually suggest collapsing the two constructs together because of the similarities between them (Archer, 2001). RA interferes with friendships and threatens exclusion from peer groups (Henington, Hughes, Cavell, & Thompson, 1998) Acts of RA may be verbal or nonverbal and include spreading rumors as a form of retaliation, excluding others from
12 play or encouraging others to exclude peers, and social exclusion through gossip. RA has serious implications for social and emotional maladjustment in both males and females across a wide range of age groups, beginning in preschool (Crick, Casas, & Mosher, 1997) and extending into college (Loudin, Loukas, & Robinson, 2003) Importantly, findings suggest that the social psychological implications for college students who engage in relational aggression are highly notable These students demonstrate rejection by peers, display fewer prosocial acts than non aggressors, and exhibit antisocial externalized behavior. Additionally, students in this developmental stage who engage in RA also exhibit borderline personality features (e.g., self destructive behavior, bulimia, anger management problems) (Werner et al., 1999). Interestingly, Crick (1996) found that RA predicts social maladjustment above and beyond what overt aggression predicts alone. Thus, without considering RA as a distinct construct, many researchers would not be able to identify students who would be considered aggressive and likely would not be able to account for a major contributing factor to social maladjustment (i.e., RA). RA research has primarily involved determination of prevalence rates by gender and social outcome expectancies of perpetrators (e.g., Zalecki & Hinshaw, 2004). Researchers hypothesize that RA impacts the psychological well being of females more than males because of the nature of female relationships (Merrell, Buchanan, & Tran, 2006). For example, Frith (2004) noted that female friendships are characterized by intimacy, trust, self disclosure, and rely heavily on supportiv e features. Because female friendships are typically more intense and the value placed on them is high (compared to males), the impact of RA is more significant for females. Along the same lines, Goldstein and Tisak (2004) found that males reported diffe rent outcome expectancies for
13 aggression than females. Females reported that they would feel worse if victimized and rated victimization as more damaging and hurtful to relationships than males did. In addition to the aforementioned variables, the literat ure base on RA also has focused some attention on cognitions that youth embrace in relation to the purpose and outcomes of aggression and victimization (Werner & Nixon, 2005). For example, Werner and Nixon (2005) found that youth who held positive beliefs about RA were more likely to report themselves as aggressors. Notably a number of factors such as minority status, disability status, and socioeconomic status that may bear upon the experience of RA victimization still remain unclear. Further research i s needed to address not only the kinds of psychopathology associated with RA victimization but also to provide a forum for victim voice to be heard. By hearing the voice and personal experiences of students who have been victimized, researchers may begin t o truly understand the experience of relational victimization. Additionally, research is needed to address specific risk or protective factors that serve to heighten or diminish the long term outcomes for victims of RA. Victims of Aggression The current way that many schools attempt to deal with the ramifications of aggression and violence on campus is either to take punitive measures (e.g., suspension, expulsion) and/or focus efforts on universal school based prevention and intervention programs to targ et perpetrators (Batsche & Porter, 2006). One of the reasons that educational organizations may prefer to focus on perpetrators is because the externalized behavior typical of perpetrators is more overt and easier to target (i.e., perpetrators are a more recognizable threat to students and faculty). In contrast, victims of RA in the
14 school system may become marginalized because they do not pose an immediate threat to others and are not as visible as perpetrators. Additionally, school officials often fail to recognize the long term ramifications for victims (Elias & Zins, 2003). While attempts at school wide programming are commendable, schools often lose sight of the importance interventions play for victims. For example, even if a school official is abl e to stop an aggressor from targeting specific students, the emotional experience of the victimizing act may continue to linger for those students who were previously targeted. Adolescents report being victimized to some degree by peers anywhere from 30 % 50% of the time in a typical school year (Evans, Marte, Betts, & Silliman, 2001). Herein lays one rationale for school shootings (e.g., Columbine, Littleton): victims are left to analyze and evaluate their own emotions and thoughts related to their vict imization (Garbarino & DeLara 2002). Rarely are systematic programs present in schools to help victims; instead, the more common course of action is for individuals to seek out school based mental health counseling, which is not available in all schools. Therefore, while helping bullies is crucial, more attention must be given to victims, as they are (a) difficult to identify because they typically present with internalizing problems, and (b) likely to have poorer psychological adjustment in late adolesc ence through adulthood compared to same aged peers who do not experience victimization (Christiansen & Evans, 2005; Parker & Asher, 1987). Some researchers also have found that victims are predisposed to externalizing problems (Peskin et al., 2007). Vict im classification 1 One of the few studies in the victimization literature that used minority youth participants was a study conducted by Graham, Bellmore, and 1 Note. Victimization in this section perta ins mainly to direct and some indirect forms of aggression. However, relational victimization is specifically excluded from this profile synopsis.
15 Juvonen (2003). These authors categorized African American ( n = 350) and Latino ( n = 435) middl e school participants who experienced victimization in three ways: self report victim, peer report victim, and true victim. Self report victims are students who report themselves as victims. Self report victims tend to have more painful memories and expe riences of victimization, but their reports are not always verifiable because they are highly subjective. Peer report victims are students who are nominated by others as a likely target of aggression. Peer report victims usually are the students who are e asily identifiable by a peer group as the class clown, the nerd, or the weirdo. Finally, true victims are students who are identified as victims based on self and peer measurement instruments. True victims are considered to be real victims beca use both objective and subjective measures point to some level of victimization the student has experienced. Self identified victims reported just as much psychopathology as true victims and in some cases, even more pathology (Graham et al., 2003). In t his study, more females reported themselves as self identified victims. Importantly, other researchers have found that negative short term consequences of victimization are only found for females (e.g., change of peer group, sadness, decreased academic per formance) and not for males (Paul & Cillessen, 2003). These findings illustrate the need for gender specific research and interventions in this area Interestingly, if self reports were not used in the Graham et al. (2003) study to interpret the data, 19 2 students would have been misclassified as neither a bully nor a victim. Thus, self reports may indicate true psychological maladjustment, whereas peer reports may provide information on social maladjustment. Considerations for the
16 methodology that will be utilized in the current study partially stem from the robust findings from the Graham et al. (2003) study resulting from the use of self report measures. Goldbaum and colleagues (2003) also developed a classification system for victims based on a stu dy of middle school students ( N = 1,241): non victims (low levels of victimization), desisters (high levels of victimization that decreased over time), late onset victims (increasing levels of victimization), and stable victims (consistently high levels of victimization over time). Stable victims suffered the worst inter and intrapersonal problems, suggesting that the cumulative effects of victimization over time intensify the harm to victims in the form of anxiety, withdrawal, and somatization. The stab ility of victimization, according to Paul et al. (2003), suggests that over a four year time frame, victimization was equally stable in elementary and middle school settings (correlation exceeding .70 between years). This study underscores the importance o f intervening early. Victim profile Bullies and victims differ from one another in a number of ways. Regardless of type of victim classification, most victims tend to be introverted, passive, self blaming, sensitive, and overly quiet and cautious (Mynar d & Joseph, 1997). More serious problems associated with victimization include suicidal ideation, Post Traumatic Stress Disorder (PTSD), violence against perpetrators (Boney McCoy & Finkelhor, 1995), and internalizing disorders (e.g., anxiety and depressi on) (Kochenderfer Ladd & Ladd, 2001). Between 5 % and 10% of victims actually aggress against their perpetrators because they do not have the skills to manage the interaction otherwise (e.g., lack of
17 communication skills to resolve the problem, lack of pro blem solving skills to determine the best approach to resolve the conflict) (Pellegrini, Bartini, & Brooks, 1999). In the school setting, victims are sometimes targeted because they are perceived as unpopular, belong to a rejected peer group (e.g., skat ers, Goths), and prefer to be by themselves rather than with others (England & Petro, 1998). Clearly, the personality traits of victims predispose them to a higher likelihood of experiencing internalizing disorders. For example, a study conducted by Kalti ala Heino, Rimpela, Rantanen, and Rimpela (2000) with adolescents in Finland found that among 17,643 participants, victimization was related to anxiety (17.2%), depression (42%), eating disorders (4.8%), and psychosomatic problems (21.5%). However, bullie s were found to be more prone to excessive drinking (66.3%) and substance use (31.3%) compared to victims (12.1%; 7.7%, respectively). The authors in this study suggested that victims may attract negative attention because they are not able to protect or defend themselves from abuse, based on their core personality structure. Aside from the commonalities in personality traits that many victims share, victimization also may be accounted for by distortions made in cognitive processing. Camodeca and Goossens (2005) studied common distortions made during social information processing (SIP) (Crick & Dodge, 1994) in Dutch elementary aged children ( N = 242). Victims cognitions were analyzed based on interpretation of intent made during an aggressive act, the ki nds of goals selected to respond to the perpetrator/victim, and their perceived self efficacy to resolve the situation. Results indicated that both bullies and victims responded more emotionally to situations that posed a conflict more so than non victim/ non bully peers. Deficits in each step of the SIP were evidenced by
18 bullies and victims. However, two major differences were found for victims: (1) victims trusted others less than bullies because of their continued harassment, and (2) victims were the on ly participants in the study that noted they felt unable to cope with their sadness. If victims did attempt to deal with the situation, they preferred to use aggressive tactics. Given these findings, an important question that still remains unanswered i s why victims often default to aggression. One reason may be that their frustration level is so high that they simply aggress (i.e., frustration aggression hypothesis) (Berkowitz, 1989). Research also indicates that a very small percentage of student byst anders intervene to defend the victim physically. Thus, because of the lack of support victims receive, they may feel the need to protect themselves via a physical altercation (Salmivalli, Lagerspetz, Bjorkqvist, Osterman, & Kaukiainen, 1996). Another pos sible explanation may be prosocial skill deficits such that victims are unable to respond adaptively (Fox & Boulton, 2006). A study by Borg (1998) also examined the emotional reactions of 9 to 14 year old victims This study found that 38% felt like seeki ng revenge, 37% felt anger, 37% felt pity for self, 25% claimed they were not bothered, and 24% of the students felt helpless. Unfortunately, many victims did not turn to others for help. If help was sought, younger victims tended to seek help more than adolescents. A nother study conducted by Hunter, Boyle, and Warden in Scotland (2004) intended to extend the results of Borg (1998) by examining the role of age, gender, cognitive appraisals, and emotional reactions in help seeking behaviors of Caucasian children ranging in age from 9 to 14 year olds ( N = 830). The most important variable
19 that predicted students help seeking behavior in this study was gender ; more females sought social support than males. Also, the more severe the emotional reaction, the more social support was sought. Females specifically stated that social support from primarily friends (followed by family members) was the best strategy to relieve painful emotions. While the Hunter et al. (2004) study is promising because it showed t hat female victims did seek support, it is difficult to generalize the findings from this study to other populations. It is unclear if the findings would hold true in an American public school system, for older females, or for at risk youth. Additionally, this study found that victims relied on friends and family for social support and not anyone in the school system. A few hypothesized possibilities for the Hunter et al. (2004) findings include: (1) students may not feel a strong enough bond with teachers or mental health professionals to trust them with personally sensitive information; (2) there is a lack of services for victims and they are keenly aware of this; (3) teachers may not know how to support their students; thus, failed attempts to help victi ms creates a climate whereby students know that the most helpful services possible are not available in a school setting. Notably, with regard to hypothesis 3 above, researchers have found that teachers do not intervene on behalf of victims due to lack of awareness of what victims experience and how to adequately identify them (Atlas & Pepler, 1998). There is an unfortunate mismatch between who students turn to and where the problem behavior actually occurs. In an ideal world, one would hope that student s could resolve problems they are facing in the context in which the problem(s) is taking place (e.g., school) especially since many students may need a trained professional (e.g., psychologist, counselor) who can provide support. However, students may not even realize that support from mental health
20 professionals is (a) available and/or (b) necessary. Studies such as Hunter et al. (2004) support the notion that school based support is not perceived as being as central as support from family and peers. Ri sk factors for victimization 2 Contextual (i.e., school and home), interpersonal (i.e., friendship and peer status), and intrapersonal (e.g., academic achievement, self efficacy) are three major domains of risk factors that impact the development of victim status for physical and indirect aggression. Research has shown that early risk factors are similar for both males and females, including an externalized and an internalized component (Paul & Cillessen, 2003); yet, risk factors change based on gender dur ing late childhood and early adolescence. Considering that children do not develop in isolation, ecological risk factors need to be taken into account when discerning what promotes the development of victim status. Contextual risk factors A number of di fferent researchers (e.g., Espelage et al., 2000; Farrington, 2005; Shields & Cicchetti, 2001) have found that child rearing practices and the home environment contribute to bullying behavior. Specifically, punitive and harsh parenting styles characterized by poor supervision, erratic discipline, and rejection produce children who bully during their childhood, adolescence, and adulthood. Adults who were bullied during childhood often rear children who become bullies (Smith & Farrington, 2004). Thus, in the context of the family, bullying behavior may be considered a learned behavior because of the modeling set forth by parents. Victims also come from homes where disciplinary tactics such as punitive punishment are utilized frequently. However, what differs between the home environments of bullies and victims is the level of hostility and rejection emitted by parents; more specifically, 2 Note: Relational victimization not included.
21 bullies experience higher levels of hostility and rejection in their homes compared to non bullies. Children who are victims in the home develop insecure attachment styles with their family members and these attachment problems beget school based problems with peers as well (Levy & Orlans, 1998). Stevens, De Bourdeaudhuij, and Oost (2002) also addressed how the family contex t, specifically child rearing practices (e.g., autonomy, punishment), family problem solving strategies, and overall family functioning (e.g., cohesion, expressiveness, control, moral emphasis), contributed to victimization status within the school setting Fifth and sixth grade students in Belgium ( n = 1,719) and their parents ( n = 1,401) were included in the study. Of the total participants, 17.5% of the children were identified as victims. Compared to parents reports, victims reported lower levels of bo nding and personal relationships with their relatives, an inability to express emotions to parents, and subjection to tremendous control and discipline in their households. While the researchers did not provide any hypotheses for this finding, this researc her believes that victims may have distorted perceptions of reality, based on previous research related to deficits made during the process of social information processing (see Camodecca & Goossens, 2005). Moreover, students who experience victimization a t school may in fact generalize their perception of victim status to all sub systems of which they are a part. While Stephens et al. (2000) argued that the reports by this particular sample were expected given their developmental stage, the current resea rcher questions the rationale used by these authors because in this study the mean age was 11.5 years, which is considered preadolescence. During this stage, behavior typified by adolescents is slowly emerging, and thus the results may not be completely a ccounted for by age. Furthermore,
22 this researcher suggests that the negative perceptions of family may indicate two different characteristics of victims. The first may be that, in fact, victims families tend to be less enmeshed and more focused on punitiv e consequences instead of problem solving and emotional expression. The second possibility may be that victims have distorted cognitions of the home environment (in addition to distortions about others) that may fuel their own separation and lack of attach ment to the family. The best point of entry for intervention is thus unknown and remains somewhat ambiguous. Psychologists have a choice to attempt to change a victims cognitions (e.g., individual therapy) or the dysfunctional family relations and process es (e.g., family therapy) both ways ultimately culminating in boosting a victims resilience against aggressors at school. Veenstra, Lindenberg, Oldehinkel, De Winter, Verhulst, and Ormel (2005) were among the first researchers to conduct a multivariate analysis of factors that contribute to victimization. This study provided insight into which variables (within the domain of family, school, and intrapersonal characteristics), when combined, contributed significantly to predicting victimization. A sub sample of data was taken from The Tracking Adolescents Individual Lives Survey (TRAILS), ( n = 1,065) which follows preadolescents to the age of 25. The first major finding was that familial vulnerability to internalizing and externalizing disorders contr ibuted to predicting victimization. This was the first time a research team found this result. The second major finding was that the impact of parenting diminishes as a predictor variable when SES, familial vulnerability, academic performance, and prosoci al behavior are taken into account. The authors contended that parenting may be less important to victimization status in adolescent aged populations than with elementary aged children. The study also found that victims were
23 relatively positive about thei r relationship with parental figures this is contrary to other research on victimization and family relations. Overall, Veenstra and colleagues reported results that were contradictory to other victimization literature that contends that families are crit ical predictor variables. This may in part be due to the location where the study took place (i.e., Holland), the relatively large sample size, and the statistical analyses that were conducted. More specifically, the large sample size may have increased th e likelihood of detecting significance in the data. Further studies are needed to confirm these findings. Victims also experience problems in the context of school. Victims suffer from poor academic performance, report unhappiness at school, and view school as an unsafe place to attend (Smith & Shu, 2000). Nansel, Haynie, and Simons Morton (2003) obtained self report data from middle school students ( N = 930) in Maryland to address the stability of victimization across middle school and how victimiza tion impacts middle school adjustment (e.g., following rules, completing homework, involvement with school activities) and perceptions of school climate (e.g., teacher support, rule clarity, student student respect). Victim status identified among student s who had been victims on three or more occasions and aggressed against other students less than two times (at the time of assessment). In this sample, 50% of participants reported victimization and one quarter reported repeated victimization (i.e., consis tent victimization across sixth and seventh grades). Victims reported poorer school adjustment compared to bullies and non involved peers. Victims also reported lower perceptions of school climate both in the sixth and seventh grades than bullies and the comparison group. The directionality of the relationship still remains unclear from this study; however, a warm and supportive school
24 climate where rules and expectations are clear may protect students from the effects of victimization and/or deter aggress ive acts in the first place. School satisfaction, defined as a students global satisfaction with school (e.g., pleasure in school), is a factor related to overall psychological well being, attendance rates, behavior problems and school drop out (Huebner & McCullough, 2000). Verkuyten and Thijs (2002) hypothesized that school satisfaction may be linked to peer victimization. The researchers also contended that social cognition, specifically social self esteem, may serve as a mediator variable between vict imization and school satisfaction. Their hypotheses were confirmed, such that students who were victimized had lower levels of self esteem and their overall satisfaction in school was lower than non victimized peers. Teachers reactions to victims also contribute to the isolation and rejection victims feel in schools. Nesdale and Pickering (2006) attempted to determine how teachers judgments and punishment strategies were influenced by their perceptions of (a) identification with their class (i.e., how attached and committed they felt), (b) popularity status of victim, and (c) whether the victim is identified as either a good or bad student. A total of 90 experienced teachers ( M = 13.4 years of teaching) practicing in Australia were included in the study and were provided with scenarios about physical aggression among males as the basis of their answers. Findings showed that teachers who liked the aggressor, based on the classification of good student on the assessment instrument, were less likel y to be sympathetic to the victim. Similarly, teachers attributed causality of the altercation to the victim if he/she held favorable opinions about the aggressor.
25 While the Nesdale and Pickering (2006) study is limited in its generalizability, the imp lications for future research are profound. This study highlights the importance of a variety of factors that teachers take into account when deciding whether to intervene during aggressive acts. If teachers are biased towards certain children, then the ch ances of taking a victim seriously are decreased. Therefore, victims may not only be rejected by peers but also by teachers. If findings of this study are replicated, the need for training for teachers related to the impact of aggression on victims will b e paramount. An important question that remains to be answered is Does school or the home environment play a more pivotal role in the development of victimization tendencies in children and adolescents? Ahmed and Braithwaite (2004) examined this questi on by analyzing how the combination of family (i.e., authoritarian parenting, authoritative parenting, and family discord) and school variables (i.e., like for school, school control over bullying, and school hassles) contributes to victim, bully, and vict im/bully status. This was the first study that combined different contextual risk factors into one study. Participants in the study included Australian students ( n = 1401) in late elementary and early middle school as well as their parents ( n = 978). Bot h variables were found to predict group membership; however, school variables predicted membership ( R = .54) more than family variables ( R = .41). When risk factors from both contexts were combined, the accuracy of identifying group membership increased to 61%. Notably, no single variable was able to satisfactorily discriminate among all three groups (i.e., bully, victim, control). The discriminant function analysis conducted in this study found that both victims and bullies experienced problems in scho ol and at home, and both perpetrators and victims felt schools did not have any control over bullying. However,
26 victims reported more family disharmony than bullies. This study provided evidence that victimization does not occur by happenstance; instead, there is a degree of shaping and negative experience that occurs in the home that sets the foundation for harmful relationships in the school context. Additionally, the study implies that ecological interventions that include the entire family may be more effective than those that include only the victim. Interpersonal risk factors. In general, there is consensus in the research that when a peer group is more accepting of a student, the chance of being victimized is decreased. Yet, victimized students a re typically found to be less liked by their peers (Phillipsen, Deptula, & Cohen, 1999). Additionally, children with larger networks of friends receive more support and thus may be less negatively impacted by bullying. Friendship quality and social compet ence are two other interpersonal factors that Goldbaum et al. (2003) assessed in Canadian middle school students ( N = 1,241). This study was novel because the researchers tried to gain insight into the directionality of the relationship between interperso nal factors and victimization by categorizing students into four distinct groups: non victim, late onset victim, stable victim, and desister. For example, the late onset victims allowed the researchers to examine interpersonal factors that may have preced ed victimization (e.g., antecedents). Individuals in the study who reported poor quality friendships were found to be at a higher risk for continued victimization. Late onset victims and desisters both reported positive friendships prior to victimization. Additionally, as victimization increased, late onset victims reported lower levels of trust and affection for peers. Thus, many late onset victims did not seek out new friendships (i.e., withdrew) and consequently participants may have developed more pat hology
27 because the protective nature of friendships did not serve a buffering effect. Therefore, their interpersonal functioning deteriorated as the cyclical nature of victimization continued. Intrapersonal risk factors. The victimization literature has often overlooked the question of whether internalizing problems such as anxiety and depression precede, mediate, or are a consequence of victimization. Goldbaum et al. (2003) tried to tease out whether psychological maladjustment predisposes a person to vi ctimization or if the converse was true. This study found that late onset victims at Time 1 of the study reported higher levels of internalizing problems and poor peer relationships; therefore, it was suggested that late onset victims may be suffering fro m internalizing problems that predispose them to victimization. Additionally, once the bullying occurs, victims internalizing problems tend to become worse. Paul et al. (2003) also evaluated intrapersonal risk factors in fourth through seventh grade s tudents. Participants ( N = 600, predominately Caucasian) were rated according to teacher and self reports on a number of traits, including internalizing and externalizing behaviors, and social and academic self efficacy. This study corroborated findings f rom previous studies, namely that victims exhibited low social and academic self efficacy and were rated high on instruments that assessed both internalizing and externalizing behaviors. Interestingly, this study found that females reported lower levels of self efficacy across both domains compared to males. Females also exhibited higher levels of depression, negative social perceptions, and anxiety compared to males. As would be expected, males typically exhibit more externalizing behavior, which comm only is believed to be the precursor to bullying behavior. However, externalizing
28 behavior also may be a precipitant of victimization. Kokkinos and Panayiotou (2004) conducted a study in Cyprus to ascertain the relationship between disruptive behavior di sorders [i.e., Conduct Disorder (CD) and Oppositional Defiant Disorder (ODD)] and subsequent victimization in a sample of 202 adolescents. Findings indicated that while CD predicted bully status, ODD predicted victim status. The authors suggested that th e mild symptoms of ODD exhibited by elementary school children may serve as one reason why other students bully the victims (e.g., deliberately irritating and noncompliant behavior bothers others). Unfortunately, many students who present with ODD and do n ot receive intervention continue along a trajectory that may lead to CD and Antisocial Personality Disorder (Hinshaw & Lee, 2003). Thus, this study appears to provide a rationale for why many victims often turn into bullies in later years. Additionally, t his study supported prior findings that victims have lower self esteem than bullies and control children. Social skills deficits are another discernable intrapersonal risk factor common to victims. Fox and Boulton (2005) utilized a multi informant approac h to determine which social skills deficits victims display compared to non victimized peers. Additionally, the researchers wished to determine if peer, self, and teacher reports diverged on social skill ratings. A total of 330 Dutch children ( M = 10.3 yea rs old) and 12 teachers participated in this study. Agreement on only three items was found among participants; specifically, victims were perceived as scared, weak, and unhappy. Additionally, a large proportion of victims were classified by one or more participants as non assertive, distressed, and withdrawn. Overall, victims were perceived as having fewer adaptive social skills than non victims.
29 Many children are at risk for psychological, social, and interpersonal problems to some degree based on their family history, experience at school, and/or individual factors (e.g., temperament, genetic predisposition). At what point though do children cross the normative threshold of daily life stressors and become involved in a relationship that involves victimization? Presently, there is no clear or definitive answer to this question in the literature. Therefore, while a focus on risk factors is imperative so that the etiology of victimization may be better understood, risk takes on a negati ve connotation and may be more difficult to alter (e.g., in cases of parental drug abuse or a genetic predisposition for depression). Thus, an alternative to focusing on risk factors is to utilize a strengths based approach whereby a childs pre existing gifts or abilities could be cultivated to protect or enhance a childs well being. A paradigm shift moving towards a focus of protective variables in research and practice may do exactly that: improve well being. Protective Factors Protective factors may be viewed in two different ways: (1) as variables that decrease or moderate the risk of becoming victimized, and (2) as moderators of psychopathology. To date, attention paid to how protective factors play a role in victimization has been minimal. Withou t increasing our knowledge base of protective factors, preventive programming can rely only on risk reduction approaches. Social support A study conducted by Baldry and Farrington (2005) examined the role of family context (i.e., authoritative parentin g) and personal coping skills as two protective variables that may moderate the risk for victimization. The study included only Italian high school males ( N = 679), who were mainly from the upper social echelon
30 in Italy. While a majority of students in t his sample reported negative and hostile interactions with their parents, the hierarchical regression analyses confirmed that when very supportive parenting styles were intact, rates of victimization were reduced. Males who were identified in the highest r isk category for victimization benefited the most from parenting support and coping skills. One hypothesized reason for this outcome was that parents who are supportive of their children are more likely to help them problem solve difficult situations rath er than leaving their children to accomplish this task on their own (Dwairy, Achoui, Abouserie, & Faraji, 2006). Thus, reasons for victimization and how to deal with it may have been discussed in the context of the family. The literature contains stud ies that indicate that negative peer interactions serve as risk factors for victimization (Schreck, Miller, & Gibson, 2003). The literature also contains studies that provide evidence supporting that the converse is true: students who engage in positive p eer relations are less likely to be segregated out as a target for aggression, even as early as the kindergarten years (Hanish, Ryan, Martin, & Fabes, 2005). Overall, perceived support appears to be the underlying reason why peer friendships are critical for victimized children. In a longitudinal study conducted by Ladd and Burgess (2001), 396 kindergarten and first grade students were studied to address how behavioral and relational risk and protective factors impact adjustment between kindergarten and fi rst grade. The study confirmed the researchers hypotheses that peer acceptance, the number of friends one had, and a positive teacher student relationship inhibited the development of maladjustment in this sample. More importantly, positive relationship s may have compensated for the externalizing behaviors that many of the students exhibited (i.e., students who would typically become victimized because of their
31 overt behavior did not because they were supported by peers and teachers). Peer acceptance wa s found to have the most wide reaching impact for victims as it helped to increase victims attention span in class, decrease misconduct, and increase school liking. The findings in this study are consistent with an additive model framework such that the more protective factors there are, the stronger the impact on future prosocial adjustment. Connection to school While there is no research to date that suggests a direct link between experiencing victimization and feeling a strong connection to the sc hool environment (i.e., one construct within school climate), it might be hypothesized that school climate (or facets of school climate) differentiates how a child responds to relationally aggressive attacks (i.e., specifically evaluating school climate as positive serves as a buffer to victimization). Moreover, positive school climate may actually reduce the prevalence of relational aggression on school campuses because of tolerance policies and the general feeling of support generated by school faculty. A positive school climate has been found to serve as a resilience factor for students in a variety of ways, including protecting children from using tobacco, drugs, and alcohol (Suldo, Hanguaer, Witte, Mihalas, Popkave, Powell, & Hardesty, 2006), curbing de linquent behavior (Farrington, 2005), and promoting an overall sense of well being (Sellstrom & Bremberg, 2006). Additionally, other researchers have found that when a school provides a positive and safe environment and facilitates social and academic succ ess, students are more likely to be buffered from negative outcomes (Christle, Jolivette, & Nelson, 2005). Moreover, a positive school climate has been shown to decrease bullying on school campuses (Unnever & Cornell, 2004). Additionally, schools in Engla nd have found that when the school environment is perceived as safer, via direct intervention, students are
32 less likely to be victimized (Smith et al., 2000). Therefore, given that school climate has been indicated as a positive force in the lives of stu dents, it also may impact the way a student handles aggressive confrontations. Intrapersonal factors Aside from external factors previously mentioned, internal factors such as coping styles serve to increase victims resilience from further maladjustment issues. For example, Kochenderfer Ladd and Skinner (2003) examined how coping strategies (e.g., approach and avoidance) may influence victims adjustment (e.g., social problems, anxiety, depression) in 356 elementary aged children. Results revealed that both males and females were equally at risk for victimization; however, the type of coping strategy used was different for males versus females. Consequently, the specific maladjustment issues were also differentiated. For example, males were more likely to resolve conflict using an approach style whereby they tended to try to resolve problems with the perpetrator on their own. Thus, males tended to have lower levels of loneliness and social support problems because they did not isolate themselves. Howev er, the authors mentioned that when males did seek social support but did not receive the kind of support they expected or desired, loneliness levels actually increased. A possible way to decrease loneliness for males who seek help may be to provide school based therapy groups for males to increase the relationships they have with other males and so that the appropriate adult is prepared to listen and provide adequate support. On the other hand, females who sought social support were less at risk for social problems and did not report loneliness upon asking others for support. Avoidance coping was an ineffective protective factor for both genders, possibly because bullies tended to feel that they could continue the behavior without repercussion.
33 Unfortuna tely, there is a paucity of research focusing on the relationship between intrapersonal protective factors and victimization (aside from the research conducted by Kochenderfer Ladd & Skinner, 2003). The lack of research on victimization and protective fac tors is somewhat surprising, given that many psychologists conducting interventions focus mainly on individuals instead of systems level issues. Regardless, if protective factors are shown to buffer the relationship between victimization and psychopatholo gy, a new focus of interventions for practitioners may be available such that a strengths based approach may be utilized during therapeutic interventions for victims. Relational Victimization Much less is known about relational victimization than physic al and social victimization; consequently, researchers have advocated for studies on relational victimization to be conducted with diverse populations (Young et al., 2006). The purpose of this section of the chapter is to highlight the research that has b een conducted and to exemplify the necessity for further studies to be completed in this critical area of study. Psychopathology. Relational victimization, similar to physical victimization, has been linked to depression, social isolation, anxiety (Walke r & Cillessen, 2006), and PTSD (Gumpel & Kliewer, 2006). Craig (1998) studied the relationship between victimization and pathology among predominately Caucasian middle school students in Canada ( N = 546). This study found that relational victimization was a significant predictor variable for anxiety (per a self report questionnaire). Also, levels of depression were markedly higher in the victim group compared to the control group. Additionally, relational victims were found to engage in indirect aggressi on themselves as a defense
34 mechanism. Those victims who used this approach also reported higher levels of anxiety. Victims using more indirect forms of aggression may believe their anonymity is upheld and thus feel as though their aggressor will be less l ikely to retaliate against them again. Dill and colleagues addressed another component of psychopathology, negative affect (e.g., mad, sad, scared), in 731 elementary aged children via a longitudinal design. Dill et al. (2004) developed a three stage mul ti modal model whereby at Time 1 shyness and withdrawal were assessed (these two variables were deemed directly related to peer victimization), at Time 2 negative affect was assessed, and finally at Time 3 depressed mood was assessed. This study confirmed a number of hypotheses. First, negative affect increased as a result of relational victimization. Second, students who held beliefs that aggression was warranted towards them personally were likely to have a stable negative affect. Third, shyness was found to be a trait that served as an antecedent to relational victimization. Therefore, the researchers suggested social skills training and active attempts towards building students communication skills. Finally, personal beliefs (i.e., social cognitions ) about the kind of support that a child receives influences the attributions he/she makes regarding an aggressive act Toblin, Schwartz, Gorman, and Abou ezzeddine (2005) found that elementary aged aggressive victims who perceived a lack of support display ed problems with self regulation including impulsivity, hyperactivity, and emotional dysregulation. Bullies (e.g., those students who felt supported) did not experience emotional dysregulation to the same degree as victims. The aforementioned study lends c redibility to interventions that adhere to a cognitive behavioral approach so that appropriate and inappropriate behaviors are addressed, as well as maladaptive schemas about self and others.
35 Specifically, interventions such as cognitive behavioral therap y may be more appropriate for those students who engage in negative attribution bias. Notably, the researchers stated that the results may not be generalized to low income families due to the nature of the participant pool. Storch, Masia Warner, Crisp a nd Klein (2005) also addressed psychopathology among victims in a one year prospective study assessing the relationship between relational victimization, social anxiety, and phobia in a sample of urban ninth grade students attending a parochial school ( N = 144). The researchers found that regardless of gender, relational victimization predicted social phobia one year later. However, based on initial baseline data, little support was garnered for the prediction of victimization based on levels of social anx iety. Thus, it may be that victimization deters students from socially interacting with peers they perceive as aggressive and also leads students to evaluate themselves negatively. Coping style. Olafsen and Viemero (2000) addressed how fifth and sixth gr aders ( N = 510) in Finland utilized different styles of coping to deal with relationally aggressive students [i.e., five specific styles including aggression, distraction (e.g., engage in a hobby, taking a walk), self destruction (e.g., smoke, self mutilat e), stress recognition (e.g., cry, scream, ask for advice), and endurance (e.g., think about it, watch TV)]. This study found that there were no gender differences in the prevalence of victimization nor the coping style used. All victims of RA were found t o use more self destruction strategies. However, the conclusions from this study must be viewed with caution. This is because first, the researchers did not utilize an assessment instrument that directly measured relational victimization. Instead, they me rely added a question to the Olweus
36 Bullying questionnaire regarding gossip to serve as the relational victimization indicator (i.e., one item indicator). Second, this study did not find any major differences between victims and their non victim counterpar ts; thus, the conclusions made about victims of relational aggression are based more on clinical judgment rather than on rigorous statistical analyses. Summary and Future Research The adages boys will be boys and thats just how girls are are no long er justifiable excuses to condone physical and/or relational aggression (Clarke & Kiselica, 1997). As evidenced in this literature review, victimization of any sort affects youth in profound ways and cannot be dismissed as gender normative and appropriate behavior. In fact, victimization may impact the way a child is perceived by his/her peers, or it may alter a childs mental health outcomes (Lindenberg, Oldehinkel, De Winter, Verhulst, & Ormel, 2005). Inevitably, limitations in the current body of rese arch must be addressed for educators and mental health professionals to secure a comfortable place for children within the school environment, thus affording children an opportunity to engage in healthy relationships that will aid in the development of soc ial and personal skills. Based on the high prevalence rates of aggression exhibited within the school system, lack of future research in the area of relational victimization will hinder the possibility of positive future outcomes for victims. Vis vis this literature review, several gaps and limitations within the current research have been elucidated. First and foremost, a paucity of research exists on relational victimization. While a number of studies have been conducted on perpetrators (e.g., Van Acker & Talbott, 1999; Connor et al., 2003; Rodkin & Hodges, 2003), very
37 few studies address the characteristics of victims and associated mental health outcomes. Additionally, specific risk and protective factors that mediate and/or moderate the relations hip between victimization and psychopathology are minimal in number. The necessity of understanding outcomes for victims of RA in more depth is clear. Without future research, development of interventions to provide support for victims and to teach coping skills to those who have been victimized will continue to be overlooked. Second, the majority of research on general victimization in the United States tends to focus on elementary aged children (e.g., Crick, 1997; Dill et al., 2004; Nansel et al., 2003) While this is commendable, research on relational victimization also should include adolescents. Adolescents spend a majority of their time in school settings, and major developmental milestones are met during this time. Additionally, indirect forms of aggression tend to be the primary form of aggression used in schools during adolescence because the penalties for overt aggression become harsher (Underwood, Galen, & Paquette, 2001). If, in fact, because adolescents experience the brunt of indirect aggr ession from peers during this stage of development, it is important for researchers to understand if and how victimization impacts developmental milestones (e.g., social and emotional development; interpersonal relationship building). Additionally, more than 50% of the research published on victimization has been conducted in countries other than the United States (e.g., Lindenberg et al., 2005; Verkuyten et al., 2002 ). While these studies afford researchers new information on aggression and victimization the results may not be generalizable due to cross cultural differences. Therefore, more research conducted in the United States is needed because cultural nuances may account for different findings in specific populations of students.
38 Finally, resilienc y factors (e.g., protective variables) associated with any form of victimization are not fully understood because of the limited number of studies that have addressed this particular issue (e.g., Goldbaum et al., 2003; Christiansen et al., 2005). More spe cifically, no studies that have reviewed how protective variables impact RA victimization were identified in the literature. As such, it is critical to determine if there are resiliency factors that may shield children from deleterious outcomes. Current Study The current study attempted to address many of the limitations mentioned previously including the lack of participant diversity, the focus on risk factors instead of protective factors, and integrating a number of variables into an ecological model to explain relational victimization. Specifically, this study focused on adolescents for two main reasons. First, adolescence is a critical time when peer interactions inform an individuals identity formation (Paul & Cillessen, 2003). Thus, the prevale nce and variability in victimization may be higher than at other points during a persons development. Therefore, an adolescent who is victimized may engage in different coping strategies and also may ultimately demonstrate different pathology compared to an elementary school student. In fact, adolescents may exhibit completely different patterns of behavior and subsequent outcomes as a result of different values and interpretations that coincide with development. Additionally, while there is some disagreem ent in the field regarding gender specificity related to differential expressions of aggression, some researchers have found that adolescent males engage in as much relational aggression as females (Peets & Kikas, 2006). Therefore, this study was not limi ted to adolescent
39 females but also included males to further investigate how gender was related to relational victimization and presenting psychopathology. Second, the stability of remaining in the victim role becomes the most potent during adolescence (Casey Cannon, Hayward, & Gowen, 2001), and thus intervening is critical to maintain a students sense of well being throughout the course of high school and beyond. For example, if a student continues to be victimized in the high school setting, the like lihood of entering abusive relationships in the future will increase (Coie & Dodge, 1983). In an ideal world, school wide approaches to bullying prevention would have halted over 65% of aggression in schools utilizing a universal prevention approach (Leff et al., 2001); however, with the prevalence of relational victimization on campuses, psychologists are at the crucial point where targeted interventions are necessary to help reverse some of the events that have impacted victims (e.g., repeated victimizat ion, reactive aggression that leads to academic problems, teacher disapproval). While most studies have focused on participants between the ages of four through twelve, the current study extended the literature base to include a middle school sample. Als o, with increased knowledge of RA during middle school, practitioners may be able to make the transition into high school easier for victims because extra support towards building resiliency may be provided. Third, as evidenced by the aforementioned stu dies, the typical students who participate in victimization studies are middle and upper class Caucasian students who are not considered at risk. As such, an unstudied population in the study of victimization, specifically relational victimization, is po or, at risk minority youth (Young et al., 2006) because almost no researchers have developed research agendas that target this particular
40 demographic group and relational victimization. Yet, there is some evidence that this line of research would be fruitf ul; for example, Schreck, Miller, and Gibson (2003) found that African American males experienced the highest frequency of victimization when considering physical victimization and theft. Also, one research study found that African American youth engaged i n more relational and overt aggression (Phillipsen et al., 1999) than Caucasian students. Yet, Phillipsen et al. (1999) is the only study that has attempted to look at the prevalence differences across racial lines. Finally, one study found an inverse corr elation between SES and victimization, such that the lower poverty level in a family, the higher the rate of victimization (Veenstra et al., 2005) This finding supports the hypothesis that poverty may impact relational victimization. Unfortunately, poo r, at risk minority populations often receive less than adequate mental heal th services. For example, Ringe l and Sturm (2001) found that 31% of majority students received mental health services compared to only 13% of minority students. Limited practitione rs in urban areas, interventions that are not culturally sensitive, and the manner in which mental health services are marketed to urban and ethnic populations may account for the differential percentages between groups (Tolan & Dodge, 2005). Therefore, t he first step towards providing services that are culturally sensitive and appropriate for minority at risk youth is to make a direct attempt at establishing a relationship with this population. This may be done by hearing their voices through qualitative and quantitative data collection and learning more about the kinds of victimization they experience. Allowing students to establish their voice does not only provide rich data but also has been documented as an intervention in itself (Roberts & Coursol, 19 96). Notably, all qualitative work in this area has been conducted with
41 elementary aged children in other countries (e.g., Holland, Cyprus). Undertaking a study with this specific population will contribute to the literature because at this point in time t here are no studies (to the researchers knowledge) that have investigated relational victimization among low income, ethnic minority children. Fourth, a number of studies have focused on the risk factors that increase victimization, but few have focuse d on how to protect students. Instead of focusing on the negative attributes that contribute to victimization, the study took a positive strengths based approach to understanding victims. How may educators and psychologists help victims engage in the recov ery process using their own personal strengths through intervention? This study intended to answer that question by moving away from the identification of problems among or within a child (e.g., negative attributes or characteristics) that need to be chang ed or removed and instead focusing on which factors may be promoted to enhance resilience. Risk versus protective factors Why are the risks so much stronger than the protective factors in terms of predictive validity? There is the possibility that the protective factors that make a substantial contribution to resilience have not been examined to date. Because a number of ecological variables impact ones functioning, four domains of protective factors were analyzed in this study, including school, home, peers, and intrapersonal. By addressing these four domains, the researcher acknowledged that students have many different ecologies and developmental issues that impact their overall functioning. In past research, researchers typically have focused on o ne domain at a time instead of looking at a combination of protective factors. Thus, the current study used a protective protective model such that the combined effects of a number of
42 protective factors are hypothesized to negate the risk factors of victi mization (Zimmerman & Arunkumar, 1994) (e.g., gender and hope). Protective factors that were examined in the current study included school variables (e.g., teacher support, classmate support), family and best friend support variables, and intrapersonal fac tors (i.e., hope and spirituality). P ositive psychology Positive psychology focuses on strengths within humankind rather than weaknesses (c.f., Snyder et al., 2005). Although many of the constructs under the umbrella term positive psychology such as signature strengths (e.g., social intelligence, valor, honesty, leadership) have been studied since the inception of psychological practice (c.f., Allport, 1961), the positive psychology movement has caused a resurgence of interest on such topics and propo sed that empirical research should examine how increases in certain traits and states may increase adaptive functioning. Positive psychology researchers and practitioners attempt to move away from pathologizing clients and rendering diagnostic labels. The study of positive psychology rests on understanding positive emotion, positive character, and positive institutions (Seligman, 2006). The crux of treatment is founded upon acknowledging ones strengths and building upon those strengths. In essence, treatm ent does not hinge upon ameliorating a clients problems or changing a client. A different route is utilized instead: subjective well being is enhanced by focusing on past and present experiences that have a beneficial impact on the clients sense of self. Peterson and Seligman (2004) have even called this emerging positive psychology approach as un DSM.
43 There are six core virtues that serve as the foundation for prevention and intervention. They include courage, love and humanity, justice, temperance, s pirituality and transcendence, and wisdom and knowledge (Seligman, 2002). These values may be easily assessed through self report measures developed for adults and children (e.g., www.authentichappiness.com) and results are provided to the client, therapis t, and/or researcher immediately. Thus, instead of potentially causing damage to a persons identity by searching for evidence of psychological problems, this approach provides positive and immediate feedback on strengths and assets as an initial step to r elieve any potential suffering. The results serve as a basis to begin to work on uncovering the positive potential that all humans possess to some extent. The protective variables that were examined in this study fall under the umbrella of the study of po sitive psychology. Specifically, perceived social support is considered part of the positive institution; spirituality is considered a facet of positive character; and finally, hope is an example of a positive emotion. A discussion of each variable is del ineated below. Perceived social support. Perceived social support is defined as a students perceptions of general support or specific supportive behaviors which may include emotional, appraisal, and/or informational support provided by persons in their n etwork (i.e., teachers, peers, parents, and friends) (Jackson & Warren, 2000). The Basic Behavioral Science Task Force of the National Advisory Mental Health Council (1996) found that over a hundred studies have shown that social support protects people fr om life events that are stressful (e.g., divorce) and also promotes wellness in persons with schizophrenia and long term illness. Perceived social support also has been found to
44 protect students from school adjustment and adolescent personal problems. More specifically, Demaray and Malecki (2002b) found that students who perceived higher levels of social support were most likely to exhibit lower levels of aggression, depression, anxiety, and hyperactivity. The construct of social support has been examined in students who are classified as at risk and minority (e.g., Hispanic middle school students) (Demaray & Malecki, 2002a). Thus, use of this construct is substantiated because past research has targeted the population studied in this dissertation, and perc eived social support has previously been linked to decreases in pathology. Therefore, a natural moderator relationship was hypothesized to exist among relational victimization, perceived social support, and depression. Intrapersonal protective factors There are a number of intrapersonal factors that may be evaluated to determine what may best protect an individual from psychopathology and/or make an individual resilient to victimization. However, this author believed that two constructs that were wor thy of investigation were hope and spirituality. These two constructs were chosen because previous research has supported their positive impact on well being in children, adolescents, and adults. More importantly however, the ease in which hope and spirit uality may be integrated into interventions made the examination of these two constructs more promising. Hope theory involves three main components including goals, pathways thinking, and agency thinking. Goal directed behavior is the major tenet of h ope theory such that it provides the cognitive component for students to want to achieve or maintain a specific outcome (Snyder, 2000). In order to obtain a specific goal, a student needs to engage in pathways thinking, which is synonymous with generating feasible routes to obtain a goal
45 (e.g., strategies). The final process that must take place for hope to be instilled in a student is agency thinking reflection upon a students capacity to actually reach the goal. Thus, hope theory purports that goal dire cted thinking is a system of positive and negative feedback loops that provide students with varying degrees of optimism (Snyder et al., 2002). Snyder et al. (2002) noted that hope has been linked to higher achievement, better physical health, and better athletic performance. Furthermore, Amlund Hagen, Myers, and Mackintosh (2005) examined a high risk group of 65 children who had incarcerated mothers. The results of the study indicated that children who exhibited higher levels of hope also reported lowe r levels of externalizing and internalizing problems. Additionally, these authors believed that intervention efforts could be targeted at any or all of the hope pathways and goals in order to encourage higher levels of hope in children. Spirituality is a long standing phenomenon that has rarely been embraced by psychological research (Benson, Scales, Sesma, & Roehlkepartain, 2005). It should be noted that organized religion and spirituality are two distinct constructs. Organized religion refers broadly to an institution or persons that follow a specific faith. Spirituality on the other hand has been defined as a search for the sacred and insinuates a process or a pathway that a person takes (Pargament, 1999, p.12 as cited in Pargament & Mahoney, 2002 ). Spirituality typically denotes a more personal and private configuration of feelings and actions in relation to some transcendent entity (King & Boyatzis, 2004, p.3). Assessment of ones spirituality typically involves two indicators: importance and/o r salience spirituality plays in ones life and attendance at a
46 spiritual/religious event. A lengthy discussion regarding the history of religion and spirituality is beyond the scope of this paper (see Pargament & Mahoney, 2002 for an in depth discussion) What is important, however, are the numerous findings that suggest that higher levels of faith and spirituality have predicted higher levels of functioning during the grieving process (McIntosh & Spilka, 1990), lower levels of behavioral problems (Brody Stoneman, & Flor, 1996), more involvement in community activities, and a greater proclivity to engage in civic res ponsibilities (Smetana, Campione Barr & Daddis 2004 ). The Monitoring the Future Study ( Johnston, OMalley, & Bachman, 2000) found that 8 3.7% of high school seniors reported some degree of spiritual/religious affiliation; however, participation and importance declined between sixth and twelfth grades. African American youth reported the highest levels of spirituality. This finding has been supported by other researchers (c.f., Hodge, 2004; Walsh, 1999 ). This researcher hypothesized that spirituality would serve as a moderator between relational victimization and psychopathology for two main reasons: (1) minority adolescents engage (i.e., Af rican American) in spirituality more than other ethnic groups, and (2) prior studies have exemplified the significance spirituality plays on a myriad of factors including depression, hopelessness ( Mystakidou, Tsilika, & Parpa, Pathiaki, Patiraki, et al., 2 007), substance use (Rostosky, Danner, & Riggle, 2007), and exposure to community violence (Jones, 2007). Outcomes of victimization Finally, the present study addressed depression via a mixed methods approach to ascertain the quality and severity of vic tims experiences. The prevalence of depressive symptoms, in females especially, increases during
47 adolescence (Hazler & Mellin, 2004). Therefore, because past research has found that many children who are victimized suffer from depression, it was inferred that depression may be exacerbated in female adolescent victims. A study conducted by Galen and Luther (in press), as cited by Underwood (2004), found that the experience of victimization (i.e., social aggression victimization) uniquely accounted for cha nges in depression and anxiety levels among preadolescents. Additionally, past research has suggested that victims suffer from internalizing disorder, but research has failed to elaborate on the degrees to which students suffer (i.e., mild, moderate, seve re) (Seals & Young, 2003). In conclusion, this study utilized a developmental ecological framework and a strengths based approach to assess the relationship between relational victimization and depression in high risk minority youth. The study aimed to p rovide an opportunity for victims to share their experiences as a stepping stone to provide better services to underprivileged and distressed individuals.
48 Chapter 3 Method Overview This chapter describes the method used in the present stud y. Included in this chapter is a description of the setting for the data collection, the participants, procedural considerations, assessment instruments, research design, and statistical analyses. Setting Data for this study were collected from two middle schools loc ated in the same school district in Florida. In 2006 2007, this district had a total of 274 schools (of which 48 were middle schools) serving 193,480 student s Of the 193,480 students, the majority were Caucasian (82,959), followed by Hispanic (52,153), A frican American (42,570), Multiracial (9,830), Asian/Pacific Islander (5,400), and American Indian (568). The school district ranks the quality of their schools based on letter grades ranging from A (best possible grade) to F (worst possible grade). T able 1 provides information specific to the schools that participated in this study. Table 1 Characteristics of Participating Schools School Name School Grade for 2005 2006 and 2006 2007 years Met Annual Yearly Progress % of minority students % of stud ents obtained free and reduced lunch Middle School A (pilot) B D No 93 93 Middle School B B, C No 90 80
49 The original plan for data collection was to include only Middle School A. This school was chosen because it had a high percentage of students w ho were categorized as minority (i.e., any racial/ethnic group other than Caucasian) as well as a high percentage of students receiving free or reduced lunch. The overall atmosphere of Middle School A was typical of other urban schools; for example, the teacher and student turnover rates at this middle school were high. On average, two to three acts of violence occurred each week (based on number of violent incident reports on www.fldoe.gov). Most of the student body was African American. Notably, a new principal had joined the staff in 2005 2006 and was trying to revitalize the school, including the Exceptional Student Education (ESE) program. However, at the onset of data collection, the principal left the school for personal reasons. Given the change in administration and the subsequent turmoil that his departure created, a decision was made to collect data at another school to avoid a number of confounding variables that could not be controlled at School A. Therefore, this researcher contacted nine schools in the district with similar demographics in an effort to gain participation. Middle School B was the only school that decided to participate. Middle School A thus became the pilot school, and Middle School B served as the primary data collecti on site. While Middle School B is also considered an urban school, some distinct differences existed between School A and B. For example, Middle School B had a higher proportion of Latino students compared to Middle School A. Additionally, the number of violent incidents reported were fewer in number at Middle School B than at Middle School A. On average, a violent incident occurred only once every 10 days at Middle School B. The administrative arrangement at Middle School B was also different
50 because th e same principal provided leadership to the school for quite a few years. This particular principal was focused on increasing academic outcomes and life success for his students, thus a number of changes had been instituted during the year of data collect ion, including teacher accountability, bell schedule changes, and involvement with the community. Anecdotal observations made during data collection seemed to indicate that these changes had resulted in considerable tension among faculty and staff at the s chool. Participants Participants in this study included middle school students enrolled in grades six through eight. The target number of students in this study was approximately 500; however, the final number of participants was considerably lower ( n = 188). A total of 820 consent forms were distributed at the primary data collection site (i.e., Middle School B). Of these, 235 were returned (a 28.66% response rate). Of the consent forms that were returned, 27 indicated they did not want to participate in the study, and five indicated consent for the survey administration only. At the time of data collection, the re searcher was unable to locate 52 students who had parent consent to participate because these students were either (a) absent from school or (b) were not granted permission by their teacher to leave the classroom because of testing or class lecture. As such, a total of 35 students completed the surveys at the pilot school (i.e., Middle School A), and a total of 153 students completed the survey s at the primary data collection site (i.e., Middle School B). The data from the pilot school were not collapsed with the data obtained from the primary data collection site; therefore, the data from the pilot school were mainly used to determine readabili ty and whether there were any glitches in the administration
51 process that needed to be rectified prior to the second administration. Thus the statistical analyses were based only on participants from Middle School B. In order not to have any students fe el excluded during data collection, all students, regardless of race/ethnicity, were allowed to participate in the data collection, However, only students who were identified as minority students were included in the data analysis. The data that were co llected on Caucasian students will be used in future studies but were not analyzed as part of this study. Note however that only three participants were not included because of ethnic status. The one group of students who were excluded from data collectio n was students with severe developmental disabilities, specifically students who had compromised intellectual functioning. This group of students was not excluded because of possible problems with readability and interpretation of the instruments. Addition ally, logistical issues including the extensive personnel needed to support the completion of the survey and the extra funding to hire individual school support personnel was not available. Students were required to obtain informed consent from a legal gu ardian prior to participation (i.e., active consent). Additionally, participants were required to provide their assent prior to survey administration and individual interviews. A detailed explanation of student demographic characteristics is provided in C hapter 4. Procedure s Pre data collection. Once each school verbally agreed to participate, a formal letter was signed by each principal to document their support. The researcher then submitted an application to the University of South Floridas Institu tional Research Board (IRB) to garner approval for the study. Additionally, an application was also
52 submitted to the Department of Assessment and Evaluation of the specific county where this study was conducted. Approval was obtained from both entities. At the time of submission, consent forms in English (see Appendix A) and student assent forms for the survey study (see Appendix B) were submitted. The primary consent forms provided the opportunity for parents to decide whether their child was able to par ticipate (a) in the entire study including interview; (b) just the survey study; or (c) not at all. The different consent options were developed to try to increase the response rate by providing parents with a choice to allow their child to only participat e in the survey, if that was more comfortable for parents, rather than consenting to the entire study. Since Middle School B was composed primarily of Hispanic students, a brief letter explaining the study was provided in Spanish (see Appendix C). All stud ents received both an English and Spanish form, regardless of their racial background, to increase the probability of parents receiving a letter in their native language. Of note, it was not feasible to create a fully translated consent form in the time a vailable to complete the study before the school year ended, A meeting with the principal (guidance counselor at pilot school) was arranged to discuss how the schools would like to handle the logistical issues of data collection (e.g., dates and times). T he researcher proposed to both administrators that she would meet with students briefly in their classrooms to describe the study and pass out the consent forms. Specifically, the researcher suggested one specific subject area (e.g., electives) class to t arget for the consent process. The rationale for targeting a specific subject area was based upon the fact that the likelihood of obtaining an adequate size would be
53 increased if the researcher targeted a class that all students at the school were required to take. Middle School A chose to have the researcher enter all elective classrooms throughout the day to pass out the consent forms. Middle School B arranged to have the researcher enter each Language Arts classroom at the school. Additionally, the p rincipal at Middle School B assigned a contact person (the head of the Language Arts Department) with whom the research could work on logistical issues. Prior to meeting with students in their classrooms, the researcher requested to meet with all teachers in the subject area that was chosen to discuss the study in general terms and what their participation would entail. The pilot school was unable to designate a time for this to occur; thus, an email was sent to all teachers explaining the purpose, process and procedures. A meeting at School B was held one week prior to the commencement of data collection. Approximately 50% of the Language Arts teachers were at the meeting. A subsequent email was sent out to all Language Arts teachers informing them of the study. The researcher and a first year graduate student visited each elective classroom at the pilot school to seek consent for participation. This process took approximately two full days over a two week period. The primary researcher visited all clas srooms at School B three days a week for two weeks to inform students about the study and to pass out consent forms. During this process, students were informed about the general purpose of the study, what would be expected of them if they were to take par t of the study, and what the incentives would be. After each classroom discussion, which took approximately five to ten minutes total, teachers were provided with a manila envelope in
54 which to place the returned forms. Teachers were also given ample amoun ts of candy to give to students who returned their forms. Any candy that was leftover served as a donation to each teacher involved in the study. The researcher returned three times a week for two weeks after each discussion to remind students and teachers about the forms and to collect them. Forms were also returned directly to the Language Arts Department Head. Once the forms were collected, the researcher made a roster of students who obtained consent to participate for each school. The researcher provi ded the final roster to school administration. Before conducting the pilot survey, the Freys Readability Test was conducted on the entire survey to determine the reading level of the instrument. The entire survey was deemed to be at a 3 rd grade reading l evel. Based on the fact that the target population was middle school students, the researcher believed that three grades below the sixth grade re ading level would be sufficient A total of 450 consent forms were distributed at Middle School A; however, very few parent co nsent forms were obtained at this pilot school ( n = 35 ). Since the number of students who returned the forms was minimal at the pilot school ( n = 35), one day was scheduled for data collection. Following administration of the measures, 3 of the 35 students were asked general questions about what they thought about them. These students reported that they were easy, kinda fun, and weird. Students reported that they had difficulty understanding a question on the spirituality measure referring to the concept of higher power. A standard response to any question requiring the definition for a higher power was decided among the researcher and assistants (i.e., a higher power
55 is something or someone that you believe helps or controls t he world like God or Buddha). At School B, six dates were set to collect data. Students with parent consent to participate were asked to go to the cafeteria during either first or fifth period over the course of three weeks. By moving out of the classroo m, students (a) appeared more focused, (b) space was provided to spread out, thus increasing the likelihood of honest responses, and (c) those who were not involved in the study remained in their classrooms and received instruction. Also, in a smaller sett ing, persons involved with data collection were better apt at managing behavioral issues that arose because the ratio the number of students who required aid per researcher was lower. Data collection The researcher and one graduate student (and one un dergraduate student who helped twice) traveled to the primary site to collect the data. Each research assistant (and researcher) had a roster, assent forms, and surveys with them at each station. Students entered the cafeteria and created three lines to ch eck in with the researchers. Each member of the team highlighted the students name and marked their identification number next to their name (based on the number on the front of the survey). This procedure was instituted to (a) keep track of who particip ated in the study, and (b) have the opportunity to contact students after the survey for individual interviews, based on their responses. After each student was checked in, a survey packet was provided to each student, including the child assent form and t he demographics form (see Appendix J). Provided that students at School B were predominately Hispanic, and many of them were recent immigrants, two specific data collection days were set aside for ESOL
56 students. This decision was made because the undergr aduate research assistant spoke Spanish and could help translate any items if necessary. Moreover, the last two data collection days were set aside to provide additional time in case students reading was compromised because of language barriers. A total o f four students needed additional support from the undergraduate assistant because they did not understand some of the survey items. The order in which the measures were administered was randomized to eliminate any potential order effects Counter balanc ing the measures also protected confidentiality because students were completing different portions of the survey at various times. Moreover, students were seated at separate tables in the cafeteria and talking was not allowed to further protect student r esponses. When students completed and turned in the survey packet the research team scanned each survey for any missing items or items that had two or more responses. If a student made a mistake, the number of the item was circled, and the student was ask ed to return to his/her seat to finish. Finally, when students turned in their surveys, they received two pieces of candy. After the entire data collection was completed, three students were chosen randomly and were provided a $25 gift certificate to Best Buy. Students who turned in consent forms but were not allowed to participate in the study were also included in the total sample pool where names were drawn for the prizes. Additionally, the Language Arts teacher who collected the most consent forms rece ived a $20 gift certificate to Starbucks. This incentive was instituted because of a suggestion by the assistant principal.
57 Post data collection Student survey numbers were entered into a protected Excel sheet so that the researcher was able to track st udent names with identification numbers. To protect confidentiality, the rosters were shredded once the information had been entered into the database. The researcher entered the data into a separate Excel sheet where the only identifiable student informat ion was a number. Data were scanned to determine if any participant Christmas treed (randomly providing responses without reading the questions) the survey and/or if a survey had less than a 70% response item rate. Only one survey was eliminated from the study because of responses appearing haphazard. Once the data were entered and basic statistical analyses were completed, the researcher identified two different groups: students who scored high on victimization and low on depression and students who sco red high on victimization and high on depression. Students who scored high on the victimization scale had obtained scores that were at least one and a half standard deviations above the mean score for the entire sample. Indication of pathology either hi gh or low was determined by the T score obtained on the depression inventory. Any student who scored over T = 70 on the CDI S was considered part of the high pathology group because this is the criterion in the CDI S manual to label someone as experienci ng depression in the Clinically Significant range. Any student who received a score of T = 38 or below were considered in the low pathology group. These two different groups were the population targeted for qualitative inquiry. Interviews In order to further understand the quantitative data, the researcher conducted individual interviews with eight students to qualitize the data and provide
58 further insight into the importance or lack thereof of the proposed moderator variables in this study. Each i nterview was audio taped and transcription was completed by first and second year graduate students in a School Psychology program in Florida. In the consent letter, parents were alerted to the fact that their child may be asked to further participate in a n individual interview; thus, a new consent procedure did take place at this point in time. There were a total of seven steps that were completed prior to conducting the research interviews. This process included (1) selecting a sample, (2) designing the interview format, (3) developing questions, (4) selecting and training interviewers, (5) doing a pilot test of the interview procedures, (6) conducting the interviews, and (7) analyzing the interviews (Stewart & Cash, 1997). As previously stated, the samp le of students to be selected was determined by victim status and depression score. This strategy is known as extreme case sampling, which is a type of purposive sampling. The interview format was a standardized open ended interview which involved a pre se lected set of questions as to limit researcher bias. The interview guide (see Appendix J) specified the questions and the sequence in which they should be asked. However, qualitative approaches often take a recursive approach such that the approaches tend to be flexible because they often depend on guidance and information provided by participants. Therefore, it was necessary at times to alter the original interview questions when the participants thoughts and ideas were not congruent with the initial qu estions. This researcher was the primary interviewer and a first year School Psychology student conducted two interviews. A total of eight interviews were held (i.e., four for each type of participant).
59 While the variables were evident for the quantitati ve analyses (e.g., depression, victimization), at the time the interview protocol was developed there was no support for the creation of a priori themes or codes for the qualitative portion of this study. Since the ultimate goal was to give students voice about their experiences, this researcher did not want to limit the discussion or possible topics covered during the interviews. Measures The Childrens Depression Inventory Short Form (CDI S). Depression was assessed using the Childrens Depression Inventory Short Form (CDI S; Kovacs, 1985). The original CDI is a 27 item questionnaire that assesses major aspects of depression including cognitive, behavioral, and affective symptoms. The CDI S measures the same aspects of depression; however, the tota l number of items on the instrument is 10 (see Appendix D). Each item asks an individual to describe his/her feelings about a variety of issues in the past two weeks, on a scale ranging from zero to two. A score of zero indicates an absence of the symptom, whereas a score of two indicates that the symptom is present most of the time. The author of both forms noted that comparable results are obtained when using either instrument. The current study used the CDI S instead of the CDI to decrease any type of response bias that would have occurred if the entire survey was too long and cumbersome for students. The CDI and CDI S are noteworthy among other self report questionnaires for depression because they have been used in studies of both clinically referred and non referred youth. The scales also include a wide range of symptoms (other than mood) (Compas, 1997). Internal consistency for the CDI has been found to be adequate and high (e.g., a = .80; Smucker, Craighead, & Green, 1986). Test retest reliability also has been found to be
60 relatively high; however, in some non clinical samples, depressed mood fluctuates more often and thus the stability coefficient is sometimes skewed because students scores may change dramatically from baseline assessment to comp arison assessments (Saylor, Finch, Spirito, & Bennett, 1984). Overall, the CDI and CDI S are widely used measures that have been found to have adequate psychometric properties with a variety of populations, including the population that was surveyed in thi s study ( Graham, Bellmore, & Mize, 2006; Nangle, Erdley, Newman, Mason, & Carpenter, 2003; Storch, Noc k, Masia Warner, & Barlas, 2003 ). Notably, at the end of each day of data collection, this researcher scanned all the CDI S forms to see if any student responses were particularly high, such that a student would require services immediately. No students appeared in need of immediate service. However, an Excel spreadsheet was made with students names that scored in the Clinically Significant range for depression, and the list was provided to the school psychologist at Middle School B. The Child and Adolescent Social Support Scale (CASSS). The Child and Adolescent Social Support Scale (CASSS; Malecki, Demaray, & Elliott, 2000) was originally develope d to ascertain adolescent participants perceptions of social support from five sources: parent(s), teacher(s), classmate(s), a close friend, and persons at school (see Appendix E). Each sub scale contains 12 items that measure four different types of supp ort: emotional, appraisal, informational, and instrumental. Students read each statement and rate how often they perceive a particular supportive behavior on a six point scale. The higher the global and sub scale scores are, the more social support a stude nt perceives.
61 Malecki, Demaray, and Elliot (2000) and Malecki and Elliott (1999) found high construct validity for the CASSS because strong correlations existed with other established measures of social support. Likewise, when the CASSS was correlated wit h other self report measures of similar yet different psychological constructs (e.g., self concept, social skills), lower correlations were found. The authors reported a test retest reliability of .78 for the total score. Internal consistency levels were found to be very high, with coefficient alpha scores ranging between .92 and .95 for each subscale. This researcher used a modified version of the CASSS such that the scale persons at your school was not utilized. Assuming that one assesses the construc t on face validity alone, it appears that there may be a high degree of overlap between the scale persons at your school and the scales classmates and teachers. Therefore, this researcher contacted the scale author to determine if, in fact, more data existed to explain whether or not an overlap was present in her dataset. Dr. Demaray provided this researcher with a factor structure matrix, a correlation matrix, and a total variance explained matrix (based on a sample of over 3,000 students in elementa ry, middle, and high school). According to the factor structure matrix, persons at your school shared no loading with parents, minimal loading with close friend, and high loadings with teacher(s) and classmates sub scales. Furthermore, moderate c orrelations were found between the sub scale persons at your school and all other sub scales. Finally, the variance accounted for by the sub scale persons at your school was small ( R = .04). Therefore, this researcher decided to exclude persons at yo ur school sub scale based on the information provided by Dr. Demaray. In trying to keep the survey as short as possible (in order to maximize honest responding), it made sense to eliminate a scale that
62 evidenced high correlations with other sub scales and contributed minimally to the overall explained variance of the instrument. Spirituality Assessment Instrument (SSA) The Spirituality Assessment Instrument is a brief, self report measure developed by this researcher (see Appendix F). The items were ada pted from the Spiritual Involvement and Beliefs Scale (SIBS; Hatch et al., 1998) which is a self report inventory that assesses spirituality across various religious traditions. Items included in the scale were developed with the intention of the instrumen t being used across racial and ethnic groups; therefore, the author tried to use the most general language possible so that the instrument was appropriate for a wide variety of individuals. The instrument taps into a number of underlying principles includi ng the ability to apologize, to forgive, to pray, and to trust in the unknown. This researcher chose to develop a new instrument for a variety of reasons. First, other commonly used instruments such as The Spiritual W ell Being Scale (SWBS; Bufford, Paloutz ian, & Ellison, 1991 ) have a number of psychometric limitations (Ledbetter, Smith, Fischer, Volser Hunter, & Chew, 1991). Additionally, the SWBS has a more narrowly defined focus of spirituality, such that many of the items have a Judeo Christian overtone. Since this researcher was interested in spirituality and not religiosity, adapting questions from the SIBS was an attempt at developing an assessment instrument more consistent with a broad framework of spirituality (i.e., the authors consulted with leade rs of various denominations including Islam, Hindu, Buddhist, and Christian to try to determine which common aspects of spirituality were akin to each religion). The new instrument also assessed both behaviors and cognitions associated with spirituality.
63 The third reason a new instrument was constructed was because the SIBS was standardized on a relatively small sample ( N = 77) of adults. This sample is not similar to the demographics of participants in this study. Finally, there was no norm referenced ins trument available, to this researchers knowledge that had been developed for adolescents. Clearly, this is one of the major limitations in spirituality research today: the non existence of appropriate culturally and developmentally sensitive instruments. Notably, on the new scale, items one and five ask the same question. Item one was repeated to measure consistency of participant responding. A lie index was not developed for other measures utilized in this study because the other measures had been previ ously developed by other researchers, and normative data had already been substantiated. Based on the lie index, 42.7% of the answers on items 1 and 5 were an exact match; conversely, 57.3% of the answers had a lack of agreement. A majority of the particip ants responses obtained a one point difference (34.87%), followed by a two point difference (13.82%), three point difference (7.2%), and finally a four point difference (1.32%). Because this scale is in the preliminary stages of development/validation and there is some support for a tendency towards inconsistent responding, the results obtained using the SSA should be viewed with caution. The Childrens Hope Scale (CHS). The Childrens Hope Scale (CHS; Snyder, Hoza, Pelham, Rapoff, & Ware, et al., 1997) i s a dispositional self report scale that assesses two elements of hope: agency and pathways thinking related to goal attainment for children aged 8 to 16 years. The scale was initially used with a homogenous sample of Caucasian children aged 9 through 14 y ears in Oklahoma ( N = 372). The standardization sample included a variety of children with special needs including boys
64 with a diagnosis of Attention deficit/Hyperactivity Disorder (ADHD) and children with chronic illnesses (i.e., sickle cell anemia, arthr itis, cancer). The initial CHS consisted of 12 items; however, the current scale includes six items. Three items address pathways thinking and three items address agency thinking. Children are asked to answer questions based on a five point scale with resp onse options ranging from all of the time to none of the time. Items include statements such as I think I am doing pretty well and I think things I have done in the past will help me in the future. The measure is shown in Appendix G. The median score f or internal consistency on the CHS was .77, and the test retest correlation was .71. To ascertain a convergent validity coefficient, parents were asked to rate their childrens hope. Snyder et al. (1997) found a moderate correlation between self and parent reports ( r = .38). Discriminant validity was assessed by comparing the CHS with Kazdin et al.s (1983) Hopelessness Scale and the WISC III. The results indicated that there was a small (sometimes) negative correlation among the measures ( r = .24, and .03 ; respectively). Valle, Huebner, and Suldo (2004) conducted another evaluation on the CHS with a more heterogeneous sample (i.e., African American and low SES) of older adolescents ( N = 460) ranging in age from 15 to 19 years old. The coefficient alpha sc ore was higher than the original study conducted by Snyder and colleagues (1997) ( a = .84). A confirmatory factor analysis was completed to test the theoretical underpinnings of the CHS. A two factor model was supported (GFI = .96; CFI = .95). Criterion validity was established by comparing the CHS with various measures including life s atisfaction (SLSS), social support (CASSS), problem behaviors (YSR), life events (LEC), and
65 temperament (JEQ A). T tests were conducted to determine whether there were any differences reported on the CHS relative to SES, gender, and race. No statistically significant differences were found for SES or gender. However, a significant difference was found for race, such that there was a difference between scores for African American children and Caucasian children. However, Valle et al. (2004) noted that the ef fect size for this difference was small ( d = 0.12). Valle et al. (2004) also replicated the study on middle school students aged 10 to 14 years ( N = 531). All psychometric properties were in the moderate to high range; for example, the coefficient alpha o btained was .83, and criterion validity was established with life satisfaction ( r = .49), perceived social support ( r =.59), and temperament ( r = .18), and behavior, both internalized and externalized ( r s = .33 and r = .32, respectively). Gender and race were found to have significant differences, but the effect sizes were small. Given that both studies established excellent psychometric properties for the CHS, this researcher felt confident that the scale was an appropriate measure to assess the construct of hope in children and adolescents. Furthermore, while the instrument was initially meant to be administered to students aged 5 16 years, Valle et al. (2004) provided evidence that the instrument may be used in early and late adolescence thus meeting the needs of the current research study. Furthermore, the instrument had been used with African American students (Valle et al., 2004), and the findings were adequate. Social Experiences Questionnaire (SEQ). The Social Experiences Questionnaire (SEQ; Crick & Grotpeter, 1996) is one of the most widely used measures to assess overt victimization (three items), relational victimization (five items), and prosocial behavior
66 (five items) (c.f., Storch et al., 2003; Storch & Esposito, 2003) (see Appendix H). The sca le consists of 15 items that are rated on a scale ranging from 1 (never) to 5 (all the time). Initial psychometric data collected by Crick et al. (1996) found adequate internal consistency across factors ( a = .77 to .80). However, this study only included third through sixth graders, and thus the adequateness of the measure cannot be generalized to older students. Storch and colleagues (c.f., Storch, Crisp, Roberti, Wagner, & Masia Warner, 2005) addressed the issue of the psychometric properties and the val idity of the SEQ scores for use with adolescents. A total of 1,178 adolescents from predominately middle class Caucasian backgrounds were used in this study. Initial results revealed a gender difference in overt victimization (i.e., that males had signific antly higher scores than females). However, there were no gender differences in scores on the relational victimization sub scale. This finding is consistent with the literature that suggests males engage in more overt forms of bullying behavior (Storch et al., 2003). The internal consistency was high for the relational victimization and prosocial behavior scales ( a = .82; a = .77, respectively); however, only a moderate alpha was found for the overt victimization scale ( a = .50). Correlations were found between overt and relational victimization ( r = .58, p < .001) and relational victimization and prosocial behavio r ( r = .42, p <.001). Overall, the major limitation of using this specific instrument was that it had not been used with a minority sample. Yet, at the same time (as stated in chapter 2), most of the research on victimization has not been conducted with m inority groups. Therefore, the lack of psychometric support of this scale for assessing minority students was
67 expected given the past and current trend in the types of participants that are in relational victimization studies. Regardless, the instrument ha s been widely used and maintains the most support by experts in the field to identify whether someone experienced relational aggression. Data Analysis Analysis software The quantitative analyses were conducted using Excel, SPSS, and MPlus (Muthen & Muth en, 1998). The qualitative data were transcribed into a word document. To further systematize the analysis of the qualitative data, the Atlas software program was utilized. A professor in the Department of Special Education at the University of South Flori da provided training to this researcher on the software. By entering the qualitative data into such a program, complex relationships that may have not been visible to the naked eye were found. Furthermore, this kind of software has support for revealing ps ychological states associated with different experiences, which is exactly what this researcher assessed in this study (Gottschalk & Bechtel, 1995). By using software packages such as Atlas, the researcher was better apt at determining the between method triangulation; consequently, resulting in a better understanding of how the data converged, diverged, or was complementary (Kelle, 2001). Furthermore problematic issues related to interpretive rigor were eliminated (Lincoln & Guba, 2000) when using Atlas. To improve rigor, inter rater reliability was established with another graduate student at an 85% agreement rate. This person also identified themes and categories, and completed the same analysis as the main researcher. Descriptive statistics Means, me dians, standard deviations, and values for skewness and kurtosis were computed for the individual scores across measures. This
68 was illustrated quantitatively in the form of tables. Extreme outliers were detected and made note of to possibly include in th e qualitative portion of the study. However, extreme outliers were not dropped from this study because the focus of the study was, in part, addressing extreme outliers, also known as persons who experienced very high levels of victimization and/or were e xperiencing severe depression. Relationships among predictor and outcome variables The Pearson product moment correlation coefficient (PPMCC) was used in this study to see the degree of overlap between each construct. A correlation matrix was used to rep resent the findings. Additionally, the correlation matrix provided insight into issues with multicollinearity among variables. Comparison of groups Given that there were so many different combinations of characteristics that this research agenda was inte rested in understanding, a variety of analyses were run to compare groups by ethnicity and grade. The analyses that were run included a one way analysis of variance (ANOVA) and multivariate analysis of variance (MANOVA). Mul tiple regression Multiple regre ssion allows a researcher to look at the relationship between predictor variables and a dependent variable and provides an established method for determining how much variance a specific predictor variable accounts for any change in the outcome variable (e .g., R ). In order to understand the equations that drove the testing of the model, a pictorial representation of the proposed model may help the reader conceptualize the following series of equations (see Figure 1). The model suggests that depending on t he degree of victimization (e.g., score of four or five indicates experience of victimization is strong)
69 varying levels of depression will result. However, this study hypothesized that a number of moderator variables, both individually and in combination w ith one another, would impact the relationship between victimization and depression. The unit of analysis in this study was an individual students depression score (i.e., outcome variable). The moderator variables included hope, spiritua lity, and perceived support. Note that gender, grade, and ethnicity were not driving variables that predicted depression; however, this researcher hypothesized that gender, grade, and ethnicity may have had a direct relationship with the degree to which on e was spiritual and/or experienced hope. Thus, ultimately this model was used as the basis for understanding potential three way interaction effects and is consistent with the process for testing moderator models as proposed by Baron and Kenny (1986). Fi nally, the assumptions that underlay this specific analysis were tested including linearity, normality, multicollinearity, and homoskedasticity. Chapter 4 provides further detail on the outcomes of the assumption testing. Figure 1 Conceptual moderator model for depression RA Victimization Depression Moderator Variables (Spirituality, Hope, Support) Grade Race Gender RA Victimization Depression RA Victimization Depression Moderator Variables (Spirituality, Hope, Support) Grade Race Gender
70 Rationale for model testing Give n this is the first time, to this researchers knowledge that the aforementioned variables were analyzed together, there was a potential for the relationships between variables to be weak. Therefore, a priori, the researcher could not review other studies to determine the probable correlation coefficient that would result in this particular study. Moreover, the correlation results obtained from this study indicated that the relationships between variables were moderate. According to the literature on modera tor versus mediator models, researchers suggest that mediator models be conducted when a strong relationship is established a priori and during the course of the study (Frazier, Tix, & Barron, 2004). Therefore, from a methodological standpoint, this partic ular study lent itself better to a moderator analysis. The second reason why a moderator analysis was conducted was based on the conceptual understanding of relational victimization and depression. Given that the research portrays a mixed picture of relat ional aggression and victimization prevalence rates, this researcher expected that there would, in fact, be differences among males and females. As such, gender was anticipated to have a differential impact on the relationship between relational victimizat ion and depression. Furthermore, females have been known to have a higher incidence rate of depression beginning in adolescence. Given that this population was mainly composed of adolescents, the researcher expected that again, gender would be an influenti al variable underlying the key constructs in this study. The type of study that would discern whether gender attenuated the relationship between the predictor and outcome variables is a moderator like study.
71 Finally, ethnic group affiliation was an impor tant factor in this study, as much of the work to date on relational victimization has focused on either Caucasian persons or participants not located in the United States. Three different ethnic groups were analyzed in this study, and one of the goals of this study (i.e., the driving theoretical research question) was to elucidate for whom the relationship between the predictor and outcome variables was either stronger or weaker. This kind of research question is aligned with a moderator variable test, rat her than a mediator test which would answer the how and why types of questions (Wong, Beutler, & Zane, 2007). Along the same lines, questions related to paths and causality typically require some type of intervention implementation and tend to have the highest power when the study is experimental in nature (Vujik, van Lier, Crijnen, & Huizink, 2006). This study was non experimental, did not have any type of randomized control design, and an intervention was not implemented. Therefore, because of the des ign of the study and the driving conceptual questions posed by the researcher, a moderator test was deemed more appropriate than a mediator test. Chapter 4 Results Overview 3 In this chapter, the results of both the qualitative and quantitative aspec ts of the study are reported. This chapter begins with a section titled Treatment of the Data, which discusses data entry and verification techniques, as well as descriptive information in detail (e.g., central tendencies) for student demographic inform ation and for each 3 The data in this study were collected in the last two month s of the school year
72 specific survey instrument. Differences in self reports between students from different minority groups were examined to determine whether the groups should be collapsed into one group (e.g., minority) or should remain separated (e.g., H ispanic). A section devoted to this decision making process is also included in this chapter. Subsequently, the process of determining reliability for each of the measures is discussed. Following these two sections, the results for each research question posed in Chapter 1 are described. Treatment of the Data The data were entered into an Excel spreadsheet by both the researcher and a first year graduate school psychology student after each administration of the survey. Each score was entered for every participant on each item. Missing data were coded as a blank space in the Excel document. If a participant circled two scores for one item, the mean score was calculated and inputted. After all data were entered, a separate database was created for minor ity participants to ensure that only minority students were included in the data analysis. The researcher checked the data by randomly picking various participants ID numbers and subsequently matching the data in the database to the paper and pen survey completed by the participant. If an error was found, the data point was changed in the computer to reflect the correct answer. Additionally, the researcher checked the subsequent assessment protocol (by participant ID number) to make sure that the previou sly coded assessment was correct; for example, if the protocol was MM180, MM179 was checked for errors. The percent of error during data entry was minimal: a total of 2% of the data points entered were incorrect. Additionally, extreme values were checked across each participant for each item to ensure that the data were either (a) entered incorrectly or (b) the case was exceptional (i.e., outlier). A total of one case was
73 dropped from the entire data subset because Christmas treeing was readily apparent and the protocol was detected as an outlier in SPSS. Reliability of M easures A thorough analysis was conducted for each measure to determine the reliability of the scores utilized in this study. Given that the main construct under review (relational vict imization) was hypothesized to be influenced by gender and ethnicity, two tables, Table 1 and 2 in Appendix L provide more detailed information about the reliability data for each measure. Reliability statistics for ethnicity by gender were not examined be cause the sample sizes in the cells were too small and the variance was zero for some cases. Therefore, reliability coefficients were not able to be calculated. CDI S The overall reliability for the CDI S was .82, which is consistent with the reported re liability coefficient in the literature (a = .80). The CDI S was slightly more reliable for females ( a = .84) than for males ( a = .74). The reliability also differed based on ethnicity such that African American and Hispanic participants obtained higher re liability coefficients (a = .82 and .83, respectively) compared to students in the mixed race group (a = .73). There was a wide range for the item to total statistics indicating that some items correlated more closely to the total score than others (0.00 t o .72). The average inter item correlation coefficients for the total sample, sample separated by gender, and sample separated by race were in the moderate to small range for all items. SEQ The reliability of the SEQ for the total sample was virtually the same (a = .84) as previously reported (a = .77 to .80). The scores from the instrument were more reliable for males (a = .84) than females (a = .74). Additionally, the scores appeared more reliable for African American (a = .80) and Hispanic students ( a = .81) than mixed
74 race students (a = .72). The range for item to total correlations was wide such that, for mixed race students, a relationship barely existed between some of the items and the total score (i.e., r = .13), and conversely was quite large f or males overall (i.e., r = .88). CHS The reliability for the CHS in this study (a = .85) was comparable to other reported studies (a = .84). The alpha coefficients were relatively similar across gender and race. All of the reliability coefficients were i n an acceptable range. The item to total correlations had a more restricted range than the aforementioned measures such that the relationship was higher across all gender and ethnic groups ( r = .41 to .81). CASSS. The reliability of the scores on the CAS SS (48 items) was high (a = .97). Additionally, the obtained reliability coefficient was higher than the reliability purported by the developers of the measure (a = .78). There were no differences in the reliability coefficients across gender or ethnic gro ups. The average inter item correlation was .34, indicating that the items were related but not to the extent that would cause concern for future statistical analyses. SSA Since this six item instrument was developed for this particular study, there were no other studies with which the reliability statistics could be compared. Overall, adequate and high reliability (a = .81) was achieved; however, the reliability coefficients were higher for females (a = .82) compared to males (a = .75). High reliability was also found for African American and Hispanic students (a = .82); however, inadequate reliability for mixed race students (a = .65) was obtained. Also, the average inter item correlation found for the data on mixed race students was almost zero ( r = .04 ) indicating that there was almost no identifiable relationship among the items on the measure.
75 Considering the instrument for all other participants, moderate inter item correlations were achieved. Confirmatory f actor a nalysis (CFA) Prior to discussing the results gleaned from this study, a brief exploration of the factor structures of the CASSS and SEQ will be discussed. As a reminder, the subscale, persons at your school was eliminated because of the strong loading it had with two other subscales (i. e., teachers and classmates). According to Dr. Malecki (personal correspondence, November, 2006), the other four factors had little to no relation with one another ( N = 3,000). The current study, using CFA and maximum likelihood estimation confirmed that there were four distinct continuous latent variables or factors underlying the CASSS. T he fit indices found for this instrument ( RMSE A = 0.09; CFI = 0.79) were not as high as found by the developers of the scale ( RMSEA = 0.05; CFI = .90) (personal correspo ndence with Dr. Maleckis graduate assistant, June 2007). Notably, if the model obtained a good fit, the RMSEA score would be 0.05 or less and the CFI score would be above 0.90. A number of the items cross loaded onto other factors. For example, three it ems that composed the teacher subscale (i.e., items 13, 17, and 21) also loaded onto the parent subscale (several modification indices were above 3.84 which indicates model data misfit ) Moreover, two items on the best friend factor also had high loadings on the parent subscale (i.e., items 44 and 45). Some of the error s (uniqueness) for the observed variables were also correlated with one another which was less than favorable since error terms are expected to be (modification indices ranged from 3.33 to 26 .89) random. Again, these findings are important when thinking about the gestalt that the data present.
76 Yet, this study only included a small sample size and thus the lack of fit and complete support for the factor structure may be attributed to this limit ation. A three factor model was tested using CFA for the SEQ. The three factors included relational victimization, prosocial behavior, and overt victimization. The model was estimated using maximum likelihood estimation. The fit statistics indicated less than acceptable fit; however, they were somewhat adequate given the small sample size used to run the CFA ( CFI = 0.86; RMSEA = 0.08). The developer of this instrument was contacted to ascertain the most recent fit statistics. The research lab contacted th is researcher and stated that information from a CFA is not on file and that the lab could not provide further information. Two factors, relational victimization and overt victimization, were highly related ( r = .86). Additionally, the modification indices indicated that there were a number of items that contributed to the misfit to this the model. The range of modification indices were from 5.96 to 40.56. Since the primary focus was not on CFA, the researcher decided not to further delve into more detailed information about the factor structures and cross loadings. Demographics. When considering the complete data subset in this study, 42 students were classified as African American, 85 students were classified as Hispanic, and 26 students were classified as mixed race. Most students reported living with their mother and father (36.6%), followed closely by students living in a single home headed by their mother (28.1%). Only 2.0% of students reported living in a home with their father only, while 23.5% of stu dents lived with a biological parent and a stepparent. A small percentage of students lived in foster care placement (0.7%) and 9.2% of students reported other for their living situation.
77 The average age of students who participated was 12.94, or almost 13 years old ( SD = 1.13). The youngest student who participated was 11 years old and the oldest student who participated was 16 years of age. While this would typically signify a sample largely composed of seventh graders, the majority of the sample was s ixth graders (42%), followed by seventh graders (32%), and then eighth graders (26%). Retention is the most likely culprit for the above average age of this population. Based on participants self reports, most of the students were average or above averag e achievers. Results indicated that 15.7% of students reported that they received mostly As, 37.9% of students reported obtaining mostly As and Bs, 11.1% of students reported obtaining mostly Bs, 21.6% reported obtaining mostly Bs and Cs, and 8.5% of students reported obtaining mostly Cs. The other 13.7% of students reported a grade point average (GPA) of lower than 1.74. Students in this sample spent relatively little time on their homework, such that 44.4% of students reported that they spent less than one hour per week on their homework. An average of one hour to less than three hours of homework completion was reported by 35.9% of the sample, followed by a range of three hours to less than five hours per week by 9.2% of the sample. Some students reported spending from five hours to less than ten hours per week on their homework (7.2%), while 2.7% of the students in this sample reported spending ten hours or more on their homework per week. Students also reported spending few hours participating i n after school activities such as band and sports. Thirty four percent of students noted that they never participate in after school activities over the past few months, followed by 23.5% of students who reported engaging in after school activities once or twice in the past month. A few students reported that they engage in these types of activities once per week (7.2%) and
78 several times per week (17.6%), in the past month. Finally, some students reported that they engage in daily after school activities fo r less than one hour (5.9%) and for more than one hour (11.1%). This particular sample declared association with many friends, given that 58.6% of the sample said they had ten or more friends. Interestingly, 24.3% of the sample reported they did not know h ow many friends they had. Finally, the overall percentage of students decreased when the choices of number of friends also decreased: seven to ten friends (8.5%), five to seven friends (4.6%), two to four friends (3.3%), and one friend (0.7%). In ord er to determine if there were any categorical variables that were particularly different across races, a chi squared analysis was completed for two demographic variables: gender and free and reduced lunch. While the sample consisted of more females than ma les, a significant difference for gender was not evident across the minority groups, ? (2, N = 152) = 0.18, p > .01. Additionally, 82.9% of the students stated that they received free and reduced lunch, and the majority of students who received this servi ce were African American. A statistically significant difference did not exist for students of different minority membership and their attainment of free and reduced lunch, ? (2, N = 152) = 2.5, p > .05 (see Table 3). Note that non minority participants were only included in this percentage and that the percentage may be higher and more representative of the school population as a whole, if Caucasian students were included in data analysis.
79 Table 3 Analysis of Categorical Variables for Differences in the Participant Sample Variable Total ( N = 153) African American ( N = 42) Hispanic ( N = 85) Mixed ( N = 26) ? Gender Male Female 33.6 % 66.4 % 35.7% 64.3% 33.3% 66.7% 30.8% 69.2% 0.18 Free and reduced lunch 82.9 % 90.5 % 81.0 % 76.9 % 2.58 Defining Characteristics of Each Survey Instrument Perceived social support Each variable was exa mined in three different ways by looking at descriptive statistics separated by gender, ethnicity, and ethnicity by gender. Examination of the data in this manner was important because this study relied heavily upon issues of gender differences and minorit y status. Perceived social support had an overall mean of 4.58 ( SD = 0.86) for males and 4.82 ( SD = 0.81) for females, p > .05. These scores indicated that a majority of the participants in the sample perceived that they were supported either most of the time or all of the time. The skewness and kurtosis values indicated a fairly normal distribution of scores for males and females ( skew = 0.25, 0.52; kurtosis = 0.80, 0.28, respectively). The effect sizes for the difference in mean scores between mal es and females was small for the parent ( d = 0.05 ), teacher ( d = 0.08 ), and classmate ( d = 0.20 ). A large effect size was found for the best friend subscale ( d = 0.80 ). See Table 4 for further information.
80 Table 4 Characteristics of the Social Suppor t Measure (CASS S ) by Gender Scale Total ( N = 153) Males ( N = 51) Females ( N = 102) Effect size Parent M 4.89 4.95 4.85 0.05 SD 1.05 0.97 1.09 Skewness 0.94 0.95 0.92 Kurtosis 0.18 0.82 0.04 Teacher M 4.69 4.65 4.72 0.08 SD 1.01 0.85 1.08 Skewness 0.73 0.39 0.82 Kurtosis 0.09 0.35 0.10 Classmate M 4.33 4.18 4.41 0.20 SD 1.12 1.11 1.13 Skewness 0.51 0.32 0.63 Kurtosis 0.34 0.61 0.09 Best Friend M 5.05 4.56 5.30 0.80 SD 0.99 1.07 .85 Skewness 1.02 0.27 1.61 Kurtosis 0.16 0.99 2.43 Total Scale M 4.74 4.58 4.82 0.29 SD 0.83 0.86 0.81 Skewness 0.43 0.25 0.52 Kurtosis 0.52 0.80 0.28 Note Scale range = 1 (Never) to 4 (Most of the Time) to 6 (Always). Effe ct size = (Male mean female mean)/Pooled SD.
81 Table 5 affords the reader a different look at the data. According to this table, the parent subscale may be categorized as the scale with the most non normal score distributi on across all ethnic groups. M ix ed race and African American participants reported the most support from parents ( M = 4.98, SD = 0.89; M = 5.12, SD = 0.92, respectively) compared to Hispanic students, p > .05. Hispanic students perceived the most support from their best friend ( M = 5.12, SD = 0.97) compared to students in other ethnic groups, p > .05. Overall, all students across ethnic groups felt the least supported by classmates. There were a number of small and moderate effect sizes across the ethnic groups. No large effect sizes were evident for any of the sub scales on the CASSS, based on ethnic membership.
82 Table 5 Characteristics of the So cial Support Measure (CAS S S) by Ethnicity Scale African American ( N = 42) Hispanic ( N = 85) Mixed Race ( N = 26) Effect Size Parent M 5.12 4.74 4.98 SD 0.92 1.13 0.89 Skewness 1.59 0.76 0.39 Kurtosis 3.10 0.38 1.42 0.36 a 0.22 b 0.15 c Teacher M 4.85 4.69 4.45 SD 0.78 1.06 1.15 Skewness 0.12 0.82 0.48 Kurtosis 0.86 0.06 0.39 0.16 a 0.22 b 0.43 c Classmate M 4.37 4.31 4.35 SD 1.30 1.09 0.95 Skewness 0.53 0.55 0.33 Kurtosis 0.96 0.06 0.00 0.05 a 0.04 b 0.02 c Best Friend M 5.10 5.07 4.91 SD 0.97 0.97 1.11 Skewness 0.88 0.97 1.36 Kurtosis 0.66 0.15 1.71 0.03 a 0.16 b 0.19 c Total Scale M 4.86 4.70 4.67 SD 0.82 0.84 0.81 Ske wness 0.48 0.50 0.10 Kurtosis 0.81 0.31 0.64 0.19 a 0.04 b 0.23 c Note Scale ranging from 1 (Never) to 4 (Most of the Time) to 6 (Always). a Effect size = ( African American (AA) Mean Hispanic (HS) Mean)/Pooled SD ; b Effect size = (HS Mean mixed race (MR) Mean )/ Pooled SD ; c Effect size = (AA Mean MR Mean)/Pooled SD. Males across all ethnic groups obtained mean scores that appeared lower on all subscales of perceived social support compared to females across al l ethnic groups, but this finding was not statistically significant. Based on the total mean score for females in
83 this sample, a persons best friend was perceived as providing the most support. Males on the other hand differed: African American males repo rted that their teacher provided the most support, while Hispanic and Mixed race students perceived their parents as providing the most support. Again, while these findings are interesting, the differences were not deemed statistically significant (see Tab le 6). Table 1 in Appendix K provides a detailed account of the effect sizes for these differences. As evident by Table 1 in Appendix K, the largest differences existed between African American females and Hispanic males on the best friend support subscale ( d = 0.91) and between African American females and mixed race males ( d = 1.00), p > .05. Furthermore, a large effect size was found between African American females and Hispanic females on the parent support subscale ( d = 0.74), p > .05.
84 Table 6 Chara cteristics of the So cial Support Measure (CASSS) by Ethnicity and Gender Scale Total ( N = 153) M F African American ( N = 42) M F Hispanic ( N = 85) M F Mixed race ( N = 26) M F Parent M 4.95 4.85 4.66 5.37 5.07 4.58 5.08 4.94 SD 0.97 1.09 1.21 0.60 0.83 1.23 0.91 0.90 Skewness 0.95 0.92 0.99 0.97 0.59 0.60 0.47 0.39 Kurtosis 0.82 0.04 0.72 0.21 0.70 0.81 1.72 1.43 Teacher M 4.65 4.72 4.81 4.88 4.56 4.74 4.56 4.40 SD 0.85 1.08 0.82 0.77 0.87 1.15 0.90 1.26 Skewness 0.39 0.82 0.14 0.28 0.81 0.87 0.44 0.53 Kurtosis 0.35 0.10 1.18 0.55 0.13 0.32 1.72 0.59 Classmate M 4.18 4.41 4.25 4.44 4.17 4.38 4.07 4.47 SD 1.12 1.13 1.42 1.25 1.00 1.14 1.00 0.93 Skewness 0.32 0.63 0.55 0.52 0.36 0.67 0.66 0.78 Kurtosis 0.61 0.09 1.23 0.81 0.18 0.16 0.55 1.51 Best Friend M 4.56 5.30 4.61 5.38 4.54 5.33 4.53 5.08 SD 1.07 0.85 1.14 0.75 1.06 0.83 1.15 1.07 Skewness 0.27 1.61 0.05 1.40 0.30 1.44 0.83 1.85 Kurtosis 0.99 2.43 1.88 1.34 0.78 1.07 0.90 4.05 Total Scale M 4.58 4.82 4.58 5.02 4.59 4.76 4.56 4.72 S D 5.21 0.81 1.01 0.67 0.81 0.86 0.83 0.81 Skewness 0.25 0.52 0.11 0.32 0.53 0.54 0.29 0.25 Kurtosis 0.80 0.28 1.88 0.30 0.26 0.42 0.49 0.37 Note Scale ranging from 1 (Never) to 4 ( Most of the Time) to 6 (Always). M = M ales. F = F emales.
85 While the subscale data are interesting within themselves, the initial research questions were developed to measure perceived social support as one distinct construct, rather than four different kin ds of social support. Moreover to ensure credibility in measuring perceived social support as one construct, a correlation analysis was completed to determine if there was a rationale for combining the measures or separating them. The analysis indicated th at there was a moderate to high overlap between the total mean scores on each subscale (see Table 7). Thus, the decis ion was made to continue to conceptualize and analyze perceived social support as one construct. Table 7 Correlation Matrix of Subscales on Perceived Social Support Measure (CASS S ) (N = 153) Scale 1. 2. 3. 4. 5. 1. Parent 1 2. Teacher .45* 1 3. Classmate .51* .51* 1 4. Best Friend .46* .44* .68* 1 5. Total .76* .75* .86* .81* 1 p < .01. Spirituality. The next variable und er review is spirituality, as assessed by the SSA. The mean score for all participants in the sample was 4.04 ( SD = 0.81). The data for the total sample deviated slightly from a normal distribution. A statistically significant difference was found for the total score on the spirituality measure, F (2, 104.87) = 4.49, p < .05, such that different ethnic groups purported different levels of spirituality. African American students reported the highest levels of spirituality in their lives ( M = 4.33, SD =
86 0.68); however, the data obtained from this sub sample of students was not normally distributed ( skew = 1.57, kurtosis = 2.65). Hispanic students purported the influence of spirituality in their lives ( M = 3.96, SD = 0.85) less than African American students bu t more than Mixed race students ( M = 3.81, SD = 0.74). According to post hoc testing, there was a statistically significant difference on mean scores reported by African American ( M = 4.34, SD = 0.68) and Latino students ( M = 3.96, SD = 0.85), p < .05. The effect size for this difference was 0.48, which is a medium effect size. The other statistically significant difference in mean scores was found between African American ( M = 4.34, SD = 0.68) and mixed race participants ( M = 3.81, SD = 0.74), p < .05. The effect size for the di fference in mean scores was 0.74 based on calculations using from the pooled standard deviation (Rosnow & Rosenthal, 1996). This is considered a medium to large effect size (see Table 8). Table 8 Characteristics of the Spirituality Measure (SSA ) by Ethnicity Scale Total ( N = 153) African American ( N = 42) Hispanic ( N = 85) Mixed race ( N = 26) Effect Size SSA M 4.04 4.33 3.96 3.81 SD 0.81 0.68 0.85 0.74 Skewness 1.07 1.57 0.92 0.46 Kurtosis 1.16 2.65 0.65 0.89 0.46* a 0.18 b 0.74* c Note. Scale ranging from 1 (None of the time) to 3 (Some of the time) to 5 (All of the time). a Effect size = ( African American (AA) Mean Hispanic (HS) Mean)/Pooled SD; b Effect size = (HS Mean mixed race (MR) Mean)/ Pooled SD; c Effect size = (AA Mean MR Mean)/Pooled SD Results of ANOVA indicated a statistically significant difference, p < .05.
87 The distribution of the scores on the SSA by gender indicated that the data were more normally distributed for gender than for ethn icity. Females purported slightly more influence of spirituality in their lives than males ( M = 4.09, SD = 0.85; M = 3.95, SD = 0.73, respectively), p > .05 (see Table 9). Table 9 Characteristics of the Spirituality Measure (SSA ) by Gender Scale Total ( N = 153) Male ( N = 51) Female ( N = 102) Effect Size SSA M 4.04 3.95 4.09 SD 0.81 0.73 0.85 Skewness 1.07 0.79 1.22 Kurtosis 1.16 0.91 0.47 0.17 Note. Scale ranging from 1 (None of the time) to 3 (Some of the time) to 5 (All of the time). Effect size = (Male Female)/Pooled SD. African American females espoused the highest mean sc ore for spirituality, while Mixed race females reported the lowest mean score for spirituality, p > .05. Notably, the data obtained from both African American and Mixed race females had a high kurtosis value ( kurtosis = 3.81, 3.04; respectively). African American and Hispanic males reported higher mean levels of spirituality compared to Mixed race males (see Table 10), p > .05. Table 2 in Appendix K provides information on the effect size differences among males and females of different ethnicities. The la rgest effect sizes were evident between African American females and all other persons who participated in the study. On the other hand, African American males did not statistically differ in terms of their mean scores when compared with all other types of participants in this study.
88 Table 10 Characteristics of the Spirituality Measure (SSA ) by Gender and Ethnicity Scale Total ( N = 153) M F African American ( N = 42) M F Hispanic ( N = 85) M F Mixed race ( N = 26) M F SSA M 3.95 4.09 3.99 4.53 3.95 3.96 3.85 3.80 SD 0.73 0.85 0.89 0.45 0.72 0.91 0.43 0.86 Skewness 0.79 1.22 0.79 1.64 0.90 0.93 0.38 1.32 Kurtosis 0.91 1.40 0.27 3.81 1.55 0.41 0.13 3.04 Note. Scale range from 1 (None of the time) to 5 (All of the time). M = Males. F = Females. Depression The next variable under review is the dependent variable, depression, as measured by the CDI S. The raw score is difficult to interpret without a clinical manual; therefore, the raw s cores were converted to T scores for the sake of conceptualization and understanding. Of the 153 participants, 5.6% of students met the established criterion on the CDI S to be deemed Clinically Significant for depression (i.e., T = 70 or higher). Also, 5.2% of students in the sample scored in the Borderline range for depression (i.e., T = 60 69). In light of this finding, females mean T scores ( M = 48.43, SD = 10.69) compared to males ( M = 45.18, SD = 6.19) and the total sample ( M = 47.35, SD = 9.54) were not statistically significant. See Table 11.
89 Table 11 Characteristics of the Depression Measure (CDI S ) by Gender Scale Total ( N = 153) Male ( N = 51) Female ( N = 102) Effect Size CDI S M 47.35 45.18 48.43 SD 9.54 6.19 10.69 Skewness 1.72 1.47 1.50 Kurtosis 2.94 1.43 1.87 0.34 Not e T score may range from 39 to 100+. Effect size = (Male Female)/Pooled SD. Hispanic participants reported the highest scores on the CDI S ( M = 49.25, SD = 10.72). The mean T scores for mixed race students ( M = 45.04, SD = 6.90) and African American students (M = 44.93, SD = 7.49) were below the mean for the total sample (see Table 12). The difference in depression T scores produced a statistically significant difference, F (2, 13826.64) = 3.95, p < .05. Post hoc testing provided evidence that the sta tistical difference may be attributed to T score differences between African American ( M = 44.93, SD = 7.49) and Hispanic students ( M = 49.25, SD = 10.72), p < .05. The effect size for this difference was small to medium according to Cohen (1988) ( d = 0.4 4).
90 Table 12 Characteristics of the Depression Measure (CDI S) b y Ethnicity Scale African American ( N = 42) Hispanic ( N = 85) Mixed race ( N = 26) Effect Size CDI S M 44.93 49.25 45.04 SD 7.49 10.72 6.90 Skewness 1.60 1.54 1.62 Kurtosis 1.56 2.05 2.43 0.44* a 0.42 b 0.02 c Note T score may range from 39 to 100+. a Effect size = ( African American (AA) Mean Hispanic (HS) Mean)/Pooled SD; b Effect size = (HS Mean mixed race (MR) Mean)/ Pooled SD; c Effect size = (AA Mean MR Mean)/Pooled SD. *Results of ANOVA indicated a statistically significant difference, p < .05. Hispanic females obtained the highest mean on the depression inventory ( M = 51.95, SD = 12.00), whereas mixed race males obtained the lowest mean score ( M = 41.88, SD = 3.23). While there were distinct differences, stati stical significance was not achieved. Except for Hispanic males, all the other sub groups evidenced some degree of non normality based on the skewness and kurtosis displayed in Table 13. The most notable values for non normal data were the large kurtosis v alues for mixed race males and African American males. Finally, Table 3 in Appendix K provides the effect sizes between all groups.
91 Relational victimiza tion The next measure under consideration is the SEQ, which measures various forms of aggression and prosocial behavior. Females reported more relational victimization than males ( d = .30) (see Table 14). However, these differences were not statistically significant. All students (except African American males) reported more relational victimization than physical and verbal victimization, p > .05. Females in all ethnic groups also reported more relational victimization than their male counterparts p > .05 Hispanic females reported the most relational victimization while mixed race males reported the least relational victimization, p > .05. The normality of the mixed race student data is a bit concerning, given that all the skewness and kurtosis values wer e above one and two. The non normality of the data for the mixed race student group was expected because it was the least homogeneous group compared to the other two ethnic samples. Table 13 Descriptive Statistics of the Depression Measure (CDI S) by Ethnicity and Gender Scale Total ( N = 153) M F African American ( N = 42) M F Hispanic ( N = 85) M F Mixed race ( N = 26) M F CDI S M 45.18 48.43 46.20 44.22 45.57 51.05 41.88 46.44 SD 6.19 10.69 7.19 7.69 6.13 12. 00 3.23 7.68 Skewness 1.47 1.50 1.78 1.71 0.44 1.26 1.82 1.27 Kurtosis 1.43 1.87 2.65 1.85 0.32 0.87 3.60 1.10 Note T score may range from 39 to 100+. M = M ales. F = F emales.
92 Table 14 Characteristics of the Victimization Measure (SEQ) by Gender Scale Total ( N = 153) Male ( N = 51) Female ( N = 102) Effect Size Relational M 2.35 2.17 2.45 SD 0.97 1.03 0.93 Skewness 0.75 1.17 0.59 Kurtosis 0.15 0.94 0.03 0.29 Total M 1.94 2.00 1.92 SD 1.03 1.09 1.00 Skewness 1.21 1.12 1.28 Kurtosis 0.89 0.57 1.18 0.08 Note. Means derived from scale that ranged from 1 (never) to 3 (sometimes) and 5 (all the time). Total refers to verbal and physical victimizati on (relational excluded). Effect size = (Male mean Female mean)/Pooled SD. There were a number of moderate to large effect sizes found for participants when ethnicity and gender were taken account together. Large differences were found between African American males and mixed race males ( d = 0.72), Hispanic males and mixed race males ( d = 0.75), and Hispanic females an d mixed race males ( d = 0.92). See Tables 3 and 4 in Appendix K for all effect sizes estimated on the SEQ by gender and ethnicity. Hope. The final measure that will be reviewed is the scale that assessed levels of hope in the participant sample (CHS). Th e mean score on the CHS for the total sample was 4.00 ( SD = 0.80). This means that students in this sample espoused relatively high levels of hopeful thinking and behaviors. The data were fairly normally distributed with a slight negative skew and kurtosis (see Table 15). The mean score for males in this sample
93 was slightly lower than the mean for the total sample ( M = 3.81, SD = 0.78), p > .05. The mean score for females, on the other hand, was slightly higher than for the total sample ( M = 4.10, SD = 0.79 ), p > .05. The distributions for males and females were again slightly non normal with negative skew and kurtosis values. There was a moderate difference between how males and females scored on the CHS ( d = 0.46). African American participants scored above the mean ( M = 4.10, SD = 0.71) and higher than Hispanic ( M = 3.98, SD = 0.84) and mixed race students ( M = 3.90, SD =0 .82) on the CHS, p > .05. The data obtained from mixed race students had a kurtosis value that was less than favorable ( kurtosis = 1.43) compared to the normality of the distribution of the other two ethnic groups (see Table 16). The largest difference wa s evident for Hispanic and African American students scores for hope ( d = 0.65), p > .05. Table 15 Characteristics of the Hope Me asure (CHS ) b y Gender Measure Total ( N = 153) Male ( N = 51) Female ( N = 102) Effect Size CHS M 4.00 3.81 4.10 SD 0.80 0.78 0.79 Skewness 0.66 0.13 0.96 Kurtosis 0.27 1.18 0.58 0.37 Note Score range from 1 (none of the time) to 3 (some of the time) to 5 (all the time). Effect size = (Mean males Mean females)/Pooled SD
94 Table 16 Characteristics of the Hope Measure (CHS ) b y Ethnicity Scale African American ( N = 42) Hispanic ( N = 85) Mixed race ( N = 26) Effect Size Hope M 4.10 3.98 3.90 SD 0.71 0.84 0.82 Skewness 0.46 0.78 0.30 Kurtosi s 0.92 0.10 1.43 0.15 a 0.10 b 0.27 c Note Score range from 1 (none of the time) to 3 (some of the time) to 5 (all the time). a Effect size = ( African American (AA) Mean Hispanic (HS) Mean)/Pooled SD; b Effect size = (HS Mean mixed race (MR) Mean)/ Pooled SD; c Effect si ze = (AA Mean MR Mean)/Pooled SD. Finally, African American females purported the highest levels of hope ( M = 4.35, SD = 0.52) and mixed race males experienced the lowe st levels of hope ( M = 3.56, SD = 0.62), p > .05. Yet, even though the mean score was lower for mixed race students, the levels of hope they purported were still experienced a lot of the time according to the scale on the CHS. For further information, se e Table 17. Overall, the data for all participant groups were relatively normally distributed except for some slightly inflated negative kurtosis values for African American males ( kurtosis = 1.14) and African American females ( kurtosis = 1.19). The effe ct sizes comparing gender and ethnicity are displayed in Table 6 in Appendix K.
95 Table 17 Characteristics of the Hope Measure (CHS ) b y Ethnicity and Gender CHS Total ( N = 153) M F African American ( N = 42) M F Hispanic ( N = 8 5) M F Mixed race ( N = 26) M F M 3.81 4.10 3.64 4.35 3.97 3.99 3.56 4.06 SD 0.78 0.79 0.78 0.52 0.81 0.86 0.62 0.87 Skew 0.13 0.96 0.43 0.44 0.64 0.86 0.75 0.76 Kurtosis 1.18 0.58 1.14 1.19 0.66 0.46 0.27 1.01 Note Means derived from Likert type scale that ranged from 1 (none of them time) to 3 (some of the time) and 5 (all the time). M = M ales. F = F emales. To help the reader understand how participants self reports in this study compare to other persons who have used these measures previously, Table 18 was developed. The data presented in Table 18 were chosen from studies that had similar participant demograph ics (or as close as possible) to this study. Data are only presented on the CASSS, CHS, and CDI S, and SEQ because the SSA has never been used before in other research studies. Table 18 Sum Scores and Standard Deviations Obtained in Current Study Compare d to Referred Sample Measure Study Cited Current Study Mean Current Standard Deviation Referred Mean/Sum Score Referred Standard Deviation CHS Valle, Huebner, and Suldo (2004) 22.68 a 24.67 b 4.66 a 4.68 b 28.89 a 28.26 b 5.70 5.47 CASSS Demaray and Ma lecki (2003) 228.02 39.30 240.45 52.73 CDI S Mather and Cartwright Hatton (2004) 47.34 9.57 53.34 12.53 SEQ Crick and Grotpeter (1996) 13.68 4.76 10.32 c 3.59 Note Gold standard is defined as a referred sample. a Males b Fem ales.
96 Research Question 1 R esults The first research question asked, How many students in this sample experienced relational victimization, according to the SEQ? Did more females or males report victimization? To what degree did this sample experience relational victimization (e.g ., different levels on the SEQ)? Of the students surveyed in this study, 24.84% reported that they experienced one or more of the types of relational victimization sometimes (i.e., answered positively to any question on the SEQ greater than 3). However this researcher wanted to know what type of relational victimization was most prevalent for students in this sample. Therefore, a frequency count also was completed for each item that was constructed to measure relational victimization (i.e., items 1, 2, 4, 5, 7). Thirty percent of students reported that they are left out on purpose when it is time to do an activity. Similarly, 38.8% of students reported that another student tries to get back at you by not allowing group membership. Higher percentages w ere found for students who reported that other students lie to make other kids not like them (i.e., 63.2%). Students also experienced other students making threats (34.9%) as a means of group exclusion. Finally, many students purported that other students at school say mean things in order to keep others from liking them (46.1%). Table 19 shows the percentage of students in the total sample and by race who reported their experience with various forms of victimization. Statistical differences did not exist for any of the items related to relational victimization, based on chi square analyses. The specific numerical results from the chi square analyses also are shown in Table 19.
97 The data suggested that females reported significantly more relational victimization (50.50%) compared to males (15.68%). While the differences are apparent, examining the degree to which males and females differed in term s of the intensity of their experiences provides a more sound representation of their experiences and can potentially be grounds for intervention development. Table 20 provides such evidence that females experienced some forms of relational victimization o n a more frequent basis than males. Specifically, females engaged in saying more mean statements compared to Table 19 Percentage of Sample Reporting Various Forms of Relational Victimiz ation Total ( N = 153) African American ( N = 42) Hispanic ( N = 85) Mixed race ( N = 26) ? SEQ Items % % % % 1 (Left out) a 30.3 26.2 29.8 38.5 1.17 2 (Group exclusion) b 38.8 35.7 42.9 30.8 1.45 4 (Lies) c 63.2 64.3 60.7 69.2 0.65 5 (Threats) d 34.9 31.0 36.9 34.6 0.44 7 (Mean statements) e 46.1 54.8 45.2 34.6 2.67 Note Sca le ranges from 1 to 6 for each item. Any item denoted as a 3 or above was included in the frequency count. a A re you left out on purpose when it is time to do an activity? ; b D oes a kid who is mad at you try to get back at you by not letting you be i n their group anymore? ; c H as another kid told lies about you to make other kids not like you or be at you? ; d D oes another kid say they wont like you unless you do what they want you to do? ; e D oes another kid try to keep others from liking you by s aying mean things about you?
98 males. Males and females alike experienced students lying and using exclusionary tactics (i.e., leaving person out of a group) against them to ruin relationships frequently (i.e., sometimes, almost all of the time all the time). This study provided evidence to suggest that the kind of relational victimization differed by gender (rather than assuming that males just do not engage in this kind of beha vior). Table 20 Percentage of Males and Females Who Reported Varying Degrees of V ictimization (N = 152 ) SEQ Items Never Almost Never Sometimes Almost All the Time All the time ? % % % % % 1 (Left out) a M F 39.2 46.5 23.5 26.7 31.4 17.8 0.00 5.9 5.9 3. 0 7.08 2 (Group exclusion) b M F 43.1 40.6 29.4 14.9 15.9 26.7 3.9 8.9 7.8 8.9 6.70 4 (Lies) c M F 27.5 19.8 19.6 11.9 33.3 28.7 7.8 15.8 11.8 23.8 6.64 5 (Threats) d M F 51.0 45.5 21.6 15.8 15.7 16.8 5.9 11.9 5.9 9.9 2.74 7 (Mean statements) e M F 39 .2 34.7 29.4 11.9 15.7 32.7 3.9 8.9 11.8 11.9 10.85* Note M = Males. F = Females. p < .05 a A re you left out on purpose when it is time to do an activity? ; b D oes a kid who is mad at you try to get back at you by not letting you be in their group an ymore? ; c H as another kid told lies about you to make other kids not like you or be at you? ; d D oes another kid say they wont like you unless you do what they want you to do? ; e D oes another kid try to keep others from liking you by saying mean thing s about you? Research Question 2 Results Research question 2 asked, Which coping strategies/mechanisms (i.e., spirituality, hope, perceived social support) are used and/or cited most frequently by minority youth in high risk midd le school students?
99 Since the scales were continuous (i.e., CHS, CASSS, SSA) but had slightly different scales (e.g., ranges from 1 5 and 1 6), the scores were transformed into three categorical variables: low, medium, and high. Note that for the CASSS, scores of five and six were combined for the sake of comparing each instrument on the same scale. The low group consisted of persons who scored a one or two on each scale. The medium group consisted of persons who scored a three. Finally, the high group c onsisted of persons who scored either a four or five. Table 21 displays the percentage of participants who cited any use of the aforementioned coping strategies. Table 21 Categorical Representation of Coping Strategies for All Participants (N = 152) Low Medium High Measure/Sub scale % % % Perceived social support Total Parent Teacher Classmates Best Friend 2.0 6.6 7.9 11.2 2.6 21.7 15.1 17.1 24.3 16.2 76.3 78.3 75.0 64.5 81.2 Hope 15.8 28.9 55.3 Spirituality 10.5 34.9 54.6 As the Table 21 suggests, best friends were perceived as pro viding the most support, compared to the other constructs measured by the CASSS. Parents and teachers closely followed (based on percentage) best friends in terms of how students perceived the level of support received from persons in their environment. F inally, participants
100 reported the classmates as providing the least amount of support. The reason for this may be related to the prevalence of bullying on campus. While mean differences among participants have been reported previously, a synthesis of the various coping mechanisms and gender and ethnic differences were funneled into a new table to help the reader view whether any differences among the three different races and their purported coping strategies were apparent. As Table 22 indicates, females i n all ethnic categories espoused higher levels of perceived social support; of which, African American females reported the most perceived social support, among all of the groups, p > .05. Interestingly, African American males, compared to all other racial groups and gender, reported the least amount of social support. The findings for the second coping strategy, spirituality, indicated that African American females again reported the highest levels of spirituality, followed by African American males, p > 05. The Hispanic and Mixed race groups reported relatively the same level of spirituality. Finally, internalized hope was found to be higher in all females across ethnic lines. However, African American females espoused more hope than persons in the mixed race and Hispanic groups. Table 22 Comparisons of Mean Scores for Coping Strategies by Gender and E thnicity African American ( N = 46) Hispanic ( N = 86) Mixed race ( N = 26) Male Female Male Female Male Female Perceived social support 4.58 5.02 4.59 4.76 4.56 4.72 Spirituality 3.99 4.53 3.95 3.96 3.85 3.80 Hope 3.64 4.35 3.97 3.99 3.56 4.02 Note Range for perceived social support was 1 6; range for hope and spirituality was 1 5.
101 While there were clearly differences in purported coping strategies that may have a clinical impact, a determination of statistical significance was also important to ascertain. Therefore, a multivariate analysis of variance (MANOVA) was conducted to determine whether there were any statistically significant int eractions involving gender and race for any of the coping strategies (see Table 23). Findings from the MANOVA indicated that there were no main effects for gender or race when considering all coping strategies. Moreover, there were no interaction effects b etween gender and race for any of the coping strategies. Table 23 MANOVA Summary Table for Coping Strategies (N = 153) Source Dependent Variable Df Type III SS Mean Square F P ? Gender (G ) Hope Spirituality Perceived social support 1 1 1 4.48 0.74 1. 70 4.48 0.74 1.70 7.31 1.20 2.46 .12 .28 .14 0.94, p <.01 Race (R ) Hope Spirituality Perceived social support 2 2 2 0.58 3.37 0.51 0.29 1.69 0.26 0.47 2.72 0.37 .62 .07 .69 0.95 R x G Hope Spirituality Perceived social support 2 2 2 3.38 2.10 0.51 1.69 1. 05 0.25 2.76 1.70 0.37 .07 .19 .69 0.94 E rror Hope Spirituality Perceived social support 147 147 147 90.07 90.95 101.53 0.61 0.62 0.69
102 Research Question 3 Results Research question 3 asked, Does victim status predict higher rates of depression among minority youth in a high risk middle school? Specifically, do higher scores on the SEQ predict higher depression scores on the CDI S? Prior to conducting a simple linear regression model (or the moderator models), the researcher wanted to see if t here was any relationship between the predictor, moderator, and outcome variables (see Table 24). According to the Pearsons product moment correlation analysis, perceived social support had a strong positive relationship with spirituality ( r = .69) and ho pe ( r = .66). Perceived social support also had a negative and moderate relationship with depression ( r = .44) depression and a negative and small relationship with victimization ( r = .17). Spirituality, on the other hand, had a stronger negative relatio nship to depression ( r = .46) and a strong positive relationship with hope ( r = .65). Depression had a moderate relationship with victimization ( r = .46) and negative and moderate relationship with hope ( r = .40). Table 24 Correlations Among Predictor Moderator, and Criterion Variables ( N = 152) Measure 1 2 3 4 5 1.Perceived social support 1 2. Spirituality .69** 1 3. Depression .44 ** .46** 1 4. Victimization .17** .09 .46** 1 5. Hope .66 ** .65** .40** .03 1 ** p < .01. Give n that a positive relationship was found between victimization and depression, it seemed reasonable to proceed to determine whether a significant predictive relationship existed. However, prior to proceeding with this analysis (or the subsequent
103 moderator analyses), the researcher wanted to ensure that the basic assumptions of multiple regression were met: (1) normal distribution of variables, (2) linear relationship between independent and dependent variables, (3) reliability of variables, and (4) equal er ror variance (e.g., homoskadesticity). While scores on the CDI S were slightly non normal due to larger than one skewness and kurtosis values, multiple regression is robust to variations of normality in the dependent variable. The second assumption, as cla rified by Table 24, indicated that a linear relationship existed between depression and victimization. Whil e the second assumption was met, the issue of highly related moderator variables posed a problem to the analysis and interpretation of the regression models. When multicollinearity is a factor in a regression model, the proportion of variance attributed to various predictor and/or moderator models may be inflated. Thus, one technique to compensate for this particular issue in the dataset is to center t he variables by subtracting the mean score from each individual score. Additionally, since the scales for the continuous variables were different (i.e., scales ranging from 1 5 and 1 6), the validity of using centering became a more plausible approach. Th us, this researcher centered all the variables and transformed the scores prior to running the simple and multiple regression models presented below. The output when running the centered model was comparable to running the regression analyses without cente ring. Therefore, the researcher decided to use the un centered model for ease of t ranslation of the results (Krom r e y & Foster Johnson, 1998). All variables achieved average and/or above average Cronbachs alpha coefficients, suggesting that the variables were reliable. Finally, the last assumption, normal and random distribution of the error variance was violated, suggesting that the
104 residuals were correlated. Specifically, the error variance for victimization and depression became increasingly larger as victimization and depression scores increased, such that upon observation of the data an upward linear trend was evident, rather than a random plot. This issue of homoskadesticity, or the correlation of error terms, has been dealt with in the literature by either taking the log of each score in a data set or taking the square root of each mean score. However, when the scores were transformed using the log function, the slight variations in normality became more profound. Thus, the data were not ultimately t ransformed and remained in their original state. The results from the regression analysis established a significant positive predictive relationship between relational victimization and depression F (1, 151) = 31.94, p < .0001 The variance in depression accounted for by relational victimization was 21.7%. Research Question 4 Results The fourth research question asked, Which protective factor(s) (i.e., spirituality, hope, perceived social support) served as a moderator(s) between victimization and dep ression among minority youth in a high risk middle school? Did this relationship differ based on gender, ethnic group (e.g., Hispanic versus African American), grade, and/or school? In order to determine if various moderator relationships existed, a mini mum of six new interaction term variables were constructed in SPSS for each moderator variable. Each moderator variable was analyzed from the bottom up, such that the most complex three way interaction was assessed for significance, before two way interact ions, and finally prior to determination of significance for main effects. Therefore, this section will
105 follow a format whereby an analysis of each variable and the cumulative influences of gender, race, and grade combined with each primary moderator varia ble will be displayed. Spirituality as a moderator variable Spirituality did not serve as a moderator variable between relational victimization and depression in this study. There were no significant two or three way interactions. However, there was a main effect for gender when victimization and spirituality values were held constant, F (3, 1967.37) = 36.76, p <.001, a main effect for ethnicity, F (4, 1519.59) = 50.79, p < .001, and a main effect for grade, F (4, 1490.88) = 27.88, p < .001. The first main effect alludes to the fact that males and females reported different depression scores, when the values for victimization and spirituality remained constant. The second main effect suggests that depression scores differed based on race. Finally, the third main effect supported that persons in different grades reported different scores on the depression inventory. Appendix M provide s a detailed outline of how spirituality was tested as a model. Perceived social support as a moderator variable A regression model with a three way interaction was not supported by this data such that gender, ethnicity, or grade combined with perceived social support did not significantly moderate the relationship between victimization and depression. However, a two way interact ion emerged from the data, F (1 1501.59) = 32.19, p < .001, supporting the statement that higher levels of perceived social support moderated the relationship between relational victimization and depression. For example, with low levels of hope, victimizat ion was related to higher levels of depression. As hope increased however, the relationship between victimization and depression was attenuated. The incremental R change value wa s small (? R = .02).
106 This finding (i.e., the two way interaction) also was still significant when race was held constant, F (1 1254.33) = 24.25, p < .001 and when grade was held constant, F (1 1477.28) = 27.44, p < .001 (see Appendix M). To promote easier interpretation of this moderator effect, see Figure 2. Based on the figure, support served to lower ones depression score across all categories of victimizations status (i.e., low, medium, high). However, a person who ex perienced the most benefit from perceived social support was one who experienced high levels of victimization and high levels of perceived social support. For example, the difference between someone who experienced high victimization and low perceived soci al support and someone who experienced high victimization and high social support was approximately 35 points on the depression inventory. Based on Figure 2, it appeared that the more victimization one experienced, the more perceived social support served as a protective factor. Figure 2 Perceived Social Support as a Moderator b etween Relational Victimization and Depression 0 20 40 60 80 100 120 Depression Score High Support Moderate Support Low Support High Medium Low Victimization Status
107 Hope as a moderator variable A three way interaction emerged for the variable hope such that the combination of gender and hope moderated the relationship between victimization and depression, F (1, 1967 .37) = 36.76, p < .001. The incremental R change was 2.8%. Given the complexity of this relationship two figures were developed to show how low levels, moderate levels, and high levels of purported hope varied by gender. Figure 3 exemplifies varying level s of hope espoused by females. Based on the data displayed in Figure 3, females with low hope experienced the highest levels of depression when victimized. As purported hope increased, depression scores decreased for females. The difference between scores for depression was over 50 points when hope was accounted for as a moderator variable, such that the scores significantly decreased as hope increased. 0 2 0 40 60 80 100 120 1 2 3 4 5 6 7 8 9 Victimization Category Depression Low Medium High Medium Hope High Hope Low Hope Figure 3. The Contribution of Hope as a Moderator for Depression in Females
108 Males also slightly reaped the benefi t of hope as a moderator variable between relational victimization and depression. However, when one compares Figure 3 and Figure 4, the benefit is not as profound when compared to females. This finding may be largely in part due to the restricted range of scores in which males reported on the depression inventory. When grade level was added as a potential combination moderator with hope, a two way interaction emerged from the regression model, F (1 1473.33) = 27.31, p < .001, such that when grade was he ld constant, hope served as a moderator between depression and relational vict imization. The R change was small (? R = .049) for this model. The same relationship held true when race was entered as a variable, F (1 1272.63) = 24.90, p < .001. The R change value was slightly less when race was held constant (? R = .042) rather than gender. Research Question 5 Results The fifth research question asked, How do students experience relational victimization? What are students perceptions as to why they are victimized? Are students able to verbally define how they cope with relationa l aggression? Who do 0 20 40 60 80 100 120 1 2 3 4 5 6 7 8 9 Victimization Category Depression Low Medium High Figure 4. The Combination of Hope as a Moderator for Depression in Males Low Hope Moderate Hope High Hope
109 victims specifically feel supported by, if anyone? How does the victimization impact their well being? The overall purpose of the qualitative interviews was to provide a thicker and richer understanding of the quantitative data. The o riginal intent of the qualitative questions was to elicit responses from students regarding their experience of relational victimization. However, by using a grounded theory method, this researcher found a number of poignant issues that were provided by pa rticipants that were not directly related to the original questions. Thus, this section has a two fold purpose: (1) to answer the research questions as posed in Chapter 1 and (2) to provide information that was gleaned during the interviews pertinent to mi ddle school students lives and the environments in which they take part. Student demographics. Prior to a discussion related to the qualitative questions proposed in Chapter 1, a brief discussion of the characteristics of the interviewees is important to provide context and insight into their answers. A total of eight individual interviews were completed. Four of these students were classified as experiencing high levels of depression ( T = 70 or higher) and high levels of victimization (in the top quartil e on the SEQ), and the other four students were classified as experiencing little to no depression ( T < 35) and high levels of victimization (in the top quartile on the SEQ). All but two of the students were sixth graders; additionally, all of the students who were interviewed in the high depression category were females and all the students who were interviewed in the low depression category were males. This differentiated gender split was not done purposefully by this researcher as students were chosen c ompletely based on their scores (the researcher did not look at gender or ethnicity when making lists of
110 potential interviewees). A detailed explanation for what may account for the compositional differences of the two groups is discussed in Chapter 5. Thr ee of the four females were Hispanic and one female was African American. Two of the males were Hispanic and two of the males were African American. Four of the students in this sample reported that they were bilingual. Two of the students moved to Tamp a from Puerto Rico, one student moved to Tampa from Maryland, and the other five students were born and raised in the Tampa Bay Area. Three of the students reported living in a single parent home, and seven of the students reported siblings living in the h ome. Six of the students shared that they were in honors and/or advanced classes. All but one student reported positive personality attributes including, I am respectful, I obey teachers, I am talented, and I treat other people how I would like to be treated. The one student in the sample that was overly negative about herself shared that she has been diagnosed with bipolar disorder and was on medication for the disorder. Perceptions of victimization Participants had a difficult time pinpointin g why other students personally picked on them. However, when the researcher asked the students to share in general why students were bullied, participants were able to extrapolate. The number one reason that students attributed bullying to was mean stud ents. Students interviewed described mean students as, and the mean ones just get an attitude for no reason just cause you say hi; well, its been kinda tough cause there are a lot of kids who just push around and act mean to you; and when theyre in a bad mood they take it out on other people. The next two most frequently cited reasons were
111 low self esteem and appearing to others as an easy target. For example, one female student reported, Well, they take it out on those that they think, that they see, like they see me, and they see that I probably have a low self esteem because at the beginning of the school year I dont know a lot of people so Im very quiet and I sit looking down. So they see that, and they say, oh, shes an easy target. Individual students reported the following as reasons why students victimize others: hatin, no reason, and peer influence. Additionally, students reported that they were picked on by others because of specific personal characteristics. For example one African American male was flexible, enjoyed dancing, and had the ability to do the splits. However, students continually made homosexual innuendos to this student and bullied him because of his natural dance abilities. The student shared, Theres s omesome people umsome people umthink that Im gay cuz I do aI can do a split. Theyumsome people um thats about it. They just keep callin mecallin me gay and thats all. Location and frequency of victimization While initially the qualitativ e research questions did not afford a discussion of the location and frequency of victimization, this researcher felt that it was important to provide information that students shared regarding these two topics. For the most part, students reported that v ictimization happened in all areas of this particular campus including the classroom, the gym, the hallways, behind the school, the bathrooms, the lunch line, and in the portables. The frequency to which students experienced victimization differed widely a mong the interviewees. Some students stated that bullying happened everyday (5 students); while other students stated that it happened three times per week (2 students) and two times per week (2 students).
112 For the most part, there was a degree of stabilit y in terms of how often students were victimized during the week. Moreover, there were no particular places that were considered hot spots for victimization to occur because it was occurring everywhere. Safety issues. Provided that victimization report edly occurred on this campus, understanding student perceptions of safety was imperative to obtain a comprehensive view of the context in which the students were trying to learn and socialize. Students appeared to have mixed emotions about the level of saf ety on campus --some students stated that they felt safe; however, after further discussion and probing all students stated that they did not feel safe. Thus, a degree of ambiguity existed among all participants related to perceptions of safety on and off campus. Six reports were made regarding feelings of safety on this campus, five reports were made related to feelings of fear and lack of safety, and three reports were too ambiguous to categorize. Students reported that they did not feel safe because of weapons brought to school ( I heard that once they brought a gun to school and that scared me when I came here. So I didnt feel safe.), bullies and gang members (Sometimes Im just afraid that theyll just come and bring something to hurt me [referring to a bully]; What makes me feel not so safe is bullies here and gang members and all that) and lockdowns ( Because like they have a lot of lockdowns, if stuff happens in the neighborhood they do, but it also makes me feel unsafe, because like why shou ld we have to go on lockdown?). One African American student also reported that she feels physically safe but not emotionally safe as a result of the bullies on campus (Well, I dont think, like someones gonna come in here with like a shotgun or somethi ng, but uh, but emotionally, I dont feel that good.).
113 On the other hand, students also reported feelings of safety on this campus because of lockdowns I feel pretty safe cause there are a lot of people who are, you know, like if somebody comes and th ey dont know, they go on lockdown and they lock all the doors and I feel really safe, older sibling support ( I know most everybody from my big brother. My big brother is in the eight, so, when he came, he was in the seventh when he came here, so he must have met a lot of people. And then when they meet him, when I walk by him as hes like this is my little brother, and everybody just starts talking to me.), the student resource officer (SRO) on campus, and the principal ( UhhI likecuz thethe princip al he cares and he umhe tries to get the bad people outta the school and send them to another school. ). Also, one student stated that, the teachers are always watching; however, this statement was not representative of any other students feelings who participated. Coping strategies A glimpse at Figure 5 below provides a graphical representation of all the coping strategies purported by students. The number one coping strategy that students reported was talking about and sharing the problem with som eone else. Some direct quotations from the interviews include: Ill talk to my friends and they just tell me to get over it. Itll be okay. Yeah, like my best friend Ive know since first grade, I talk to her about everything and we keep it all a secr et. My friends say that umwhy dont you just go talk to him and tell him to stop umtalkin bouttalkin about you and everything. I be like, man, they keep playing, and I aint gonna do nothing, you know, and R. be like just dont do it, just worry a bout it. The second coping strategy that was reported most frequently was relying on friends to make you feel better after an act of victimization occurred. For example, one Hispanic female interviewee stated, Well, my friends, they make me laugh and the y just try to
114 cheer me up. So it makes me feel happy. Another African American male student shared, If I like, get hurt, R. will come help me. And my friend J., he will come help me. And then well stop it and then we would start playing again. The final poignant quote made by an African American male related to sharing information about the victimization experience was, Uh, because itll keep my mind off of it or I can talk to a friend about whats happening and make some jokes, and that, and l augh, and thatll make me feel better. Because, um, if Im ever mad, I cant stay mad that long. The third most frequently reported coping strategy was the use of ignoring, as cited by six of the eight students. Students shared that they either placed their head on a table, pretended they did not hear the bullies, and/or actively ignored students by attending to school work. Two other coping strategies that were identified by more than two different individuals included taking out anger on another perso n (e.g., I take it out on other people when I get really mad, even though it bothers me, I still do it.) and spending time alone, away from everyone (e.g., and people will see that that means Im sad, and I will just be completely quiet for awhile.).
115 Figure 5 Coping strategies reported by participants CF:Coping Strategy Talk and share Friends Convince friends Confront bully Hit Alone time Ignore Take it out on others Religion was not reported by students initially as a coping tool or a protective factor against victimization. However, when students were directly asked about their faith and how their faith may ha ve altered the way they thought about bullying, many students resonated with their faith as a coping mechanism. Seven of the eight students reported some religious affiliation, all of which were a sect of Christianity. Some students reported attending chu rch on a weekly or bi weekly basis while others reported attendance once in awhile. Two students reported that while they were affiliated with a denomination, their religion and/or spirituality had no effect on their life. The other six students reported t hat religion and/or spirituality was beneficial to them in a number of ways including feeling better about oneself (e.g., Its good and we need to believe him because you feel better.), focusing on more positive life events ( e.g., It makes me be a bette r person just to take some bad things off of my mind. I think about that the
116 Godthat God had my back and just dont worry bout it Im gone get there.), and aiding in general wellness (e.g., It helps my insides.). Persons who support victims 4 Parent s upport was the most frequently cited support system (15) for students under distress from victimization. Interestingly, other family members were also noted to provide support including grandparents (2) and siblings (5). Other persons, outside of the famil y constellation, who were cited as support figures included friends (13), boyfriends (2), and guidance counselors (4). Notably, none of the students reported ever using the guidance counselor for support but stated that this person would be a viable option if necessary. For example, one female student stated, Oh, I never have talked to a guidance counselor, but if I get mad at, when I dont want to talk to my mom or my friend about it, I just go to, I will just go to a guidance counselor. Also, students never discussed teacher support unless prompted. For the most part, students did not believe that teachers or administrators would help them, as exemplified by the following: They [bullies] justthey just talk regularly. Its just that I dont think the teachers just dont pay attention to them. Perceptions of teachers. Two major themes that emerged from the data were positive and negative teacher attributes. While the central focus of this study was not related to student perceptions of teachers, this topic is an important one to explore because students often felt that they could not approach a teacher about bullying because of the poor relationships already imbued in the school setting. In general, students had poor relationships and/or negative feeli ngs about teachers when (a) teachers engaged in unfair practices, (b) teachers blamed students without hearing their side of the story, and 4 Number in parentheses refers to number of times mentioned in all eight interviews.
117 (c) teachers would not provide additional help on school work. Some poignant remarks made by students include: Som etimes I get mad at them so, I dont really want to talk to my teachers or anything. yelling and if you ask her for help, she gets mad. I s somebody talkin back there, she just blames it on like me and my friends sometimes. Because Mr. S. took my book bag so I cant do no work and if I dont do no work he, hes trying to give me an F. Everything, everything I do he get mad and be like, today when I had my hood, my hoodie on my shirt, head, he kept getting mad cause I had it on my head bu t we was outside walking to the cafeteria and he just always get mad at me The reader should be reminded that for the most part, students in this study felt supported by their teachers almost all of the time. Therefore, the question must be asked wheth er the students who participated in the qualitative interviews had different relationships with their teachers compared to the rest of the sample; it may be that, those persons who experienced victimization generalized that feeling or belief towards other relationships. Additionally, the quantitative analysis did not specifically ask students to think about the context of aggression and victimization during survey administration; whereas, the qualitative interviews revolved around this topic area directly. Therefore, students may have responded differently to the CASSS instrument if they were asked to think about situational factors that may change their perceived levels of support. As such, it is important to place the negative comments in perspective and c onsider that there are a variety of factors including instruments and personal bias that may account for the differences in the qualitative and quantitative findings.
118 The role of victimization on well being. Out of the eight interviews conducted, no stude nts stated that they were inoculated from the deleterious effects of bullying behavior. However, some students initially stated that they were not bothered by harassment; yet, within two to three minutes during the interview process, students changed their story and began to delve deeper into their emotions about experiencing victimization. Table 25 provides a list of all the different emotions experienced by students. Additionally, Table 25 provides a variety of direct quotations from participants to extra polate on the code. As evident by the data presented in Table 25, students experienced a wide array of emotions, all of which were negative and considered to affect mood. Students appeared to experience emotions to different degrees across a continuum o f mild to severe feelings from sadness to depression and from mad to sheer anger. The intensity of emotions experienced typically related to the degree to which bullying was experienced.
119 Table 25 Emotional Responses to Bullying B ehavior Code S upporting Quotations Angry just get really, really angry. He punched me in the cheek, and it really bothered me because I just didnt know. Annoyed Its gettin on my nerves. U hh, they keep, they keep coming up to me and telling me stuff about, stupid stuff and Ill be getting mad about that, I dont, I dont want to hear it. Cry I may cry sometimes, like why do these people misjudge me, or why, you know, do these people talk about me? I know Im not a bad person, I know Im very kind and careful about what I do, but it still affects and sometimes you just have to let people be people, you know Depressed Uh, like, uh sometimes I just feel empty inside, but I have like many, many problems, uh, that like, depression, and uh, soand I have manic depression and clinical depression because of the bullying and its uh, and when all of them act like that it doesnt help me. Empty Um, I feel very, um, empty inside. But, eventually, if they do it a lot over time, I will be depressed. But if its the first time, I will just be annoyed. Feel Bad Bad. Cause they dont need to say Im a sixth grade, because that dont matter. Frustrated Sometimes I get real frustrated, even with the medication and Ill yell back at them, but once they see how angry I am, they tend to quiet down a bit. Hide Feelings I try not to show it so I wont get in a fight. Hurts even though it hurts sometimes. That makes me fell, uh, pretty bad. And I have my own problems, and I dont really appreciate thatand it reall y gets me frustrated because I get so easily angered. And their, um and their insults actually hurt a lot a lot more, but um, luckily I do not have their, the classes with them, so I get to avoid them lots. Mad Ohh, they keep, they keep coming up to me and telling me stuff about, stupid stuff and Ill be getting mad about that, I dont, I dont want to hear it. No impact They call me big ears but I know that already so I dont really care. Outcast I just act like the outcast and just stay indepen dent for a while. Sad Sad. Cuz I never even doneI aint even say nothing to them and they just saying stuff about me and everything -just bringing up stuff for no reason. Upset Some of the girls here, they would prefer to mess with you than do their schoolwork. It doesnt. Im not gonna lie. Sometimes it does get to me because I am a child, and peers really affect you at this age. Worried I just worry about that when Im trying to do my schoolwork.
120 Chapter 5 Discussion Summary of Stud y The present study was conducted to explore relational victimization and depression in high risk, minority adolescents. This study was novel because it was one of the first studies to explore (a) the role of positive protective factors in moderating de pression, (b) how relational victimization was experienced by minority middle school students, and (c) how a mixed methods design could provide a deeper level of insight into relational aggression in this population. Findings showed that relational victimi zation was a statistically significant predictor of depression. Moreover, hope and perceived social support were found to moderate the relationship between victimization and depression. This chapter summarizes the results from Chapter 4, discusses implicat ions of the results, examines limitations, and suggests directions for future research. Incidence of Relational Victimization Relational victimization was defined in this study as an intentionally hostile act that could be either covert or overt and used to destroy friendships/relationships. Relational victimization was assessed via the most widely used measure available today (SEQ). The SEQ has been used in a variety of studies including research on relational victimization in the context of drug use (Sul livan, Farrell, & Kliewer, 2006) and life satisfaction (Martin & Huebner, 2007). The reported prevalence of relational victimization in this sample was approximately 25%. This percentage exceeded prevalence rates reported by students on other school campus es ( Kaltiala Heino et al., 2000; Nansel et al., 2001 ). Notably, although one other study found that mixed race
121 students experienced more victimization (Stein, Dukes, & Warren, 2007), the current study did not find statistically significant differences in r elational victimization across ethnic groups. This difference may have emerged because Stein and colleagues (2007) identified three distinct groups of victims (i.e., bully, victim, and bully victim), which the current study did not do. Although the over all mean score for relational victimization did not statistically differ by gender or ethnicity, the mean scores obtained in this study were consistent with Storch and colleagues (2003), who also found a somewhat elevated mean score for relational victimiz ation for minority students. Studies with other populations (e.g., rural Caucasian and African American students) of students found lower total mean scores. For example, Martin et al. (2007) found that the mean score for relational victimization was signif icantly lower than found in this study. Also, studies by Crick and Bigbee (1998) and Crick and Grotpeter (1996) found lower mean scores for victimization. Thus, one may infer that the participant demographics in this study, be it either ethnic composition (population included Hispanic and mixed race participants), low SES, or attending an urban school, may have attributed to higher reports of victim status. Frequency and intensity of relational victimization may be a more prominent aspect of the lives of mi nority youth living in urban environments than their majority peers in other types of environments. The type of relational victimization most widely cited by this sample included making mean statements, telling lies, and using overt group exclusionary tactics. However, the only statistically significant difference found was that females made more mean statements than males. What is quite interesting across the board is the fact that all
122 three groups, regardless of race or gender, stated that the most pr evalent form of relational victimization that occurred in school was lying. This quantitative finding was also supported by each qualitative interview such that each student stated that they had other students make up stories about them. Rationale for ly ing One reason that participants may have reported more lying is because lying is one of the most direct ways to ruin a persons relationships with others or defame a persons integrity. In schools, students find out that others are lying about them thro ugh rumors and gossip that can spread quickly. Moreover, if a person was not aware that they were a target of bullying, lying is a surefire way to let a person know. Clearly, there is no denial that someone is trying to hurt you when another person approac hes you with false accusations. Thus, if a student tries to stay out of the milieu of interfacing with aggressive students on campus, he/she is unable to do so when lies are spread. In essence, a person who is lied about is thrown into the mix of the aggre ssor victim relationship without choice. Therefore, bullies may use lying as the quickest and most direct approach at letting a victim know he/she is a target of aggression. The second reason lying may be used so often is because it almost always elicits a response from a middle school student. Since sense of self is not fully developed during middle school, lying may be a tactic that is more detrimental to emotional well being and social emotional development because identify formation is still taking pl ace (Erickson, 1963). As such, if a student does not have his/her ego strength intact at this point, the student will have no choice but to respond by crying, trying to save face, or becoming mutually aggressive. All of these behaviors are natural ways to deal with the cognitive dissonance of believing one thing about self and hearing another. Students of this age do
123 not have the skills to cope and/or know all the possible ways to deal with another persons lying because personal development is still growin g rapidly. Thus, the kind of response that students obtain from lying may be more noteworthy than other more covert behaviors. In essence, the aggressor is obtaining the desired reaction and may use this to his/her benefit to make the victim look weak or i nadequate. The third reason that lying may be the most prevalent form of relational aggression used is because students (the aggressors) feel as though they need to protect themselves from victimization. By taking hold of the reins and facing down, stu dents are likely to show others that they will spread rumors rather than be the target of them. A study conducted by Pugh Lilly, Neville, and Poulin (2001) found that African American females engaged in antisocial behaviors for the two reasons mentioned ab ove. More importantly, the antisocial behaviors that they engaged in were cited to be influenced by their perception of how supportive and/or hostile the environments around them were. Given that this sample is housed in a neighborhood that includes by gan gs, prostitution, drugs, and violence, many students may not feel supported. Thus, one of the major reasons why they may engage in more relational victimization than other groups of students is the environment in which they live and how the environment sup ports persons who exhibit a hard exterior via both physical and relational aggression. Relationship b etween Victimization and Depression Relational victimization and depression were found to be moderately correlated. Interestingly, when considering rel ational victimization and physical victimization combined (i.e., total victimization), a strong relationship existed with depression. Additionally, both forms of victimization were positively correlated with each other. It
124 appears from the data that perso ns who are victimized in multiple fashions may have a higher likelihood of becoming depressed. Juvonen and colleagues (2000) believed that students who are repeatedly victimized fall into a trap of self blame which contributes to psychological maladjustmen t. Consider a child who has rumors spread about him, has had boys punch him in the locker room, and is never invited to parties. This student would likely be more depressed because he has experienced multiple modalities of victimization. Unfortunately, he cannot compartmentalize one incident or one type of victimization to be a fluke experience because of the repeated victimizing incidents. Instead, this student blames himself, experiences maladjustment, and ultimately has difficulties in school. Aside f rom establishing of a moderate positive relationship between relational victimization and depression, this study supported previous findings that physical victimization (Cullerton Sen & Crick, 2005; Klomek, Marrocco, Kleinman, Schonfeld, & Gould, 2007) and relational victimization ( La Greca & Harrison, 2005; Storch & Esposito, 2003 ) are predictive of internalizing disorders. This study found a significant yet moderate predictive relationship between both total victimization and relational victimization with depression. Total victimization accounted for 25% of the total variance, and relational victimization accounted for 18% of the total variance. This finding supports the work of Prinstein and colleagues (2001), as well as the work of La Greca and Harrison (2005), who found that 21% of the total variance in depressive symptoms was accounted for by social functioning (i.e., victimization and social status).
125 Perpetrators of victims What is extremely important to understand about the victimization and the subsequent depression experienced by these students is that the victimization that was noted came from persons who were relatively unknown to the victim. Each student during the qualitative interview shared that, for the most part, the students who aggress ed upon them were not in their social circle and that they did not have a previous relationship with the student(s). This is intriguing given that this experience has effects as profound as developing internalizing symptomology based on behaviors emitted b y someone who is unknown. Victimization from the general peer group (and not just close friends or those in romantic relatio nships (Goldstein & Tisak, 2004 ) ) influences students experience of depressive symptomatology The fact that students unknown to b ullies are often the ones victimizing students (and ultimately contributing to depression), necessitates school based primary prevention programs that are evidenced based (Young & Raffaele Mendez, 2003) The issue, however, with implementing primary prevention programming is that the number of programs that target relational aggression and victimization are few in number, (Van Acker & Talbott, 1999), not empirically validated (e.g., Allies in Action, Owning Up, Creating a Safe So cial Climate in Our Schools), and fail to achieve a curriculum that can be implemented across settings (August, Lee, Bloomquist, Realmuto, & Hektner, 2003) Coping Mechanisms Perceived social support Both males and females reported fairly high perceived social support from parents, best friends, and teachers. Perceived social support also was found to moderate the relationship between relational victimization and depression. Perceived social support as a moderator variable for psychopathology has been found in
126 previous studies whereby perceived social support was negatively related to clinical maladjustment, and positively related to personal adjustment and school adjustment (Demaray, Malecki, Davidson, Hodgson, & Rebus, 2005). Notably, in this study, the percentage of students who reported high social support from classmates was significantly l ower than the support reported for other persons measured on the CASSS. The qualitative interviews provided additional information about this finding: the students who were interviewed noted that classmates were often unwilling to be kind and act in prosoc ial ways. Additionally, because the prevalence of bullying was so high on this campus, it makes sense that students felt the least support from classmates. Lack of classmate support has important implications for the learning environment and the developme nt of psychopathology. First, classmate support has been found to be a significant individual predictor of emotionality, including social stress, depression, self esteem, and anxiety (Demaray et al., 2005). Moreover, classmate support has been shown to hav e a preventive effect on running away from home among adolescents in Hong Kong (Cheung, Liu, & Lee, 2005). Finally, isolation and lack of support has been linked to violent school behavior (e.g., Columbine and Littleton school shootings) ( Jhally & Katz 19 99). As previous research suggests, classmate support is important to social adjustment for adolescents. While clearly not everyone will ever feel completely supported, an important question that still remains unknown to researchers and practitioners is h ow students could feel more supported in light of the aggressive behaviors exhibited on school campuses. In other words, what kinds of interventions would help facilitate cohesiveness among middle school students? One possible approach that has been sugges ted is peer based
127 group learning. While peer group learning seems to be a quick and easy answer because it has been proven to increase intrinsic motivation (Song & Grabowski, 2005) and improve academics and relationships in elementary school children (Calh oon, Al Otaiba, Greenberg, King, & Avalos, 2006), peer based learning may not necessarily enhance the feelings of support from classmates. For example, a final product may be developed and presented to the class; however, working together in a group on an academic task may do nothing to eradicate the bullying behavior that may not be noticeable to a teacher. Along the same lines, peers in middle school may feign liking one another in front of the teacher; however, behind this faade, a climate of hostility and bullying may be present. Forcing middle school adolescents to work together without considering a number of variables (e.g., gender differences, working styles, ethnic nuances) may in fact be more deleterious to a victim who may beget more harm and vic timization in a group setting. Further studies must address why students feel unsupported by the peer group and what tactics or programming would help to build support. Finally, what needs to still be considered is why the students in this particular stud y reported high levels of perceived social support yet engaged in bullying and/or were victims of relational aggression. The literature (Parcel & Dufur, 2001) suggests that parental monitoring and teacher support are directly linked to an individuals adhe rence to social norms. Yet, even students who reported very high levels of perceived social support, reported engagement in antisocial behaviors which would suggest that they would act in prosocial and healthy ways. Also noteworthy is the fact that student s who were victimized reported high mean scores for parent support. This finding also conflicts with a study conducted by Stevens et al. (2002) where the researchers found that victims
128 tended to purport a lack of parental support. The incongruent findings from the aforementioned studies and this research dissertation may be explored via a qualitative study to determine the nature of the adult supportive relationships and how adult role models impact antisocial or prosocial behaviors. Hope Hope, defined as holding positive beliefs about particular goals regardless of circumstances (negative or positive), is considered an emotion that provides one with motivation to act, to accomplish something, or to be inspired. Hope was found to be a significant moderat or for both males and females in this sample such that the higher levels of hope one espoused, the more attenuated the relationship was between relational victimization and depression. However, hope had a stronger influence on female depression scores comp ared to male depression scores (i.e., 35 points decrease versus 10 points decrease, respectively). The exact reason for the differential responses based on gender is unknown; however, a few hypotheses regarding this have been developed by this researcher. First, the base rate of depression w as low for males in this sample Thus, if a clinical population (in a residential setting, for example) was used, the variance in depression scores may have been higher. Consequently, the true effects of hope may have be en more pronounced and easier to detect in a sample that actually experiencing Clinically Significant depression. Second, males in the general population report and experience depressive symptoms in different ways than females ; f or example, male depres sion often resembles aggressive disorders and extreme reactions to psychosocial stressors ( McGrath, 2002 ; Mo ller Leinkuhler Heller, & Paulus, 2007). Third, the CDI S does not tap into non normative kinds of depressive behaviors; thus, the detection of dep ression for middle
129 school males may have not been as adequate as desired. An instrument that may be recommended for future research would be the Gotland Scale of Male Depression (Bech, 2001), which detects depression symptoms that are both atypical and typ ical. Finally, since the levels of hope in males were found to be less than females, males may in fact not experience the same internalization of hope as females to reap this constructs benefits. Males who have a history of experiencing poverty and disenf ranchisement with education have been known to experience less hope (Seaton, 2007). Putting aside the different effects hope had for males and females on depression, hope may have been such a strong moderator variable because of the nature of the constru ct. Persons with hope tend to have agentic and pathways thinking (Snyder, 1995). This kind of mentality and cognitive beliefs may make persons with higher levels of hope more resilient to some of the symptomatology associated with depression such as anhedo nia. If a person continues to have hope and believes in obtaining certain goals, they may not experience such extreme forms of depression. Additionally, hope is often linked to and/or augmented by faith and religion (Watts, Dutton, & Gulliford, 2006). Pa rticipants in this sample who espoused hope may have also consciously or unconsciously tapped into their religion or called upon divine intervention to provide support during the aftermath of victimization. This study found a moderate correlational relatio nship between hope and spirituality, making the hypothesis that faith may be linked to high levels of hope plausible. However, what still remains unclear is why spirituality did not serve as a moderator variable in this study if the constructs are, in fact interrelated.
130 Interestingly, students who participated in the qualitative interviews never verbalized any comments directly connected to hope (by itself as an intrapersonal factor) or hope in relation to victimization. The closest comments that suggeste d a student had a degree of hope were related to life goals and how to obtain them. When each of the surveys were matched with the qualitative interviews, four of the eight students interviewed reported high levels of hope on their surveys. Yet, those same students had a difficult time formulating how they would attain their goals when asked specifically about them. This may have occurred because pathways thinking may still be developing in early adolescence. Also, there is the potential that students did n ot answer the CHS honestly, thus explaining the incongruence between the survey and the qualitative interview. Spirituality Spirituality was not found to moderate the relationship between relational victimization and depression. This result was surprisin g given that spirituality has been a strong predictor, mediator, and moderator for persons of color for variables including of quality of life (Utsey, Bolden, & Williams 2007), partner violence (Mitchell, Hargrove, Collins, Thompson, Reddick, & Kaslow, 20 06), civic involvement (Smetana & Metzger, 2005), and suicidal ideations and attempts (W alker, Utsey, & Bolden, 2005). Nevertheless, students reported relatively high levels of spirituality. Specifically, African American females reported the most spirit uality compared to the other participants in this study. This finding is similar to other studies that report that persons of African American decent typically posses high levels of faith and spirituality ( Abernethy, Houston, Mimms, & Boyd Frankin, 2006; B oyd Franklin, 2003 ). What is noteworthy, however, is the fact that six of the eight students reported that their religion
131 helped them to cope with daily life hassles. Therefore, there is some indication that spirituality was in fact an important construct and contributor to resilience. There are a number of possible reasons why spirituality did not serve as a statistically significant moderator variable in this study. First, the measure used in this study (SSA) was never tested on a group of middle school students; instead it was adapted from the SIBS, a well known measure used for adults. Thus, there was no normative data to compare the results from this study to, nor was the instrument field tested to determine if it was tapping the construct of spiritua lity. Second, many students had a difficult time conceptualizing one of the items that used the word higher power. It is possible that students had difficulty with this item because they had not heard of this term, despite the fact that they may have had an understanding of the concept of a higher power. Participants also may have used other terms synonymous with this term that were not explicitly asked about in the study (e.g., God, Allah, Buddha). Finally, some students may have not attained full devel opment of the construct of spirituality. Fowler (1981) suggests that students move from a fantasy like understanding of religion and spirituality to a more concrete understanding. Furthermore, Fowler (1981) states that persons come to terms with the deepes t level of spiritual and religious awakening in late adolescence and early adulthood. Important Gender Implications The prevalence of depression in middle school females has been found to be higher than for males across numerous studies in the United St ates and abroad ( e.g., Mittendorfer Rutz, 2006; Rastad, Ulfberg, & Sjdn, 2006; Prinstein, Borelli, Cheah, Simon, & Aikins 2005 ). Prior to adolescence, rates of depression among boys and girls
132 are roughly equal. However, beginning at approximately age 1 4 and continuing into adulthood, females experience depressive disorders at twice the rate of males (Jacques & Mash, 2004; Koplewicz, 2003). As many as one out of four females will experience a depressive episode during her adolescent years (Jacques & Mas h, 2004). This study supported other studies such that more females than males experienced depression. Moreover, the range of depression experienced by males was restricted to T scores that remained in the Borderline Range, whereas scores for some females surpassed the criterion for Clinically Significant. The relationship between victimization and depression was stronger for females, suggesting that victimization comes at a higher cost for females rather than males. As Leadbeater, Boone, Sangster, and Mat hieson (2006) found, boys appear to be able to keep their self worth intact even though bantering takes place among them. Females, on the other hand, fail to do so and appear to internalize the relationally aggressive behavior as a reflection of characteri stics of self. Notably, Klomek et al. (2007) found that victimized males actually internalize emotions about victimization to some degree, as do females. This study found a direct link between the frequency of bullying and the higher likelihood of suicid al ideation and suicide attempts for males. On the other hand, females experienced issues related to suicidality when they had any type of victimization occur. Frequency was not a determinant in predicting suicidality for females in the study by Klomek and colleagues (2007). While the debate of whether more females or males experience victimization is important in its on right, this researcher believes what is even more critical to understand are the ramifications that each gender experiences post victimiza tion. The findings from
133 Klomek et al. (2007) and this study about participant responses and thoughts about their personal victimization have important implications for both policy and intervention. While most interventions for relational victimization are gender neutral, this study supports the notion that interventions may be more beneficial if they are designed to address the different experiences males and females have with victimization. For example, an intervention for females may focus on decreasing suicide risk by using techniques from the S.O.S. program (Aseltine & DeMartino, 2004). Additionally, female centered interventions may provide cognitive behavioral techniques to manage depression ( Spielmans, Pasek, & Mc F all, 2007). Moreover, the very large impact that hope had in moderating depression for females should be considered during intervention development. One way to build hope is to help students believe in the attainment of personal excellence. Stoeber and Rambow (2007) suggested that when stu dents strived for perfection, performance in school improved (vis vis increasing students levels of hope), depressive symptoms were decreased, and subjective well being was promoted. Researchers and practitioners should focus on activities that build ho pe rather than focusing solely on how to manage depression or victimization. These activities could be implemented at the individual, group, or systems level. The previous suggestion is aligned with the current positive behavior support (PBS) in schools, w hich is attempting to broaden its conceptualization of school wide interventions from solely behaviorally focused to encompassing positive psychology (i.e., hope, subjective well being), community psychology, and cultural psychology (Carr, 2007) tenets. O n the other hand, male interventions may focus on how to manage overt aggression as a response to relational victimization. Since males are often programmed to
134 believe that their strength during identity development is linked to violent behavior (Gilligan, Garbarino, Gilligan, & Thompson, 1999), interventions for males may focus on other ways masculinity could be developed and sustained. Also, male centered programming may help males to learn how to express their emotions about the victimization, rather tha n falling into the trap of either silence or rage. Many schools discuss a zero tolerance approach to deal with bullying. Some schools (the school in this study, too) have signs in the shape of a driving stop sign that state No Bullying Zone. While these signs have a presence that insinuates an administration that does not condone bullying, they are still a surface level attempt at curbing the behavior. Moreover, as students mentioned in their interviews, teachers either ignore the behavior or condone it as just a mere nuisance. This is not acceptable. Finally, when teachers interact aggressively with other teachers, they set an example for students that adults engage in the same behavior and there are no repercussions. The reality is that the signs in sch ools and the lack of care by teachers to stymie bullying are not curtailing the behavior especially since victimization is so prevalent on school campuses. A policy change needs to be instituted at the systems level given the established relationship betwe en relational victimization and depression. Change at any level takes time, energy, and buy in by all parties involved (Kratochwill & Stoiber, 2002); however, when a systems level change is implemented (rather than targeting an individual), the likelihood that change will actually occur and remain stable is higher (Curtis & Stollar, 2002). Teachers, parents, students, and administrators need to be aware of the kind of behavior that is occurring on campus and not turning a blind eye to it. This recommendati on is consistent with research conducted by Nesdale and Pickering (2006). Thus, if a policy
135 was instituted with all parties, feedback and support such that the entire ecosystem at the school understands the repercussions of victimization and why there are consequences, the chances of victimization occurring at such alarming rates may be lessened. This kind of policy execution is necessary given the findings that females experience extremely poor consequences of victimization. Limitations Sample size There are a number of important limitations that must be recognized about this study. First, the sample size was smaller than expected. With the small sample size, it was more difficult to detect statistically significant findings for the moderator model s. While perceived social support was found to serve as a moderator for relational victimization and depression, it is expected that grade and ethnicity would also have been moderator variables if the sample size was larger. Moreover, the effect sizes and the R values may have been larger if the s ample size was adequate. Possibly if a large enough sample size would have been obtained, the researcher may have been able to evaluate the cumulative effects of the moderator variables on depression. Future rese arch may attempt to discern whether hope and spirituality, for example, serve as dual pro tective factors and as such aid in the moderation of depression scores. Survey instruments The second major limitation of this study was the instruments used. While all of the instruments (except for the spirituality questionnaire) had excellent reliability and validity, they may have not been appropriate for this sample. For example, many of the students in the qualitative interviews reported support from a family me mber other than a parent (e.g., aunt, grandparent, sibling). Latino students specifically reported aide from grandmothers and siblings, which is consistent with the literature on who
136 provides much family support in Hispanic family structures (Goodman & Sil verstein, 2005). Thus, students may have completed an entire sub scale haphazardly because it was not relevant to them. The development of another edition of the CASSS may be important in light of the qualitative findings elicited from minority students. As mentioned previously, the CDI S may have not been the best instrument for inner city students who may experience and/or exhibit depressive symptoms in different ways. During the survey administration, many students reported that none of the answers typ ified how they felt. Thus, students were forced to pick an answer that may have not been truly representative of their experience. Furthermore, the SSA did not appear to resonate with students. While the reliability of the instrument was more than adequate there were quite a few questions about the meaning of the items that participants posed. Additionally, over half of the participants purported two different scores for the same question posed on two different items. Therefore, the actually validity of th e instrument remains in question. Again, this limitation may be directly linked to the lack of statistically significant findings for spirituality as a moderator variable. Finally, if time and permission were granted, the surveys should have been translat ed into Spanish. The number of Hispanic students who participated in this study was high, and many of them were English language learners. Therefore, while some of the students may have obtained proficiency in English, they may have resonated better with S panish versions of the surveys. Clearly, when translating any type of survey instrument other limitations arise due to cultural understanding and cultural perception of the items. However, the integrity of the data would have probably been better facilitat ed
137 if students were able to choose which language they preferred to complete the assessment instrument in. Member checking By the time the data analysis began, the students who participated in the study were dismissed from school for summer break. As su ch, there was no opportunity to conduct member checks with the individuals who participated in the interviews. Therefore, the interpretation of the interviews was solely on the basis of this researcher rather than meeting with each student to confirm that the analysis was correct. Furthermore, after reading the transcripts of each interview, the researcher had a number of follow up questions to elucidate more information pertinent to the study and to clarify some statements. However, this was not possible a nd should be conducted in future studies. Future Research This study has laid the groundwork for a number of subsequent studies that may be completed to further understand the role of relational victimization, depression, and protective factors that promo te resilience in middle school students. The first suggested research path that one may consider is to complete a qualitative study prior to a quantitative study with students in order to understand how they actually conceptualize relational victimization. While measures have been developed by researchers with their own interpretation of this specific kind of victimization, students seem to conceptualize it differently. Students appear to couple relational victimization with social and verbal victimization (per interviews with participants). Thus, the concept of different types of aggression (e.g., social, verbal, relational, physical) may be more of a theoretical issue rather than a reflection of how students view aggression. While the theoretical differenc es
138 in social, relational, and verbal aggression are important, applied researchers must be cognizant that students themselves cannot easily differentiate among these concepts. Assuming a qualitative study (e.g., focus groups, field study) is completed, mea sures may be able to be developed that are more aligned with students experiences. Second, future researchers may want to examine other possible coping mechanisms and/or moderator variables that were not evaluated in this study. For example, many of the students in this study were recent immigrants from Central and South American countries. Therefore, level of acculturation may play an important role in how students internalize others picking on them or the actual prevalence of bullying behavior (Yu, Huan g, Schwalbe rg, Overpeck & Kogan 2003) Some Hispanic students in this study may have reported more victimization because they were not used to this kind of behavior in their native countries. Moreover, Hispanic and African American students may in fact ha ve experienced the same degrees of victimization but held different mental models about the appropriateness of the behaviors. Third, other outcome variables are important to examine. It is implausible to think that victimization only impacts depression. As other researchers have shown, relational victimization increases stress, anxiety, and school maladjustment ( Leadbeater et al., 2006; Sullivan et al., 2006 ). Thus, this study may be replicated with the same moderator variables to determine how they might alter the relationship between relational victimization and other forms of pathology. When this is completed, more comprehensive interventions may be developed that would cover the gamut of possible resulting emotions, cognitions, or behaviors that occur post victimization.
139 Fourth, the role that school climate plays is essential to the understanding of relational victimization because of the prominence of this act on school campuses. Assuming a school has a tense and hostile environment; it is probable th at the cultivation of aggression is prominent among students. While school climate and bullying have been studied previously in relation to victimization ( Smith et al., 2000; Unnever & Cornell, 2004 ), school climate as a moderator variable has never been r eviewed; nor has school climate been thoroughly investigated with regard to relational victimization. Fifth, the gender of the perpetrator may be an important variable that alters the degree to which a victim experiences pathology. A recent study conducte d by Felix and McMahon (2006) found that when a male was a perpetrator of all forms of aggression, the victim experienced more extreme forms of pathology. This study did not consider the gender of the perpetrator and it would be interesting to determine if the significant findings for hope and perceived social support as moderator variables would be stronger when perpetrator characteristics were accounted for. The potential to create a structural equation model (SEM) may be appropriate for this kind of anal ysis given that there would be a number of latent variables that would compose perpetrator characteristics. Sixth, teacher related variables that either promote more bullying on campus and/or decrease the prevalence of relational victimization are also essential to research. While some researchers consider teacher variables as part of school climate, other researchers have developed an entire body of literature pertinent to teacher characteristics and student outcomes (c.f., Darling Hammond, 2004 ; Wentze l, 1997 ; Pianta & Nimetz, 1991; Vieno et al., 2005). This researcher believes that a study solely on teacher attributes and behaviors would be a provocative study to undertake to determine if
140 teachers may have biased accounts on what occurs in the classroo m and how differential expectations and biases ( Puig, Lambert, Rowan, Winfrey, & Lyubansky, et al., 2003 ) may interact and condone student harassment. Since students reports in this study indicated that teachers did in fact play a part in condoning the behavior in the classroom, there may be many underlying issues between the student and teacher relationship that should be explored. However, since only eight students r eported this, conclusions cannot be comfortably drawn about teachers influence on relational aggression and future studies are necessary. Seventh, because of the ever changing role of victim and aggressor it may have been likely that when the survey was administered some students were in the midst of being a target of aggression and thus feeling more distress. On the other hand, students may have been assessed when they were in the role of aggressor and as such, may have not reported such extreme emotions commensurate with depression. In order to compensate and control for timing of assessment as a confounding variable, a longitudinal analyses may be a more appropriate type of study to see how victimization and depression scores change over time. Final Tho ughts As this study found, relational victimization is not a phenomenon that is confined to middle class Caucasian communities and schools. Relational victimization is an act that perpetrates males and females across ethnic lines for mixed race, African American, and Hispanic students. Given that students who experienced more victimization also experienced more severe depression, it is clear that victimization has implications for students mental health outcomes. This study was the first to find that hop e and perceived
141 social support moderate this relationship. The next step must be to determine how these findings may be developed into a program or future research endeavors to make use of the important contributions of this study. Positive psychology vari ables should be interwoven and used in interventions and/or future research to further enhance the well being of middle school students. The information gleaned from this study is a stepping stone for making the lives of adolescents more fruitful and enjoy able during some of the most important years of their lives. Students should no longer live in fear at school, but enjoy the social and learning environment of the middle school experience.
142 References Abernethy, A.D. Houston, T.R. Mimms T ., & Boyd Franklin, N. (2006). Using prayer in psychotherapy: Applying Sue's differential to enhance culturally competent care. Cultural Diversity & Ethnic Minority Psychology, 12 101 114. Ahmed, E., & Braithwaite, V. (2004). Bullying and victimizati on: cause and concern for both families and schools. Social Psychology of Education, 7 35 54. Allport, G. W. (1961). Pattern and Growth in P ersonality New York: Holt, Rinehart and Winston Amlund Hagen, K., Myers, B.J., & Mackintosh, V.H. (2005). Ho pe, social support, and behavioral problems in at risk children. Ameri can Journal of Orthopsychiatry, 75 211 219. Archer, J. (2001). A strategic approach to aggression. Social Development, 10 267 271. Aseltine, R. H., & DeMartino, R. (2004). An outc ome evaluation of the SOS s uicide prevention program. American Journal of Public Health, 94 446 451. Atlas, R.S., & Pepper, D.J. (1998). Observations of bullying in the classroom. Journal of Educational Research, 92 86 99. August, G. J., Lee, S. S., Bloomquist, M. L., Realmuto, G. M., & Hektner, J. M. (2003). Dissemination of an evidence based prevention innovation for aggressive children living in culturally diverse, urban neighborhoods: the Early Risers effectiveness study. Prevention Science, 4 271 286. Baldry, A.C., & Farrington, D.P. (2005). Protective factors as moderators of risk factors in adolescence bullying. Social Psychology of Education, 8 263 284.
143 Baron, R. M., & Kenny, D. A. (1986). The moderator mediator variable distinct ion in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology, 51 1173 1182. Basic Behavioral Science Task Force of the National Advisory Mental Health Council (1996). Basic B ehavioral Science Research for Mental Health: Family processes and social networks. American Psychologist, 51 622 630. Batsche, G., & Porter, L. (2006). Bullying. In G. Bear & K. Minke (Eds.), Childrens Needs III: Development, Prevention, and Interve ntion (pp. 25 48). Bethesda: The National Association of School Psychology. Bech, P. (2001). Male depression: Stress and aggression as pathways to major depression. In A. Dawson and A. Tylee (Eds.), Depression Social and Economic Timebomb (pp.63 66 ). London: British Medical Journal Books. Benson, P. L., Scales, P. C., Sesma, A. Jr., & Roehlkepartain, E. C. (2005). Adolescent spirituality. In K. A. Moore & L. Lippman (Eds.). What do children need to flourish? Conceptualizing and measuring indicato rs of positive development (pp. 25 40). New York: Springer. Berkowitz, L. (1989). The frustration aggression hypothesis: Examination and reformulation. Psychological Bulletin, 106 59 73. Boney McCoy, S., & Finkelhor, D. (1995). Psychosocial sequel ae of violent victimization in a national youth sample. Journal of Consulting and Clinical Psychology, 63 726 736. Borg, M.G. (1998). The emotional reactions of school bullies and their victims. Educational Psychology, 18 433 444.
144 Boyd Franklin N. (2003). Race, class, and poverty In F. Walsh (Ed.), Normal family processes: Growing diversity and complexity (3rd ed.) (pp.260 279). New York: Guilford Press. Brody, G.H., Stoneman, Z., & Flor, D. (1996). Parental religiosity, family processes, and youth competence in rural, two parent African American families. Developmental Psychology, 32 696 706. Bufford, R.K., Paloutzian, R.F., & Ellison, C.W. (1991). Norms for the Spiritual Well Being Scale. Journal of Psychology and Theology, 19, 56 70 Calhoon, M.B., Al Otaiba, S., Greenberg, D., King, A., & Avalos, A. (2006). Improving reading skills in predominantly Hispanic Title 1 first grade classrooms: The promise of peer assisted learning strategies. Learning Disabilities Research & Pract ice, 21 261 272. Campbell, J. J., & Frabutt, J. M. (1999). Familial antecedents of children's overt and relational aggression. Paper presented at the Biennial Meeting of The Society for Research in Child Development, Albuquerque, NM. Camodeca, M. & Goo ssens, F.A. (2005). Aggression, social cognitions, anger, and sadness in bullies and victims. Journal of Child Psychology and Psychiatry, 46 186 197. Carr, E.G. (2007). The expanding vision of positive behavior support: Research perspectives on hap piness, helpfulness, and hopefulness. Journal of Positive Behavior Interventions, 9 3 14.
145 Casey Cannon, S., Hayward, C., & Gowen, K. (2001). Middle school girls reports of peer victimization: Concerns, consequences, and implications. Professional School Counseling, 5 138 147. Cheung, C.K., Liu, S.C., & Lee, T.Y. (2005). Parents, teachers, and peers and early adolescent runaway in Hong Kong. Adolescence, 40 403 424. Christiansen, E.J., & Evans, W.P. (2005). Adolescent victimization: Testing models of resiliency by gender. Journal of Early Adolescence, 25 298 316. Christle, C.A., Jolivette, K., & Nelson, C.M. (2005). Breaking the school to prison pipeline: Identifying school risk and protective factors for youth delinquency. Excepti onality, 13 69 88. Clarke, E.A., & Kiselica, M.S. (1997). A systematic counseling approach to the problem of bullying. Elementary School Guidance and Counseling, 31, 310 325. Cohen, J. (1988). Statistical power analysis for the behavioral sciences ( 2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates. Coie, J.D., & Dodge, K.A. (1983). Continuities and changes in childrens social statues: a five year longitudinal study. Merrill Palmer Quarterly, 29 261 282. Compas, B.E. (1997). Depression in children and adolescents, In E. Mash & L. Terdel (Eds.), Assessment of Childhood Disorders (pp.204 206). New York: The Guildford Press. Connor, D. F., Steingard, R. J., Anderson, J. J., & Melloni, R. H. (2003). Gender differences in reactive and pr oactive aggression. Child Psychiatry and Human Development, 33 279 294.
146 Craig, W. (1998). The relationship among bullying, victimization, depression, anxiety, and aggression in elementary school children. Personality Differences, 24 123 130. Cres well, J. W. (1994). Research Design: Qualitative and Quantitative Approaches. Thousand Oaks, CA: SAGE. Crick, N.R. (1997). Engagement in gender normative versus nonnormative forms of aggression: links to social psychological adjustment. Developmental Ps ychology, 33 610 617. Crick, N.R. (1996). The role of overt aggression, relational aggression, and prosocial behavior in the prediction of childrens future social adjustment. Child Development, 67 2317 2327. Crick, N.R., & Bigbee, M.A. (1998). Relational and overt forms of peer victimization: A multi informant approach. Journal of Consulting and Clinical Psychology, 66, 337 347. Crick, N. R., Casas, J. F., & Mosher, M. (1997). Relational and overt aggression in preschool. Developmental Ps ychology, 33 579 588. Crick, N. R., & Dodge, K. A. (1994). A review and reformulation of social information processing mechanisms in children's social adjustment. Psychological Bulletin, 115 74 101. Crick, N.R., & Grotpeter, J.K. (1996). Childrens t reatment by peers: victims of relational and overt aggression. Developmental Psychopathology, 8 367 380.
147 Crick, N.R., & Nelson, D.A. (2002). Relational and physical victimization within friendships: nobody told me thered be friends like these. Jo urnal of Abnormal Child Psychology, 30 599 607. Cullerton Sen, C., & Crick, N.R. (2005). Understanding the effects of physical and relational victimization: The utility of multiple perspectives in predicting social emotional adjustment. School Psyc hology Review, 34 147 160. Curtis, M.J. & Stollar, S.A. (2002). Best Practices in System Level Change. In A. Thomas and J. Grimes (Eds.), Best Practices in School Psychology, IV Bethesda, MD: NASP Publications. Darling Hammond, L. (2004). Inequalit y and the right to learn: Access to qualified teachers in Californias public schools. Teachers College Record, 106 1936 1966. Demaray, M. K. & Malecki, C. K. (2002 a ). Critical levels of perceived social support associated with student adjustment. S chool Psychology Quarterly, 17, 213 241. Demaray, M.K., & Malecki, C.K. (2002b). The relationship between perceives social support and maladjustment for students at risk. Psychology in the Schools, 39 305 316. Demaray, M.K., Makecki, C.K., Davidso n, L.M., Hodgson, K.K., & Rebus, P.J. (2005). The relationship between social support and school adjustment: A longitudinal analysis. Psychology in the Schools, 42 691 706. Dill, E.J., Vernberg, E.M., Fonagy, P., Twemlow, S.W., & Gamm, B.K. (2004). Neg ative affect in victimized children: The roles of social withdrawal, peer rejection, and attitudes toward bullying. Journal of Abnormal Child Psychology, 32 159 173.
148 Dwairy, M., Achoui, M., Abouserie, R., & Faraji, A. (2006). Parenting styles, indi viduation, and mental health of Arab adolescents. Journal of Cross Cultural Psychology, 37 262 272. Elias, M.J., & Zins, J.E. (2003). Bullying, other forms of peer harassment, and victimization is the schools: Issues for school psychology research an d practice. Journal of Applied School Psychology, 1 9 1 5. England, E.M., & Petro, K.D. (1998). Middle school students perceptions of peer groups: Relative judgments about group characteristics. Journal of Early Adolescence, 18 349 373. Erickson, E.H. (1963). Childhood and society (2 nd ed.). New York: Norton. Evans, W.P., Marte, R.M., Betts, S., & Silliman, B. (2001). Adolescent suicide risk and peer related violent behaviors and victimization. Journal of Interpersonal Violence, 16, 1330 1348 Farrington, D.P. (2005). Childhood origins of antisocial behavior. Clinical Psychology and Psychotherapy, 12 177 190. Felix, E. D. & McMahon, S. D. (2006). Gender and multiple forms of peer victimization: How do they influence adolescent psychosoci al adjustment? Violence and Victims 21 707 724 Fowler, J.W. (1981). Stages of faith: The psychology of h uman development and the quest for m eaning HarperCollins Publishers: New York, NY. Fox, C.L., & Boulton, M.J. (2005). The social skills proble ms of victims of bullying: Self, peer, and teacher perceptions. British Journal of Educational Psychology, 75 313 328.
149 Frazier, A.P. Tix, P.A., & Barron, E.K. (2004). Testing moderator and mediator effects in counseling psychology research. Journal of Counseling Psychology 51 115 34. Frith, H. (2004). The best of friends: The politics of girls friendships. Feminism & Psychology, 14 357 360. Garbarino, J. & DeLara, E. (2002) And Words Can Hurt Forever: How to Protect Adolescents from Bullyi ng, Harassment, and Emotional Violence N ew Y ork: Free Press. Gilligan, C., Garbarino, J., Gilligan, J., & Thompson, M. (1999). Boys to men: Questions of violence. Harvard Education Letter Retrieved on June 26, 2007, from http://www.edletter.org/past/ issues/1999 ja/forum.shtml#talk. Goldbaum, S., Craig, W.M., Pepler, D., & Connolly, J. (2003). Developmental trajectories of victimization: Identifying risk and protective factors. The Journal of Applied School Psychology, 19 139 156. Goldstein, S.E. & Tisak, M.E. (2004). Adolescents outcome expectancies about relational aggression within acquaintanceships, friendships, and dating relationships. Journal of Adolescence, 27 283 302. Goodman, C.C., & Silverstein, M. (2006). Grandmothers raising g randchildren: Ethnic and racial differences in well being among custodial and coparenting families. Journal of Family Issues, 27 1605 1626. Gottschalk, L.A., & Bechtel, R. (1995). Computerized measurement of the content analysis of natural language f or use in biomedical and neuropsychiatric research. Computer Methods and Programs in Biomedicine, 47, 123 130.
150 Graham, S., Bellmore, A.D., & Mize, J. (2006). Peer victimization, aggression, and their co occurrence in middle school: Pathways to adjustm ent problems. Journal of Abnormal Child Psychology, 34 363 378. Greco, L.A., Freeman, K.E., & Dufton, L. (2007). Overt and relational victimization among children with frequent abdominal pain: Links to social skills, academic functioning, and health service use. Journal of Pediatric Psychology, 32 319 329. Gumpel, T., & Kliewer, W. (2006 ). Peer victimization at school, post traumatic stress disorder symptoms, and adjustment in Israeli adolescents. Poster accepted for presentation at the Biennia l Meeting of the Society for Research on Adolescence, San Francisco, CA : March 2006. Hanish, L.D., Ryan, P., Martin, C.L., Fabes, R.A. (2005). The social context of young childrens peer victimization. Social Development, 14 1 19. Hatch, R.L., Berg, M.A., Naberhaus, D.S., & Hellmich, L.K. (1998). The Spiritual Involvement and Beliefs Scale: development of a new instrument. J ournal of Family Practice, 46 476 487. Hawker, D.J., & Boulton, M.J. (2000). Twenty years research on peer victimization and psychosocial maladjustment: A meta analytic review of cross sectional studies. Journal of Child Psychology and Psychiatry, 41 441 455. Hazler, R.J., & Mellin, E.A. (2004). The developmental origins and treatment needs of female adolescents with depression. Journal of Counseling and Development, 82 18 24.
151 Henington, C., Hughes, J. N., Cavell, T. A., & Thompson, B. (1998). The role of relational aggression in identifying aggressive boys and girls. Journal of School Psychology, 36 457 477 Hinshaw, S.P., & Lee, S.S. (2003). Conduct and oppositional defiant disorders. In E. Mash and R. Barkley (Eds.), Child Psychopathology, 2 nd edition, (p.144 198). New York: The Guildford Press. Hodge, D.R. (2004). Spirituality and people with mental illness: Developing spiritual competency in assessment and intervention. Families in Society: The Journal of Contemporary Social Services 85 36 44. Huebner, E.S., & McCullough, G. (2000). Correlates of school satisfaction among adolescents. Journa l of Educational Research, 93 331 335. Hunter, S.C., Boyle, J.M., & Warden, D. (2004). Help seeking amongst child and adolescent victims of peer aggression and bullying: The influence of school stage, gender, victimization, appraisal, and emotion. Br itish Journal of Educational Psychology, 74 375 390. Ingraham C.L., & Oka E. R. ( 2006 ). Multicultural iss ues in evidence based interven tions. Journal of Applied School Psychology, 22 127 149 Jackson, Y., & Warren, J.S. (2000). Appraisal, social support, and life events: Predicting outcome behavior in school age children. Child Development, 71 1441 1457. Jacques, H.K., & Mash, E.J. (2004). A test of the tripartite model of anxiety and depression in elementary and high school boys and girls. Journal of Abnormal Child Psychology, 32 13 25.
152 Jhally, S & Katz, J. (1999). Tough Guise Violence, Media & the Crisis in Masculinity [Video] Media Education Foundation. Johnston, L.D., O'Malley, P.M., & Bachman, J.G. (2000). Monitoring the future national survey results on drug use, 1975 1999. Volume I: Secondary school students NIH Publication No. 00 4802. Rockville, MD: National Institute on Drug Abuse. Jones, J.M. (2007). Exposure to chronic community violence. Journal of Black Community Psy chology, 33 125 149. Kaltiala Heino, R., Rimpela, M. Rantanen, P., & Rimpela, A. (2000). Bullying at school an indicator of adolescents at risk for mental disorders. Journal of Adolescence, 23 661 674. Kelle, U. (2001). Sociological explanations b etween micro and macro and method triangulation. In N. Fielding & M. Schreier (Eds.), Qualitative and quantitative Forschung: Ubereinstimmungen und Divergenzen [online]. Available: www.qualitative research.net/fqs/fqs/htm. King, P.E., & Boyatzis, C.J. (2004). Exploring adolescent spiritual and religious development: Current and future theoretical and empirical perspectives. Applied Development Science, 8 2 6. Klomek, A.B., Marrocco, F., Kleinman, M., Schonfeld, I.S., & Gould, M.S. (2007). Bullyi ng, depression, and suiciduality in adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 46 40 49.
153 Kochenderfer Ladd, B. & Ladd G.W. (2001). Variations in peer victimization: Relations to childrens maladjustment. In J. Juvonen & S. Graham (Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 25 48). New York: The Guildford Press. Kochenderfer Ladd, B. & Skinner, K. (2003). Childrens coping strategies: Moderators of the effects of peer victimization? Developmental Psychology, 38 267 278. Kokkinos, C.M., & Panayiotou, G. (2004). Predicting bullying and victimization among early adolescents: Associations with disruptive behavior disorders. Aggressive Behavior, 30 520 533. Koplewic z, H.S. (2002). More than moody: Recognizing and treating adolescent depression. New York: G.P. Putnams Sons. Kovacs, M. (1985). The Childrens Depression Inventory. Psychopharmacology Bulletin, 21 995 998. Kratochwill, T.R., & Stoiber, K.C. (2 002). Evidence based interventions in school psychology: conceptual foundations of the procedural and coding manual of division 16 and the society for the study of school psychology task force. School Psychology Quarter ly, 17 341 389. Ladd, G.W., & B urgess, K.B. (2001). Do relational risks and protective factors moderate the linkages between childhood aggression and early psychological and school adjustment? Child Development, 72 1579 1601. La Greca, A.M., & Harrison, H.M. (2005). Adolescent peer relations, friendships, and romantic relationships: Do they predict social anxiety and depression? Journal of Clinical Child and Adolescent Psychology, 34, 49 61.
154 Leadbeater, B.J. Boone, E.M. Sangster, N.A. & Mathieson, L.C. (2006). Sex differe nces in the personal costs and benefits of relational and physical aggression in high school Aggressive Behavior, 32 409 419. Ledbetter, M.F., Smith, L.A., Fischer, J.D., Vosler Hunter, W.L., & Chew, G.P. (1991). An evaluation of the construct vali dity of the Spiritual Well Being Scale: a confirmatory factor analytic approach. Journal of Psychological Theology, 19 94 102. Leff, S.S., Power, T.J., Manz, P.H., Costigan, T.E., & Nabors, L.A. (2001). School based aggression prevention programs for young children: current status and implications for violence prevention. School Psychology Review, 30 344 353. Levy, T.M., & Orlans, M. (1998). Attachment, trauma, and healing: Understanding and treating attachment disorder in children and famili es. Washington, D.C.: CWLA Press. Lincoln, Y.S., & Guba, E.G. (2000). Paradigmatic controversies, contradictions, and emerging confluences. In N.K. Denzin & Y.S. Lincoln (Eds.), Handbook of qualitative research (2 nd ed., pp.163 188). Thousand Oaks, C A: Sage. Loudin, J. L., Loukas, A., & Robinson, S. (2003). Relational aggression in college students: examining the roles of social anxiety and empathy. Aggressive Behavior, 29 430 439. Martin, K.M., & Huebner, E.S. (2007). Peer victimization and proso cial experiences and emotional well being of middle school students. Psychology in the Schools, 44 199 208. McGrath, E. (2002, July/August). Teen Depression Boys.
155 McIntosh, D., & Spilka, B. (1990). Religion and physical health: The role of persona l faith and control beliefs. Research in the Social Scientific Study of Religion, 2, 167 194. Merrell, K.W., Buchanan, R., & Tran, O. (2006). Relational aggression in children and adolescents: A review with implications for school settings. Psychology i n the Schools, 43 345 360. Merrell, K.W. (2001). Assessment of childrens social skills: Recent developments, best practices and new directions. Exceptionality, 9 3 18. Miranda, J., Bernal, G., Lau, A., Kohn, L., Hwang, W.C., & LaFromboise, T. (20 05). State of the science on psychosocial interventions for ethnic minorities. Annual Review of Clinical Psychology, 1 113 142. Mitchell, M.D., Hargrove, G.L., & Collins, M.H. (2006). Coping variables that mediate the relation between intimate part ner violence and mental health outcomes among low income African American women. Journal of Clinical Psychology, 62 1503 1520 Mittendorfer Rutz, E. (2006). Trends of youth suicide in Europe during the 1980s and 1990s -gender differences and implicati ons for prevention. Journal of Men's Health & Gender, 3 250 257. Mofidi M., DeVellis R.F., Blazer D.G., DeVellis B.M., Panter A.T., & Jordan J.M., Spirituality and depressive symptoms in a racially diverse US sample of community dwelling adults. Journ al of Nervous and Mental Disorders, 94, 975 977.
156 Moller Leimkuhler, A.M., Heller, J., & Paulus, N.C. (2007). Subjective well being and male depression in male adolescents. Journal of Affective Disorders 98, 65 72. Muthen, L.K., & Muthen, B.O. (1 998). MPlus Statistical Users Guide. Available at http://www.statmodel.com/download/usersguide/Mplus%20Users%20Guide%20 v 4.pdf Mynard, H. & Joseph, S. (1997). Bully/vic tim problems and their association with Eysencks personality dimensions in 8 to 13 year olds. British Journal of Educational Psychology, 67 51 54. Mystakidou, K., Tsilika, E., Parpa, E., Pathiaki, M., Patiraki, E., Galanos, A., & Vlahos, L. (2007) Exploring the relationships between depression, hopelessness, cognitive status, pain, and spirituality with advance cancer. Archives of Psychiatric Nursing, 21 150 161. Nangle, D.W., Erdley, C.E., Newman, J.E., Mason, C., & Carpenter, E.M. (2003). P opularity, friendship quantity, and friendship quality: Interactive influences on childrens loneliness and depression. Journal of Clinical Child and Adolescent Psychology, 32, 546 555. Nansel, T.R., Haynie, D.L., & Simons Morton, B.G. (2003). The asso ciation of bullying and victimization with middle school adjustment. Journal of Applied School Psychology, 19 45 61. Nansel, T.R., Overpeck, M., Pilla, R.S., Ruan, W.J., Simons Morton, B., & Scheidt, P. (2001). Bullying Behaviors Among US Youth: Pr evalence and Association With Psychosocial Adjustment, 285 2094 2100.
157 Nesdale, D., & Pickering, K. (2006). Teachers reactions to childrens aggression. Social Development, 15 109 127. Olafsen, R., & Viemero, V. (2000). Bully/victim problems and coping with stress in school among 10 to 12 year old pupils in Aland, Finland. Aggressive Behavior 26 57 65. Osterman, K., Bjorkqvist, K., Lagerspetz, K., M.J., Kaukiainen, A., Landau, S. F., Fraczek, A., et al. (1998). Cross cultural evidence of female indirect aggression. Aggressive Behavior, 24 1 8. Pakaslahti, L., Spoof, I., Peltola Asplund, R.L., & Jarvinen Keltikangas, L. (1998). Parents social problem solving strategies in families with aggressive and non aggressive girls. Aggressive Behavior, 24 37 51. Parcel, T.L., & Dufur, M.J. (2001). Capital at home and at school: Effects on child social adjustment. Journal of Marriage and Family Therapy, 63 32 47. Pargament, K.I. & Mahoney, A. 2002. Spirituality: The Discovery and Conser vation of the Sacred. C. R. Snyder & S. J. Lopez Eds., Handbook of Positive Psychology pp. 646 659. New York: Oxford University Press. Parker, J.G., & Asher, S.R. (1987). Peer relations and later personal adjustment: Are low accepted children at r isk? Psychological Bulletin, 102 357 389. Paul, J.J., & Cillessen, A.H.N. (2003). Dynamics of peer victimization in early adolescence: Results from a four year longitudinal study. Journal of Applied School Psychology, 19 25 43.
158 Peets, K., & Kika s, E. (2006). Aggressive strategies and victimization during adolescence: Grade and gender differences, and cross informant agreement. Aggressive Behavior 32, 68 79. Pellegrini, A.D., Bartini, M., & Brooks, F. (1999). School bullies, victims, and agg ressive victims: Factors relating to group affiliation and victimization in early adolescence. Journal of Educational Psychology, 91 216 224. Peskin, M.F., Tortolero, S.R., Markham, C.M., Addy, R.C., & Baumler, E.R. (2007). Bullying and victimizati on and internalizing symptoms among low income Black and Hispanic students. Journal of Adolescent Health, 40 372 375. Peterson, C., & Seligman, M.E.P. (2004). Character strengths and virtues: A handbook and classification Washington, D.C.: American Psychological Association and Oxford University Press. Phillipsen, L.C., Deptula, D.P., & Cohen, R. (1999). Relating characteristics of children and their friends to relational and overt aggression Child Study Journal, 29 269 289. Pianta, R.C., & Nimetz, S.L. (1991). Relationships between children and teachers: Associations with classroom and home behavior. Journal of Applied Developmental Psychology 12 379 393. Prinstein, M.J., Boergers, J., & Vernberg, E.M. (2001). Overt and relational aggre ssion in adolescents: Social psychological adjustment of aggressors and victims. Journal of Clinical Child Psychology, 30 479 491.
159 Prinstein, M.J. Borelli, J.L. Cheah, C.S.L. Simon, V.A. & Aikins, J. W. (2005). Adolescent girls' interpersonal vulnerability to depressive symptoms: A longitudinal examination of reassurance seeking and peer relationships. Journal of Abnormal Psychology, 114 676 688. Pugh Lilly, A. O., Neville, H. A., & Poulin, K. (2001). Defending ourselves: Black girls' pe rceptions of self reported delinquency. Psychology of Women Quarterly, 25 145 154 Puig, M., Lambert, M., Rowan, G., Winfrey, T., Lyubansky, M., Hannah, S. & Hill, M. (2003). Behavioral and emotional problems among Jamaican and African American childr en, ages 6 to 11: teacher reports versus direct observations. Journal of Emotional and Behavioral Disorders, 7 240 250. Rastad, C., Ulfberg, J., Sjdn, P. O. (2006). High prevalence of self reported depressive mood during the winter season among Swedi sh senior high school students. Journal of the American Academy of Child & Adolescent Psychiatry, 45 231 238. Rauste von Wright, M. (1989). Physical and verbal aggression in peer groups among Finnish adolescent boys and girls. International Journal o f Behavioral Development, 12 473 484. Ringel, J.S. & Sturm, R. (2001). National estimates of mental health u tilization and expenditures for c hildren in 1998. Journal of Behavioral Health Services Research 28 319 333.
160 Roberts, W.B., & Coursol, D. H. (1996). Strategies for intervention with childhood and adolescent victims of bullying, teasing, and intimidation in school settings. Elementary School Guidance and Counseling 30 204 212. Rodkin, P.C., & Hodges, E.V. (2003). Bullies and victims in th e peer ecology: for questions for psychologists and school professionals. School Psychology Review, 32 384 400. Roosa, M.A., & Gonzales, N.A. (2000). Minority issues in prevention: Introduction to the special issue. American Journal of Community Psyc hology, 28 145 148. Rosnow, R. L., & Rosenthal, R. (1996). Computing contrasts, effect sizes, and counternulls on other people's published data: General procedures for research consumers. Psychological Methods, 1, 331 340. Rostosky, S.S., Danner, F ., & Riggle, E.D.B. (2007). Is religiosity a protective factor against substance use in emerging adulthood? Only if you're straight! Journal of Adolescent Health 40 440 447. Salmivalli, C., Lagerspetz, K., Bjorkqvist, K., Osterman, K., & Kaukiainen, A (1996). Bullying as a group process: Participant roles and their relations to social status within the group. Aggressive Behavior, 22 1 15. Schreck, C.J., Miller, J.M., & Gibson, G.L. (2003) Trouble in the school yard: A study of the risk factors of victimization oat school. Crime and Delinquency, 49 460 484. Seaton, E. (2007). Exposing the invisible: Unraveling the roots of rural boys violence in schools. Journal of Adolescent Research, 23 211 218.
161 Seals, D., & Young, J. (2003). Bullyin g and victimization: Prevalence and relationship to gender, grade level, ethnicity, self esteem, and depression. Adolescence, 38 735 747. Seligman, M.E.P. (2006). Award for distinguished scientific contribution. American Psychologist, 61 772 788 Seligman, M.E.P. (2002). Authentic happiness: Using the new positive psychology to realize your potential for lasting fulfillment New York, NY: Free Press. Shields, A. & Ciccetti, D. (2001). Parental maltreatment and emotion dysregulation as risk fa ctors for bullying and victimization in middle childhood. Journal of Clinical Child Psychology, 30 349 363. Smetana, J. G., Campione Barr, N., & Daddis, C. (2004). Developmental and longitudinal antecedents of family decision making: Defining health behavioral autonomy for African American adolescents. Child Development 75 1418 1434. Smetana, J.G., & Metzger, A. (2005). Family and religious antecedents of civic involvement in middle class African American late adolescents. Journal of Research o n Adolescence, 15 325 352 Smith, C.A., & Farrington, D.P. (2004). Continuities in antisocial behavior and parenting across three generations. Journal of Child Psychology and Psychiatry, 45 230 247. Smith, P.K., & Shu. (2000). What good schools can do about bullying: findings from a survey in English schools after a decade of research and action. Childhood, 7 193 212.
162 Smucker, M. R., Craighead, W. E., Craighead, L. W., & Green, B. J. (1986). Normative and reliability data for the Children's De pression Inventory. Journal of Abnormal Child Psychology, 14 25 39. Snowden, L.R., & Yamada, A.M. (2005). Cultural differences in access to care. Annual Review of Clinical Psychology, 1 143 166. Snyder, C.R., Hoza, B., Pelham, W.E., Rapoff, M., War e, L., Danovsky, M., Highberger, L., Rubinstein, H., & Stahl, K.J. (1997). The development and validation of the Childrens Hope Scale. Journal of Pediatric Psychology, 22 399 421. Snyder, C. R., Rand, K. L., & Sigmon, D. R. (2002). Hope theory: A m ember of the positive psychology family. In C. R. Snyder & S. Lopez (Eds.), Handbook of positive psychology (pp. 257 276). New York: Oxford University Press. Snyder, C.R. (1995). Conceptualizing, measuring, and nurturing hope. Journal of Counseling an d Development, 73 355 360. Snyder, C.R. (2000). The past and future of hope. Journal of Social and Clinical Psychology, 19 11 28. Song, H.D., & Grabowski, B.L. (2005). Stimulating intrinsic motivation for problem solving using goal oriented conte xts and p eer g roup composition. Educational Technology Research and Development, 54, 445 466. Spielmans, G.I. Pasek, L.F. & Mcfall, J.P. (2007). What are the active ingredients in cognitive and behavioral psychotherapy for anxious and depressed childr en? A meta analytic review. Clinical Psychology Review, 27 642 654.
163 Stein, J.A., Dukes, R.L., & Warren, J.L. (2007). Adolescent male bullies, victims, and bully victims: A comparison of psychosocial and behavioral characteristics. Journal of Pediat ric Psychology, 32, 273 282. Stevens, V., De Bourdeaudhuij, I., & Van Oost, P. (2002). Relationships of the family environment to childrens involvement in the bully/victim problems at school. Journal of Youth and Adolescence, 31 419 428. Stewart, C.J., & Cash, W.B., Jr. (1997). Interviewing: Principles and practices (8 th ed.). Madison, WI: Brown & Benchmark. Stoeber, J., & Rambow, A. (2007). Perfectionism in adolescent school students: Relations with motivation, achievement, and well being. Perso nality and Individual Differences, 42, 1379 1389. Storch, E.A., Crisp, H., Roberti, J.W., Bagner, D.M., & Masia Warner, C. (2005). Psychometric evaluation of the Social Experience Questionnaire in adolescents: Descriptive data, reliability, and factor ial validity. Child Psychiatry and Human Development, 36 167 176. Storch, E.A., & Esposito, L.E. (2003). Peer victimization and posttraumatic stress among children. Child Study Journal, 33 91 98. Storch, E.A., Masia Warner, C., Crisp, H., & Klein, R.G. (2005). Peer victimization and social anxiety in adolescence: A prospective study. Aggressive Behavior, 31 437 452. Storch, E.A., Nock, M.K., Masia Warner, C., & Barlas, M.E. (2003). Peer victimization and social psychological adjustment in Hi spanic and African American children. Journal of Child and Family Studies, 12 439 452.
164 Suldo, S. M., Hangauer, J., Witte, R. B., Mihalas, S. T., Popkave, K., Powell, H. A., & Hardesty, R. B. (2006, March). Relationship between school climate and s ubstance use in early adolescents. Paper accepted for presentation at the Annual Conference of the National Association of School Psychologists, Anaheim, CA. Sullivan, T.N., Farrell, A.D., & Kliewer, W. (2006). Peer victimization in early adolescence: A ssociation between physical and relational victimization and drug use, aggression, and delinquent behaviors among urban middle school students. Development and Psychopathology, 18, 119 137. Tapper, K., & Boulton, M. (2000). Social representations of p hysical, verbal, and indirect aggression in children: sex and age differences. Aggressive Behavior, 26 442 454. Tashakkori, A., & Teddle, C. (2003). Handbook of Mixed Methods in Social and Behavioral Research Thousand Oaks: Sage Publications. Taub, J. (2002). Evaluation of the Second Step violence prevention program at a rural elementary school. School Psychology Review, 31 186 201. Toblin, R.L., Schwartz, D., Gorman, A.H., & Abou ezzeddine, T. (2005). Social cognitive and behavioral attribute s of aggressive victims of bullying. Applied Development Psychology, 26 329 346. Tomada, G., & Schneider, B.H. (1997). Relational aggression, gender, and peer acceptance: invariance across culture, stability over time and concordance among informant s. Developmental Psychology, 33 601 609.
165 Tolan, P. H., & Dodge, K. A. (2005). Childrens mental health as a primary care and concern: A system for comprehensive support and service. American Psychologist, 60 601 614. Underwood, M.K. (2004). Glare s of contempt, eye rolls of disgust and turning away to exclude: Non verbal forms of social aggression among girls. Feminism & Psychology, 14 371 375. Underwood, M.K., Galen, B.R., Paquette, J.A. (2001). Top ten challenges for understanding gender and aggression in children: Why cant we all just get along? Social Development 10 248 266. Unnever, J., & Cornell, D. (2004). Middle school victims of bullying: Who reports being bullied? Aggressive Behavior, 30 373 388. Utsey, S.O., Bolden, M.A. & Williams, O. (2007). Spiritual well being as a mediator of the relation between culture specific coping and quality of life in a community sample of African Americans. Journal of Cross Cultural Psychology, 38 123 136. Valle, M. F., Huebner, E. S., & Suldo, S. M. (2004). Further validation of the Childrens Hope Scale Journal of Psyc hoeducational Assessment, 22 320 337. Van Acker, R., & Talbott, E. (1999). The school context and risk for aggression: Implications for school based prevention and intervention efforts. Preventing School Failure, 44 12 28.
166 Veenstra, R., Lindenberg, S., Oldehinkel, A.J., De Winter, A.F., Verhulst, F.C., & Ormel, J. (2005). Bullying and victimization in elementary schools: A comparison of bullies, victims, b ully/victims, and uninvolved preadolescents. Developmental Psychology, 41 672 682. Vieno, A., Perkins, D.D., Smith, T.M., & Santinello, M. (2005). Democratic school climate and sense of community in school: A multilevel analysis. American Journal of Community Psychology, 36 327 341. Verkuyten, M., & Thijs, J. (2002). School satisfaction of elementary school children: The role of performance, peer relations, ethnicity, and gender. Social Indicators Research, 59 203 228. Walker, A.R., & Cilles sen, A.H.N. (2006). Physical and relational victimization in adolescence: A longitudinal examination of sex differences and behavior correlates. Poster accepted for presentation at the Biennial Meeting of the Society for Research on Adolescence, San F rancisco, CA : March 2006. Walker, R.L., Utsey, S.O., & Bolden, M.A. (2005). Do sociocultural factors predict suicidality among persons of African decent living in the U.S.? Archives of Suicide Research, 9 203 217. Walsh, F. (1999). Religion and spirit uality: Wellsprings for healing and resilience. In F. Walsh (Ed.), Spiritual resources in family therapy (pp. 3 27). New York: Guildford Press. Watts, F., Dutton, K., & Gulliford, L. (2006). Human spiritual qualities: Integrating psychology and reli gion. Mental Health, Religion, and Culture, 9 277 289.
167 Wentzel, K.R. (1997). Student motivation in middle school: The role of perceived pedagogical caring. Journal of Educational Psychology, 89 411 419. Werner, N.E., & Crick, N.R. (1999). Relational aggression and social psychological adjustment in a college sample. Journal of Abnormal Psychology, 108 615 623. Werner, N.E., & Nixon, C.L. (2005). Normative beliefs about relational aggression: An investigation of the cognitive bases of adolescent aggressive behavior. Journal of Youth and Adolescence, 54 229 243. Wong, E.C., Beutler, L.E., & Zane, N.W. (2007). Using mediators and moderators to test assumptions underlying culturally sensitive therapies: An exploratory example. Culturally Diverse Ethnic Minority Psychology, 13 169 177. Yoon, J.S., Hughes, J.N., Cavell, T.A., & Thompson, B. (2000). Social cognitive differences between aggressive rejected and aggressive non rejected children. Journal of School Psychology, 38 551 570. Young E.L., Boye, A.E., & Nelson, D.A. (2006). Relational aggression. Understanding, identifying, and responding in schools. Psychology in the Schools, 43 297 312. Young, E. L., & Raffaele Mendez, L. M. (2003). The mental health professional's role in u nderstanding, preventing, and responding to student sexual harassment. Journal of Applied School Psychology, 19 7 23. Yu, S. M., Huang, Z. J., & Schwalberg, R. H. (2003). Acculturation and the health and well being of U.S. immigrant adolescents. Journa l of Adolescent Health, 33 479 488.
168 Zalecki, C.A., & Hinshaw, S.P. (2004). Overt and relational aggression in girls with attention deficit hyperactivity disorder. Journal of Clinical Child and Adolescent Psychology, 33 125 137. Zimmerman, M.A., & Arunkumar, R. (1994). Resiliency research: Implications for schools and policy. Society for Research in Child Development, 8 1 19.
170 Appendix A: Parent Active C onsent Form Dear Parent or Caregiver: This letter provides information about a research study that will be conducted at XX Middle School by Stephanie Mihalas, M.A., an advanced doctoral candidate from the University of South Florida. The goal of this st udy is to determine the effect of students interpersonal relationships on their social and emotional wellness ; ultimately culminating in the development school based interventions to improve adolescents well being. Who Is Conducting This Study : Steph anie Mihalas, M.A., an advanced doctoral candidate at the University of South Florida (USF) in her fourth year of coursework. The study and its planning is in cooperation with the principal and administrators of XX Middle School to ensure that the study p rovides information that will be helpful to this school. Why Your Child Is Being Requested To Participate : This study is being conducted as part of dissertation research entitled, The Impact of Interpersonal Relationships on Well Being. Your child is being asked to participate because he or she is a student at XX Middle School. It is expected that approximately 600 students will participate, as the entire school is being asked to do so. Why Your Child Should Participate : The purpose of this study is to learn more about what leads to happiness and health during the teenage years! The information that is collected from students may help increase the overall knowledge of risk and protective factors that lead to social and emotional wellness during m iddle school. In addition, information from the study will be shared with the principal at XX in order to increase his knowledge of what students consider to be the strengths and weaknesses of their schooling and other life experiences. Information from t his study may provide a foundation from which to improve the schooling experiences and well being of students at XX. Please note all students who participate will be entered into a drawing for one of several gift certificates ($25 each). What Participat ion Requires : If your child is given permission to participate in the study, he or she will be asked to complete several paper and pencil questionnaires. These surveys will ask about your childs thoughts, behaviors, and attitudes towards school, teache rs, classmates, family, and life in general. Completion is expected to take your child between 30 and 60 minutes. Stephanie Mihalas, M.A. will personally administer the questionnaires at XX, during regular school hours (specifically during an elective co urse), to small groups of students who have parent permission to participate. Participation will occur during one class period, one time during this school year. Participation of the surveys may be terminated if any student(s) behave rowdy or do not fo llow the directions read aloud by the researchers. In addition to completing surveys, your child may be asked to participate in one brief (30 45 minutes) individual interview with Stephanie Mihalas, M.A. The individual interview will occur during regu lar school hours (during an elective period), and will consist of asking your son or daughter more information about how he/she answered the survey. The interview will be tape recorded for transcription purposes. If your child is chosen to participate in t his part of the study, he/she will receive a $10 gift certificate. Please Note : Your decision to allow your child to participate in this research study must be completely voluntary. You are free to allow your child to participate in this research study or to withdraw him or her at any time. If you choose not to participate, or if you withdraw at any point during the study, this will in no way affect your relationship with XX, USF, or any other party. Confidentiality of Your Childs Responses : There is minimal risk to your child for participating in this research. Stephanie Mihalas and USF research assistants will be present during administration of the questionnaires in order to provide assistance to your child if he or she has any questions or conc erns. Additionally, school guidance counselors will be available to students in the unlikely event that your child becomes emotionally distressed while completing the measures. Your childs privacy and research records
171 Appendix A (continued) will be ke pt confidential to the extent of the law. Authorized research personnel, employees of the Department of Health and Human Services, and the USF Institutional Review Board and its staff may inspect the records from this research project, but your childs in dividual responses will not be shared with school system personnel or anyone other than us and our research assistants. Access to your own childs responses will not be made available to protect your childs confidentiality and ensure honest responding. Yo ur childs completed questionnaires will be assigned a code number to protect the confidentiality of his or her responses. Only Stephanie Mihalas will have access to the locked file cabinet stored in a locked office that will contain: 1) all records linki ng code numbers to participants names, and 2) all information gathered from school records. Please note that although your childs specific responses on the questionnaires will not be shared with school staff, if your child indicates that he or she inten ds to harm him or herself and/or scores high on the depression inventory, we will contact district mental health counselors to ensure your childs safety. What Your Childs Responses May be Used For : The information from this study to inform educato rs and psychologists about the effects of various kinds of interpersonal relationships on students well being. The results of this study may be published. However, the data obtained from your child will be combined with data from other students in the pu blication. The published results will not include your childs name or any other information that would in any way personally identify your child. Questions? If you have any questions about this research study, please contact Stephanie Mihalas, M.A. at (813) 971 1838. If you have questions about your childs rights as a person who is taking part in a research study, you may contact a member of the Division of Research Compliance of the University of South Florida at 813 974 9343. Want Your Child to Pa rticipate? To permit your child to participate in this study, complete the attached consent form and have your child turn it in to his or her elective period teacher. Sincerely, Stephanie Mihalas, M.A. Advanced Doctoral Candidate, Psychological a nd Social Foundations ------------------------------------------------------------------------------------------------------------------------------Consent for Child to Take Part in this Research Study (Student Name: _______________________________) I agree to my child participating in both portions of the study, if he/she is chosen to do so. ___________________ ______________________ __________ Signature of Parent Printed Name of Parent Date I agree to my child participating in the survey por tion of the study only. ___________________ ______________________ __________ Signature of Parent Printed Name of Parent Date I do NOT give consent for my child to partake in any part of this study. ___________________ ______________________ ____ ______ Signature of Parent Printed Name of Parent Date Investigator Statement I have carefully explained to the subject the nature of the above research study. I hereby certify that to the best of my knowledge the subject signing this consent form unde rstands the nature, demands, risks, and benefits involved in participating in this study. _______________________ _______________________ ___________ Signature of Investigator Pr inted Name of Investigator Date
172 Appendix B: Child Assent F orm Hello! Today you will be asked to take part in a research study by filling out several surveys. I am doing this study to find out how friendships and other interpersonal relationships impact students social and emotional wellness Who I Am : I am S tephanie Mihalas, M.A., an advanced doctoral candidate at the University of South Florida. I am working with your principal to make sure this study provides information that will be helpful to your school. Why I am Asking You to Take Part in the Study : T his study is part of a project titled The Impact of Interpersonal Relationships on Student Well Being. You are being asked to take part in it because you are a student at XX Middle School. Why You Should Take Part in the Study : We need to learn more about what leads to happiness and health during the teenage years! The information that I gather may help us better understand which kinds of thoughts and behaviors teens hold, leads to emotional wellness during middle school. Also, information from thi s study will be shared with the school staff at XX Middle School to help them understand what students consider to be the strengths and weaknesses of their experiences at school and in life. However, any information that you provide will be strictly confid ential: no information reported to XX faculty will be associated with your name. Please note you will be entered into a raffle to win a $25 gift certificate. Filling Out the Surveys : These surveys will ask about your thoughts, behaviors, and attitudes to wards school, teachers, classmates, family, and life in general. I expect it will take between 30 and 60 minutes to fill out all the surveys. Answering More Questions in Individual Interviews : Students who respond in certain ways to the survey questi ons will be selected to take part in individual interviews with myself. The individual interviews will consist of asking follow up questions to your answers from the survey. If chosen to be individually interviewed, you have the choice of participation. The interviews will be kept confidential. The individual interviews will only take about 30 45 minutes, and will occur during regular school hours several weeks from today. If you are chosen and agree to participate you will receive a $10 gift certificat e. Confidentiality (Privacy) of Your Responses : I do not expect that t here will be more than minimal risk to you for taking part in this research. I will be here to help the entire time you are filling out the surveys in case you have any questions o r concerns. Your school guidance counselors are also on hand in case you become upset. Your privacy and research records will be kept confidential (private, secret) to the extent of the law. People approved to do research at USF, people who work for th e Department of Health and Human Services, and the USF Institutional Review Board may look at the records from this research project, but your individual responses will not be shared with people in the school system or anyone other than us and our research assistants. Your completed surveys will be given a code number to protect the privacy of your responses. Only we will have access to the locked file cabinet stored at USF that will contain: 1) all records linking code numbers to names, and 2) all informa tion gathered from school records. Please note that although your specific responses will not be shared with school
173 Appendix B: continued staff, if you indicate you plan to harm yourself, we will let district mental health counselors know in order to ma ke sure you are safe. Please Note : Your involvement in this study is completely voluntary. By signing this form, you are agreeing to take part in this research. If you choose not to participate, or if you wish to stop taking part in the study at an y time, you will not be punished in any way. If you choose not to participate, it will not affe ct your relationship with XX Middle School USF, or anyone else. What Well Do With Your Responses : I plan to use the information from this study to let oth ers know t he effects of students social relationships with others and how it impacts emotional wellness The results of this study may be published. However, your responses will be combined with responses from other people in the publication. The publishe d results will not include your name or any other information that would in any way identify you. Questions? If you have any questions about this research study, please raise your hand now or at any point during the study. Also, you may contact me late r at (813) 971 1838 (Ms. Mihalas) or after class. If you have questions about your rights as a person who is taking part in a research study, you may contact a member of the Division of Research Compliance of the University of South Florida at 813 974 563 8 or the Florida Department of Health, Review Council for Human Subjects at 1 850 245 4585 or toll free at 1 866 433 2775. Thank you for taking the time to take part in this study. Sincerely, Stephanie Mihalas, M.A. Advanced Doctoral Candidate Psy chological and Social Foundations --------------------------------------------------------------------------------------------------------------------Assent to Take Part in this Research Study I freely give my permission to take part in this study. I understand that this is research. I have received a copy of this letter and assent form for my records. ________________________ ________________________ ____________ Signature of child Printed name of child Date taking part in the study Statem ent of Person Obtaining Informed Assent I certify that participants have been provided with an informed assent form that has been approved by the University of South Floridas Institutional Review Board and that explains the nature, demands, risks, and ben efits involved in participating in this study. I further certify that a phone number has been provided in the event of additional questions. ________________________ ________________________ ___________ Signature of p erson Printed n ame of person Date o btaining assent obtaining assent
174 Appendix C: Informational Letter in Spanish Estimados Padres de Memorial Middle School, Hola! Esta carta es para informarle sobre una investigacin en la escuela Memorial Middle School. El director, Mr. Copeland, y sus em pleados se han puesto de acuerdo para participar en unas investigaciones con la Srta. Stephanie Mihalas. La Srta. Mihalas es una candidata avanzada de nivel de doctorado en Psicologa Escolar en la Universidad del Sur de la Florida (USF). Se les esta pidie ndo la participacin de todos los estudiantes para las investigaciones. No obstante, la participacin de su hijo(a) no es mandatoria. En resumen, esta investigacin evaluara una variedad de factores a los cuales se enfrentan los estudiantes de escuelas in termediarias. Estos factores incluyen sus amistades y pensamientos sobre sus maestros y escuela. La meta de la investigacin es entender mejor como promover el bienestar de adolescentes en las escuelas intermediarias. En especfico, los resultados de la in vestigacin se les entregaran a la escuela Memorial para informar a los administradores sobre lo que los estudiantes crean que son las virtudes y defectos de su escuela. La siguiente hoja contiene un esquema detallado sobre la investigacin y describe exa ctamente lo que se espera de su hijo(a). Favor de leer el documento. Si tiene cualquier pregunta, la informacin de contacto esta incluida. Gracias por su tiempo en leer esta informacin. Mejores deseos, Stephanie Mihalas, M.A. Dear Memorial Middle Schoo l Parents, Hello! This letter is intended to inform you about a study that is taking place at Memorial Middle School. The principal, Mr. Copeland, and his staff have agreed to participate in a study with Ms. Stephanie Mihalas, an advanced doctoral candida te in School Psychology at the University of South Florida. Every student is being asked to participate in this research study; however, your childs participation is not mandatory. In brief, the study is looking at a number of factors that middle school students typically deal with including friendships and thoughts about teachers and school. The goal of the study is to better understand how to promote the well being of adolescents in middle school. Specifically, the findings from the study will be provi ded to Memorial to inform administrators about students perceptions of strengths and weaknesses at their school. The following sheet provides a detailed outline about the study and exactly what is expected of your child. Please read through the form. If you have any questions, contact information is listed. Thank you for taking the time to read this information. Best wishes, Stephanie Mihalas, M.A.
175 Appendix D: The Childrens Depression Inventory Short Form Publisher does not permit re print of this m easure. If you have any questions, please contact Western Psychological Services (WPS) at 800 648 8857.
176 Appendix E: The Child and Adolescent Social Support Scale CASSS Over the next two pages, you will be asked to respond to sentences about some form of support or help that you might get from a parent, a teacher, a classmate, and a close friend. If the statement describes the person(s) never, you would circle the number . If the statement represents the person a lways, you would circle the number . Read each sentence carefully and respond to them as honestly. There are no right or wrong answers. My Parent(s) Never Almost Never Some of the time Most of the time Almost always Always 1. show they are proud of me. 1 2 3 4 5 6 2. understand me. 1 2 3 4 5 6 3. listen to me when I need to talk. 1 2 3 4 5 6 4. make suggestions when I dont know what to do. 1 2 3 4 5 6 5. give me good advice. 1 2 3 4 5 6 6. help me solve problems by giving me information. 1 2 3 4 5 6 7. tell me I did a good job when I do something well. 1 2 3 4 5 6 8. nicely tell me when I make mistakes. 1 2 3 4 5 6 9. reward me when Ive done something well. 1 2 3 4 5 6 10. help me practice my activities. 1 2 3 4 5 6 11. take time to help me decide things. 1 2 3 4 5 6 12. get me many things I need. 1 2 3 4 5 6 My Teacher(s) 13 cares about me. 1 2 3 4 5 6 14 treats me fairly. 1 2 3 4 5 6 15. makes it okay to ask questions. 1 2 3 4 5 6 16. explains things that I d ont understand. 1 2 3 4 5 6 17. shows me how to do things. 1 2 3 4 5 6 18. helps me solve problems by giving me information. 1 2 3 4 5 6 19. tells me I did a good job when Ive done something well. 1 2 3 4 5 6 20. nicely tells me when I make mista kes. 1 2 3 4 5 6 21. tells me how well I do on tasks. 1 2 3 4 5 6 22. makes sure I have what I need for school. 1 2 3 4 5 6 23. takes time to help me learn to do something well. 1 2 3 4 5 6 24. spends time with me when I need help. 1 2 3 4 5 6 My Classmates 25. treat me nicely. 1 2 3 4 5 6 26. like most of my ideas and opinions. 1 2 3 4 5 6 27. pay attention to me. 1 2 3 4 5 6 28. give me ideas when I dont know what to do. 1 2 3 4 5 6 29. give me information so I can learn new thi ngs. 1 2 3 4 5 6 30. give me good advice. 1 2 3 4 5 6 31. tell me I did a good job when Ive done something well 1 2 3 4 5 6
177 32. nicely tell me when I make mistakes. 1 2 3 4 5 6 33. notice when I have worked hard. 1 2 3 4 5 6 34. ask me to join activities. 1 2 3 4 5 6 35. spend time doing things with me. 1 2 3 4 5 6 36. help me with projects in class. 1 2 3 4 5 6 My Close Friend 37. ...understands my feelings. 1 2 3 4 5 6 38. sticks up for me if others are treating me badly. 1 2 3 4 5 6 39. helps me when Im lonely 1 2 3 4 5 6 40. gives me ideas when I dont know what to do. 1 2 3 4 5 6 41. gives me good advice. 1 2 3 4 5 6 42. explains things that I dont know what to do. 1 2 3 4 5 6 43. ...tells me he or she likes what I do. 1 2 3 4 5 6 44. nicely tells me when I make mistakes. 1 2 3 4 5 6 45. nicely tells me the truth about how I do on things. 1 2 3 4 5 6 46. helps me when I need it. 1 2 3 4 5 6 47. shares his or her things with me. 1 2 3 4 5 6 48. takes time to help me solve my problems. 1 2 3 4 5 6 Appendix E (continued)
178 Appendix F: Spirituality Scale Adolescents SSA The six sentences below describe how children think about themselves and how they do things in general. Read each sentence carefully. For each sentence, please think how you are in most situations. Place circle the number that describes YOU the best. If the statement never describes you, circle the . However, if the statement describes you all the time, circle the . There are no right or wrong answers. None of the time A little bit of the time Some of the time A lot of the time All of the time 1. My life has a purpose. 1 2 3 4 5 2. Participating in spiritual activities makes me feel like I have a purpose in life. 1 2 3 4 5 3. I b elieve in a higher power. 1 2 3 4 5 4. I examine what I say and do to make sure they are consistent with my spiritual values. 1 2 3 4 5 5. My life has purpose. 1 2 3 4 5 6. I am thankful for all that has happened to me and in my life. 1 2 3 4 5
179 Appendix G: The Childrens Hope Scale CHS The six sentences below describe how children think about themselves and how they do things in general. Read each sentence carefully. For each sentence, please think how you are in most situatio ns. Place circle the number that describes YOU the best. If the statement never describes you, circle the . However, if the statement describes you all the time, circle the . There are no right or wrong answers. None of the time A little bit of th e time Some of the time A lot of the time All of the time 1. I think I am doing pretty well. 1 2 3 4 5 2. I can think of many ways to get things in life that are most important to me. 1 2 3 4 5 3. I am doing just as well as other kids my age. 1 2 3 4 5 4. When I have a problem, I can come up with lots of ways to solve it. 1 2 3 4 5 5. I think the things I have done in the past will help me in the future. 1 2 3 4 5 6. Even when others want to quit, I know that I can find ways to solve the problem. 1 2 3 4 5
180 Appendix H: Social Experiences Questionnaire Self Report SEQ S The following are some things that happen to some kids. For each question, circle the number that describes how often this happened to you over the last month. If your answer is never you would circle the and if your answer is more like all the time you would circle a . How often Never Almost Never Sometimes Almost all the time All the time 1. are you left out on purpose when it is time to do an ac tivity? 1 2 3 4 5 2. does a kid who is mad at you try to get back at you by not letting you be in their group anymore? 1 2 3 4 5 3. does another kid give you help when you need it? 1 2 3 4 5 4. has another kid told lies about you to make other kids n ot like you or be at you? 1 2 3 4 5 5. does another kid say they wont like you unless you do what they want you to do? 1 2 3 4 5 6. does another kid try to cheer you up when you feel sad or upset? 1 2 3 4 5 7. does another kid try to keep others fro m liking you by saying mean things about you? 1 2 3 4 5 8. do you get hit by another kid at school? 1 2 3 4 5 9. do other kids share things with you? 1 2 3 4 5 10. do you get pushed around or shoved? 1 2 3 4 5 11. does another kid do something that makes you feel happy? 1 2 3 4 5 12. does another kid say something nice to you? 1 2 3 4 5 13. does another kid yell at you and call you mean names? 1 2 3 4 5 14. do kids yell or curse at you? 1 2 3 4 5 15. do other kids say they will beat you up i f you dont do what they want you to do? 1 2 3 4 5
181 Appendix I: Semi Structured Interview Questions (Initiate small talk; build rapport; allow time for interviewer to discuss personal background.) Start tape recorder: 1. I would like you to sh are some information about yourself. Feel free to share whatever you like (e.g., favorite music, grade, place of birth, wishes, goals) 2. Overall what has your experience been like her at XX Middle school? a. Prove for i. Interpersonal relationships with students ii. R ules and policies iii. Feelings of safety iv. Academic success/failure 3. You noted on your survey that you have experienced other kids bothering you at school or in your neighborhood. a. Are you comfortable sharing more about this experience? i. Probe for: 1. Frequency of att acks 2. Type of attacks 3. Occurrence 4. Types of students involved 4. How did this experience make you feel inside? a. Probe for feelings of: i. Anger ii. Sadness iii. Loneliness iv. Rage v. Depression vi. Aggression vii. Nothing viii. Fear 5. Is there anything that makes you feel better after a student hu rts/bothers you? a. Probe for strategies, techniques, cognitions 6. Who do you turn to for support when you need it? a. What kinds of situations do you find that you need support? b. Do you think you can use your school/ people that work at your school for support? If so, who? 7. Does spirituality and/or religion play a role in your life? a. If no: Do no prompt. Go to next question. b. If yes: Allow student to answer this question in any way desired. Most important to get free flow of answers. Only prompt allowed: Tell me more about that. 8. Is there anything else you would like to share about anything we have spoken about today? Do you have any comments about the survey you filled out previously?
182 Appendix I (continued) 9. Any other thoughts or comments? End tape recorder. Provide student with gift certificate. Have student sign form that he/she received certificate.
183 Appendix J: Student Demographics Form ID # _________________ __ ______________________________________________________________________ PLEASE READ EACH QUESTION AND CIRCLE ONE ANSWER PER QUESTION: 1. I am in grade: 6 7 8 2. My gender is: Male Female 3. How old are you? 10 11 12 13 14 15 16 4. Do you receiv e free or reduced lunch? Yes No 5. My race/ethnic identity is: a. African American/Black d. Hispanic/Latino b. Asian American or Pacific Islander e. Native American/Indian c. Caucasian/White f. Mixed Race or Other (please specify:) _______________ 6. My cum ulative GPA is: a. Mostly As (3.75 to 4.00) e. Mostly Cs (1.75 to 2.24) b. Mostly As and Bs (3.25 to 3.74) f. Mostly Cs and Ds (1.25 to 1.74) c. Mostly Bs (2.75 to 3.24) g. Mostly Ds (1.0 to 1.24) d. Mostly Bs and Cs (2.25 to 2.74) h. Mostly Ds and Fs (< 1.0) 7. On average, how much time per week do you spend doing your homework: a. Less than 1 hour e. From 10 hours to less than 15 hours b. From 1 hour to less than 3 hours f. From 15 hours to less than 20 hours c. From 3 hours to less than 5 hours g. From 2 0 hours to less than 25 hours d. From 5 hours to less than 10 hours h. 25 hours or more 8. In the past few months, how often did you participate in after school activities such as clubs, band, sports, or student government? a. Never d. Several times per week b. O nce or twice e. Daily, less than an hour c. O nce per week f. Daily, more than an hour 9. Who do you live with at home (siblings excluded)? a. Mother and father d. Parent and stepparent b. Mother e. Foster care placement c. Father f. Other (please explain ): ______________ ____________________________________________________________________________ 10 Approximately how many friends do you have (both school and neighborhood friends)? a. 1 friend d. 7 10 friends b. 2 4 friends e. 10+ friends c. 5 7 friends f. Not really sure
184 Appendix K: Effect Sizes for Survey Instruments Table 1 Effect Siz es for Race and Gender for Perceived Social Support on the CASSS Parent Support H ispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.4 2 0. 07 0.3 8 0. 27 African American female ( n = 27) 0. 41 0. 74 0. 43 0.59 Hispanic male 0. 44 0.0 1 0. 15 Hispanic female 0. 42 0.31 Mixed race male 0.16 Teacher Support Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.09 0. 06 0. 29 0. 38 African American female ( n = 27) 0.39 0.13 0.40 0.48 Hispanic male 0.17 0.00 0.15 Hispanic female 0.1 6 0.2 9 Mixed race male 0.14 Classmate Support Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.07 0 .11 0.14 0.19 African American female ( n = 27) 0.24 0.05 0.31 0.03 Hispanic male 0.19 0. 10 0. 31 Hispanic female 0.28 0. 08 Note Effect sizes derived from subtracting row demographic from column demographic ((e.g., AA male HS male)/Pooled SD).
185 Appendix K (continued) Table 1 continued Effect Siz es for Race and Gender for Perceived Social Support on the CASSS Best Friend Support Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.06 0.80 0.07 0.43 African American female ( n = 27) 0.91 0.0 6 1.00 0.34 Hispanic male 0.87 0.01 0.51 Hispanic female 0.92 0.28 Mixed race male 0.50 Note Effect sizes derived from subtracting row demographic from column demographic ((e.g., AA ma le HS male)/Pooled SD). Table 2 Effect Sizes for Race and Gender for Spirituality on the SSA Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.05 0.03 0.18 0.22 African American female ( n = 27) 0.96 0.72 1.53 1. 13 Hispanic male 0.01 0.15 0.19 Hispanic female 0.13 0.18 Mixed race male 0.07 Note Effect sizes derived from subtracting row demographic from column demographic ((e.g., AA male HS male)/Pooled SD).
186 Appendix K (continued)
187 Table 3 Effect Sizes for Ethnicity and Gender on the CDI S Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.10 0.43 0.70 0.03 African American female ( n = 27) 0.19 0.63 0.33 0.29 Hispanic male 0.52 0.65 0.13 Hispanic female 0.81 0.41 Mixed race male 0.68 Note Effect sizes derived from subtracting row demographic from column demographic ((e.g., AA male HS male)/Pooled SD). Table 4 Effect Sizes for Relational Victimization by Ethnicity and Gender on the SEQ Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.11 0.23 0.33 0.12 African Americ an female ( n = 27) 0.18 0.15 0.41 0.04 Hispanic male 0.34 0.17 0.23 Hispanic female 0.60 0.11 Mixed race male 0.52 Note Effect sizes derived from subtracting row demographic from column demographic ((e.g., AA male HS male)/Pooled SD).
188 Appendix K (continued) Table 5 Effect Sizes for Total Victimization by Ethnicity and Gender on the SEQ Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.27 0.30 0.87 0.51 African American fem ale ( n = 27) 0.16 0.17 0.50 0.11 Hispanic male 0.00 0.72 0.28 Hispanic female 0.73 0.30 Mixed race male 0.42 Note Effect sizes derived from subtracting row demographic from column demographic ((e.g., AA male HS male)/Pooled SD). Table 6 Effe ct Sizes for Hope by Ethnici ty and Gender on the CHS Hispanic male ( n = 28) Hispanic female ( n = 56) Mixed race male ( n = 8) Mixed race female ( n = 18) African American male ( n = 15) 0.41 0.41 0.57 0.51 African American female ( n = 27 ) 0.56 0.47 1.46 0.43 Hispanic male 0.02 0.58 0.11 Hispanic female 0.51 0.08 Mixed race male 1.86 Effect size = (Mean males Mean females)/Pooled SD.
189 Appendix L: Reliability Estimates for All Survey Instruments Note Depression was measured by the CDI S; relational victimization was measured by the SEQ; hope was measured by the CHS; perceived social support was measured by the CASSS; and spirituality was measured by an instrument developed for this stu dy (SSA). Table 1 Reliability Statistics for All Measures Utilized in Research Study by Gender Measure Total items Reliability Statistics Total ( N = 153) Males ( N = 51) Females ( N = 102) Depression 10 a .82 .74 .84 Average Inter Item Correlation .33 .24 .35 Range of Item to Total Correlation .30 to .62 .26 to .56 .29 to .68 a .84 .89 .74 Average Inter Item Correlation .43 .61 .35 Relational Victimization 5 Range of Item to Total Correlati on .41 to .72 .54 to .88 .36 to .65 a .85 0.82 0.86 Average Inter Item Correlation .49 0.44 0.51 Hope 6 Range of Item to Total Correlation .58 to .70 .54 to .74 .55 to .74 a 0.97 Average Inter Item Correlation 0.36 Perceived social support 47 (one item dropped) Range of Item to Total Correlation .47 to .72 0.97 0.41 .35 to 83 0.96 0.35 .47 to .69 a .81 .75 .82 Average Inter Item Correlation .40 .34 .43 Spirituality 6 Range of Item to Total Correlation .42 to .64 .32 to 0.61 .4 6 to .68
19 0 Appendix L (continued) Note Total number of items for each measure and total reliability coefficients are the same as presented in Table 1. Table 2 Reliability Statistics for All Measures Utilized in Research Study by Ethnicity Measure Reliability Statistics African American ( N = 42) Hispanic ( N = 85) Mixed race ( N = 26) Depression a .82 .83 .73 Average Inter Item Correlation .34 .34 .21 Range of Item to Total Correlation .42 to .72 .29 to .69 .06 to .67 a .80 .81 .72 Average Inter Item Correlation .45 .45 .34 Relational Victimization Range of Item to Total Correlation .48 to .75 .46 to .72 .13 to .71 a .80 .87 .86 Average Inter Item Correlation .40 .54 .51 Hope Range of Item to Total Correlation .44 to .65 .61 to .73 .41 to .81 a .97 .97 .96 Average Inter Item Correlation .38 .37 .34 Perceived social support Range of Item to Total Correlation .33 to .80 .49 to .70 .20 to .84 a .82 .82 .65 Average Inter Item Correlation .44 .43 .04 Spirituality Range of Item t o Total Correlation .43 to 0.66 .46 to .66 .29 to .55
191 Regression Analyses of Depression Using Hope and Gender as Moderator Variables (N = 152) Predictor Model B SE B R 2 ? R 2 Model 1: Relational Victimization 4.68 0.73 0.47** .22 Model 2 : Re lational Victimization Hope 4.67 4.83 0.65 0.78 0.47** 0.40** .38 .16 Model 3 : Relational Victimization Hope Female 4.47 5.23 3.71 0.64 0.78 1.31 0.45** 0.43** 0.18** .41 .03 Model 4: Relational Victimization Hope Female Relational Victimization x Hope 14.86 0.85 2.6 5 2.43 3.88 2.36 1.34 0.89 1.48** 0.07 0.13* 1.16** .44 .03 Model 5: Relational Victimization Hope Female Relational Victimiz ation x Hope Relational Victimization x Female 13.70 0.90 0.90 2.38 1.55 3.99 2.36 3.27 0.89 1.31 1.36** 0.07 0.04 1.13** 0.22 .44 .00 Appendix M: Moderator Model Summaries
192 * p < .01. p < .0 5. Model 6: Relational Victimization Hope Female Relational Victimization x Hope Relational Victimization x Female Hope x Female 12.22 2.19 10.10 2.11 1.78 2.90 4.05 2.46 7.07 0.90 1.30 1.66 1.22** 0.18 0. 50 1.01* 0.25 0.62 45 .01 Model 7: Relational Victimization Hope Female Relational Victimization x Hope Relational V ictimization x Female Hope x Female Relational Victimization x Hope x Female 2.11 3.13 30.98 0.21 18.95 6.71 3.97 6.08 3.42 19.88 1.38 7.89 4.66 1.80 0.21 0.26 1.53 0.10 2.70 1.44 2.38** .47 02 Appendix M (continued)
193 Regression Analyses of Depression Using Spirituality and Gender as Moderator Variables (N = 152) Predictor Model B SE B R 2 ? R 2 Model 1 : Relational Victimization 4.68 0.73 0.47** .22 Model 2 : Relational Vic timization Spirituality 4.43 5.19 0.64 0.76 0.44** 0.44** .41 .19 Model 3 : Relational Victimization Spirituality Female 4.23 5.37 3.00 0.63 0.75 1.27 0.42** 0.45** 0.15** .43 .02 Model 4: Relational Victimization Spirituality Female Relational Victimization x Spirituality 11.82 0.63 2.24 1.77 4.08 2.63 1.33 0.94 1.18** 0.05 0.11 0.84 .44 .01 Model 5: Relational Victimization Spirituality Female Relational Victimization x Spirituality Relational Victimization x Female 10.70 0.64 0.95 1.70 1.41 4.21 2.62 3.24 0.94 1.31 1.07** 0.05 0.05 0.75 0.20 .44 .00 Appendix M (continued)
194 * p < .01. p < .0 5. Model 6: Relational Victimization Spirituality Female Relational Victimization x Spirituality Relational Victimization x Female Spirituality x Female 10.08 0.09 3.69 1.59 1.48 1.19 4.31 2.74 7.48 0.96 1.31 1.72 1.00* 0.01 0.18 0.75 0.21 0.25 .45 .01 Model 7: Relational Victimization Spirituality Female Relational Victimization x Spirituality Relational Victimization x Female Spirituality x Female Relational Victimization x Spirituali ty x Female 5.15 2.83 15.86 0.47 9.22 3.37 1.78 6.98 4.11 23.04 1.57 8.72 5.36 1.98 0.21 0.26 1.53 0.10 2.70 1.44 2.38** .45 .00 Appendix M (continued)
195 Regression Anal yses Using Perceived Social Support and Gender as Moderators of Depression (N = 152) Predictor Model B SE B R 2 ? R 2 Model 1 : Relational Victimization 4.68 0.73 0.47** .22 Model 2 : Relational Vic timization Perceived Social Support (PSS) 4.15 4.56 0.64 0.77 0.41** 0.39** .37 .15 Model 3 : Relational Victimization PSS Female 3.88 4.90 3.51 0.66 0.77 1.32 0.9** 0.42** 0.17** .40 .03 Model 4: Relational Victimization PSS Female Relational Victimization x PSS 16.90 1.65 2.70 2.65 4.10 2.17 1.31 0.83 1.68** 0.14 0.13* 1.35** .43 .04 Model 5: Relational Victimization PSS Female Relational Victimization x PSS Relational Victimization x Female 15.82 1.97 2.42 2.70 2.22 4.12 2.17 3.25 0.82 1.29 1.58** 0.17 0.1 2 1.37** 0.32 .45 .01 Appendix M (continued)
196 * p < .01. p < .0 5. Model 6: Relational Victimization PSS Female Relational Victimization x PSS Relational Victimization x Female PSS x Female 14.89 2.68 7.04 2.52 2.06 1.92 4.19 2.24 8.30 0.83 1.30 1.55 1.48** 0.23 0.35 1.28** 0.29 0.48 .46 .01 Model 7: Relational Victimization PSS Female Relational Victimization x PSS Relational Victimization x Female PSS x Female Relational Victimization x PSS x Female 7.58 0.72 26.78 1.02 15.92 4.92 2.82 5.98 2.99 21.54 1.21 8.26 4.31 1.66 0.76 0.06 1.33 0.52 2.27 1.23 1.93 .45 .00 Appendix M (continued)
197 Regression Analyses of Depression Using Hope and Grade as Moderator Variables (N = 15 2) Predictor Model B SE B R 2 ? R 2 Model 1 : Relational Victimization 4.68 0.73 0.47** .22 Model 2 : Relational Vic timization Hope 4.70 4.83 0.65 0.78 0.47** 0.40** .41 .19 Model 3 : Relational Victimization Hope Grade 4.66 4.83 0.38 0.66 0.78 0.77 0.46** 0.40** 0.03 .38 .00 Model 4: Relational Victimization Hope Grade Relational Victimization x Hope 17.72 2.73 0.78 3.08 3.75 2.27 0.75 0.87 1.76** 0.23 0.07* 1.47** .43 .05 Model 5: Relational Victimization Hope Grade Relational Victimization x Hope Relational Victimization x Grade 20.34 2.89 1.82 3.22 1.09 4.16 2.27 1.95 0.88 0.75 2.02** 0.24 0.15 1.53* 0.30 .43 .00 Appendix M (continued)
198 * p < .01. p < .0 5. Model 6: Relational Victimization Hope Grade Relational Victimization x Hope Relational Victimization x Grade Hope x Grade 19.89 4. 40 5.75 3.13 1.12 0.96 4.18 2.69 4.25 0.88 0.76 0.9 1.08** 0.36 0.49 1.49** 0.31 0.36 .44 .01 Model 7: Relational Victimization Hope Grade Relational Victimization x Hope Relational Victimization x Grade Hope x Grade Relational Victimization x Hope x Grade 10.86 0.80 6.29 1.05 3.59 1.83 1.09 9.84 5.79 12.62 2.23 4.71 2.91 1.08 1.08 0.67 0.53 0.50 0.98 0.69 1.25 .44 .00 Appendix M (continued)
199 Regression Analyses of Depression Using Spirituality and Grade as Moderator Variables (N = 152) Predictor Model B SE B R 2 ? R 2 Model 1 : Relational Victimization 4.68 0.73 0.47** .22 Model 2 : Relational Vic timization Spirituality 4.43 5.19 0.64 0.76 0.44** 0.44** .41 .19 Model 3 : Relational Victimization Spirituality Grade 4.39 5.19 0.32 0.64 0.76 0.75 0.44** 0.45** 0.03 .41 .00 Model 4: Relational Victimization Spirituality Grade Relational Victimization x Spirituality 14.25 0.95 0.51 2.31 3.91 2.52 0.74 0.90 1.42** 0.08 0.4 1.01* .43 .02 Model 5: Relational Victimization Spirituality Grade Relational Victimization x Spirituality Relational Victimization x Grade 15.81 1.05 1.15 2.38 0.69 4.26 2.52 1.94 0.91 0.75 1.57** 0. 09 0.10 1.13* 0.19 .44 .01 Appendix M (continued)
200 * p < .01. p < .0 5. Model 6: Relational Victimization Spirituality Grade Relational Victimization x Spirituality Rela tional Victimization x Grade Spirituality x Grade 15.7 3.29 7,37 2.33 0.82 1.45 4.24 2.92 4.58 0.90 0.75 0.97 1.00* 0.01 0.18 0.75 0.21 0.25 .44 .00 Model 7: Relational Victimization Spirituality Grade Relational Victimization x Spirituality Relational Victimization x Grade Spirituality x Grade Relational Victimization x Spirituality x Grad e 15.45 3.12 6.93 2.27 0.66 1.35 0.04 9.46 5.99 14.17 2.14 4.99 3.24 1.14 1.54 0.26 0.59 1.07 0.18 0.51 0.04 .44 .00 Appendix M (continued)
201 Regression Analyses Using Pe rceived Social Support and Grade as Moderators of Depression (N = 152) Predictor Model B SE B R 2 ? R 2 Model 1 : Relational Victimization 4.68 0.73 0.47** .22 Model 2 : Relational Vic timization Perceived Social Support (PSS) 4.15 4.56 0.66 0.77 0.41** 0.39** .37 .15 Model 3 : Relational Victimization PSS Grade 4.05 4.65 0.80 0.67 0.78 0.78 0.40** 0.40** 0.07 .43 .06 Model 4: Relational Victimization PSS Grade Relational Victimization x PSS 19.37 2.99 1.16 3.14 4.05 2.13 0.75 0.82 1.93** 0.26 0.10 1.60** .43 .00 Model 5: Relational Victimization PSS Grade Relational Victimization x PSS Relational Victimization x Grade 20.22 2.88 0.40 3.13 0.50 4.25 2.14 1,95 0.82 0.75 2.01** 0.25 0.00 1 .59** 0.14 .43 .00 Appendix M (continued)
202 Model 6: Relational Victimization PSS Grade Relational Victimization x PSS Relational Victimization x Grade PSS x Grade 20.62 4.91 5.90 3.14 0.75 1.12 4.26 2.72 5,24 0.82 0.76 0.93 2.05** 0.42 0.50 1.60** 0.20 0.47 .44 .01 Model 7: Relational Victimization PSS Grade Relational Victimization x PSS Relational Victimization x Grade PSS x Grade Relational Victimization x PSS x Grade 6.26 2.38 14.68 0.23 7.39 2.30 1.65 9.48 5.08 13.23 1.91 4.87 2.60 0.98 0.62 0.20 1.25 0.12 2.01 1.27 2.07 .45 .01 * p < .01. p < .0 5. Appendix M (continued)
203 About the Author Stephanie Tamara Mihalas received her Ph.D. in School Psychology at th e University of South Florida. Her area of emphasis focused on mental health outcomes in urban minority youth and relational victimization among middle school students. Stephanie also attended an APA accredited predoctoral internship at the Kennedy Krieger Institute at the Johns Hopkins School of Medicine where she completed two rotations in the Child and Family Therapy Clinic and the Behavior Management Clinic. Stephanie has published articles and book chapters related to relational aggression, middle scho ol substance use, and methodological considerations in School Psychology. Stephanie has a passion for enhancing emotional well being and advocating for disadvantaged youth at the individual and systems levels.