This item is only available as the following downloads:
1 EXAMINING DIF F ERENT FORMS OF PEER AGGRESSION AND VICTIMIZATION AND THEIR RELATIONS WITH SOCIAL, PSYCHOL O GICAL, AND SCHOOL FUNCTIONING By JENNIFER E. ROSADO MUNOZ A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 3
2 201 3 Jennifer E. Rosado Muoz
3 To my parents and husband
4 ACKNOWLEDGMENTS I would like to thank my parents and husband for all of their support throughout my academic career. I would also like to thank my advisor, Dr. Brenda Wiens whose guidance and support facilitated the completion of this dissertation
5 TAB LE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTI ON ................................ ................................ ................................ .... 11 Peer Aggression and Victimization ................................ ................................ ......... 14 Gender Differences in Peer Aggression and Victimization ................................ ...... 21 Consequences of Aggression and Victimization ................................ ..................... 24 Social Functioning ................................ ................................ ............................ 25 Internalizing Problems ................................ ................................ ...................... 28 School Impairment ................................ ................................ ........................... 30 Measurement Issues in Studies of Peer Aggression/Victimization and Psychosocial Functioning ................................ ................................ .................... 32 Aims and Hypotheses ................................ ................................ ............................. 36 Aim One ................................ ................................ ................................ ........... 37 Aim Two ................................ ................................ ................................ ........... 38 Aim Three ................................ ................................ ................................ ......... 38 Aim Four ................................ ................................ ................................ ........... 39 2 METHODS ................................ ................................ ................................ .............. 41 Participants ................................ ................................ ................................ ............. 41 Measures ................................ ................................ ................................ ................ 42 Demographic Informat ion ................................ ................................ ................. 42 Self Report of Social Functioning ................................ ................................ ..... 42 Anxiety ................................ ................................ ................................ .............. 43 Self Report of Peer Aggression and Victimization ................................ ............ 45 ........................... 46 Cyber Aggression and Victimization Measure ................................ .................. 47 School Functioning ................................ ................................ ........................... 48 Procedures ................................ ................................ ................................ ............. 49 3 RESULTS ................................ ................................ ................................ ............... 55 Descriptive Analys es ................................ ................................ ............................... 55 Aim One ................................ ................................ ................................ ........... 59 Aim Two ................................ ................................ ................................ ........... 61 Aim Three ................................ ................................ ................................ ......... 63 Aim Four ................................ ................................ ................................ ........... 68
6 4 DISCUSSION ................................ ................................ ................................ ......... 81 Implications ................................ ................................ ................................ ............. 95 Limitations ................................ ................................ ................................ ............... 98 Future Directions ................................ ................................ ................................ .... 99 APPENDIX A DEMOGRAPHIC INFORMATION ................................ ................................ ......... 102 B 1 03 C CYBER AGGRESSION AND VICTIMIZATION MEASURE ................................ .. 104 D PARENT/GUARDIAN CONSENT LETTER ................................ .......................... 107 E STUDENT ASSENT FORM ................................ ................................ .................. 109 F TEACHER CONSENT FORM ................................ ................................ ............... 110 LIST OF REFERENCES ................................ ................................ ............................. 111 BIOGRAPH ICAL SKETCH ................................ ................................ .......................... 121
7 LIST OF TABLES Table page 2 1 Sample demographic characteristics ................................ ................................ .. 53 2 2 Total sample mean and standard deviations of measures used in study ........... 54 3 1 Percent of sample experiencing aggression and victimization at least once this year by gender ................................ ................................ ............................. 73 3 2 Means and standard deviations of study measures by gender ........................... 74 3 3 Summary of regression statistics for aim 2 a: Association between traditional aggression and social functioning ................................ ................................ ....... 75 3 4 Summary of regression statistics for aim 2 a: Association between cyber aggression and social functioning ................................ ................................ ....... 75 3 5 Summary of regression statistics for aim 2 b: Association between traditional victimization and social functioning ................................ ................................ ..... 75 3 6 Summary of regression statistics for aim 2 b: Association between cyber victimization and social functioning ................................ ................................ ..... 76 3 7 Summary of regression statistics for aim 3 a and 3 c: Anxiety Disorders Index from the MASC ................................ ................................ ......................... 76 3 8 Summary of regression statistics for aim 3 a and 3 c: Total Physical Symptoms of Anxiety from the MASC ................................ ................................ 76 3 9 Summary of regression statistics for aim 3 a and 3 c: Total Social Anxiety from the MASC ................................ ................................ ................................ ... 77 3 10 Summary of regression statistics for aim 3 b and 3 c: Anxiety Disorders Index from the MASC ................................ ................................ ......................... 77 3 11 Summary of regression statistics for aim 3 b and 3 c: Total Physical Symptoms of Anxiety from the MASC ................................ ................................ 77 3 12 Summary of regression statistics for aim 3 b and 3 c: Total Social Anxiety from the MASC ................................ ................................ ................................ ... 78 3 13 Summary of regression statistics for aim 4 a: GPA ................................ ........... 78 3 14 Summary of regression statistics for aim 4 a: Total number of suspensions and referrals ................................ ................................ ................................ ....... 78 3 15 Summary of regression statistics for aim 4 b: GPA ................................ ........... 79
8 3 16 Summary of regression statistics for aim 4 b: Total number of suspensions and referrals ................................ ................................ ................................ ....... 79 3 17 Summary of regression statistics for aim 4 c: GPA ................................ ........... 79 3 18 Summary of regression statistics for aim 4 c: Total number of suspensions and referrals ................................ ................................ ................................ ....... 79 3 19 Summary of regression statistics for aim 4 c: GPA ................................ ........... 80 3 20 Summary of regression statistics for aim 4 c: Total number of suspensions and referrals ................................ ................................ ................................ ....... 80
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EXAMINING DIF F ERENT FORMS OF PEER AGGRESSION AND VICTIMIZATION AND THEIR RELATIONS WITH SOCIAL, PSYCHOL O GICAL, AND SCHOOL FUNCTIONING By Jennifer E. Rosado Muoz August 201 3 Chair: Brenda Wiens Major: Psychology Peer victimization is a prominent occurrence amongst youth. Previously, peer victimization was generally confined to sch ool premises. Due to technological advances and frequent use of cell phones and internet by adolescents, peer victimization now defies school boundaries, with some evidence indicating an increase in the prevalence rates of cyber victimization Peer victimi zation in all forms can result in psychosocial and academic consequences. Youth who are victims of aggression are more likely to be depressed, anxious, have poorer social functioning, and have a lower grade point average compared to youth who are not victi mized. Further, perpetrators are also more likely to have poorer social functioning, poorer school attendance, receive suspensions and referrals, and have lower grade point averages than non aggressive peers. Gathering information from multiple sources, such as students and teachers, is information from self and teacher report to address areas in need of further attention in the literature, particularly with regards to cyb er aggression and victimization. Prevalence
10 of traditional aggression and victimization was higher compared to cyber aggression and victimization, consistent with findings in the literature, with little unique cyber peer victimization. While youth reports of traditional and cyber aggression and victimization did not predict social functioning, reports of both traditional and cyber victimization predicted anxiety symptoms. Moreover, cyber aggression were significantly related to school functioning as measure d by GPA and total number of suspensions and referrals combined, while traditional victimization and was significantly related to GPA These results indicate that youth who experience traditional and cyber victimization are more likely to experience increased anxiety as well as poorer school functioning, therefore, anxiety and school functioning should be addressed when schools are contemplating implementation of peer aggression prevention and intervention efforts. Findings from this study, indicate a higher prevalence of traditional aggression and victimization, as well as significant overlap between these constructs, suggesting that academic environments should continue to put forth effort to counteract traditional bullying as this type of aggression continues to be more prevalent.
11 CHAPTER 1 INTRODUCTION Peer aggression and victimization are significant concerns among youth. According to research, these behaviors occur across cultural and socioeconomic groups. Prevalence of peer aggression in prior research ranges from 4 to 36%, with victimization rates r anging from 9 to 32% from mostly urban and suburban schools, as well as from a cross national survey including youth from over 30 countries in grades four through twelve ( a w, & Sawyer, 2009, Rodkin & Berger, 2008; Spriggs, Iannotti, Nansel & Haynie, 2007). Peer aggression and victimization are most prevalent within the middle school environment, with one study of youth from urban (58%), suburban (28%), and rural (15%) environments in Maryland schools reporting approximately 37.9% of middle school students being involved in frequent peer victimization ( Another study focusing on a suburban middle school population reported 21.6% of youth bullied others within the past month and 3.8% indicated bullying others within th e past week (Branson & Cornell, 2009). Examining bullying by ethnicity in one study revealed 10.8% and12.1% of Caucasians, as well as 16.6% and 11.7% of African Americans and 15.4% and 13.6% of Hispanic youth reported physically bullying others and experie ncing victimization, respectively (Wang, Iannotti, & Nansel, 2009). Moreover, regarding verbal bullying, 35.9% and 36.6% Caucasian, 45.9% and 35.7% African American, and 36.5% and 37.1% of Hispanic youth reported verbal bullying and victimization, respecti vely. Lastly, 28% and 42.5% Caucasian, 29.8% and 36.5% African American, and 25.7% and 40.2% of Hispanic youth reported relational aggression and victimization respectively.
12 There is a subset of research focusing on youth who report both engaging in bullying and being a victim of bullying (often referred to as bully victims). According to Wang, Iannotti, & Nansel (2009) who examined bullying and victimization in a nationally r epresentative sample of over 7500 adolescents (grades 6 10) from all over the United States, 5.1% of their sample endorsed both physically bullying others and being a victim of physical bullying; being a physical bully victim was more likely for males (7.2 %) than females (3.2%). In addition, 20.3% of the sample endorsed verbally bullying others and being a victim of verbal bullying (22.6% male & 18.2% female), while 16.7% of the sample reported engaging in relational bullying, as well as experiencing relati onal victimization (15.4% males and 18% females). Peer aggression has typically taken the form of physically, verbally, and relationally aggressive behaviors. Within the past ten years, due to the advancement of technology, peer aggression has become a mo re multifaceted condition. Aggression and victimization now occur by electronic and internet means as well. Therefore, cyber aggression is currently a significant concern for youth. According to the literature, the prevalence of cyber aggression in some st udies rivals the rates of traditional (i.e., physical, verbal, and relational) aggression. Currently, studies involving middle school students from the southern United States, as well as a cross national study involving grades 6 through 10 found the preval ence of cybe r aggression ranges from 4% to 35% with cyber victimization rates ranging from 5 % to 14 % ( Dempsey, Sulkowski, Nichols, Storch 2009 ; Hinduja & Patchin, 2008; Kowalski & Limber, 2007 ; S ourander, Klomek, Ikonen, Lindroos, Luntamo, & Koskelainen, et al., 2010 ; Williams & Guerra, 2007 ; Wang, Iannotti, & Nansel, 2009 ). Particularly, one study focusing on middle school
13 students (grades 6 8) indicated 3.6% of students engaged in cyber bullying and 15.1% reported cyber victimization (Kowalski & Limber, 2007). Some researchers have argued that the prevalence has increased since 2004, where Ybarra and Mitchell (2004) found that 19% of their population sample reported engaging in cyber aggression and 4% reported being victimized online. Cyber bullies who a lso experience cyber victimization were found to range from 4.5% to 9.5% of student samples (Kowalski & Limber, 2007; Wang et al., 2009). According to the Wang et al. (2009) study, after surveying over 7,000 adolescents in grades 6 through 10 from all over the United States, 4.5% of the sample qualified for a classification of both cyber bully and victim; the percentage was similar for males (4.9%) and females (4.2%). Lower rates of reported cyber aggression and victimization in some studies compared to tr aditional aggression and victimization rates could be due to a variety of reasons. According to Olweus (2012), the prevalence of cyber aggression and victimization is actually quite low, which could relate to the degree of overlap between traditional aggre ssion and cyber aggression. Despite availability of electronic media most youth continue to engage in more traditional forms of aggression. Further, using a lower bound minimum of at least twice a month could eliminate youth who have cyber aggressed or exp erienced victimization that may have just occurred once (Olweus, 2012). Lastly, cyber bullying measures widely vary within the literature, which could impact the prevalence rates. P eer victimization in all forms negatively impacts psychosocial functionin g (Spriggs et al., 2007). Adverse outcomes occur both for those who aggress and for those who are the victims of aggression. Substance use, suicide, depression, anxiety,
14 and school impairment are negative outcomes cited within the literature with these out comes affecting youth into adulthood ( Beran & Li, 2005; Hawker & Boulton, 2000; Mitchell, Ybarra, & Finkelhor, 2007 ). Although peer aggression and victimization have been widely studied, and research in the more recent area of cyber aggression is growing, there are areas in need of further attention. Many studies in the literature have defined a cutoff point where teasing turns into peer aggression, however, few studies have examined the impact of peer aggression as a continuous variable (Craig, 1998; Storc h, Brassard, & Masia Warner, 2003; Peskin, Tortolero, & Markham, 2006; Rodkin & Berger, 2008). Further, few studies have examined multiple informants as a source of information in measuring traditional aggression and victimization With regards to cyber a ggression, the means used to assess cyber aggression and victimization var ies and often only a few questions are used to assess cyber behavior. Measurement of the behaviors with more comprehensive questionnaires would be helpful, particularly for examining gender differences and relationships between these behaviors and youth funct ioning. Finally, while adverse outcomes in the domains of social functioning, anxiety, and academic and school functioning have been more widely examined for traditional forms of aggression and victimization, the relationship between cyber aggression and v ictimization and these domains is in need of further study These areas were the focus of this study. Peer Aggression and Victimization According to Olweus (1993), the definition of bullying incorporates three important components: 1) unwanted aggressive b ehavior 2) repeated behavior over an extended period of time and 3) a power differential between the aggressor and victim. Bullying is a more severe and chronic form of peer aggression. The construct of peer
15 aggression has been defined as harmful behaviors repeatedly directed towards another youth through the use of physical, verbal, or relational aggression (Crick & Grotpeter, 1996; Rys & Bear, 1997). Notably, due to overlap in definitions both terms are used interchangeably within the literature, blurring the boundary between peer aggression and bullying. For the purpose of this study, the term peer aggression will be used as it denotes a continuum of less frequent teasing to more frequent and severe bullying. Noticeably lacking from the definition of bull ying is motivation. While there is a definition of the factors that comprise bullying there is a definite lack of understanding of why particular youth engage in bullying. According to Rigby (2012), children ranging in age from 8 to 18 endorsed the followi ng reasons for bullying: I was annoyed, retribution, for enjoyment, others were also bullying, the victims were weak, to demonstrate how tough I was, and to acquire money or items. Overall, minimal attention has been given to understanding and elucidating the specific reasons why youth bully others (Rigby, 2012). Factors that may play a role in bullying include moral disengagement, identified process where an individual ration alizes his/her aggression (Pornari & Wood, 2010). Also, social information processing theory by Crick and Dodge (1994) stated that there are deficits in the processing of social information in aggressive individuals (Pornari & Wood, 2010). Even though moti vation for bullying is not within the scope of this study, further understanding motivation for bullying behaviors could potentially impact interventions in schools. Peer aggression is comprised of two broad subtypes, overt and indirect aggression ( Berger 2007; Kuppens, Grietens, Onghena, & Mich ie ls, 2009 ; Nixon, 2001;
16 Pornari & Wood, 2010). Physical, behavioral, and verbal behaviors are encompassed within the definition of overt aggression. Hitting, pushing, and kicking are examples of physical aggressio n, stealing a lunch constitutes behavioral aggression, and threatening to beat someone up, name calling, and yelling comprise verbal aggression ( Kuppens et al., 2009; Pornari & Wood, 2010). Indirect aggression is a more subtle form of aggression, which is also known as relational aggression, social aggression, or psychological aggression (Kumpulainen, Rasanen, Henttonen, Almqvist, Kresanov, Linna, et al., 1998; Storch et al., 2003). Due to its subtle nature, relational aggression is often more difficult to detect. Within the literature relational aggression has been defined as behavior intended to harm peer relationships or threaten to damage relationships (Craig, 1998; Crick & Grotpeter, 1996). Relational aggression involves spitefully spreading negative co mments or rumors and socially excluding others with the intent to harm the relationships of others create feelings of rejection, threaten to end friendships, and exclude others from groups (Kuppens et al., 2009). Estimations from research indicate that 1 in 10 youth are chronically victimized by peers (Tokunaga, 2010). This research indicated that r easons for peer victimization vary and mostly depend on individual characteristics, peer group dynamics, and the climate of the school. Youth who are perceived as being inadequately prepared to defend themselves are more likely to experience traditional peer aggression. Developing a reputation as being physically incapable, being known for acquiescing to the demands of peers, or being rejected by the peer cohort increases the likelihood that a youth will be victimized (Hodges, Malone, & Perry, 1997; Olweus, 1993; Schwartz, Dodge, & Coie, 1993). Socially incompetent peers are also at increased risk for being
17 victimized. Further, socially withdrawn behavior, such a s shyness, also i ncreases the likelihood that a youth will be victimized (Cook, Williams, & Guerra, 2010). Cyber Aggression and Victimization In addition to traditional peer aggression, cyber aggression is another form of aggression that has developed wit hin the last 20 years. Within the literature cyber aggression is known by many terms, such as cyber bullying, electronic bullying, and online social cruelty (Pornari & Wood, 2010; Kowalski & Limber, 2007). Just as there are multiple terms for the construct there are multiple definitions available. To illustrate, cyber aggression has been defined as a purposeful, aggressive act that a group or individual repeatedly engages in, via electronic means, against an individual who is unable to defend himself or he rself easily (Smith, Mahdavi, Carvalho, Fisher, Russell, & Tippett, 2008). Further, the construct has also been defined as aggressive and repeated information being delivered though electronic format for the intended purpose of harming others (Dooley, Pyza lski, & Cross, 2009). For the purpose of the current research study, the definitions will be combined. Cyber aggression will be defined as an intentional aggressive act via electronic format used for the sole intent of harming others (Smith et al., 2008; D ooley et al., 2009). This construct encompasses the use of mobile phones and internet based communications as mediums for relational and verbal aggression as physical aggression cannot occur through these media. The increasing availability of cell phones, with the majority of phones currently coming equipped with internet capability, allows for easy internet access (Mishna, Saini, & Solomon, 2009; Pornari & Wood, 2010). Up to 97 % of adolescents ages 12 18 use the internet (Kowalski & Limber, 2007). The hig h prevalence of internet use indicates that the majority of youth are vulnerable to victimization through electronic means. This
18 advancement of technology has altered the way individuals communicate in a social context. In addition to direct, face to face interaction, people communicate via e mail, blogs, social networking sites, instant messaging, chat rooms, text messages, and webcams (Mishna et al., 2009). All of these communication styles can occur in the palm of your hand allowing for negative social i nteractions to become more pervasive. There is some question as to whether traditional aggression and cyber aggression are similarly overlapping constructs or contain unique characteristics with some overlap Existing l iterature appears to support both cl aims. On the one hand, cyber aggression has been characterized by some researchers as a mere extension of traditional peer aggression with minimal unique qualities (Erdur Baker, 2010; Li, 2005; Li, 2006; Olweus, 2012; Raskauskas & Stoltz, 2007). One study found overlap rates between traditional and cyber aggression to range from 88% to 93%, indicating small percentages of youth who experience cyber aggression in the absence of other types of peer aggression (Olweus, 2012). Conversely, one study fou nd some o verlap between the two constructs, but noted that the similarities between the two constructs were small with the study ultimately supporting each construct being comprised of distinct characteristics (Erdur Baker, 2010). Furthering the support of traditio nal and cyber aggression being distinct factors, Dempsey Sulkowski, Nichols, and Storch, (2009) illustrated through confirmatory factor analysis that cyber victimization is a separate factor apart from overt and relational aggression. Notably, there was a weak relationship between all aggression types. Additionally, cyber aggression accounted for adverse outcomes, such as social anxiety, above and beyond traditional means of peer aggression Supporting it as an independent factor containing unique componen ts,
19 Green (2006) proposed three criteria delineating the factors associated with traditional peer aggression that cyber aggression violates: peer aggression being relegated mostly to school environments, the victims being aware of who is bullying them, and there being a power differential between the bully and victim. The availability of electronic media has increased the prevalence of peer aggression occurring outside of school boundaries. Peer aggression is no longer solely relegated to the school yard. Additionally, many youth have constant access to electronic media, thus distinguishing it from traditional peer aggression, as traditional peer aggression is generally contained to specific context s (e.g., the school lunchroom, playground, etc.) Instead cyber aggression allows for victimization to follow the child throughout the day whether they are at school or elsewhere. To illustrate, Mishna, Saini, and Solomon (2009) found that youth described how they are now subject to peer aggression within the co nfines of their home where before they perceived themselves to be safe. In the same study, children described feeling particularly invaded when dealing with peer aggression within their home environment. Thus, due to advances in technology, victimization c an occur all day for some students making it virtually inescapable. Anonymity is a nother way in which cyber aggression differs from traditional peer aggression. Cyber aggression can occur anonymously. Youth are able to create pseudo social networking acco unts, text peers, and write anonymously on blogs in order to victimize peers. Anonymity allows youth to engage in behaviors they would not otherwise engage in since fear of repercussions is minimal (Mishna et al., 2009). Conversely, while there is anonymit y with some cyber aggression in other instances it
20 is quite obvious who is aggressing For example, many youth make no attempt at hiding their identity on social networking sites when engaging in cyber aggression ( Mishna et al., 2009). Thus youth can cho ose to be identified as aggressors online or they can easily choose to remain anonymous when engaging in cyber aggression. Cyber aggression also differs from traditional aggression as often there is a lack of power differential between bullies and victims when online. In more traditional forms of peer aggression the perpetrator generally has either more physical or social power over the victim. Online, this power differential between aggressor and victim tends to be non existent. Anonymity can allow those who perceive themselves to be weaker in person to be on equal footing in cyberspace. The victim can turn into the aggressor with the touch of a button. While a power differential has been indicated to exist in cyberspace, it is typically associated with co mputer knowledge instead of physical and social power (Patchin & Hinduja, 2006). Via the internet and cell phone, verbal and relational aggression occurs in many forms. Spreading rumors, name calling, and circulating compromising photographs and/or videos are ways that an individual can be victimized (Slonje & Smith, 2008). Cyber aggression can also occur in the form of coercion, where youth, especially males, can peer pressure females into exposing themselves over web cams or taking inappropriate pictures on their cell phones and sending them via text message (Mishna et al., 2009). This coercion can come in the form of black mail. For example, females from one study indicated that if they did not expose themselves to a male peer, that peer threatened to re veal their secrets (Mishna et al., 2009). Further, youth sometimes provide friends with their passwords to internet sites, allowing peers to masquerade as
21 friends and aggress on pose as other pee rs can also victimize the person they are imitating by posting hurtful or embarrassing content in the form of comments or pictures (Mishna et al., 2009). Youth who engage in or are victims of, traditional aggression are significantly more likely to engage in cyber aggression (Pornari & Wood, 2010). Cyber aggression is a format allowing youth who are victims of traditional aggression to retaliate as the internet provides a buffer between perpetrator and victim whereas a victim is less likely to retaliate when face to face with their aggressor when traditional peer aggression occurs, especially if the victimization is through physical means. Gender Differences in Peer Aggression and Victimization General models of aggressive behaviors tend to be linked to gender, with physical and overt aggression being the most widely researched (Spieker, Campbell, Vandergrift, Pierce, Cauffman, Susman, & Roisman, 2012). When attempting to understand the experiences of aggressors and victims in traditional and cyber aggres sion, gender is an important factor to consider. To facilitate the understanding of the role gender plays in peer aggression and victimization, gender disparities have been examined by researchers within the field. The majority of research attention has f ocused on physical (overt) aggression, thus more is known about the gender disparities (Spieker et al., 2012). Overall, when examining overt forms of traditional aggression such as physical aggression, males are more often the culprits and recipients of th is type of aggression while victims include both males and females (Erdur Baker, 2010 ; Rodkin & Berger, 2008). As toddlers, males exhibit more physical aggression compared to females when examining typically developing children, with decreased aggression o ccurring around the age of three (Tremblay, Japel, Perusse, McDuff, Boivin, Zoccolillo
22 et al.,1999). Risk factors such as low socioeconomic status, increased maternal symptoms of depression, and decreased maternal warmth predict increased and consistent ph ysical aggression (Spieker et al., 2012). Persistent physical aggression could lead to decreased social functioning and rejection by normative peers, further exacerbating aggressive behaviors as these youth have not developed appropriate means of interacti ng with peers (Spieker et al., 2012). Theoretically, research lacks understanding of the developmental process regarding relationally aggressive behaviors (Spieker et al. 2012). Ev idence, dating back to the 1920 s traditionally characterized females as perp etrators of relational aggression, as well as victims, with current researcher continuing to support this viewpoint (Archer, 2004; Craig, 1998 ; Pornari & Wood, 2010 ). While some research continues to provide evidence that females are more likely to engage in and experience relational aggression and victimization, understanding how this gender difference occurs is less well known and researched. Preliminary data emerged suggesting high maternal control, low maternal empathy, and maternal negative affect were associated with the development of relational aggression in preschool age males and females (Brown, Arnold, Dobbs, & Doctoroff, 2007; Casas Weigel, Crick, Ostrov, Woods, Jansen Yeh et al., 2006; Curtner Smith Culp, Culp, Scheib, Owen, Tilley et al., 200 6). Focusing specifically on females, evidence indicated maternal callousness and low maternal sensitivity placed females at risk for relational aggression, but not males (Spieker, 2010). Craig (1998) postulated that f emales have a greater likelihood of e ngaging in relational forms of aggression as the importance of peer relationships is more consistent with their social goals. Similarly, Crick et al. (1999) suggests that girls tend to become more troubled
23 when relationship difficulties occur. Re lational aggression, which could be characterized as reactive, defensive, or proactive, could potentially disrupt social relationships and enhance social status thus perpetuating the cycle (Marini, Dane, Bosacki, YLC CURA, 2006 ; Spieker, 201 2 ). Inconsistent findi ngs regarding gender discrepancies in cyber aggression behaviors predominate within the literature (Erdur Baker, 2010; Sourander et al., 2010). Considering the nature of cyber aggression, as well as the potential social ramifications of being victimized, f emales are the lead contenders to be characterized as cyber aggressors. Cyber aggression is similar to relational aggression as the victim is being bullied in an indirect manner. Additionally, the aggression is occurring within the social realm as others a re often privy to the victimization online. Research pertaining to relational aggression indicates females are more likely to be perpetrators, thus it would follow that they might also engage in more cyber aggression than males. Supporting this viewpoint, females have been shown to be perpetrators of cyber aggression and to be victimized more than males (Kowalski & Limber, 2007; Marini et al., 2006 ). Dempsey et al. ( 2009 ) also noted that a larger proportion of females in their sample reported being cyber victimized. However, according to the results from this study the noted relationship between cyber victimization and gender was weak. While some evidence indicates gender disparitie s, other studies have failed to identify gender differences in perpetrators and victims of traditional and cyber aggression (Patchin & Hinduja, 2006). One study found no gender differences for all overt, relational, and cyber victimization (Pornari & Wood, 2010). Notably, cyber aggression and victimization was assessed via only six questions total in this study.
24 Another study found no significant differences between gender groups for relational victimization, indicating that both genders were victimized equ ally (Storch et al., 2003). Given discrepancies in findings, further research is needed with regards to gender differences in cyber aggression and victimization. Consequences of Aggression and Victimization linked to multiple adverse outcomes, including both externalizing and internalizing difficulties. To illustrate, aggression and victimization has been shown to be associated with substance use, youth violence, physical health symptoms, delinquency, poor academic functioning, poor school attendance, low self worth, psychosocial difficulties, anxiety (generalized and social), and depression (Hawker & Boulton, 2000; Kumpulaine n et al., 1998; Spriggs et al., 2007). Similar to traditional forms of bullying, youth who were bullied online also reported experiencing anxiety, depression substance use, and difficulty concentrating in school (Beran & Li, 2005; Mitchell et al., 2007). In addition to acute symptoms of psychosocial distress and internalizing difficulties, those who bully and are bullied are also at risk of experiencing long term psychosocial and psychological difficulties later in life (Nijmeijer, Minderaa, Buitelaar, Mulli gan, Hartman, & Hoestra, 2008). For example, loneliness, depressed mood, suicidal behavior criminal behavior, and anxiety, especially social anxiety in adulthood are difficulties victimized adults and adults who were bullies have experienced (Gladstone, B arker, & Malhi, 2006; Klomek, Sourander, Niemela, Kumpulaine n, Piha, Tamminen, et al., 2009).
25 As t he present study will focus on social functioning, anxiety, and school functioning as they are related to traditional peer aggression and victimization, as well as cyber aggression and victimization research in these specific areas will be highlighted below It should be noted that in research on these three areas of functioning, it is not always possible to disentangle whether peer aggression and victimization are precursors to poor functioning in these areas, or whether poor functioning in these areas predisposes a child to being a bully or victim, as many studies are not able to directly examine causality due to the me thodologies employed. Social Functioning Social functioning is negatively impacted in aggressive and victimized youth. and peer interactions, thus impacting overall soci al competence. Social competence is defined as the ability to successfully carry out social interactions and maintain interpersonal relationships, as well as being able to express and construe various forms of communication (Larson, Whitton, Hauser, & Alle n, 2007; Ford, 1982). Social functioning is the behavioral component of social competence ; it is the observable part of social behavior (Nixon, 2001). Prosocial behaviors, such as sharing and being kind, comprise good social functioning, which in turn tend s to generate friendships. Friendships are comprised of two individuals who jointly participate in the development and maintenance of the friendship allowing for peers to learn cooperation, conflict resolution, and negotiation as well as contributing to t he overall adjustment of youth (Hoza, 2007; Newcomb, Bukowski, & Bagwell, 1999; Greco & Morris, 2005). These skills allow youth to carry out appropriate social interactions and navigate the social realm of life. Further, a ppropriately developed psychosocia l functioning in children is
26 outcomes (Renk & Phares, 2004). Thus, social functioning is an important aspect of the being (No rmand, Sch n eider, & Robaey, 2007). psychological health (Greco & Morris, 2005). To illustrate, having difficulties with peers is predictive of future delinquency, substan ce use, school dropout, academic difficulties, and psychopathology ( Hoza Mrug, Gerdes, Hinshaw, Bukowski, Gold, et al., 2005; Normand et al., 2007). Thus, peer acceptance and rejection both have a qualitative this study will be on peer rejection via peer victimization. Victimized youth are characterized as having poor social relationships with peers et al., 2007). These youth tend to have fewer friends and are ofte n more rejected than youth who are not victimized (Spriggs et al., 2007 ; Nijmeijer et al., 2008). In turn, youth who experience psychosocial difficulties are more likely to be unpopular and rejected or victimized by peers (Dill, Verberg, Fonagy, Twemlow & Gamm, 2004; Hodges, Boivin, V itaro, & Bukowski, 1999). The lack, or poor quality of, peer interactions provides a barrier to these youth learning and developing prosocial skills; these youth are likely to lack the appropriate skills necessary to develop po sitive peer relationships. Y outh who bully others are also likely to lack the prosocial skills necessary to develop appropriate social relationships (Smokowski & Kopasz, 2005). Unpopular and disliked youth are more likely to be entering into friendships w ith other rejected peers,
27 such as antisocial peers or peers who are more likely to engage in criminal behavior (Normand et al., 2007). Two studies found that amongst middle school youth, bullies, a nd those who were both bullies and victims were less likel y to indicate that friendships were important (Dempsey et al., 2009 ; H owever peer relationships in one of these studies were assessed wit h two questions via self report. While some aggressive youth may lack prosocial skills and ha ve few friends, n ot all aggressive youth are unpopular or disliked. To illustrate, there are youth who have high social standing, such as athletes, who can be popular although they engage in aggressive behavior. A general reason for this is due to the soci al power athletes have. Athletes tend to be leaders and considered attractive, thus they hold social sway and can pressure others into doing things (Rodkin & Berger, 2008). These researchers noted that people who are wealthy, attractive, and hold positions of leadership are more likely to hold positions of power, even in the school setting. S tudents who fit into these categories may be deemed as popular even though they may not be excessively liked by others. Thus, a ggressive students may be able to achieve a high social status and attain popularity (Rodkin & Berger, 2008). There is a dearth of information regarding the social risk factors associated with cyber aggression. One study examining these factors was conducted in Finland. According to this study, being a victim of cyber aggression, as well as being a combined cyber bully and a cyber victim was predictive of difficulties with peers and emotional problems such as depression, anxiety, and loneliness (Sourander et al., 2010). Thus, further research is needed regarding the relationships between cyber aggression/victimization and youth social functioning.
28 Internalizing Problems Anxiety is an adverse outcome associated with peer aggression. Moderate levels of anxie ty and depression were found in a school sample of victims of traditional bullying (Bond, Carlin, Thomas, Rubin, & Patton, 2001 ; Gladstone et al., 2006 ; Sourander et al., 2000). In another study of adolescents ages 13 to 16, Storch, Brassard, and Masia War ner (2003) found that overt and relational victimization were associated with significant fear of being evaluated negatively by others, somatic symptoms, and avoidance of social situations. Moreover, the study indicated that multiple victimization experien particular, relational victimization is the form of aggression that is most associated with symptoms of social anxiety. Taken together, these findings indicate that youth who experience direct and relati onal victimization are at risk for experiencing significant psychosocial difficulties. Peer victimization may also be related to anxiety in adulthood. In one study utilizing a clinical population of adults, it was found that being bullied in childhood was significantly related to high levels of anxiety ( Bond et al., 2001). However, the participants were adults recalling events that occurred in their childhood, which is subject to bias. In line with research on traditional peer victimization, research has al so found that cyber victimization was associated with social anxiety (Dempsey et al., 2009). Even when studies control for social functioning and socioeconomic factors, repeated victimization has been significantly connected to self reported symptoms of a nxiety (Bond et al., 2001). Specifically, victims of verbal and relational peer aggression reported higher levels of anxiety (Craig, 1998). Research in this area has additionally shown lems in creased as
29 they aged; high school bullies and victims were more likely to endorse internalizing difficulties than those in middle and elementary Research is lacking regarding the causality between peer victimization and an xiety Thus, it is undetermined if the relationship is bidirectional or if either variable precedes the other. Further, there are several vulnerability factors associated with the development of anxiety, exacerbating the difficulty in determining w hether experiencing peer victimization is a causal factor in the development of anxiety To illustrate, familial factors play a role in the onset of anxiety. Youth whose parent or parents have a specific type of anxiety disorder are three to five times more likel y to develop an anxiety disorder compared to youth whose parents do not have an anxiety disorder (Merikangas, 2005). Further, there are temperament and personality factors such as behavioral inhibition, which is an increased psychological response when fac ed with novel input from the environment, that precede the development of anxiety (Merikangas, 2005). Thus, genetic factors and temperament, as well as environmental exposures such as parenting and exposure to life stressors, also play a role in the develo pment of anxiety (Merikangas, 2005). Environmental exposures pertaining to life stressors will be central to this study. Life stressors include peer victimization, which can be classified as an experience that threatens the perceived well being of others. Lack of peer acceptance has been associated with social anxiety. Children who are disliked or rejected by peers are at risk for the developmental of emotional and behavioral difficulties (Greco & Morris, 2005). Disliked or rejected children could potenti ally become withdrawn and decrease their social interactions exacerbating their poor social functioning and increasing their anxiety in social situations. Conversely,
30 these children could attempt to increase their social interactions in inappropriate ways causing further rejection and victimization (Greco & Morris, 2005). When conceptualized as a life stressor, p eer victimization may play a role in the onset of anxiety symptoms and anxiety disorders Assessing for anxiety varies widely within the peer victi mization literature with self report being the predominant method of assessment However, the type of self report measure used varies Several studies within the literature assess internalizing symptoms with questions generated by the researchers instead of a standardized measure. To illustrate, several studies measured anxiety via three and four question social anxiety, versus the anxiety spectrum (Dempsey et al., 2009; Storch et al., 2003). Thus, the inclusion of more comprehensive measures of anxiety symptoms would help further research in this area. School Impairment Poor social functioning, as seen in aggressors and victims, negatively affects academic functioning. It is hypothesized that peer victimization could increase anxiety, interrupt concentration within the classroom, and interfere with youth acquiring and retaining information ( Nansel Overpeck, Pilla, Ruan, Simons Morton, & Scheidt 2001; Sharp 1995). Fur ther, rejected and victimized children in the classroom tend to have a poorer academic self concept and are more likely to rely on teachers for assistance on class work compared to children with normal social functioning (Flook, Repetti, & Ullman, 2005; Me rcer & DeRosier, 2008). Victimized youth also tend to avoid attending school, which results in increased absences and diminished opportunities to develop
31 academic abilities within the classroom (DeRosier Kupersmidt, & Patterson, 1994; Eaton Kann, Kinchen, Shanklin, Ross, & Hawkins et al., 2008). In addition to victimized youth demonstrating poorer academic functioning compared to non victimized and socially accepted peers, aggressors are also more likely to exhibit poorer academic functioning than their n onaggressive peers (Hinshaw, 1992). Overall underachievement, lower grade point average, and school failure of aggressive youth could be partly attributed to increased conflict with teachers and school personal (DeRosier & Lloyd, 2010). Aggression tends to disrupt classrooms resulting in disciplinary action (Coie & Dodge, 1998). Youth who are disruptive in the classroom and who have received school suspensions have decreased opportunities to build academic skills and can fall further behind (DeRosier & Llo yd, 2010). Therefore, diminished academic functioning has been linked to both aggressor and victim status. One study assessing the academic performance of minority youth indicated that poor academic functioning for Caucasians and Hispanics was associated w ith bully, bully victim, and victim status (Spriggs et al., 2007). Conversely, for African American youth, research indicated that negative school factors were not associated with peer aggression behaviors (Spriggs et al., 2007). Another study using a mult i culturally diverse sample found a significant association between poor school functioning and perpetration of traditional aggression, such that those who were nominated as aggressors, by peers as well as self report, were more likely to receive behaviora l referrals in school (Branson & Cornell, 2009). Further, the researchers indicated that those who were nominated by peers as aggressive were more likely to receive detentions and suspensions, as well as have poorer academic functioning. Cyber
32 aggression a nd victimization and its relationship to academic and school functioning were not a focus of this study. Cyber aggressors and victims of cyber aggression have been found to experience more school impairment, such as more often skipping school, receiving detentions or suspensions, and being found with a weapon on school premises (Ybarra, Diener West, & Leaf, 2007). The prevalence of delinquency is significantly higher in youth who report experiencing online peer victimization, which can further exacerbate poor academic functioning (Mitchell et al., 2007), as behaviors such as skipping school and getting into trouble at school can negatively impact overall academic performance. Overall, relatively fewer studies in this area examine the relationship of peer v ictimization and peer aggression, especially cyber victimization and aggression, to measures of academic functioning. Additionally, it is difficult to determine a directional relationship between poor academic functioning and peer victimization as there is a lack of longitudinal research determining causality. Thus, further research on these relationships would be beneficial. Measurement Issues in Studies of Peer Aggression/Victimization and Psychosocial Functioning Self report is widely used to assess self concept, social functioning, psychological functioning, and bully and victimization status. There are several benefits to self report questionnaires when studying peer victimization, such as youth are better equipp ed to discuss their victimization status than other informants since they are the ones being victimized (Ladd & Kochenderfer Ladd, 2002). Thus, self report may provide victimizati on and aggression status, as well as social functioning and anxiety, via self
33 report measures. Although the use of self report measures in peer victimization research has been shown to be valid, there are drawback s to this type of measurement. Disadvantag es of self report measures pertain to factors that decrease the validity of the report. To illustrate, youth differ in their interpretation of aggressive behavior as well as their willingness to identify themselves as a bully and or victim (Ladd & Kochende rfer Ladd, 2002). Youth may be reluctant to nominate themselves as perpetrators of peer aggression due to the social stigma associated with these behaviors or may underreport victimization status (Branson & Cornell, 2009; Griffin & Gross, 2004). For exampl e in one study it was found that only 15 of the 416 students in the sample reported themselves to be perpetrators of peer aggression even though 77 youths were nominated by peers as perpetrators (Cornell & Brockenbrough, 2004). In addition to under report aggressive behavior on self report surveys (Branson & Cornell, 2009). Cyber aggression has also been measured through self report. However, self report measures assessing cyber aggression and vic timization also have a number of limitations. There appear to be some cyber events that tend to occur as a single event, such as exposure to a personally embarrassing photo or video that can be extremely distressing to an individual and spread rapidly to a larger group of youth (Olweus, 2012). However, the definition of cyber bullying in some studies traditionally involves experiencing at least two events, thus potentially misclassifying a group of youth who have cyber aggressed or who have experienced cybe r aggression at least once in a particularly harmful way (Olweus, 2012). In addition to a potentially poor criterion limit of cyber aggression, there are few established measures of cyber aggression. Given the
34 lack of established measures in this area, som e researchers have devised their own measures for their studies (Kowalski & Limber, 2007 ; Pornari & Wood, 2010). The devised measures vary in length and specificity of items, but generally tend to be brief, creating another problem which is the use of limi ted items to assess the construct. For example, one study used a questionnaire comprised of three items assessing cyber aggression and three items assessing cyber victimization (Pornari & Wood, 2010). In another study examining relations between cyber aggr ession/victimization and psychosocial and psychological adverse outcomes, only four self report questions were used to assess cyber victimization (Dempsey et al., 2009). Drawbacks to self report have been dealt with by including multiple informants when a ssessing peer aggression and victimization, as well as social functioning. A popular method is utilizing peer reports as an alternate source of information. Higher rates of concordance among multiple informants are anticipated when informants are reporting behaviors occurring in a context where they are able to directly observe the behaviors (Achenbach, McConaughy, & Howell, 198 2 ). To illustrate, higher agreement school functi oning and social status as teachers offer a unique perspective in that they can observe children interacting with their peers (Achenbach et al., 198 2 ; Ladd & Kochenderfer Ladd, 2002). Lesser agreement among informants is expected when parents are reporting assessing the correlation between se lf and peer report found that as peers age the multi informant approach became more reliable, providing support for the use of multiple informants when available (Ladd & Kochenderfer Ladd, 2002). Measuring social
35 functioning through a multi informant approach allows unique pieces of information to be obtained regarding peer functioning. The various informants provide overlapping and independent informati on regarding peer rejection (Ladd & Kochenderfer Ladd, 2002). Self report measures allow researchers an inside view of how the child perceives him or herself, while other informants provide information on how the child is viewed by others, as well as how t he child is functioning within a certain domain. In order to account for the varying perspectives, s ome researchers have used self and peer reports in combination to identify students are bullied and victimized (Branson & Cornell, 2009; Graham, Bellmore, & Juvonen, et al., 2003). Branson and Cornell (2009) compared the utility of self report versus peer report surveys. While this study assessed peer aggression and victimization, as well as school climate, academic functi oning, teacher tolerance, and depression, only traditional peer aggression and victimization were the focus. Peer nominations for traditional peer aggression correlated moderately with disciplinary infractions, detentions, suspensions, and lower grade poin t average while self reported bullying correlated with aggressive attitudes, disciplinary referrals, suspensions, and GPA ( Branson & Cornell, 2009; Cornell & Brockenbrough, 2004 ). When the magnitude of the correlations was compared, self report measures ha d a higher association with aggressive attitudes, while disciplinary referrals were more correlated with peer report than self report. Despite the potential benefits of using peer reports, use of peer report was ultimately not feasible within the current study. Although peer reports are considered the gold standard, research indicates that self report provides valuable information when
36 as a multiple perspective viewpoint, teacher report of social functioning as well as aggression and victimization status, was used in conjunction with self report. Teacher report is commonly used in conjunction with peer report within the literature to assess aggressive behavior and social functioning (DeRosier & Lloyd, 2010; Kumpulainen et al., 1998; Kuppens et al., 2009; Ladd & Kochenderfer Ladd, 2002). Kuppens et al. (20 09) noted that teacher reports of physical and verbal aggression were stable across time (two years), while reports of relational aggression were not as stable over time. Relational aggression is more difficult to detect than overt aggression, therefore th ese findings appear consistent with the research presented on this topic. While teacher reports for some behaviors may not be stable over time, one study indicated a significantly greater response agreement between teachers and peers compared to parent and peer, mother and father, and teacher and parent informants (Renk & Phares, 2004). Further, this study indicated that teacher and peer measures would have greater concordance when the measures are completed within the academic environment. Aims and Hypotheses While prior research in this area has shown links between peer aggression/victimization and poorer student social, psychological, and school functioning, fewer studies have examined both traditional and cyber forms of peer aggression/victimizati on in the same study with regards to how these constructs relate to multiple domains of student functioning. Moreover, there are several areas in the literature on cyber aggression/victimization that could use further attention. First, measuring cyber aggr ession and victimization with more comprehensive measures would be helpful, as prior studies have often measured these constructs with just a few
37 items. Second, there are discrepancies regarding gender differences in cyber aggression and victimization, thu s further examination of gender differences is needed, particularly with more comprehensive measures. Similarly, further examination of the relationships between cyber aggression/victimization and youth social and school functioning are needed, as research is sparse in these areas. Another area in need of further research is the relationship of both traditional and cyber victimization to youth anxiety. While prior studies have shown links between different forms of peer victimization and anxiety, fewer stud ies have used comprehensive and standardized measures of anxiety, particularly in the area of cyber victimization. Lastly, this study focused specifically on gathering information from students in a rural area, as the majority of studies focus on urban and suburban populations. Aim One The first aim of this study measure d traditional aggression via self and teacher report and victimization within the population via self report as well as cyber peer aggression and victimization through self report Additi onally, gender differences within the population for both traditional and cyber victimization and aggression were assessed. Hypothesis 1 a Consistent with prevalence rates in the literature, it was expected that females would more often report being per petrators of relational aggression with male reports dominating in physical aggression. Further, it was expected that on teacher report measures, more females would be nominated as perpetrators of relational aggression with more males being nominated as ph ysical aggressors.
38 Hypothesis 1 b It was expected that males would more likely endorse being the victims of physical aggression by peers and females would endorse being the victims of relationally aggressive behaviors Further, it was expected that on teacher report measures, more females would be nominated as victims of relational aggression with more males being nominated as victims of physical aggression. Hypothesis 1 c In addition, as cyber aggression shares some features of traditi onal relational aggression, it was hypothesized that females would report perpetrating more cyber aggression when compared to males. Aim Two The second aim of this study examined relationships between youth social functioning, as measured by self and tea cher report, and traditional and cyber peer aggression and victimization as measured by self report Hypothesis 2a. It was hypothesized that greater self reported aggressive behaviors (traditional and cyber) would be significantly associated with lower s ocial functioning composite scores, derived from self and teacher report Hypothesis 2b. It wa s hypothesized that greater self reported peer victimization (traditional and cyber) would be significantly associated with lower social functioning composite score s, derived from self and teacher report Aim Three The third aim of this study examined how t raditional and c yber v ictimization were related to anxiety using a standardized measure encompassing a wide range of anxiety symptoms (Multidimensional Anxie ty Scale for Children, MASC), as existing research has indicated a positive relationship between anxiety and victimization.
39 Hypothesis 3a. It was hypothesized that greater self reported traditional v ictim ization would be significantly positively associate d with greater self report ed symptoms from the Anxiety Disorders Index as well as total P hysical S ymptoms of A nxiety and total S ocial A nxiety from the MASC Hypothesis 3 b It was also hypothesized that greater self reported experiences of cyber victimiza tion would be significantly positively associated with greater self reported symptoms from the Anxiety Disorders Index scale, as well as total P hysical S ymptoms of A nxiety and total S ocial A nxiety from the MASC Hypothesis 3 c Extant literature has established a moderating effect of gender between peer victimization and anxiety such that females report more symptoms of anxiety than males. Therefore, it was hypothesized that gender would moderate the relationship between peer vic timization and the anxiety variables such that the relationship between peer victimization and the anxiety variables would be stronger for females than males Aim Four There is a dearth of research focusing on the association between cyber aggression/vict imization and academic functioning, school absences, suspensions, and disciplinary infractions. Thus, the fourth aim of this study assess ed the relationship between self reported traditional and cyber peer victimization and aggression and school functionin g as measured by self reported GPA and number of suspensions and referrals Hypothesis 4a. Research supports that overtly aggressive youth tend to demonstrate lower achievement compared to nonaggressive peers (DeRosier & Lloyd).
40 Thus, i t was hypothesized that greater self reported physical peer aggression would be significantly associated with poorer school functioning. Hypothesis 4b. As cyber aggression is more relational in nature, it was hypothesized that self reported cyber aggression would not be sig nificantly related to measures o f school functioning Hypothesis 4c. It was hypothesized that greater self reported traditional and cyber victimization would be significantly associated with poorer school functioning as measured by GPA, but would not be significantly associated with number of suspensions and referrals
41 CHAPTER 2 METHODS Participants Students enrolled in grades seventh and eighth were recruited from a public middle school in the Columbia County School District, a rural county in North Central Florida, to participate in the study. There were no exclusionary criteria; the children of parents who provided active consent were eligible to participate in the study. A total of 693 seventh and eighth graders were eligible to participate in this study based on numbers from class rosters obtained from the school just prior to the study Consent forms were provided for each classroom in the seventh and eight h grades. Out of 693 eligible students, 274 students (39.5%) returned consent forms. Out of the 274 consent forms returned, 241 (88% consent rate) of them indicated consent to participate in the study (137 seventh and 104 eighth graders), while 33 of the r eturned forms declined to participate in the study (19 seventh graders and 14 eighth graders). Of the 241 whose parents provided consent to participate, 185 students participated on the day of the survey; reasons that consented students did not participate were primarily due to being absent on the day of survey administration (it was right after a holiday weekend), although a few declined assent for the study. Specific numbers could not be calculated in each of these categories, as the researchers did not a sk teachers to consistently note how many consented students in their class were absent or declined assent. Seventh and eighth grade teachers were also recruited to participate in the survey. Thirty teachers were eligible to participate. Of those teachers eligible, 19 teachers (63.3%) actually participated the day of the survey. Several classrooms did not have a teacher
42 participate due to a substitute teacher being in the classroom on the day of survey administration. Measures Demographic Information Demogr aphic information pertaining to grade placement, ethnicity, grades, collected for each student ( see Appendix A for demographic form). The demographic form was the secon d sheet in each packet, behind the assent form, and was completed prior to the completion of the other study measures. Please refer to Table 2 1 for the demographic information for this study. Self Report of Social Functioning The Piers Self Concept Scale Second Edition (Piers Harris 2) 18) self perception of their behaviors and attitudes (Piers & Herzberg, 2002) Youth are asked to endorse yes or no to Harris 2 has a total of concept in six domains. Domain scale scores and a total scale score can be derived from the measure by converting the raw scores into T scores on th e profile sheet the measure provides. The domain scales within the measure include: Behavioral Adjustment, Intellectual and School Status, Physical Appearance and Attributes, Freedom from Anxiety, Popularity, and Happiness and Satisfaction. The Popularity scale was the focus of this study. The Popularity scale of the Piers is performing soc ially ( Piers & Herzberg, 2002) .The mean T Score for the Popularity
43 scale for the southern region of the United States is 50.7. Please refer to Table 2 2 for the mean T scores and standard deviations for the current sample. The Piers Harris 2 is a cultural ly diverse measure with its psychometric properties being tested with significant ethnic populations such as African Americans and Hispanics (Piers & Herzberg, 2002). Alpha coefficients for the domain scales range from .60 to .84, and the Total Self Conce pt scale coefficients range from .89 to .93, indicating a moderate to strong internal consistency (Piers & Herzberg, 2002). The For the current study, the alpha coefficient for t he Popularity scale was .82, which is similar to the alpha range stated in the manual. While test retest reliability coefficients are available for the original Piers Harris measure this data is not available for the second edition Anxiety The Multidi mensional Anxiety Scale for Children (MASC) was developed by John March in 1997 and was used in this study to measure youth report of anxiety (March Parker, Sullivan, Stallings, & Conners 1997). The MASC is a 39 item self report measure assessing anxiety symptoms on a likert scale (0= never true about me 1= rarely true about me 2= sometimes true about me 3= often true about me ) in children eight to nineteen years of age. This measure consists of four scales and three indexes, which have been identified thr ough factor analyses: Physical Symptoms of Anxiety scale (12 items), Social Anxiety scale (9 items), Harm Avoidance scale (9 items), Separation/Panic scale (9 items), Anxiety Disorders Index, Total Anxiety Index, and Inconsistency Index. Three of the six a nxiety scales ( Physical Symptoms scale, Social Anxiety scale, and Anxiety Disorders Index ) were used for the purposes of this study.
44 These scales were selected for specific reasons. The Physical Symptoms scale was chosen as it measures symptoms such as fee ling dizzy or sick to your stomach which are common anxiety symptoms youth often endorse experiencing T he Social Anxiety scale, which assesses worries regarding public performance and being humiliated are also symptoms frequently report ed by youth Addi tionally, symptoms of social anxiety m ight logically be associated with exper iences of peer victimization. Lastly, the Anxiety Disorders Index was chosen, as high scores on this scale differentiated youth who would meet criteria for an anxiety disorder, ba sed on their responses, from youth who would not meet criteria. The MASC is a culturally diverse measure with its psychometric properties being tested in countries such as Taiwan and Iceland. According to a community based sample of children, the Cronbach of the measure range from .73 to .89 indicating a moderate to strong internal consistency (Baldwin & Dadds, 2007). Additionally, this research indicated that the measure has good test retest reliability with average correlation coefficients for a three week and three month period of .79 and .93 respectively. Further, the MASC is moderately correlated with the all the subscales indicating that fema les endorse greater anxiety on all scales and ind ex es compared to males (Baldwin & Dadds, 2007) thus there are separate gender norms Within the current study, the alpha coefficients were similar to previous studies with the exception of the Anxiety Diso rders Index, indicating that the subscales from the measure performed as expected with regards to the total Social Anxiety and the total
45 for the total Physical Symptoms of Anxiety scale .86 for the Social Anxiety s cale, and .67 for the Anxiety Disorders Index. Self Report of Peer Aggression and Victimization The Adolescent Peer Relations Instrument (APRI ) was developed by Robert Parada, Ph.D. at the University of Western seventeen) self perception of their behaviors within bullying and victimization domains. Youth are asked to endorse a response on a six point likert scale ranging from 1) never to 6) ever y day The APRI is a thirty s ix item questionnaire that evaluates three types of peer aggression and victimization (physical, social, and verbal). Six subscale and two total scale scores can be derived from the measure by summing the items for each subscale and total score. The subsca les (six items each) within the measure include Physical, Social, and Verbal Bullying as well as Physical, Social, and Verbal Victimization (hereafter referred to as physical, social, and verbal aggression, and physical, social, and verbal victimization), and the total scores are total Bully and total Victim. Youth who endorse a score of eighteen for either of the bullying or victimization total scores are interpreted as never having experienced bullying or engaged in bullying others. The APRI has been corr elated with constructs of youth functioning such as depression, anger management, and self concept. In particular, it was noted that verbal and social victimization are factors most associated with depression, aggressors are most associated with externaliz ed anger, and both aggressors and victims lack good self concepts (Mar s h, Nagengast Morin Parada, Craven, & Hamilton, 2011). Strong alpha coefficients were indicated for this measure, with high internal consistency for the subscales ranging from 0.81 t o 0.89 when used with upper
46 elementary students and 0.82 to 0.93 for an adolescent population (Finger, Yeung, Craven, Parada, & Newey, 2008; March, et al., 2011). The t otal Bully and total Victim 93 and 0.94, the total Bully and total Victim scale were .94 and .97 respectively. Alpha coefficients for the subscales in this study were: physical aggression scal e .82, social aggression scale .90, verbal aggression scale .89, physical victimization scale .93, social victimization scale .90, and verbal victimization scale .94. The alpha coefficients for the sample in this study are commensurate with previous studie s. Sociometric Teacher Report of Students Social Functioning Teachers completed ratings of how much each student in their classroom is liked or disliked by their peers (Appendix B ). Ratings were made on a 7 point scale from 1) very much disliked to 7) very much liked. The raw score obtained from teacher ratings was converted into a T score to create a social functioning score For this study, a social functioning composite was created by averaging the T scores derived from the self report of the Piers H social functioning (De Rosier & Lloyd, 2010). aggression and victimization status. Teachers were asked to in dicate whether youths within their first period classroom engaged in physical, verbal, or relational aggression, as well as whether they were the victim of any of these types of peer aggression. For data analyses, students who we re reported as being an agg ressor or a victim, respectively, w ere assigned a one (reported as being a bully or victim, respectively) or zero (not reported as being a bully or victim, respectively).
47 Cyber Aggression and Victimization Measure Due to the lack of availability of compre hensive cyber aggression and victimization measures, a measure was adapted for the current study from one created by Janicke and colleagues (D. Janicke, personal communication, November 14, 2012) to examine the relationship between cyber aggression/victimi zation, child weight status, and psychosocial functioning. The Cyber Aggression and Victimization measure (Appendix C ) obtains information cyber world in four different areas. Research conducted on this me asure indicated that victims of cyber aggression experienced more parent reported internalizing symptoms than youth who were not victimized. Further, those who endorsed cyber aggression indicated higher levels of self reported depression and were more like ly to endorse engaging in traditional aggression compared to non perpetrating peers. The measure is comprised of four parts. The first part of the measure requests information regarding the frequency of use of varying electronic and internet media formats To illustrate, peers are asked to rate their frequency of use within the past month of email, text messaging, and social networking sites on a frequency scale ranging from 1) never to 4) every day Part two of the measure requests youth to endorse the fr equency with which they have experienced threatening or embarrassing events in electronic or online formats on a scale ranging from 1) never to 4) several times a week Next, the youth is asked to identify how often they have experienced consequences due t o threatening or embarrassing situations on a scale of 1) never to 4) always The last part of the measure asks the youth to report the aggressive activity he or she has engaged in via electronic or internet formats on a frequency scale ranging from 1) ne ver to 4) several times a week Total Cyber Aggression and total Cyber
48 Victimization scores can be derived from the measure by summing items asking about aggression and victimization, respectively. For the purposes of the current study, several items (ite m numbers 9, 10, and 38 ) were added to this measure to further assess youth habits regarding use of electronic media and whether they have engaged in cyber aggression using their true identity or anonymously. The following is one example of the items that were added to the measure, H ow much do your parents monitor your use of social networking sites, Janicke and colleagues (D. Janicke, personal communication, November 14, 2012) did alphas for the subscales from the measure due to all alphas falling under .60. For the current sample strong alpha coefficients were indicated for subscales from this measure that were the focus of the current study. The Cyber Aggression scale (items 31 t Cyber Victimization scale (items 11 to 18) had an alpha of .93. This indicates that the measure exceeded expectations with this sample compared to the previous samples. One potential reason for better internal consistency in the current sample is differences in age ranges of study participants. While the current study focused on seventh and eighth graders, prior work with this scale included children as young as age 8, who may use social networking sit es and cell phone text messaging less than older youth. In addition, the current study included students in the age range where peer victimization is at its peak. School Functioning Academic and school functioning was assessed via self report of grades a s well as number of behavior referrals and suspensions received, respectively. On the
49 demographic form, students reported their current grades (A, B, C, D, or F) in core classes (math, language arts, science/physical science, and civics/US history). These reported grades were transformed into an overall GPA for core classes by assigning each grade a numerical value (i.e., A=4, B=3, C=2, D=1, and F=0), summing those numerical values across the four core classes, and then obtaining the mean value across class es. Students also indicated how many suspensions and referrals they received during the course of the school year. Behavior referrals and suspensions were summed for each student for use in analyses. Procedures A pproval from the Institutional Review Board (IRB 02 ) was obtained prior to conducting the study Following approval, students in grades seven and eight were recruited from a public middle school (the middle school with the largest enrollment) in the Columbia County School District that had exhibited interest in learning more about the prevalence of traditional and cyber aggression and victimization amongst their students and how the school could best address these behaviors Since this project was conducted in collaboration with the school, some spec ific items of interest to the school were included in the measures that the students complete d although these items were not a focus of analyses for this study. Active consent was needed in order for the student to participate in the study. A consent lett er detailing the purpose of the study (Appendix D ) was sent home with each student to present to their parents or guardians P arents either indicated that they did or did not want their child to participate in the survey, and the form was returned to schoo l and given to the first period teachers. The teachers sent the consent forms to the front office on the identified day where a graduate student was waiting to collect them. The classrooms that had 100% return rate
50 for consent forms, regardless of whether the parent(s)/guardians consented or did not consent for the child to participate in the study, receive d a doughnut/bagel party after the study was complete. Only one classroom qualified for the doughnut/bagel party. Study surveys were administered on a da te agreed upon with the school during the first class of the day in late May 2012 ; survey administration within participating classrooms was supervised by teachers, and the researchers were on site for consultation if needed. Teachers were provided with a packet for each student participating in the study in their class. Each packet contain ed the measures the student was to complete and return to the teacher as well as a student assent form (see Appendix E ) The student assent form was at the top of each student packet. The order of the questionnaires was counterbalanced in each packet, such that the initial questionnaire after the assent form and demographic form varied among students to help control for orde r effects and questionnaire fatigue. On the packet were instructions delineating what each student was to complete and how it was to be completed. Additionally, there were instructions regarding the assent form. Before students began completing the packet, they either assented or declined assent to the study. Those students who declined to participate were to return the packet to the teacher. Teachers were instructed to notify students whose parents consented to the study that there would not be any discipl inary sanctions or impacts on their grade if they d id not participate. Each student packet was pre assigned an ID number (which link ed the forms together) and was passed out to the students the day of the survey. The ID numbers we re completely arbitrary a nd were not able to be
51 Students were provided with a blank cover sheet, allowing them to cover their responses as they complete d items. They were told that their responses would be anonymous and would not be able to be identif ied. Following assent to the student, s tudents completed the demographic form first and then continued with the survey While the students were completing their questionnaires, the teachers were also completing a teacher report questionnaire about each st Teachers were provided with a consent form to read (Appendix F ); by completing questionnaires they were consenting for the study. Due to this survey being anonymous, each student packet had a sticky note attached to the front co ntaining the packet ID number. While students were completing the survey, teachers collect ed these sticky notes one at a time. The teacher wrote the ir survey form, complete d the survey, and place d the sticky note for that student on top of the survey form. Then, the teacher follow ed the same procedure for all students participating in the survey, completing one survey at a time. Teachers were instructed not to write student names on any forms. Completion time for t he teacher report measures was estimated to take 15 25 minutes to complete. If the teacher decline d to participate, then they simply left their survey forms blank and return ed them to the front office with the student surveys, when they were complete. Teac hers who did participate receive d a five dollar gift card to Starbucks following completion of the study. Completed m easures were placed in a manila envelope. The researchers were on site if any questions or concerns ar o se during the completion of the mea sures. No students reported experiencing distress during the completion of the study measures, and no significant problems were reported with survey administration. All classes
52 finished the survey within the first period and did not need additional time. O nce a class had completed the measures a student brought the sealed envelopes to the front office for the researchers to collect.
53 Table 2 1. Sample demographic characteristics Variable Percent of sample ( N =185) Gender Males 44.9 Females 55.1 Ethnicity Caucasian 66.8 African American Hispanic Other 19.6 4.9 8.6 GPA 4.0 3.0 3.9 2.0 2.9 1.0 1.9 0.0 0.9 16.4 35.0 32.2 13.6 2.8 Suspensions Yes 8.7 No Number of Suspensions None 1 2 3 4 5 6 or more Referrals 91.3 90.8 4. 9 1.6 1.6 1.0 None 76. 8 1 2 3 4 5 6 or more 1 1.4 8. 6 1.0 2.1 Who Child Lives With* Mother 88.1 Father 63.2 Grandparents 10.8 Relative 14.1 Step mother Step father 5.9 12.4 Percentages listed may total more than 100% as students were allowed to select more than one option.
54 Table 2 2. Total sample mean and standard deviations of measures used in study Variable Mean SD Range APRI Total Bully 24.16 9.95 18 108 Physical Bully 7.60 3.12 6 36 Social Bully Verbal Bully 7.46 3.34 6 36 9.08 4.32 6 36 APRI Tota l Victim 28.60 16.76 18 108 Physical Victim 8.50 .07 6 36 Social Victim Verbal Victim 9.54 5.96 6 36 11.05 7.17 6 36 Cyber Aggression .62 2.05 0 21 Cyber Victimization MASC Anxiety Disorders Index Total Physical Symptoms Total Social Anxiety 1.06 3.12 0 24 47.13 11.36 25 75 44.47 9.27 34 81 49.74 11.82 32 79 PH2 Popularity Scale GPA Total Suspensions and Referrals Social Functioning Composite 49.96 10.77 24 68 2.86 .92 .0 4.0 84 2. 6 1 0 20 49.88 8.87 18.8 64.4
55 CHAPTER 3 RESULTS All data analyses were conducted using the Statistical Package for the Social Sciences 20.0 (SPSS). Descriptive Analyses Descriptive analyses were conducted to determine the prevalence of traditional and cyber aggression and victimization behaviors within the sample. For the APRI measure a total score of 18 means no aggression or victimization ; anything over 18 for the total Bully and Victim scales indicates that the youth has engaged in aggression or been a victim of aggression at least some of the time. Subscale scores (physical and social aggression and victimization) on the APRI exceeding 6 indicate that youth have eithe r aggressed or been victimized at least sometimes. Mean T scores for both the total aggression and victimization scales, as well as the subscales, indicated that youth reported aggressing or experiencing victimization at least some of the time. Overall, 7 3.6 % of students reported engaging in some type of traditional aggression toward peers at least once this past school year on the APRI When examining those students who reported engaging in traditional aggression, 44. 3 % reported physical aggressing, 42.1% endorsed relational aggressing, and 69. 4 % endorsed verbal aggressing on peers. Of those aggressors, 73.4 % were male and 73.7 % female. See Table 3 1 for the breakdown of percentages of traditional and cyber aggression and victimization types by gender for the overall sample. Overall, 7 5.7 % of youth reported being traditionally victimized by peers ( 70.9 % male, 79.8 % female). With regards to type of victimization 43.1 % endorsed physical victimization, 61.8 % endorsed relational victimization, and 68.4 % repor ted experiencing verbal victimization. In total,
56 verbal aggression seemed the most prominent form of aggression and victimization endorsed by students on the APRI. For the Cyber Aggression and Victimization measure descriptive analyses revealed that 20.6% of students reported cyber aggressing on peers (11.4% males, 27.7% females), while 30.7% students endorsed experiencing cyber victimization (20.0% male, 39.4% female). W hen a cutoff criterion score of 24 was used for the Total Bully and Total Victim scales of the APRI measure in order to select students engaging in more frequent aggression or experiencing more frequent victim ization 36.5% of youth endorsed bullying others while 41 .0 % endorse d experiences of victimization When this cutoff criterion was app lied the prevalence of traditional aggression and victimization decreased and was just slightly higher than ranges reported in the literature for studies utilizing cutoff scores to identify bullies and victims It should be noted that the APRI measure was created as a continuous measure and does not have a set cutoff criteria for defining bullies and victims Preliminary Analyses of Variance (ANOVA) were conducted to examine gender and ethnic differences in self report measures that were the focus of analyses, in order to determine whether these demographic variables should be controlled for in subsequent analy ses. Presented in Table 3 2 are the overall mean T scores and standard deviations, as well as mean T Scores and standard deviations by gender, from the Multidimensional Anxiety Scale for Children (MASC) and Piers Self Concept Scale Second Edition (Piers Harris 2 ). Also presented are the overall mean total scores from the A dolescent P eer R elations I nstrument (APRI) and t he Cyber Aggre ssion and Victimization measure, as well as the mean total scores and standard
57 deviations by gender. Analyse s of Variance (ANOVA) revealed significant gender differences between males and females on the Social Anxiety scale of the MASC F (1, 182) = 10.13, p < .01, R 2 = .23 with more females reported experiencing social anxiety compared to males. Analyses did not reveal gender differences for the Physical Symptoms of Anxiety scale F (1, 181) = .403, p = .53, R 2 = .05, and Anxiety Disorders Index F (1, 182) = 49, p = .48, R 2 = .04 from the MASC or the Popularity Domain F (1, 178) = 3.38, p = .07, R 2 = .14 from the Piers Harris 2. Aim one of this study examined gender differences for various measures from the APRI and Cyber Aggression and Victimization measure (see Aim one findings below for details of these analyses) and revealed that there were gender dif ferences for some of these variables. Due the relations between gender and several study variables gender was entered as a demographic control variable in block one for all of the primary aims in this study utilizing hierarchical regressions. For the AN for the self report measures, results did not reveal significant differences between ethnicity groups for the MASC scales Physical Symptoms of Anxiety F ( 1,181 ) = 1. 10 p = 30 R 2 = 01 Social Anxiety F ( 1,181 ) = .46 p = 50 R 2 = 00 and Anxiety Disorders Index F ( 1,181 ) = .02 p = 88 R 2 = 00 as well as the Piers Harris Po p ularity Domain F ( 1,177 ) = .17 p = 68 R 2 = .00 I n addition, the scales from the APRI Total Bully [ F (1, 170) = 1.53, p = .22, R 2 = .09, Physical Bully F (1, 180) = 2.74, p = .10, R 2 = .12, Social Bully F (1, 180) = 1.89, p = .17, R 2 = .10, Total Victim F (1, 170) = .03, p = .87, R 2 = .00, Physical Victim F (1, 178) = .03, p = .87, R 2 = .00, and Social Victim F (1, 175) = .01, p = .94, R 2 = .00, as well as the Cyber Aggression F (1, 177) = 1.42 p = .24 R 2 = .09 and Cyber
58 Victimization F (1, 176) = 2.68, p = .10, R 2 = .12 scales from the Cyber Aggression and Victimization measure used in this study did not reveal significant differences between ethnicity groups T hus ethnicity was not included as a demographic control variable in the analyses for the primary aims of this study. As would be expected given that this is a community sample of adolescents, the mean T scores for the MASC (65 or higher) and the Piers Harris 2 (60 or higher or 39 or lower) did not fall in the Clinical range; however, the ranges indicate that some individua ls did report clinically significant symptoms Primary Analyses Prior to conducting the hierarchical regressions for primary analyses, normality testing was conducted revealing variables that were not normally distributed. Several corrections via transformations were conducted to correct the non parametric data. Following each transformation of normality the data was re checked. Despite these corrections the data continued to remain not normally distributed. Thus, bootstrapped hierarchical regressions were used for the primary aims of this study. Bootstrapping is an analysis wh ere robust estimates of standard errors are obtained. The standard error coefficient is the standard deviation of the mean when the mean is randomly drawn from the sample an infinite number of times. Further, it is an alternative to parametric estimates wh en the normality of the data is questioned. Due to the number of analyses conducted for the primary study aims presented below, there could be a higher likelihood that Type 1 errors could occur with less stringent significance levels (i.e., p < .05). Howe ver, given that there are few studies examining some of the relationships being examined in the current study (particularly
59 with regards to cyber aggression and victimization), p values of less than .05 were accepted as significant to help guard against mi ssing significant findings and inform directions for further research. Notably, findings with p values of less than .01 can likely be interpreted with greater confidence, whereas findings with p values of only less than .05 should be interpreted with more caution. Aim One Analyses 1 a: Gender differences in physical and relational aggression For this aim, two Analyses of Variance (ANOVA) were conducted comparing males and females on APRI scores for social and physical aggression, respectively, assessing wh ether females self reported more relational aggression than males and males self reported more physical aggression than females. Further, one Chi Square analysis was agg ression respectively, assessing whether teachers reported more females as relational aggressors and more males as physical aggressors on the Sociometric data were conducted on a subset of the overall sample where teacher data was available. According to the results, there was a significant difference between genders on the social aggression scale of the APRI, indicating that females were more likely to endorse relational aggr ession than males, F (1, 181) = 4.80, p < .05, R 2 = .16. However the effect size was small. For the physical aggression scores on the APRI, gender differences were not significant, indicating that males and females reported similar levels of physical aggres sion against peers, F (1, 181) = .0 1 p = .94, R 2 = .00.
60 For the teacher report measure the data for students who had teacher report measures completed (110 students) was examined. The results of this aim indicated that teachers significantly reported more males as physical aggressors compared to females ( 2 = 4.73 p < .05). Conversely, there were no gender differences between ( 2 = .41 p = .52). Analyses 1 b: Gender differences in physical and relational vict imization For this aim, two Analyses of Variance (ANOVA) were conducted comparing males and females on APRI scores for p hysical and s ocial v ictimization, respectively, assessing whether males self reported greater victimization from physical aggression an d females self reported greater relational victimization. For the p hysical v ictimization scores on the APRI there was not a significant difference between males and females, F (1, 179) = .15, p = .70, R 2 = .0 1 Conversely there was a significant gender difference on the APRI s ocial v ictimization scale, indicating that females were more likely to endorse experiencing relational victimization compared to males, F (1, 176) = 10.13 p < .01 R 2 = 00 Teachers identified very few students as victims o f aggression; therefore, Chi Square analyses were not conducted for the teacher report of student victimization data. Analyses 1 c: Gender differences in cyber aggression To assess wh e ther, due to some shared features between social and cyber aggression, f emales self reported more cyber aggression than males, an Analyses of Variance (ANOVA) was conducted comparing male and female reports of Cyber Aggression on the Cyber Aggression and Victimization measure. According to the results of the analysis, males a nd females did not differ significantly in their self report of cyber aggression, F (1, 178) = 2.55, p = .11, R 2 = .12.
61 Overall, for aim one females self reported more relational a ggression and victimization compared to males. For the physical aggression and v ictimization scales there were no significant differences between males and females. A nalyses also revealed no differences between males and females for c yber a ggression. Teachers reported more males as physically aggressive, while they did not report significant differences between males and females on relational aggression. Aim Two Analyses 2 a: Associations between traditional and cyber aggression and social functioning For this aim, two bootstrap hierarchical regression analyses were conducted te sting whether greater self reported aggressive behaviors (total traditional and Cyber Aggression) were significantly associated with lower social functioning composite scores (standardized mean of self and teacher report). Block one of each regression cons isted of gender as a demographic control variable. Total traditional aggression (APRI Bully) and Cyber Aggression (Cyber Aggression and Victimization measure) scores were entered into block two, respectively. All regression analyses statistics for this aim are presented in Tables 3 3 (total traditional aggression) and 3 4 (Cyber Aggression). The overall fit for the model examining total traditional aggression was not significant ( R 2 = .01, p = .42). In block one, gender was not a significant demographic pr edictor ( B = 6.51, p = .51). The addition of total traditional aggression in block two was not significant ( B = .34, p = .21). For the regression examining total Cyber Aggression the overall model was not significant ( R 2 = .02, p = .32). In block one, gender was not a significant demographic
62 predictor ( B = 6.89, p = .49). The addition of total Cyber Aggression in block two was not significant ( B = 1.99, p = .32). Analyses 2 b: Associations between traditional and cyber victimization and social functioning For this aim, two bootstrap hierarchical regression analyses were conducted testing whether greater self reported peer victimization (traditional and cyber) was significantly associated w ith lower social functioning composite scores (standardized mean of self and teacher report). Block one of each regression consisted of gender as the demographic control variable. Total traditional victimization from the APRI (APRI Victim) and Cyber Victim ization scores from the Cyber Aggression and Victimization measure were entered as predictors in block two, respectively. All regression analyses statistics for Aim 2b are presented in Tables 3 5 (traditional victimization) and 3 6 (Cyber Victimization). The overall model including total traditional victimization was not significant ( R 2 = .01, p = .46). In block one, gender was not a significant demographic predictor ( B = 7.17, p = .47). The addition of total traditional victimization scores from the APRI in block two was not significant ( B = .18, p = .40). For the analysis including Cyber Victimization, the overall model was significant ( R 2 = .10, p < .01) following the addition of the Cyber Victimization variable in block two. However, gender was not a significant demographic predictor ( B = .90, p = .60) in block one. Additionally, the beta weight for Cyber Victimization in block two was not significant ( B = .75, p = .08).
63 Overall analyses for aim two indicated that higher scores on the APRI total aggression and victimization scales, as well as on the Cyber Aggression and Cyber Victimization scales did not predict poorer social functioning. Aim Three For this aim, six bootstrap hierarchical regression analyses were conducted which included interaction terms. The results for Aims 3a and 3b were based on beta weights for the centered variables (APRI Victim scale, total Cyber Victimization) entered in block one of each regression analysis. Results for Aim 3c were based on the addition of the interaction term in block two of each regression analysis. For three of the regressions conducted, one for each of the dependent variables (Anxiety Disorders Index, total Physical Symptoms of Anxiety, and total Social Anxiety from t he MASC), a centered demographic variable and a centered total traditional victimization (APRI Victim scores) variable were entered in block one, while the gender by total traditional victimization interaction term was added in block two. All regression an alyses statistics are presented in Table 3 7 (Anxiety Disorders Index), 3 8 (total Physical Symptoms of Anxiety), and 3 9 (total Social Anxiety). The remaining three regressions were similar except that centered gender and centered Cyber Victimization scor es (from the Cyber Aggression and Victimization measure) were entered into block one, and the gender by Cyber Victimization interaction term was entered in block two. All regression analyses statistics are presented in Table 3 10 (Anxiety Disorders Index), 3 11 (total Physical Symptoms of Anxiety), and 3 12 (total Social Anxiety). Analyses 3 a: Associations between total traditional victimization and anxiety Results from block one of the regression analyses were examined to determine whether greater self re ported total traditional victimization was significantly
64 positively associated with self reported symptoms of anxiety disorders (MASC Anxiety Disorders Index), as well as total Physical Symptoms of Anxiety, and total Social Anxiety from the MASC. Of intere st in block one of each regression were the beta weights for the centered total traditional victimization score from the APRI. For the analysis predicting the Anxiety Disorders Index from the MASC, the beta weight for the centered total traditional victimi zation predictor in block one was significant ( B = .37 p < .01). As hypothesized, the relationship between total traditional victimization and the Anxiety Disorders Index of the MASC was positive, indicating that as traditional victimization scores on the APRI increased, scores on the Anxiety Disorders Index increased (indicating more anxiety). For the analysis predicting total Physical Symptoms of Anxiety, the beta weight for the centered total traditional victimization predictor in block one was signifi cant ( B = .36, p < .01). As hypothesized, the relationship between the total traditional victimization and the Physical Symptoms of Anxiety scale of the MASC was positive, indicating that as traditional victimization scores on the APRI increased, Physical Symptoms of Anxiety increased (indicating more anxiety). For the analysis predicting total Social Anxiety, the beta weight for the centered total traditional victimization predictor in block one was significant ( B = .39, p < .01). As hypothesized, the relationship between total traditional victimization and the Social Anxiety scale of the MASC was positive, indicating that as traditional victimization scores on the APRI increased, Social Anxiety increased (indicating more anx iety). Analyses 3 b: Associations between cyber victimization and anxiety Results from block one of the regression analyses were examined to determine whether greater
65 self reported experiences of cyber victimization were significantly associated with great er self reported anxiety disorder symptoms (MASC Anxiety Disorders Index), as well as total Physical Symptoms of Anxiety, and total Social Anxiety from the MASC. Of interest in block one of each regression were the beta weights for the centered total Cyber Victimization score from the Cyber Aggression and Victimization measure. For the analysis predicting the Anxiety Disorders Index from the MASC, the beta weight for the centered Cyber Victimization predictor in block one was not significant ( B = .60, p = .28). For the analysis predicting total Physical Symptoms of Anxiety, the beta weight for the centered Cyber Victimization predictor in block one was significant ( B = 1.28, p < .01). As hypothesized, the relationship between the Cyber Victimization scale and total Physical Symptoms of Anxiety was positive, indicating that higher Cyber Victimization scores were associated with higher scores on P hysical S ymptoms of A nxiety (more anxiety). For the analysis predicting social anxiety, the beta weight for the c entered C yber V ictimization predictor in block one was significant ( B = .90, p < .05) As hypothesized, t he C yber V ictimization scale was positively associated with t otal S ocial A nxiety such that higher scores on the Cyber Victimization scale were related to higher scores on total S ocial A nxiety ( greater anxiety). Analyses 3 c: Examining gender as a moderator of the relationship between peer victimization and anxiety For this aim, the results were extracted from block two of the six bootstrap hierarchical regression analyses, which contained the interaction terms testing whether gender moderated the relationship between peer victimization
66 and anxiety such that the rela tionship between peer victimization and anxiety was stronger for females than males. For each analysis the significance of R 2 change from block one to block two indicating the additional variance explained by the interaction term, will be discussed. Overa ll model R 2 values can be found in the tables. See Table 3 7 for reg ression statistics for the analysis including total traditional victimization and the dependent variable Anxiety Disorders Index from the MASC. For this analysis, the R 2 change for block tw o, with the addition of the interaction term, was not significant ( R 2 = 01 p = .19). In block two, the interaction term gender by total traditional victimization was not significant ( B = .13, p = .12) which did not support the hypothesis of gender as a moderator. In addition, the centered gender term was also not a significant predictor in block one or two ( B = 1.13, p = .46 and B = 1.29, p = .40 respectively). See Table 3 8 f or regression statistics for the analysis including total traditional victimization and the dependent variable total Physical Symptoms of Anxiety from the MASC. For this analysis, the R 2 change for block two, with the addition of the interaction term, was not significant ( R 2 = 01 p = .23). The addition of the interaction term gender by traditional victimization in block 2 was not significant ( B = .09, p = .30), which did not support the hypothesis of gender as a moderator. The centered gender variable w as also not significant in block one or two ( B = .79, p = .56 and B = .90, p = .52, respectively). See Table 3 9 for regression statistics for the analysis including total traditional victimization and the dependent variable total Social Anxiety from t he MASC. For this analysis, the R 2 change for block two, with the addition of the interaction term, was not
67 significant ( R 2 = 01 p = .08). In block two, the gender by total traditional victimization interaction term was not significant ( B = .18, p = .05), which did not support the hypothesis of gender as a moderator. Conversely, the centered gender term in block one and two was a significant predictor ( B = 3.73, p < .05 and B = 3.51, p < .05 ). The relationship between gender and the dependent vari able was positive, indicating that females reported a higher level of social anxiety compared to males. See Table 3 10 f or regression statistics for the analysis examining total Cyber Victimization and the dependent variable Anxiety Disorders Index from th e MASC. For this analysis, the R 2 change for block two, with the addition of the interaction term, was not significant ( R 2 = .0 1 p = .19). In block two, the interaction term gender by Cyber Victimization was not significant ( B = .89, p = .35), which did not support the hypothesis of gender as a moderator. Additionally, the centered gender term in block one and block two was not a significant predictor ( B = .87, p = .62 and B = .96, p = .62, respectively). See Table 3 11 for regression statistics for the analysis examining total Cyber Victimization and the dependent variable total Physical Symptoms of Anxiety from the MASC. For this analysis, the R 2 change for block two, with the addition of the interaction term, was significant ( R 2 = 03 p < .05). However, the beta weight for the interaction term gender by Cyber Victimization in block two fell short of significan ce ( B = 1.25, p = .09), which did not support the hypothesis of gender as a moderator. T he cent ered gender term w as not a significant predictor in block one or two ( B = .16, p = .92 and B = .29, p = .84, respectively). See Table 3 12 for regression statistics for the analysis examining total Cyber Victimization and the dependent variable total Soc ial Anxiety from the MASC. For this
68 analysis, the R 2 change for block two, with the addition of the interaction term, was not significant ( R 2 = 00 p = .3 7 ). The addition of the interaction term gender by Cyber Victimization in block two was not signific ant ( B = .61, p = .41), which did not support the hypothesis of gender as a moderator. Conversely, the centered gender term was a significant predictor in block one and two ( B = 5.16, p < .05 and B = 5.22, p < .05 respectively ). The relationship between gender and total Social Anxiety was positive. Females were more likely to endorse social anxiety than males. Overall, r esults for aim three indicated that youth who reported higher rates of traditional and cyber victimization also reported high levels of a nxiety. Further, females reported experiencing significantly more social anxiety compared to males. In each regression the interaction term was not significant, indicating that gender did not moderate the relationship between total and cyber victimization and each of the dependent anxiety variables. Aim Four Analyses 4 a: Associations between physical aggression and school functioning Two bootstrap hierarchical regression analyses one for each of the dependent variables (GPA of current core classes and the total number of referrals and suspensions combined), were conducted testing whether greater self reported physical aggression was significantly associated with poorer school functioning. In each of the two bootstrap hierarchical regressions, gender was en tered into block one while the APRI p hysical a ggression scores were entered into block two. The regression analyses statistics are presented in Table 3 13 (GPA) and 3 14 (total number of referrals and suspensions combined).
69 For the analysis predicting GPA the overall model was not significant ( R 2 = .02, p = .46) Gender was not a significant predictor ( B = .21, p = .1 5 ) in block one. The addition of the physical aggression scores in block two was also not significant ( B = .02, p = .5 5 ), which did not support the hypothesis. For the analysis predicting total number of suspensions and referrals combined, the overall model was significant ( R 2 = .11, p < .0 01 ). Gender was not a significant predictor ( B = .3 4 p = 39 ) in block one. T he addition of the phy sical aggression predictor in block two fell just short of significan ce ( B = .26 p = .0 6 ) Analyses 4 b: Associations between cyber aggression and school functioning Two bootstrap hierarchical regression analyses, one for each of the dependent variables (GPA of current core classes and the total number of referrals and suspensions combined), were conducted testing whether greater self reported cyber aggression was significantly related to school functioning, with the prediction that they would not be rela ted. Gender was entered into block one of the analyses, while the total Cyber Aggression variable was entered into block two. The regression analyses statistics are presented in Table 3 15 (GPA) and 3 16 (total number of suspensions and referrals combined) For the analysis predicting GPA, the overall fit of the model was just short of significan ce ( R 2 = .04, p = .05). Gender was not a significant predictor in block one ( B =.21, p = .13 ) The addition of the Cyber Aggression predictor to block two however, was significant ( B = .07, p < .05), resulting in 3.5% of the variance being explained. The relationship between cyber aggression and GPA was negative. Lower Cyber
70 were less likely to engage in cyber aggression, which was contrary to the hypothesis that these variables would not be related. However, this finding should be interpreted with caution given that the overall model was just short of significance. For the an alysis predicting total number of suspensions and referrals, the overall fit of the model was significant ( R 2 = .1 4 p < .001). Gender was not a significant predictor ( B = 53 p = 22 ) in block one. The addition of the Cyber Aggression predictor to block two was significant ( B = 47 p < .05), resulting in 1 4 2 % of the variance being explained. There was a positive relationship between the cyber aggression predictor and the dependent variable, indicating that as reports of cyber aggression increased, so did the total number of suspensions and referrals. Therefore, youth who reported engaging in more cyber aggression also reported receiving more suspension s and referrals combined which was contrary to the hypothes is that these variables would not be related Analyses 4 c: Associations between traditional and cyber victimization and school functioning Four bootstrap hierarchical regressions, one for each of the dependent variables (GPA of current core classes and t he total number of referrals and suspensions combined), were conducted testing whether greater self reported total traditional and Cyber Victimization, respectively, were significantly associated with poorer school functioning as measured by GPA, but not s ignificantly associated with the total number of suspensions and referrals. The first two bootstrap hierarchical regression analyses contained gender in block one, while the total traditional victimization scale from the APRI (total Victim) was entered int o block two. For the two remaining bootstrap hierarchical regressions, gender was entered into block one, while
71 total Cyber Victimization was entered into block two. The regression analyses statistics are presented in Tables 3 17 (GPA with traditional vict imization as the predictor), 3 18 (total number of suspensions and referrals with traditional victimization as the predictor), 3 19 (GPA with cyber victimization as the predictor), and 3 20 (total number of suspensions and referrals with cyber victimizatio n as the predictor). For the analysis including total traditional victimization predicting GPA, the overall fit of the model was significant ( R 2 = .04, p < .05). Gender was not a significant predictor ( B =.20, p = .1 6 ) in block one. The addition of the to tal traditional victimization scale was significant in block two ( B = .01, p < .05), resulting in 3.6% of the variance being explained. As hypothesized, the relationship between the total traditional victimization scale and GPA was negative indicating tha t as self reports of traditional victimization increased, GPA decreased. For the analysis including total traditional victimization predicting total number of suspensions and referrals, the overall fit of the model was significant ( R 2 = 10 p < .001). Gender was not a significant predictor ( B = 33 p = 44 ) in block one. The addition of the total traditional victimization score in block two also was not significant ( B = .0 5 p = 10 ) For the analysis including Cyber Victimization predicting GPA, the overall fit of the model was not significant ( R 2 = .03, p = .09). Gender was not a significant predictor ( B = .22, p = .1 1 ) in block one. Similarly, the addition of the Cyber Victimization scale in block two did not result in significance ( B = .0 4, p = .06). For the analysis including Cyber Victimization predicting total number of suspensions and referrals combined the overall fit of the model was significant ( R 2
72 =. 17 p < .001). Gender was not a significant predictor ( B = 56 p = 19 ) in bloc k one. The addition of the Cyber Victimization score in block two was also not a significant predictor ( B = .3 3 p = .05) although it fell just short of significance which would provide some evidence against the hypothesis that these variables would not be related However, this must be interpreted with caution given the fact that it fell short of the .05 significance level. Overall analyses for aim four indicated that physical aggression was not significantly associated with poorer school functioning. However, greater self reported cyber aggression predicted GPA. Further, youth who reported engaging in more cyber aggression also reported receiving more suspensions and referrals combined. Regarding victimization variables, self reported traditional victi mization predicted poorer academic functioning as measured by GPA but not the total number of suspensions and referrals. C yber v ictimization was not a significant predictor of school functioning as measured by GPA and the total numbe r of suspensions and re ferrals.
73 Table 3 1. Percent of sample experiencing aggression and victimization at least once this year by gender Variable N % APRI Total Aggression Male Female Physical Aggression Male Female Social Aggression Male Female Verbal Aggression Male Female APRI Total Victimization Male Female Physical Victimization Male Female Social Victimization Male Female Verbal Victimization Male Female Cyber Aggression Male Female Cyber Victimization Male Female 131 73.6 58 73.4 73 73.7 81 44.3 41 50.6 40 39.2 77 42.1 23 28.4 54 52.9 127 69.4 54 66.7 73 71.6 1 31 75.7 56 70.9 75 79.8 78 43.1 35 43.8 43 42.6 110 6 1 8 39 49.4 71 71.7 121 68.4 49 61.3 72 74.2 37 2 0 .6 9 11.4 28 27. 7 55 30.7 16 20.0 39 39.4
74 Table 3 2. Means and standard deviations of study measures by gender Male Female Variable M SD M SD F p R 2 MASC Anxiety Disorders Index 46.48 11.58 47.66 11.21 49 48 .0 4 MASC Physical Symptoms 43.99 9.51 44.86 9.10 .40 .53 .05 MASC Social Anxiety 46.76** 11.18 52.20 ** 11.82 10.13 .0 0 .23 PH2 Popularity Scale 51.60 10.83 48.65 10.60 3.38 .07 .14 APRI Total Bully 23.03 6.84 25.06 11.82 1.85 .18 .10 APRI Physical Bully 7.58 2.22 7.62 3.69 .01 .94 .00 APRI Social Bully 6.86* 2.47 7.94* 3.84 4.80 .03 .16 APRI Verbal Bully 8.52 3.32 9.53 4.94 2.50 .12 .12 APRI Total Victim 25.87 13.79 30.88 18.66 3.90 .05 .15 APRI Physical Victim 8.24 4.32 8.53 5.61 .15 .70 .01 APRI Social Victim 8.03** 4.54 10.76** 6.66 10.13 .00 ** .00 APRI Verbal Victim 9.85* 6.08 12.04* 7.86 4.16 .04 .15 Cyber Aggression .34 1.30 .83 2.47 2.55 .11 .12 Cyber Victimization .76 3.01 1.30 3.15 1.33 .25 .09 Note. Significant p values refer to differences between gender. p < .05 ** p < .01.
75 Table 3 3. Summary of regression statistics for aim 2 a: Association between traditional aggression and s o cial f unctioning Variable B SE R 2 R 2 change Block 1 .01 .01 Gender 6.51 8. 0 5 Block 2 .01 .01 Gender 6.86 8.12 APRI Total Traditional Aggression .34 .15 Note : No values were significant. Table 3 4. Summary of regression statistics for aim 2 a: Association between cyber aggression and s o cial f un ctioning Variable B SE R 2 R 2 change Block 1 .01 .01 Gender 6.89 8.36 Block 2 .02 .01 Gender 7.71 8.61 Cyber Aggression 1.99 1.62 Note : No values were significant. Table 3 5. Summary of regression statistics for aim 2 b : Association between traditional victimization and s o cial f unctioning Variable B SE R 2 R 2 change Block 1 .01 .01 Gender 7.17 8.58 Block 2 .01 .01 Gender 7.99 8.19 APRI Total Traditional Victimization .18 .17 Note : No values were significant.
76 Table 3 6. Summary of regression statistics for aim 2 b : Association between cyber victimization and s o cial f unctioning Variable B SE R 2 R 2 change Block 1 .00 .00 Gender .90 1.71 Block 2 .10** .10** Gender .56 1.71 Cyber Victimization .75 .53 ** p < .01 Table 3 7. Summary of regression statistics for aim 3 a and 3 c: Anxiety Disorders Index from the MASC Variable B SE R 2 R 2 change Block 1 .26*** .26*** Gender Total Traditional Victimization 1.13 1.48 .37** .05** Block 2 .27 .01 Gender 1.29 1.50 Total Traditional Victimization Interaction (gender by Total Traditional Victimization) .39** .05** .13 .10 Note. Variables were centered prior to ana lyses ** p < .01 *** p < .001 Table 3 8. Summary of regression statistics for aim 3 a and 3 c: Total Physical Symptoms of Anxiety from the MASC Variable B SE R 2 R 2 change Block 1 .37*** .37*** Gender Total Traditional Victimization .79 1.18 .36** .04** Block 2 .38 .01 Gender .90 1.25 Total Traditional Victimization Interaction (gender by Total Traditional Victimization) .37** .04** .09 .09 Note. Variables were centered prior to analyses. ** p < .01 *** p < .001
77 Table 3 9. Summary of regression statistics for aim 3 a and 3 c: Total Social Anxiety from the MASC Variable B SE R 2 R 2 change Block 1 .31*** .31*** Gender Total Traditional Victimization 3.73* 1.51* .39** .04** Block 2 .32 .01 Gender 3.51* 1.52 Total Traditional Victimization Interaction (gender by Total Traditional Victimizatio n) .41** .05** .18 10 Note. Variables were centered prior to analyses p < .05 ** p < .01 *** p < .001 Table 3 10. Summary of regression statistics for aim 3 b and 3 c: Anxiety Disorders Index from the MASC Variable B SE R 2 R 2 change Block 1 .02 .02 Gender Cyber Victimization .87 1.75 .60 .62 Block 2 .03 .01 Gender .96 2.03 Cyber Victimization Interaction (gender by Cyber Victimization) .76 .71 .89 1.58 Note. Variables were centered prior to analyses. No values were significant. Table 3 11. Summary of regression statistics for aim 3 b and 3 c: Total Physical Symptoms of Anxiety from the MASC Variable B SE R 2 R 2 change Block 1 .13*** .13*** Gender Cyber Victimization .16 1.39 1.28** .54** Block 2 .16* .03* Gender .29 1.75 Cyber Victimization Interaction ( G ender by Cyber Victimization) 1.49** .68** 1.25 1.48 Note. Variables were centered prior to analyses ** p < .01 *** p < .001
78 Table 3 12. Summary of regression statistics for aim 3 b and 3 c: Total Social Anxiety from the MASC Variable B SE R 2 R 2 change Block 1 .09*** .09*** Gender Cyber Victimization 5.16* 1.71* .90* .49* Block 2 .10 .00 Gender 5.22* 2.07* Cyber Victimization Interaction (gender by Cyber Victimization) 1.00 .76 .61 1.63 Note. Variables were centered prior to analyses p < .05 *** p < .0 01 Table 3 13. Summary of regression statistics for aim 4 a: GPA Variable B SE R 2 R 2 change Block 1 .01 .01 Gender .21 .14 Block 2 .02 .00 Gender .21 .14 Physical Aggression .02 .03 Note No values were significant. Table 3 14. Summary of regression statistics for aim 4 a: Total number of suspensions and referrals Variable B SE R 2 R 2 change Block 1 .0 1 .0 1 Gender 34 37 Block 2 11 ** 10 ** Gender 35 36 Physical Aggression .2 6 .1 2 *** p < .001
79 Table 3 15. Summary of regression statistics for aim 4 b: GPA Variable B SE R 2 R 2 change Block 1 .01 .01 Gender .21 .14 Block 2 .04 .02 Gender .24 .1 3 Cyber Aggression .07* .0 3 p < .05 Table 3 16. Summary of regression statistics for aim 4 b: Total number of suspensions and referrals Variable B SE R 2 R 2 change Block 1 .0 1 .0 1 Gender 53 4 1 Block 2 1 4 *** 13 *** Gender 76 39 Cyber Aggression 47 20 p < .05 *** p < .001 Table 3 17. Summary of regression statistics for aim 4 c: GPA Variable B SE R 2 R 2 change Block 1 .01 .01 Gender .20 .1 5 Block 2 .04* .02* Gender .24 .1 5 Total Traditional Victimization .01* .00* p < .05 Table 3 18. Summary of regression statistics for aim 4 c: Total number of suspensions and referrals Variable B SE R 2 R 2 change Block 1 .0 0 .0 0 Gender 33 39 Block 2 10 *** 09 *** Gender 56 3 5 Total Traditional Victimization .0 5 .0 2 *** p < .001
80 Table 3 19. Summary of regression statistics for aim 4 c: GPA Variable B SE R 2 R 2 change Block 1 .01 .01 Gender .22 .1 3 Block 2 .03 .02 Gender .24 .14 Cyber Victimization .04 .02 Note : No values were significant. Table 3 20. Summary of regression statistics for aim 4 c: Total number of suspensions and referrals Variable B SE R 2 R 2 change Block 1 .0 1 .0 1 Gender 56 41 Block 2 17 *** 15 *** Gender .74 39 Cyber Victimization .3 3 .1 5 *** p < .001
81 CHAPTER 4 DISCUSSION Overall, the results of the current study demonstrated that traditional aggression and victimization is prevalent within this rural sample, as 73.6 % of youth reported aggressing and 75.7 % endorsed experiencing victimization at least once during the school year. This rate is higher than past research would indicate, but can be accounted for by the inclusion of any amount of peer aggression or victim iz ation. Previous research has cited 4 to 36% peer aggression rates and 9 to 31% victimization rates (Berger, 2007; O Brennan, Bradshaw, & Sawyer, 2009 ; Spriggs et al., 2007). The majority of studies categorized the aggression and victimization variables bas ed upon cutoffs, while this study kept the variables as continuous. Thus, the percentages cited include frequencies ranging from once to multiple times this past year. When a cutoff criterion of 24 was applied to the Total Bully and Total Victim scales of the APRI measure, 36.5% and 41.0% of youth reported aggressing and experiencing victimization respectively. While these percentages are similar to prevalence rates cited within the literature, they are still slightly higher The cutoff criterion of 24 mean t that youth endorsed experiencing victimization once a week on at least two items. Conversely, cyber aggression and victimization does not appear as prevalent as traditional aggression and victimization. Within this rural sample, 2 7.7 % of youth endorsed cyber aggressing and 3 9.4 % endorsed experiencing any cyber victimization. Previous studies cited 4 to 35% prevalence rates of cyber aggression and 5 to 14% cyber victimization rates (Dempsey et al., 2009; Hinduja & Patchin, 2008; Kowalski & Limber, 2007; S ourander et al., 2010; Williams & Guerra, 2007; Wang et al. 2009). The prevalence of cyber aggression was comparable to urban and suburban rates described within the
82 literature, while cyber victimization rates were higher than previous literature indicate d. This result could pertain to the lack of cutoff criterion used to measure cyber victimization. In this study, the variable was continuous, thereby resulting in more youth being identified as a victim. In general, cyber aggression and victimization rates continue to remain less prevalent compared to traditional forms of aggression. This could partially be due to the lack of established measures assessing this construct. Some research assessing these variables used only a few questions, ranging from two to four, and only counted greater frequencies, such as two or three times a month, when determining the prevalence of cyber aggression and victimization (Wang et al., 2009; Spriggs et al., 2007). Conversely, other studies evaluated cyber aggression and victi mization as a continuous variable, without a categorical cutoff (Erdue Baker, 2010; Kowalski & Limber, 2007). The lack of comparable rates between traditional and cyber aggression and victimization could additionally relate to the fact that traditional agg ression continues to be the most popular method of engaging in peer victimization. In contrast to the gender differences found with some of the peer aggression and victimization scales (discussed below), descriptive analyses revealed no ethnic differences for the physical aggression and victimization, social aggression and victimization, and total aggression and victimization scales of the APRI, as well as Cyber Aggression and Victimization scales. A prior study in the field revealed ethnicity differences in physical and verbal aggression where African Americans reported more aggression in these categories compared to Caucasians and Latinos (Wang et al., 2009). In addition, Latinos and Caucasians reported more relational victimization than African American s. No other differences were found. Caucasians were the predominant
83 ethnicity in the current sample, which may have made it more difficult to detect ethnic differences if they are present. The high percentage of Caucasians in this study was consistent with previous research (Craig, 1997; Pornari & Wood, 2010) and the population demographics of the county in which it was conducted. However, this may not be reflective of other counties within Florida or the country, and suggests the need for future studies t o include more diverse samples. The results for the first aim of this study partially supported the hypotheses that males would self report more physical aggression than females and females would self report more relational aggression than males, and that males would report more physical victimization than females and females would self report more relational victimization than males. Specifically, for the Social Aggression scale of the APRI females self reported more relational aggression compared to male s. Additionally, for the Social Victimization scale of the APRI females self reported experiencing more relational victimization compared to males. The identification of females as being more likely to engage in relational aggression and victimization was similar to other findings within the literature although there is also some research supporting a lack of gender differences in type of aggression ( Craig, 1998 ; Pornari & Wood, 2010). Existing research continues to lack understanding of the developmental process regarding relationally aggressive behaviors (Spieker, 2012). However, despite the lack of understanding of the developmental process there is substantial resear ch, dating as early as the 1920 s, indicating that gender differences exist in reference to relational aggression (Archer, 2004; Craig, 1998, Pornari & Wood, 2010). Preliminary data indicates that maternal callousness and low maternal sensitivity put females at risk for
84 relationally aggressive behaviors (Brown et al., 2007; Casas et al., 2006 ; Curtner Smith et al., 2006). The results of this study continue to support general models of aggressive behaviors indicating that females are more likely to engage in relationally aggressive behaviors compared to males (Spieker et al., 2012). In contras t, for the Physical Aggression and Victimization scales there were no significant differences between males and females. This lack of difference could be partly attributed to more females in the study compared to males. Also, e vidence exists within the lit erature indicating a lack of gender discrepancy for physical aggression and victimization ( Peskin, Tortolero, & Markham, 2006 ). This study consisted of a sample drawn from eight urban middle and high school s in Texas, with aggression and victimization bein g measured with eight and four questions respectively within the past thirty days. On the other hand, t he majority of research, regardless of methodology and the time period when research was conducted, support s gender disparities, where males are typically more likely to engage in, and are victims of, physical aggression (Erdur Baker, 2012; Rodkin & Berger, 2008 ; Zimmer Gembeck, Geiger, & Crick, 2005 ). Notably, the majority of the studies supporting gender dispa rities were conducted in urban and suburban areas or in other countries. The results of this study are not consistent with the predominant findings within the literature supporting gender disparities for physical aggression and victimization. I n addition t o the current sample containing more females than males, this lack of findings could also potentially be attributed to the nature of the rural sample included in this study which may differ from population s used in previous research studies. For this samp le, f indings such as these
85 are concerning and suggest that female engagement in physical peer aggression needs to be further studied and targeted in bullying prevention programs. Contrary to the hypothesis that there would be gender discrepancies in cyber aggression, analyses revealed no differences between males and females. While this finding is similar to some previous research findings, evidence also exists suggesting that a larger proportion of females report being perpetrators and victims of cyber agg ression (Dempsey et al., 2009; Erdur Baker, 2010; Kowalski & Limber, 2007; Marini et al., 2006; Patchin & Hinduja, 2006). The absence of findings in this study may partly relate to the continuous measurement of cyber aggression, as opposed to using cutoffs to determine aggressors. If in previous research males were more likely to endorse fewer cyber aggression episodes compared to females applying a cutoff criterion to define cyber aggressors could identify more females as aggressors. Thus, more research i s needed examining gender differences using continuous measures. The results for aim one partially supported the hypotheses that teachers would rate males as more physically aggressive and females as more relationally aggressive. Teachers reported more mal es as physically aggressive, while they did not report significant differences between males and females on relational aggression. Analyses could not be conducted examining gender differences in teacher reports of victimization, as teachers identified very few students as victims of aggression. These results could be attributed to several factors. Evidence exists within the literature supporting the stability of teacher reports of physical aggression over time, while reports of relational aggression are not consistent over time (Kuppens et al., 2009). This difference could relate to the difference between physical and relational aggression. Specifically,
86 physically aggressive behaviors are disruptive within the classroom and much more likely to be noticed, w hereas relational aggression is more difficult to detect and more likely to occur in areas where teachers are typically not present (i.e., the hallways, lunchroom, and on the bus). One study indicated that youth report teachers as lacking awareness of yout Kochenderfer Ladd & Pelletier, 2008; Smith & Shu 2000). Another study found that teachers have difficulty successfully identifying relational aggression and victimization, which could have po tentially contributed to the lack of significant findings for this aim (Bradshaw et al., 2007). In addition to these factors t he teachers who participated in this study may have had less knowledge regarding the experiences and behavior of some of the students in their class as some of the students only saw the teacher for one class period and did not have contact with that teacher for the remainder of the day. Overall, the current findings appear consistent with the research presented on this topic (R enk & Phares, 2004). For aim two examining the relationship between self reports of traditional peer aggression and victimization, as well as cyber aggression and victimization, and social functioning, the hypotheses were not supported by the results of th is study. Higher scores on the APRI Total Aggression and Victimization scales, as well as the Cyber Aggression and Cyber Victimization scales did not predict poorer social functioning as expected. This lack of significance could be attributed to the way a functioning was measured. While the social functioning composite was calculated using data from multiple informants (self and teacher), peer report was not able to be ning.
87 Additionally, teacher ratings of youth social functioning, which comprised one aspect of the social functioning composite used in analyses, were based on one item and were highly positively skewed. Highly positive ratings may have restricted the rang e of the composite measure, therefore making it more difficult to find significant results. Due to these concerns regarding teacher ratings, e xp loratory analyses examining social functioning using only the Piers Harris 2 Popularity Domain were conducted. These analyses revealed a significant relationship between self reports of traditional peer aggression ( B = .17, p < .05) and victimization ( B = .36, p < .01) and the Piers Harris 2 Popularity Domain; there was also a significant relationship for cyber a ggression ( B = .84, p < .05) but not cyber victimization ( B = .73, p = .06) As youth self report of traditional aggression and victimization and cyber aggression decreased, scores on the Piers Harris 2 increased indicating better social functioning. Th ese results suggest that the highly positive teacher ratings restricted the range of the overall composite measure thus reducing the chance of finding significant results. For aim three examining whether traditional and cyber victimization predicted sel f reported symptoms of anxiety the hypotheses for aim 3a and 3b were mostly supported Specifically, youth who reported higher rates of traditional and cyber victimization also reported high levels of anxiety. Further, females reported experiencing significantly more social anxiety compared to males. These results are consistent with previous research ( Craig, 1998 ; Dempsey et al., 2009 ). Youth who experience traditional victimization are more likely to endorse symptoms of anxiety compared to non vict imized peers. Furthermore, the results also support a relationship between self reports of cyber victimization and greater physical symptoms of anxiety and social
88 anxiety but not anxiety symptoms that are indicative of an anxiety disorder. There is a shor tage of research within the literature linking cyber victimization to symptoms of anxiety. Therefore, the current results suggest that cyber victimization is related to anxiety in line with findings for traditional peer victimization. In addition, aim 3c h ypothesized that gender would moderate the relationship between traditional and cyber victimization and each of the dependent anxiety variables. The results of this study did not support the hypotheses. The only interaction term that approached significanc e ( p = .05) was in the analysis for gender moderating the relationship between total traditional victimization and total Social Anxiety. In each regression the interaction term was not significant, indicating that gender did not moderate the relationship b etween total and cyber victimization and each of the dependent anxiety variables. For two of the anxiety scales, total Physical Symptoms of Anxiety and Anxiety Disorders Index, gender was also not a significant predictor in block one. However, for the tota l Social Anxiety scale, gender was a significant predictor in block one, even though the interaction term was not significant. The lack of significant interaction effects could be attributed to several factors, the first being that gender may not moderate the relationships between traditional and cyber victimization and symptoms of anxiety. Another reason could potentially be attributed to the MASC norms. The MASC attempts to control for gender differences by providing separate norms for males and females. Thus, the lack of significant moderating effects of gender might be attributed to the MASC T scores being gender normed In an attempt to determine the impact that the gender normed T scores may have had on the results of the regression, exploratory analy ses were conducted
89 repeating each regression for this aim using the raw scores of the MASC scales instead of the T scores. According to the results, the only interaction term that was significant ( p < .05) was in the analysis for gender moderating the rela tionship between total traditional victimization and total Social Anxiety. For the remaining regressions, the interaction termed remained non significant. Thus, the gender normed T scores appeared to have had little impact on the results. Al ternatively a s the sample in this study is a non clinical sample, the majority of participants endorsed non clinically significant levels of anxiety, thus restricting the range for this measure. Perhaps findings would have differed if the sample had been larger and had a greater range of scores on the MASC. Lastly, when examining the correlations between anxiety symptoms and reports of traditional and cyber victimization for males and females separately, it was found that for the majority of the scales the magnitude of the relationship between the variables did not differ significantly for males and females, which would support the absence of moderator effects. However, there was one exception. The correlation for males compared to females between Cyber Victimization and the Physical Symptoms of Anxiety scale was of significantly lower magnitude ( p < .05); while both correlations were positive the relationship between these variables was stronger for females compared to males. This finding suggests a possibility that ther e may also have been insufficient power to detect moderator effects of lower magnitude if they were present. Aim four examined relationships between self reported peer aggression and victimization (both traditional and cyber) and school functioning as measured by GPA of core classes and the total number of referrals and suspensions combined. It was
90 hypothesized t hat greater self reported physical aggression would be significantly associated with poorer school functioning whereas cyber aggression would not be associated with school functioning The results did not support these hypotheses. Despite evidence indicat ing that students who are more aggressive have lower academic functioning compared to non aggressive peers, the result of this aim was not commensurate with previous research findings. Similarly although research demonstrates that physical aggressors tend to receive more suspensions and are more likely to fall behind academically compared to non aggressive peers (DeRosier & Lloyd, 2010), the results of this aim were not commensurate with previous research findings although they fell just short of signific ance in the expected direction Conversely greater self more likely to engage in cyber aggression. Further, youth who reported engaging in more cyber aggression also reported r eceiving more suspensions and referrals combined. Both of these findings were contrary to the hypothesis that cyber aggression would not be related to school functioning. Although there is a dearth of literature regarding the relationship between academic functioning and cyber aggression, one study found that cyber aggressors tended to experience more school impairment such as receiving detentions and suspensions (Ybarra et al., 2007). Additionally, it was hypothesized that greater self reported traditional and cyber victimization would be significantly associated with poorer school functioning as measured by GPA, but not significantly associated with the number of suspensions and referr als. The results mostly supported the hypotheses, such that youth who reported experiencing more traditional victimization had poorer academic functioning as
91 measured by GPA. In addition, commensurate with the hypothesis traditional victimization was not a significant predictor of the total number of suspensions and referrals While research supports poorer academic functioning as measured by GPA in traditionally victimized youth, there is a lack of support for traditionally victimized youth receiving a gre ater number of suspensions and referrals compared to their non victimized peers (DeRosier & Lloyd, 2012 ; Eaton et. al., 2008). Contrary to hypotheses, cyber victimization was not a significant predictor of school functioning as measured by GPA However, a s hypothesized, cyber victimization was not significantly related to the total number of suspensions and referrals However, the relationship did fall just short of significance, suggesting that the two variables might actually have some relationship with one another contrary to the hypothesis There is a lack of research focusing on cyber victimization and measures of school functioning. However, some previous research has indicated that youth who are victims of cyber aggression often receive more frequent detentions or suspensions compared to non cyber victimized peers (Ybarra et al., 2007). However, the majority of research focusing on cyber aggression and victimization studies in general did not study cyber aggression and victimization as unique variable s apart from traditional aggression and victimization (Wang et al., 2009; Kowalski & Limber 2007). A potential reason for the unexpected significant finding that cyber aggression was related to a greater number of suspensions and referrals could be due to how suspensions and referrals were measured. This study relied on youth self report for determining the number of suspensions and referrals received. Students could have over reported or under reported the number of referrals and suspensions they recei ved
92 this school year thus introducing measurement error for this variable In the current sample, under reporting may have been a concern, as the total number of self reported referrals in this sample ( 105 ) was relatively lower than might be expected give n aggregate school records data for referrals. Taking 26.7% (percentage of seventh and eighth graders participating in the study) of referrals from school records data (183) suggests that reports of referrals should have been higher. However, this would as sume that referrals are equally distributed across students, which is not the case. Also, it is unknown how representative this sample was with regards to referrals and suspension compared to the total seventh and eighth grade student body, and it is likel y that students who get more referrals could be underrepresented in the current sample, as these students may have been less likely to have parents sign consent forms or participate. Nonetheless, the current data probably under estimate the true number o f referrals for participating students. However, when data for the current sample are broken down by gender, 58.1 % of total referrals were reported by males while 41.9 % were reported by female s which was similar to the gender breakdowns for the school dis trict. These percentages indicate that the overall sample data was consistent with what is seen at the district level with regards to gender distribution. Finally this study did not control for traditional and cyber victimization when examining the relationship of aggression to total suspensions and referrals. It could be that this relationship differs for those who are both victims and aggressors, as compared to those who are only aggressors Overall, the current findings need to be interpreted in light of the fact that analyses were conducted examining traditional and cyber aggression as separate
93 variables. However, concerns have been raised by researchers in the field of peer victimization that these may be highly overlapping co nstructs. According to Olweus (2012) the prevalence of cyber aggression is actually over emphasized within the literature, and he argues that most students who engage in cyber aggression are also perpetrators of traditional peer aggression, and that most cyber victims are also victims of traditional forms of aggression. In line with these arguments, there is research to support that traditional and cyber aggression and victimization are similarly overlapping constructs, such that cyber aggression is an ext ension of traditional peer aggression with little unique qualities (Erdur Baler. 2010; Li, 2005; Li, 2006; Raskauskas & Stoltz, 2007). To determine the extent of overlap in the current study, follow up exploratory analyses were conducted to determine if c yber aggression and victimization were independent and unique from traditional aggression and victimization. According to the results, cyber aggression was not independent from traditional aggression. There was a 100% overlap between traditional and cyber aggression variables, such that all students who indicated perpetrating cyber aggression also reported perpetrating traditional forms of peer aggression. This finding indicates that cyber aggression did not occur independently of traditional aggression w ithin this sample. Similarly, only 4.8% of the current sample endorsed uniquely experiencing cyber victimization independently of traditional victimization; the majority of cyber victims overlapped with victims of traditional peer aggression. These results support arguments and findings that cyber aggression and victimization are not necessarily unique and separable from traditional aggression and victimization. Olweus (2012) reported research findings supporting the
94 high degree of overlap ranging from 88% to 93% between traditional and cyber aggression and victimization. Overall, results from the current study indicate that children who cyber aggress and experience cyber victimization are also involved in traditional forms of peer aggression and victimizat ion, and thus may not be uniquely different from their peers who are involved only in traditional aggression and victimization. This likely explains why similar relationships were found between traditional and cyber forms of aggression and victimization an d other variables examined in this study (i.e., anxiety, school functioning). In order to examine whether cyber aggression or victimization have unique effects, it would be important to look at both traditional and cyber variables in the same analyses (e.g ., does cyber victimization contribute unique variance over and above traditional victimization when predicting anxiety). These types of analyses were not the main focus of this study, but will be important to examine in future research. Concerns regarding cyber victimization have increased due to widespread et al., 2009; Pornari & Wood, 2010). Therefore, research has increased its focus on studying cyber aggression and victimizatio n within the last 20 years. According to the literature, up to 97% of youth ages 12 18 access the internet (Kowalski & Limber, 2007). In an attempt to understand how many students access social networking sites via the internet and engage in risky cyber b ehaviors further exploratory analyses were conducted. These analyses revealed that 82.3% of the youth in this sample reported having a social networking site, indicating that at least 82.3% of the sample from this study accesses the internet. This percenta ge may not be inclusive of the students who access the
95 internet for other means (playing games) and who have or visit blog sites and other means of socializing via the internet. When examining internet risky behavior, of those who had a social networking s ite, 31.4% reported that they share their password with as the individual and engage in cyber aggressive behaviors. In addition to sharing a password, 44.4% of youth in dicated that their parents never monitor their internet use, which may potentially increase the likelihood for inappropriate online behavior. Implications Results of this study correspond with the literature documenting the association between anxiety and traditional and cyber victimization. According to the significant fin d ings of this study youth who experience traditional victimization also experience greater total anxiety disorders symptoms as well as greater p hysical anxiety s ymptoms and s ocial a nxi ety compared to youth who did not report experiencing traditional victimization. Youth who endorsed cyber victimization also reported experiencing greater p hysical anxiety s ymptoms and s ocial a nxiety compared to peers who did not report cyber victimization Overall, t hese results indicate th at youth who experience traditional and cyber victimization are more likely to experience increased anxiety compared to non victimized peers. Therefore, anxiety should be addressed when schools are contemplating implemen tation of peer aggression prevention efforts. Moreover, intervention efforts should take into account anxiety symptoms when treating traditional and cyber victimized youth. In conjunction with increased anxiety, victimized youth who are experiencing anxiet y are at risk for decreased social functioning, which could further exacerbate their anxiety in social situations (Greco & Morris, 2005). Thus, social functioning should also be monitored.
96 Regarding school functioning, cyber aggression was related to lowe increased number s of suspensions and referrals. Further, results from this study indicated that traditional and cyber victimization was significantly related to school functioning. Specifically, youth who reported experiencing more traditional victimization Similar to findings in the literature, traditionally victimized youth have decreased academic functioning (Nansel et al., 2001; Sharp 1995). Lower GPA could re sult from school avoidance allowing for decreased development of academic skills in the classroom, as well as a poorer academic self concept (Flook et al., 2005 ; Mercer & De Rosier, 2008). In addition, both youth reporting traditional and cyber victimizati on experiences also had increased numbers of suspensions and referrals. When implementing prevention initiatives for peer victimization, in addition to assessing for anxiety, school personnel should be aware of findings such as these with regards to school functioning and assess for academic and school behavioral functioning when working with youth who are victimized. Prevention efforts should also target academic functioning when working with aggressors. When assessing for peer aggression and victimizatio n, school personnel should be mindful that research supports a significant overlap between traditional and cyber aggression and victimization. Thus, when evaluating the prevalence of cyber victimization and aggression in a school, personnel should assess f or the prevalence of unique cyber victimization experiences prior to implementing more preventative efforts of anti bullying efforts could occur to focus more on c yber bullying if this type of
97 aggression is deemed as the predominant problem in the schools. Again, findings from this study, as well as research conducted by Olweus (2012), that indicate significant overlap between traditional and cyber aggression and vi ctimization constructs suggests a need for traditional peer aggression to continue to be a primary focus of prevention efforts. Thus, academic environments should continue to put forth effort and resources to counteract traditional bullying as this type of aggression continues to be more prevalent. Due to the significant rates of overlap, cyber aggression and victimization could decrease as a result of traditional aggression and victimization intervention. Contrary to popular belief, where some research ha s suggested that peer aggression has increased due to the availability of electronic media, the results from this study do no t support results of this study suggest that few unique youth were vi ctims of cyber aggression without also being victims of other traditional forms of peer aggression at school. This is an important point as media has purported that cyber aggression and victimization has increased in prevalence and has significant consequ ences. Unfortunately, this view could increase symptoms of anxiety and stress unnecessarily among parents (Olweus, 2012). However, even if cyber aggression is not as prevalent as purported, efforts to decrease this behavior may still be important. As an ex ample, results from this study indicated that 44.4% of youth believe their parents are not monitoring their internet behaviors. Providing education to parents, as well as youth, regarding internet safety could increase parental monitoring and hopefully dec rease cyber aggression and consequently cyber victimization.
98 The current study focused o n a rural sample, as there appears to be a dearth of research examining traditional and cyber aggression and victimization rates in rural populations. Findings suggest that rates of traditional and cyber aggression and victimization in this rural sample were higher compared to urban and suburban middle school rates reported within the literature, which could be attributed to the constructs being measured as continuous variables. Nonetheless, these prevalence rates highlight the importance of examining traditional and cyber peer aggre ssion and victimization in both rural and urban populations. Limitations Several limitations should be taken into account when interpreting findings from the current study. First, participant data gathered in this study came from one rural middle school, therefore findings may not necessarily generalize to other environments. A second primary limitation is that the data collected in this study is cross sectional, thereby not allowing for causal inferences regarding the direction of the relationships. Anoth Disorders Index scale from the MASC, indicating that this subscale did not have the same level of internal consistency as found in the normative sample, which may have potenti ally impacted the results containing this subscale. A further limitation of this study pertains to the method used to gather information on peer aggression and victimization. Peer aggression and victimization was measured via self report, with disadvantage s of these types of measures relating to factors that decrease the validity of the results. Youth potentially differ in their interpretation of aggression, as well as their willingness to identify themselves as an aggressor or a victim (Ladd & Kochenderfer Ladd, 2002).
99 Conversely, youth may over report their aggressive behaviors or victimization status thereby decreasing the validity of the results. Information from multiple informants was limited; particularly peer report, which is the gold standard in as was initially planned for this study, changes to methodology were made based on Institutional Review Board feedback. As a result, data on social functioning was extracted from a youth self limited, particularly for victimization status. Thus, findings for the hypotheses regarding social functioning may have differed if peer reports were available. Another limitation is that GPA and total number of suspensions and referrals combined was collected via self report. Youth report could be inaccurate for these types of data, thus collecting this informati on directly from school records could increase accuracy of the information. Future Directions The current study did not obtain information from multiple informants, particularly peer report, of social functioning. Assessing cross informant report regardin g social functioning. Thus, obtaining this information would facilitate the understanding of the effect traditional and cyber aggression has on social functioning from a mult i informant perspective. Youth in this study were not put into categories of aggressors, victims, and both aggressors and victims, which could impact the results found. However, examination of these constructs as continuous variables may be important for future studies to better understand the relationships of these behaviors and experiences with other variables.
1 00 Potential negative consequences of even minor peer victimization are more likely to be noted when peer victimization is studied as a continuous variable. This understanding could facilitate further development of prevention and intervention measures for traditional and cyber aggression and victimization. While this study assessed a variety of anxiety symptoms using a standardized measure, the s tudy was cross sectional. It could be that youth who are more inclined to experience anxiety experience more victimization, further increasing their anxiety. Similarly, youth with poor social functioning or more behavioral issues (resulting in more school referrals/suspensions) may be at higher risk for being victimized. Overall, the cross sectional data from the current study, as well as a number of other studies in the literature, makes it difficult to determine the temporal or causal relationships betwee n peer aggression and victimization and indices of youth functioning. More longitudinal research is needed to better understand the direction of these relationship s Finally, although there is significant overlap between traditional and cyber aggression a nd victimization, further research needs to be conducted examining whether cyber victimization has any unique impacts over and above the impacts of traditional peer victimization on the variables examined in this study (social functioning, anxiety, and sch ool functioning). In addition, those who are uniquely victimized or aggress through cyber means should be further studied to facilitate understanding of the characteristics of unique cyber aggressors and victims. In conjunction, motivation for engaging in aggressive behaviors deserves further study, as there is minimal information regarding motivation within the literature. Understanding why the aggression
101 is occurring could aid in the creation of more effective preventative and intervention methods.
102 APPENDIX A DEMOGRAPHIC INFORMATION ID Number _________________________ DEMOGRAPHIC INFORMATION School : ____________________________ Age: __________ Circle One of the Following: Grade : 7 th 8 th Sex : Male Female Ethnicity : White (non Hispanic) African American Hispanic Asian Native American Multi racial Other Please circle your grade in the following classes: Math (or Algebra) A B C D F Language Arts A B C D F Science /Physical Science A B C D F Civics/US History A B C D F Have you been suspended this school year? Circle one of the following : Yes No If you have been suspended this school year, please write the number of times you have been suspended below _________________ If you have received any referrals at school this year, please write the number of referrals you have received below ______________________ Who do you live with, check all that apply: Mother Father Grandparent Other relative other Step father Step mother
103 APPENDIX B ID Number _________________________ SOCIOMETRIC TEACHER REPORT OF STUDENT SOCIAL FUNCTIONING Directions: Please write the ID number of the student in the place provided. Then, rate how much the student in your class is liked or disliked by other students in his or her grade. Next, circle the aggression or victimization status of the student, if any. How much is th is student liked/disliked by other students in her/his grade? (circle one number) Scale: ______________________________________________________ ______________________ 1 2 3 4 5 6 7 Very much Disliked Somewhat Not Liked Somewhat Liked Very Disliked Disliked or Disliked Liked Much Liked Definition s for Aggressive subtypes Physical Aggression: starts fights, uses physical force to get his/her own way, hits, kicks, punches, pushes, and shoves other students Verbal Aggression: students who call others names, say mean things to others, or verbally tease others about their appearance, clothes, and or family Relational Aggression: students who say negative things about others, embarrass others, gossip about them, and spread rumors, etc Definitions for Victimization subtypes Physical Victimization: s tudents who are victims of aggression (they are picked on, hit, kicked, pushed, or shoved by others) Verbal Victimization: students who are called names, have mean things said to them, or are verbally teased by others about their appearance, clothes, and or family Relational Victimization: students who have negative things said about them, are embarrassed, gossiped about, and have rumors spread about them, etc. Please circle any of the following that apply for this student (you may circle more than one i n each row): Is this student : Physically Aggressive Verbally Aggressive Relationally Aggressive None of these Is this student: Physically Victimized Verbally Victimized Relationally Victimized None of these
104 APPENDIX C CYBER AGGRESSION AND VICTIMIZATION MEASURE
105 16 In thinking about your answers above (#8 14 ), how often do you know the person who does these things to you? (circle one) Never Sometimes Often Always 17 When you know the person in #15, are they most often a friend or not a friend? (circle one) Friend Not A Friend PART 3: HOW YOU FEEL Instructions: Based on the experiences you had above, circle ONE number per question to show how these experiences generally the questions above, circle never for questions #17 26. Because of mean, hurtful, embarrassing, or threatening messages I have received Never Sometimes Often Alway s 17. I have stayed home from school. 1 2 3 4 18. I have felt lonely. 1 2 3 4 19. I have become upset (like angry or crying) 1 2 3 4 20. I have wanted to change schools. 1 2 3 4 21. I have been afraid or worried. 1 2 3 4 22. It was harder for me to do things (such as school work) 1 2 3 4 23. I missed going somewhere, seeing people, or doing something that I would have enjoyed. 1 2 3 4 24. I have been embarrassed or felt bad about myself. 1 2 3 4 25. I have been mean to or threatened someone else by text message, over the phone or on the internet. 1 2 3 4 26. My friendships have changed. 1 2 3 4 PART 4: WHAT YOU DO Instructions: These questions ask about things you might have done over the past month. Please circle ONE answer per question. Never Once or Twice a Month One or Twice a Week Several times a week
106 27. I have sent mean or threatening text messages to someone. 1 2 3 4 28. I posted an embarrassing picture or video clip on line of someone in order to be mean or hurt them. 1 2 3 4 29. I have made mean or threatening phone calls (does not include text messaging) 1 2 3 4 30. I have sent mean or threatening emails to someone. 1 2 3 4 31. I have threatened or been mean to another person in a chat room. 1 2 3 4 32. I have picked on, made fun of, or threatened someone through instant messaging (IM) 1 2 3 4 33. I have posted mean, hurtful, or threatening comments about another person on a website (e.g., MySpace, Facebook) 1 2 3 4 34. How much do your parents monitor your use of social networking sites, blogs, etc. Never Sometimes Often Always PART 5: DURING MY 2011 2012 SCHOOL YEAR Instructions: Please circle ONE answer per question. 39. I think bullying is a problem at my school Strongly Disagre e Disagree know Agree Strongly Agree 40. I have missed school due to bullying Yes No
107 APPENDIX D PARENT/GUARDIAN CONSENT LETTER
109 APPENDIX E STUDENT ASSENT FORM
110 APPENDIX F TEACHER CONSENT FORM
111 LIST OF REFERENCES Achenbach, T. M., McConaughy, S. H., & Howell, C. T. (1982). Child/adolescents behavioral and emotional problems: Implications of cross informant correlations for situational specificity. Psychological Bulletin, 101 213 232. d oi : 10.1031//0033 2909.101.2.213 Archer, J. (2004). Gender differences in aggression in r eal world settings: a meta analytic review. Review of General Psychology, 8( 4 ), 291 322. doi: 10.1037/ 1089 26184.108.40.2061 Baldwin, J. S. & Dadds, M. R. (2007). Reliability and validity of parent and child versions of the multidimensional anxiety scale for children in community samples. Journal of the American Academy of Child and Adolescent Psychiatry, 46( 2 ), 252 260. doi : 10.1097/01.chi.0000246065.93200.a1 Band ura, A. (1986). Social foundations of thought and action: A social cognitive theory. National Institute of Mental Health; Englewood Cliffs, New Jersey: Prentice Hall, Inc Bandura, A. (1992). Social Cognitive Theory. In R. Vasta Ross (Ed ) Six theories of child development; Revised formulations and current issues. (pp. 1 60). London, England: Jessica Kingsley Publishers Bandura, A. (1999). Social Cognitive Theory: An agentic perspective. Asian Journal of Social Psychology, Special issue: Theoretical and Methodological advances in social psychology 2( 1 ) 21 41. doi: 10.1111/1467 839x.00024 Bauman, S. & Del Rio, A. (2005). Knowledge and beliefs about bullying in schools: Comparing pre service teachers in the United States and the United Kingdom. School Psychology International, 26( 4 ) 428 442. doi: 10.1177/0143034305059019 Beran, T. & Li, Q. (2005). Cyber harassment: A study of a new method for an old behavior. Journal of Educational Computing Research, 32( 3 ), 265 277. doi: 10.2190/8YQM B04H PG4D BLLH Bond, L., Carlin, J. B., Thomas, L., Rubin, K., Patton, G. (2001). Does bullying cause emotional problems? A prospective study of young teenagers. British Medical Journal, 323( 7311 ), 480 484. doi : 10.1136/bmj.323.7311.480 Boivin, M. & Hymel. S. (1997). P eer experiences and social self perceptions: A sequential model. Developmental Psychology, 33, 135 145. doi: 10.1037/0012 16220.127.116.11
112 Branson, C. E. & Cornell, D. G. (2009). A comparison of self and peer reports in the assessment of middle school bullyi ng. Journal of Applied School Psychology, 25( 1 ), 5 27 doi:10.1080/15377900802484133 Brown, S. A., Arnold, D. H., Dobbs, J., & Doctoroff, G. L. (2007). Parenting predictors of relational aggression among Puerto Rican and European American school age children. Early Childhood Research Quarterly, 22, 147 159. doi: 10.1016/j.ecresq.2206.11.002 Casas, J. F., Weigel, S. M., Crick, N. R., Ostrov, J. M., Woods, K. E., Jansen Yeh, E. A., & Huddle ston Cases, C. A. physical aggression in the preschool and home contexts, Journal of Applied Developmental Psychology, 27 209 227. doi: 10.1016/j.appdev.2006.02.003 Coie, J. D., & Dodge, K. A. (1998). Aggression and antisocial behavior. In W. Damon (Series ED.) & N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (5 th ed., pp. 779 862). New York, NY: Wiley. Coie, J. D., Dodge, K. A., & Coppotel li, H. (1982). Dimensions and types of social status: A cross age perspective. Developmental Psychology, 18( 4 ), 557 570. doi: 10.1037/0012 1618.104.22.1687 Cook, C. R., Williams, K. R., Guerra, N. G. (2010). Predictors of bullying and victimization in childhood and adolescence: A meta analytic investigation. School Psychology Quarterly, 25( 2 ), 65 83 doi: 10.1037/a0020149 Cornell, D. G. & Brockenbrough, K. ( 2004). Identification of bullies and victims: A comparison of methods. Journal of School Violence, 3, 63 87. doi: 10.1300/ J202v03n02_05 Crick, N. R. & Bigbee, M. A. (1998). Relational and overt forms of peer victimization: A multiinformant approach. Journ al of Consulting and Clinical Psychology, 66, 337 347. doi: 10.1037/0022 006X.66.2.337 Crick, N. R. & Dodge, K. A., (1994). A review and reformulation of social information Psychological Bulletin 115, 74 101. doi: 10.1002/ab.20336 Crick, N. R. & Grotpeter, J. K., (1995). Relational aggression, gender, and social psychological adjustment. Child Development, 66, 710 722. doi: 10.2307/1131945 Crick, N. R. & Grotpeter, J. K. (1996). relational and overt aggression. Development and Psychopathology, 8, 367 380. doi: 10.1017/S0954579400007148
113 & Markon K (1999). Childhood aggression and gender: A new look at an old problem. Nebraska Symposium on Motivation 45 75 141. Craig W. M. (1998). The relationship among bullying, victimization, depression, anxiety, and aggression in elementary school children. Personality and Individual Differences 24( 1 ), 123 130. doi: 10.1016/S0191 8869(97) 00145 1 Curtner Smith, M. E., Culp, A. M., Culp, R., Scheib, C., et al. bullying. Journal of Child and Family Studies, 15 181 193. doi: 10.1007/ s10826 005 9016 7 Dempsey, A. G., Sulkowski, M. L., Nichols, R., & Storch, E. A. (2009). Differences between peer victimization in cyber and physical setting and associated psychosoc ial adjustment in early adolescence. Psychology in the Schools, 46( 10 ), 962 972. doi: 10.1002/pits.20437 be havioral adjustment as a function of the chronicity and proximity of peer rejection. Child Development, 65, 1799 1813. doi:10.2307/1131295 DeRosier, M. E. academic outcomes. Reading and Writing Quarterly, 27( 1 ), 25 47. doi: 10.1080/10573569.2011.532710 Dill, E. J., Vernberg, E. M., Fonagy, P., Twemlow, S. W., & Gamm, B. K. (2004). Negative affect in victimized children: The roles of social withdrawal, peer rejection, and attitudes toward bullying. Journal of Abnormal Psychology, 32, 159 173. doi: 10.1023/B:JACP.0000019768.31348.81 Donellan, M. B., Trzesniewski, K. H., Robins, R. W., Moffitt, T. E., & Caspi, A. (2005). Low self esteem is related to aggression, antisocial behavior, and delin quency. Psychological Science, 16, 328 335. doi: 10.1111/j.0956 7976.2005.01535.x Dooley, J. J., Pyzalski, J., & Cross, D. (2009). Cyberbullying versus face to face bullying: A theoretical and conceptual review. Journal of Psychology, 217( 4 ), 182 188. doi: 10.1027/0044 3409.217.4.182 Eaton, D. K., Kann, L., Kinchen, S., Shanklin, S., Ross, J., Hawkins, J., & Weschsler, H. (2008, June). Youth risk behavior surveillance United States, 2007 (Morbidity and Mortality Weekly Report No. 57). Washington, DC: Center s for Disease Control and Prevention. Erdur Baker, O. (2010). Cyberbullying and its correlation to traditional bullying, gender and frequent and frisky usage of internet mediated communication tools. New Media & Society, 12( 1 ), 109 125. doi: 10.1177/1461444809341260
114 Finger, L. R., Yeung, A. S., Craven, R. G., Parada, R. H., & Newey, K. (2008, November). Adolescent peer relations instrument: Assessment of its reliability and construct validity when used with upper primary students. Paper presen ted at the AARE 2008 International Education Research Conference Brisbane, Australia. Flook, L., Repetti, R. L., & Ullman, J. B. (2005). Classroom social experiences as predictors of academic performance. Developmental Psychology, 41, 319 327. d oi : 10. 1037/0012 1622.214.171.1249 Ford, M. E. (1982). Social cognition and social competence in adolescence. Developmental Psychology, 18( 3 ), 323 340. doi: 10.1037/0012 16126.96.36.1993 Gilman, R., & Huebner, E.S. (2006). Characteristics of adolescents who report very high life satisfaction. Journal of Youth and Adolescence, 35( 3 ), 311 319. doi: 10.1007/s10964 006 9036 7 Gladstone, G. L., Parker, G. B., & Malhi, G. S. (2006). Do bullied children become anxious and depressed adults? A cross sectional investigation of the correlates of bullying and anxious depression. The Journal of Nervous and Mental Disease 194( 3 ), 201 208. doi: 10.1097/01.nmd.0000202491.99719.c3 Graham, S., Bellmore, A., & Juvonen, J. (2003). Peer victimization in middle school: When self and peer v iews diverge. Journal of Applied School Psychology, 19, 117 137. doi: 10.1300/J008v19n02_08 Greco, L. A. & Morris, T. L. (2005). Factors influencing the link between social anxiety and peer acceptance: Contributions of social skills and close friendships during middle childhood. Behavior Therapy, 36, 197 205 doi: 10.1016/S0005 7894(05)80068 1 Green, M.B. (2006). Bullying in school: A plea for measure of human rights. Journal of Social Issues, 62( 1 ), 63 79. d oi : 10.1111/j.1540 4560.2006.00439.x Griffin, R. S. & Gross, A. M. (2004). Childhood bullying: Current empirical findings and future directions for research. Aggression and Violent Behavior, 9, 379 400. doi: 10.1016/S1359 1789(03)00033 8 Hawker, D. S. J., & Boulton, M. J. (2000). and psychosocial maladjustment: A meta analytic review of cross sectional studies. Journal of Child Psychiatry 41( 4 ), 441 455. doi: 10.1111/1469 7610.00629 Hinduja, S. & Patchin, J. W. (2008). Cyberbullying: A n exploratory analysis of factors related to offending and victimization. Deviant Behavior, 29( 2 ), 129 156. Doi: 10.1080/01639620701457816
115 Hinshaw S. P. (1992). Externalizing behavior problems and academic underachievement in childhood and adolescence: Ca usal relationships and underlying mechanisms Psychological Bulletin, 111, 127 155. doi: 10.1037/0033 2909.111.1.127 Hodges, E. V. E., Boivin, M., Vitaro, F., & Bukowski, W. M. (1999). The power of friendship: Protection against an escalating cycle of peer victimization. Developmental Psychology, 35, 94 101. doi: 10.1037/0012 16188.8.131.52 Hodges, E. V., Malone, M. J., & Perry, D. G. (1997). Individual risk and social risk as interacting determinants of victimization in the peer group. Developmental Psycholo gy, 33, 1032 1039. doi: 10.1037/0012 16184.108.40.2062 Holt, M. K. & Espelage, D. L. (2007). Perceived social support among bullies, victims, and bully victims. Journal of Youth and Adolescence, 36, 984 994. doi: 10.1007/s10964 006 9153 3 Hoza, B. (2007). P eer functioning in children with ADHD. Journal of Pediatric Psychology, 32( 6 ), 655 663. doi: 10.1093/jpepsy/jsm024 Hoza B., Mrug, S., Gerdes, A. C., Hinshaw, S. P., et al. (2005). What a spects of p eer r elationships a re i mpaired in c hildren w ith Attention Deficit/Hyperactivity Disorder? Journal of Consulting and Clinical Psychology 73( 3 ), 411 423. doi: 10.1037/0022 006x.73.3.411 Klomek, A. B., Sourander, A., Niemela, S., Kumpulainen, K., et al. (2009). Childhood bullying behaviors as a risk for suicide attempts and completed suicides: A population based birth cohort study. Journal of the American Academy of Child and Adolescent Psychiatry, 48( 3 ) 254 261. doi: 10.1097/CHI.0b013e318196b91f Kochenderfer Ladd, B. and peer victimization. Journal of School Psychology, 46, 431 453. doi: 10.1016/j.jsp.2007.07.005 Kowlaski R. M. & Limber, S. P. (2007). Electronic bullying among middle school students. Journal of Adolescent health, 41, S22 S30. doi: 10.1016/j.jadohealth.2007.08.017 Kumpulainen, K., Rasanen, E., Almqvist, F., Kresanov, K., et al. (1998). Bullying and psychi atric symptoms among elementary school age children. Child Abuse & Neglect, 22( 7 ), 705 717. doi: 10.1016/S0145 2134(98)00049 0 Kuppens, S., Grietens, H., Onghena, P., & Michiels, D. (2009). A longitudinal study of childhood social behavior: Inter informan t agreement, inter context agreement,
116 and social preference linkages. Journal of Social and Personal Relationships, 26( 6 7 ), 769 792 doi: 10.1177/0265407509342929 Ladd, G. W. & Kochenderfer Ladd, B. (2002). Identifying victims of peer aggression from ear ly to middle childhood: Analysis of cross informant data for concordance, estimation of relational adjustment, prevalence of victimization, and characteristics of identified victims. Psychological Assessment, 14( 1 ), 76 96. doi: 10.1037//1040 35220.127.116.11 L arson, J. J., Whitton, S. W., Hauser, S. T., & Allen, J. P. (2007). Being close and being social: Peer ratings of distinct aspects of young adult social competence. Journal of Personality Assessment, 89( 2 ), 136 148. doi: 10.1080/00223890701468501 Li, Q. (2 Computers in Human Behavior, 23, 1777 1791.doi: 10.1016/j.chb.2005.10.005 Li. Q. (2006). Cyberbullying in schools: A research of gender differences. School Psychology International, 27, 157 170. d oi: 10.1177/01430343060xxxxx March, J.S., Parker, J. D. A., Sullivan, K., Stallings, P., & Conners, C. K. (1997). The multidimensional anxiety scale for children (MASC): Factor structure, reliability, and validity. Journal of the American Aca demy of Child and Adolescent Psychiatry 36( 4 ) 554 565. doi: 10.1097/00004583 199704000 00019 Marsh, H. W., Nagengast, B., Morin, A. J. S., Parada, R. H., Craven, R. G., & Hamilton, L. R. (2011). Construct validity of the multidimensional structure of bul lying and victimization: An application of exploratory structural equation modeling. Journal of Educational Psychology, 103 (3), 701 732. doi:10.1037/a0024122 Marsh, H.W., Parada, R. H., Yeung, A. S., & Healey, J. (2001). Aggressive troublemakers and victim s: A longitudinal model examining the pivotal role of self concept. Journal of Educational Psychology, 93( 2 ), 411 419. doi: 10.1037/0022 0618.104.22.1681 Marini, Z. A., Dane, A. V., Bosacki, S. L., & YLC CURA (2006). Direct and indirect bully victims: Differential psychosocial risk factors associated with adolescents involved in bullying and victimization. Aggressive Behavior 32, 551 569. doi. 10.1002/ab.20155 Marsh, H. W., Morin, A. J. S., Nagengast, B., Parada, R. H ., Craven, R. G., & Hamilton, L. R. (2011). Construct validity of the multidimensional structure of bullying and victimization: An application of exploratory structural equation modeling. Journal of Educational Psychology, 103( 3 ), 701 732. doi: 10.1037/a00 24122 McCullough, G., Huebner, E. S., & Laughlin, J. E. (2000). Life events, self concept, and being. Psychology in the Schools, 37( 3 ), 281 290. doi: 10.1002/( SICI )1520 6807( 200005 )37:3<281::AID PITS8>3.3.CO; 2 U
117 Mercer S. H., & DeRosier, M. E. (2008). Teacher preference, peer rejection, and student aggression: A prospective study of transactional influence and independent contributions to emotional adjustment and grades. Journal of School Psychology, 46 661 685. doi: 10.1016/j.jsp.2008.06.006 Merikangas, K. R. (2005). Vulnerability factors for anxiety disorders in children and adolescents. Child and Adolescent Psychiatric Clinics of North America, 14, 649 679. doi: 10.1016/j.chc.2005.06.005 Mishna, F., Saini, M., & Sol perceptions of cyber bullying. Children and Youth Services Review, 31, 1222 1228. doi: 10.1016/j.childyouth.2009.05.004 Mitchell, K. J., Ybarra, M., & Finkelhor, D. (2007). The relative importance o f online victimization in understanding depression, delinquency, and substance use Child Maltreatment, 12( 4 ), 314 324. doi: 10.1177/1077559507305996 Nansel, T. R., Overpeck, M., Pilla, R. S., Ruan, W. J., Simons Morton, B., & Scheidt, P. (2001). Bullying behaviors among US youth: prevalence and association with psychosocial adjustment. Journal of the American Medical Association 285, 2094 2100. doi:10.1001/ jama.285.16.2094 Newcomb, A. F., Bukowski, W. M., & Bagwell, C. L. (1999). Knowing the sounds: Frien dship as a developmental context In A. W. Collins & B. Laursen (Eds .), Relationships as developmental contexts (pp. 63 84). Mahwah, NJ: Erlbaum. Nixon, E. (2001). The social competence of children with Attention Deficit Hyperactivity Disorder: A review of the literature. Child Psychology & Psychiatry Review, 6( 4 ), 172 180. doi: 10.1017/S1360641701002738 Nijmeijer, J. S., Minderaa, R. B., Buitelaar, J. K., Mulligan, A., Hartman, C. A., & Hoestra (2008). ADHD and social dysfunctioning. Clinical Psychology Review, 28 692 708. doi: 10.1016/j.cpr.2007.10.003 Normand S., Schneider, B. H., & Robaey, P. (2007). Attention Deficit/Hyperactivity Disorder and the challenges of close f riendship Journal of Canadian Academy of Child Adolescent Psychiatry 16( 2 ), 67 73 differences in the social emotional problems among frequent bullies, victims, and bully/victims. Psychology in the Schools, 46( 2 ), 100 115. doi: 10.1002/pits.20357 Olweus D. (2012). Invited expert discussion pa per cyberbullying: An overrated phenomenon? European Journal of Developmental Psychology, 1 19. doi; 10.1080/17405629.2012.682358 Olweus, D. (1993). Bullying at school: What we know and what we can do Oxford, Engla nd: Blackwell Publishers.
118 Orback, I., Mikulincer, M., Stein, D., & Cohen, O. (1998). Self representation of suicidal adolescents. Journal of Abnormal Psychology, 107( 3 ), 435 439. doi: 10.1037/ 0021 843X.107.3.435 Osterman, K. Bjorkqvist, K., Lagerspetz K. M. J., Kaukiainen, A., Huesmann, L. R., & Fraczek, A. (1994). Peer and self estimated aggression and victimization in 8 year old children from five ethnic groups. Aggressive Behavior, 20, 411 428. doi: 10.1002/1098 2337(1994)20:6%3C411::AID AB2480200602%3E3.0.CO;2 4 Patchin, J. W. & Hinduja, S. (2006). Bullies move beyond the schoolyard: A preliminary look at cyberbullying: Youth violence and juvenile justice, 4 148 169. doi: 10.1177/1541204006286288 Peskin, M. F., Tortolero, S. R., & Markham, C. M. (2006). Bullying and victimization among black and Hispanic adolescents. Adolescence, 41( 163 ), 467 484. No doi Piers, E. V., & Herzberg, D. S. (2002). Piers H concept scale second edi tion manual. Western Psychological Services Los Angeles, CA. Pornari, C. D. & Wood, J. (2010). Peer and cyber aggression in secondary school students: The role of moral disengagement, hostile attribution bias, and outcome expectancies. Aggressive Behavio r, 36, 81 94. doi: 10.1002/ab.20336 Raskauskas, J. & Stoltz, A. D. (2007). Involvement in traditional and electronic bullying among adolescents. Developmental Psychology, 43 ( 3 ), 564 573. doi: 10.1037/0012 1622.214.171.1244 Raty, L. K. A., Larsson, G., Soder felt, B. A., & Wilde Larsson, B. H. (2005). Psychosocial aspects of health in adolescence: The influence of gender, and general self concept. Journal of Adolescent Health, 36( 6 ), 22 28. doi: 10.1016/ j.jadohealth.2004.10.006 Renk, K ., & Phares, V. (2004). C ross informant ratings of social competence in children and adolescents. Clinical Psychology Review, 24 ( 2) 239 254. doi 10.1016/j.cpr.2004.01.004 Rigby, K. (2012). Bullying in schools: Addressing desires, not only behaviors. Educational Psychology Review, 24( 2 ), 339 348. d oi: 10.1007/s10648 012 9196 9 Rodkin, P. C. & Berger, C. (2008). Who bullies whom? Social status asymmetries by victim gender. International Journal of Behavioral Development, 32, 473 485. doi: 10.1177/0165025408093667 Roland E. & Galloway, G. (2004). Professional cultures in schools with high and low rates of bullying. School Effectiveness and School Improvement, 15 241 260. doi: 10.1080/09243450512331383202
119 Rys, G. S. & Bear, G. G. (1997). Relational aggression and peer relati ons: Gender and developmental issues. Merrill Palmer Quarterly, 43, 87 106. No doi Schwartz, D., Dodge, K. A., & Coie, J. D. (1993). The emergence of chronic Child Development, 64, 1755 1772. doi: 10.2307/1131467 Sharp, S. (1995). How much does bullying really hurt? Education Child Psychology, 12 81 8 8. doi : 10.1177/0143034300211003 Slonje, R. & Smith, P. K. (2008). Cyberbullying: Another main type of bullying. Scandinavian Journal of Psychology, 49 (2), 147 154. doi: 10 .1111/j.1467 9450.2007.00611.x Smith, P. K., Mahdavi, J., Carvalho, M., Fisher, S., Russell, S., & Tippett, N. (2008). Cyberbulling: Its nature and impact on secondary school pupils. Journal of Child Psychology and Psychiatry, 49( 4 ), 379 385. doi: 10.1111/ j.1469 7610.2007.01846.x Smith, P.K., & Shu, S. (2000). What good schools can do about bullying: Findings from a survey in English schools after a decade of research and action. Childhood: A Global Journal of Childhood Research, 7( 2 ), 193 212. doi: 10.1177/0907568200007002005 Smokowski, P. R. & Kopasz, K. H. (2005). Bullying in school: An overview of types, effects, family characteristics, and intervention strategies. Children & Schools, 27, 101 110. doi : 10.1093/cs/27.2.101 Solberg, M. & Olweus, D (2003). Prevalence estimation of school bullying with the Olweus Bully/Victim Questionnaire. Aggressive Behavior, 29, 239 268. doi: 10.1002/ab.10047 Sourander, A., Klomek, A. B., Ikonen, M., Lindroos, J., et al. (2010). Psychosocial risk factors associat ed with cyberbullying among adolescents. Archives of General Psychiatry, 67( 7 ), 720 728. doi: 10.1001/archgenpsychiatry.2010.79 Spieker, S. J., Campbell, S. B., Vandergrift, N., Pierce, K. M., & Cauffman, E. (2012). Relational aggression in middle childhoo d: Predictors and adolescent outcomes. Social Development, 21( 2 ), 354 375. doi: 10.1111/j.1467 9507.2011.00631.x Spriggs, A. L., Iannotti, R. J., Nansel, T. R., & Haynie, D. L. (2007). Adolescent bullying involvement and perceived family, peer, and school relations: Commonalities and differences across race/ethnicity. Journal of Adolescent Health, 41, 283 293. doi: 10.1016/j.jadohealth.2007.04.009 Storch, E. A., Brassard, M. R., & Masia Warner, C. L. (2003). The relationship of peer victimization to social anxiety and loneliness in adolescence. C hild Study Journal, 33( 1 ), 1 17.
120 Tokunaga, R. S. (2009). Following you home from school. A critical review and synthesis of research on cyberbullying victimization. Computers in Human Behavior, 26, 277 287. doi: 10. 1016/j.chb.2009.11.014 Tremblay, R. E., Japel, C., Perusse, D., McDuff, P., Boivin, M., Zoccolillo, M., & Montplaisir, J. Rousseau and Bandura Revisited. Criminal Behaviour and Mental Health 9 8 23. doi: 10.1002/cbm.288 Vermeiren, R. (2003). Psychopathology and delinquency in adolescents: A descriptive and developmental perspective. Clinical P sychology Review, 23, 277 318. doi: 10.1016/S0272 7358(02)00227 1 Wang, J., Iannotti, R. J., & Nansel, T. R. (2009). School bullying among adolescents in the United States: Physical, verbal, relational, and cyber. Journal of Adolescent Health, 45, 368 375. doi: 10.1016/j.jadohealth.2009 .03.021 Williams, K. R. & Guerra, N. G. (2007). Prevalence and predictors of internet bullying. Journal of Adolescent Health, 41( 6, Suppl ), S 14 S21. doi: 10.1016/j.jadohealth.2007.08.018 Ybarra, M., Diener West, D., & Leaf, P. (2007). Examining the overlap in Internet harassment and school bullying: Implications for school intervention. Journal of Adolescence, 27, 319 336. doi: 10.1016/j.adolescence.2004.03.007 Ybarra, M. L. & Mitchell, K. J., (2004). Online aggressor/targets, aggressors, and targets: A comparison of associated youth characteristics. Journal of C hild Psychology and Psychiatry, 45(7), 1308 1316. doi: 10.1111/j.1469 7610.2004.00328.x Ybrandt, H. (2008). The relation between self concept and social functioning in adolescence. Journal of Adolescence, 31 1 16. doi: 10.1016/j.adolescence.2007.03.004 Z immer Gembeck, M. J., Geiger, T. C., & Crick, N. R. (2005). Relational and physical aggression, prosocial behavior, and peer relations: Gender moderation and bidirectional associations. The Journal of Early Adolescence 25( 4 ), 421 452. doi: 10.1177/02724316 05279841
121 BIOGRAPHICAL SKETCH Jennifer E. Rosado Muoz was born in Orlando, Florida, and is the younger of two children. She earned her Bachelor of Arts in psychology at Walla Walla Unive rsity in 2002, with a minor in b iology, and went on to earn a Masters of Arts in counseling Scholastic Achievement Award in 2005, Jennifer took a full time position as a therapist tr eatment residential facility in Pendleton, Oregon. She then took a position at Lakeside Behavioral Healthcare in Orlando, Florida as a full time Jennifer relocated to Gainesville, Florida in August of 2008 to continue her 2010 During her graduate career she received the Molly Harrower Excellence in Psychodiagnostic Assessment Award in 2011. She is currently completing her fifth year in t he clinical psychology doctoral program at the University of Florida, as a psychology philosophy degree in August of 2013.