Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Social responsibility on the Internet : a socio-ecological approach to online aggression Law, Danielle M. 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2009_spring_law_danielle.pdf [ 2.26MB ]
Metadata
JSON: 24-1.0054370.json
JSON-LD: 24-1.0054370-ld.json
RDF/XML (Pretty): 24-1.0054370-rdf.xml
RDF/JSON: 24-1.0054370-rdf.json
Turtle: 24-1.0054370-turtle.txt
N-Triples: 24-1.0054370-rdf-ntriples.txt
Original Record: 24-1.0054370-source.json
Full Text
24-1.0054370-fulltext.txt
Citation
24-1.0054370.ris

Full Text

SOCIAL RESPONSIBILITY ON THE INTERNET: A SOCIO-ECOLOGICAL APPROACH TO ONLINE AGGRESSION by Danielle M. Law B.A., University of Alberta, 2002 M.A., University of British Columbia, 2004  A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Human Learning Development and Culture) THE UNIVERSITY OF BRITISH COLUMBIA April 2009  © Danielle M. Law, 2009  ABSTRACT This series of three studies examined online aggression. More specifically, using a socioecological lens, this work assessed the interplay among individual, peer, parental, and school factors on Internet aggression. Study I used existing data to compare traditional bullying with online bullying. This study revealed that adolescents view the construct of online bullying and victimization differently from its traditional counterpart. In addition, this work highlighted some of the similarities and differences among these forms of aggression. Study II used a mixedmethod approach to examine some of the individual predictors of online aggression. Through paper and pencil questionnaires and semi-structured interviews, this work further elucidated the differences between online aggression and traditional bullying. Specifically, adolescents defined online aggression in terms of the method used for aggressing online rather than the role they played in aggressive situations. In addition, significant gender and age differences were found. Interview data also indicated that online aggression was primarily reactive in nature, as opposed to proactive, and that adolescents use both confrontational and non-confrontational aggression online. In keeping with the socio-ecological approach, Study III examined some of the parental and school factors that influence online aggression. Results from this work showed that it was not parental monitoring or limit-setting around Internet use that reduced the likelihood that participants engaged in online aggression, rather it was the amount of child self-disclosure. Directions for future research and intervention/prevention are discussed.  ii  TABLE OF CONTENTS Abstract.......................................................................................................................................... ii Table of Contents ......................................................................................................................... iii List of Tables ................................................................................................................................. v List of Figures.............................................................................................................................. vii Co-Authorship Statement ......................................................................................................... viii Chapter I: Introduction................................................................................................................ 1 Chapter 2. The Changing Face of Bullying: An Empirical Comparison Between Traditional and Internet Bullying and Victimization ............................................................... 5 2.1 Bullying................................................................................................................................. 5 2.1.1. Cyberbullying ............................................................................................................... 6 2.1.2. Measuring Bullying ...................................................................................................... 8 2.1.3. Predictors of Bullying ................................................................................................... 9 2.2. Method ............................................................................................................................... 20 2.2.1. Procedures................................................................................................................... 20 2.2.2. Measures ..................................................................................................................... 22 2.3. Results................................................................................................................................ 25 2.3.1. Question 1: Are Online and Offline Bullying Separate Constructs? .......................... 25 2.3.2. Question 2: Are There Grade and Sex Differences in Cyberbullying/victimization, Traditional Bullying, and Traditional Victimization? .......................................................... 29 2.3.3. Questions 3: Are There Differences in the Constructs Which Predict Traditional Bullying, Traditional Victimization, and Cyberbullying/victimization?.............................. 32 2.4. Discussion .......................................................................................................................... 38 2.4.1. Limitations .................................................................................................................. 45 2.4.2. Conclusion .................................................................................................................. 46 2.5. References.......................................................................................................................... 47 Chapter 3. Cyberbullying Versus Schoolyard Bullying: Same or Different? ...................... 65 3.1. Aggressive Behaviours During Adolescence..................................................................... 65 3.1.1. Proactive vs. Reactive Aggression.............................................................................. 66 3.1.2. Confrontational vs. Non-Confrontational Aggression................................................ 67 3.1.3. Relational Aggression................................................................................................. 68 3.1.5. Online Aggression ...................................................................................................... 69 3.2. Adolescent Internet Use and Online Aggression ............................................................... 70 3.2.1. Instant Messaging ....................................................................................................... 71 3.2.2. Cell Phones ................................................................................................................. 73 3.2.3. Social Networking Sites.............................................................................................. 75 3.2.4. YouTube ..................................................................................................................... 76 3.3. Predictors of Bullying Behaviours and Victimization ....................................................... 76 3.3.1. Age and Sex Differences. ........................................................................................... 77 3.3.2. Self-Esteem. ................................................................................................................ 77 3.3.3. Summary ..................................................................................................................... 79 3.4. Method ............................................................................................................................... 80 3.4.1. Participants.................................................................................................................. 81 3.4.2. Questionnaire Data – Outcome Variables................................................................... 82 3.4.2. Questionnaire Data – Predictor Variables................................................................... 83 iii  3.4.3. Questionnaire Data – Covariates ................................................................................ 84 3.5. Results................................................................................................................................ 88 3.5.1. Determining the Factor Structures of Online Aggression and Reactive/Proactive Aggression. ........................................................................................................................... 88 3.5.2. Research Question 1: What are the Sex and Grade Differences Associated with Online Aggression?............................................................................................................... 97 3.5.3. Research Questions 2 and 3: How Does Self-Esteem Predict Online Aggression? Is Online Aggression Primarily Proactive or Reactive? ......................................................... 100 3.5.4. Research Question 4: Are Adolescents Using Confrontational or Non-Confrontational Means for Engaging in Online Aggression?....................................................................... 113 3.6. Discussion ........................................................................................................................ 116 3.6.1. Conclusions............................................................................................................... 124 3.6.2. Limitations ................................................................................................................ 125 3.7. References........................................................................................................................ 127 Chapter 4. Is Monitoring Adolescent Internet Use Really the Answer? The Association Among Parenting, School, and Peer Factors on Online Aggression .................................... 138 4.1.1. Family Predictors of Bullying....................................................................................... 138 4.1.2. Peer and School Predictors ....................................................................................... 142 4.2. Methods............................................................................................................................ 144 4.3. Results.............................................................................................................................. 146 4.3.1. Determining the Factor Structure for Parenting Items.............................................. 146 4.4. Discussion ........................................................................................................................ 159 4.4.1. Conclusions............................................................................................................... 161 4.4.2. Limitations ................................................................................................................ 162 4.5. References........................................................................................................................ 163 Concluding Chapter.................................................................................................................. 170 5.1. Study I.............................................................................................................................. 170 5.2. Study II............................................................................................................................. 171 5.3. Study III ........................................................................................................................... 174 5.4. Limitations of this Work.................................................................................................. 175 5.4. Conclusions and Implications .......................................................................................... 176 5.5. References........................................................................................................................ 179 Appendices................................................................................................................................. 181 Appendix A............................................................................................................................. 181 Appendix B ............................................................................................................................. 203 Appendix C ............................................................................................................................. 210 Appendix D............................................................................................................................. 229  iv  LIST OF TABLES Table 2.1. Crosstabs analyses for Bullying, Victimization, Cyberbullying and Cybervictimization…………………………………………………………………………..  22  Table 2.2: Unweighted least squares factor analysis pattern matrix for Bullying & Victimization items…………………………………………………………………………..  27  Table 2.3: Unweighted least squares factor analysis pattern matrix for Bullying items….  28  Table 2.4: Unweighted least squares factor analysis pattern matrix for Victimization items………………………………………………………………………………………….  28  Table 2.5: Unweighted least squares factor analysis pattern matrix for Cyberbullying and Cybervictimization items………………………………………………………………..  25  Table 2.6. Main and interaction effects for Cyberbullying/victimization, Traditional Bullying, and Traditional Victimization……………………………………………………..  30  Table 2.7: Summary of hierarchical multiple regression model for Cyberbullying/victimization…………………………………………………………………. 34 Table 2.8: Summary of hierarchical multiple regression model for Traditional Bullying…  35  Table 2.9: Summary of hierarchical multiple regression model for Traditional 36 Victimization………………………………………………………………………………… Table 2.10: Summary of important items according to the Pratt index……………………..  37  Table 3.1. Correlation matrix for online aggression items…………………………………..  90  Table 3.2: Unweighted least squares factor analysis pattern matrix for all online aggression items…………………………………………………………………………….  91  Table 3.3. Correlation matrix for Proactive and Reactive items…………………………….  94  Table 3.4: Factor loadings from unweighted least squares factor analysis of Proactive and Reactive items forced to two factors………………………………………………….  95  Table 3.5: Factor loadings from unweighted least squares factor analysis of Proactive and Reactive items forced to one factors ……………………………………………………  96  Table 3.6: Summary of important proactive and reactive items according to the Pratt index………………………………………………………………………………………….  96 v  Table 3.6: Summary of hierarchical multiple regression model for Aggressive Messaging  103  Table 3.7: Summary of hierarchical multiple regression model for Hostile Websites…….  105  Table 3.8. Summary of results from the hierarchical multiple regression model for Embarrassing Photos and/or Videos…………………………………………………………  106  Table 3.9: Summary of interview themes…………………………………………………..  108  Table 4.1. Correlation matrix for parenting items…………………………………………..  148  Table 4.2: Factor loadings for the unweighted least squares factor analysis of parenting items………………………………………………………………………………………….  149  Table 4.3. Comparison of the R2-value for hierarchical multiple regression model with different orders of entry…………………………………………………………………….  151  Table 4.4. Summary of hierarchical multiple regression for Belonging & Similarity’s predictive value over & above Parenting Behaviours for Aggressive Messaging…………... 153 Table 4.5. Summary of hierarchical multiple regression for Parenting Behaviour’s predictive value over & above Belonging and Similarity for Aggressive Messaging……….  154  Table 4.6. Summary of hierarchical multiple regression for Belonging & Similarity’s predictive value over & above Parenting Behaviours for Hostile Website Development…..  155  Table 4.7. Summary of hierarchical multiple regression for Parenting Behaviour’s predictive value over & above Belonging and Similarity for Hostile Website Development. Table 4.8. Summary of hierarchical multiple regression for Belonging & Similarity’s predictive value over & above Parenting Behaviours for Commenting/Posting Embarrassing Photos/Videos………………………………………………………………… Table 4.9. Summary of hierarchical multiple regression for Parenting Behaviour’s predictive value over & above Belonging and Similarity for Commenting/Posting Embarrassing Photos/Videos………………………………………………………………..  156  157 158  vi  LIST OF FIGURES Figure 1.1. A Socio-ecological framework for examining bullying behaviours…………….. 4 Figure 2.1. Grade and sex interactions for Cyberbullying/victimization.................................  31  Figure 2.2. Grade and sex interactions for Traditional Bullying……………………………  31  Figure 2.3. Grade and sex interactions for Traditional Victimization……………………….  32  Figure 3.1. Screeplot of factor loadings for Proactive and Reactive Aggression……………  95  Figure 3.2. Grade and sex differences for Aggressive Messaging…………………………... 98 Figure 3.3. Grade and sex differences for Creating Hostile Websites……………………….  99  Figure 3.4. Grade and sex differences for Commenting/Posting Embarrassing Photos/Videos………………………………………………………………………………..  100  vii  CO-AUTHORSHIP STATEMENT The order of authorship in each manuscript was determined according to the American Psychological Association “Ethical Principles of Psychologists and Code of Conduct”, Principle 6.23 (APA, 1992)1. The co-authors, Dr. Jennifer Shapka, Dr. Shelley Hymel, and Dr. José Domene, contributed to the conceptualization and design of the studies. Data for Study I were drawn from a larger study that was developed to meet part of British Columbia’s Ministry of Education mandate for school accountability. Data for Studies II and II were collected by the first author, who also performed the analyses and prepared the manuscripts for each of the three studies. The co-authors provided feedback throughout the research and writing processes.  1  American Psychological Association. (1992). Ethical principles of psychologists and code of conduct. American Psychologist, 47, 1597-1611.  viii  CHAPTER I: INTRODUCTION The prevalence of aggression in schools has been well-documented; in fact, the issue of schoolyard bullying and aggression has received much attention in both the media and academic realms. Current work in this area has established that the implications of aggressive acts can be detrimental to both the victim’s and the perpetrator’s socio-emotional wellbeing and their academic success (e.g. Boivin, Hymel, & Bukowski, 1995; Crick, Grotpeter & Rockhill, 1999; Swearer, Song, Cary, Eagle & Mickelson, 2001). Despite the wealth of existing research on schoolyard aggression and bullying, very little systematic research has been conducted on a form of aggression that occurs over the Internet. Labeled Cyberbulling or Internet aggression, there are now countless incidences of this form of aggression reported by media, popular literature, and in conversations with administrators. This series of studies is one of the first to systematically explore Cyberbullying in an empirical way, as a preliminary step in elucidating how it is similar to and different from more traditional forms of bullying. As Information and Communication Technologies (ICTs) continue to infiltrate all aspects of our lives, understanding aggression in this context becomes imperative. We would be remiss if we assumed that aggression in virtual environments is identical to face-to-face aggression. Similarly, we cannot assume a ‘one size fits all’ in terms of prevention and intervention, or even in terms of the identification and meaning of these online aggressive acts to the perpetrators and victims. Fortunately, it is possible to draw on the large body of literature on traditional forms of aggression to generate ideas and hypotheses about what aggression conducted over the Internet may be like.  1  The use of the Internet for communicating with others has become a part of everyday life, and particularly so for adolescents (Gross, Juvonen, & Gable, 2002; Wallace, 1999). Research on Internet use has found that 94% of Canadian youth report having access to the Internet from their homes, which is a significant increase from the 79% in 2001 (MNet, 2005). Research has also found that adolescents use the Internet to seek out opportunities to interact with school-based peers (Gross et al., 2002), to overcome shyness, to experiment with identities, and to facilitate social relationships (Maczewski, 2002; Valkenburg, Schouten, & Peter, 2005). Despite these potential benefits to online socializing, it appears that adolescents are also using the Internet as an arena for bullying. Cyberbullying has been defined as a form of intentional aggression where an individual, or a group of individuals, use ICTs, such as email, websites (developed specifically for the purpose of humiliating or degrading others), Instant Messaging (IM), and/or text messaging on cell phones to inflict harm on others (through gossiping and/or by posting embarrassing things; Ybarra & Mitchell, 2004). Exacerbating this situation is the fact that adolescents often feel more comfortable communicating online and are more likely to disclose information about themselves or others over the Internet than in person (Peter, Valkenburg, & Schouten, 2005; Ward & Tracey, 2004). This means that acts of aggression over the Internet may be even more prevalent and vicious than other forms of bullying. Indeed, researchers have found that Cyberbullying is becoming increasingly prevalent among adolescents (Beran & Li, 2005; Patchin & Hinduja, 2006). An established theoretical perspective for examining traditional aggression and bullying constructs is through a socio-ecological lens (Espelage & Swearer, 2004). The  2  current work uses this framework to explore online aggression. According to this framework, it is argued that the decisions and behaviours of an individual are influenced, not by the individual alone, but by the reciprocal interplay of that individual and his family, peers, and environmental context (Bronfrenbrenner, 1979; Coie & Jacobs, 1993; Fraser, 1996; Garbarino, 2001). Consequently, from a socio-ecological perspective, bullying behaviours are not simply a product of the conflictual dyadic relationship between bully and victim; rather, bullying arises from the interaction among a number of individual, group, and contextual factors (Atlas & Pepler, 1998; Garbarino & deLara, 2002; Olweus, 1993a; Swearer & Doll, 2001; Swearer & Espelage, 2004). Specifically, research has found that bullying interactions are influenced by (a) context, such as the classroom or playground (Atlas & Pepler, 1998; Craig, Pepler & Atlas, 2000; Doll, Song, & Siemers, 2004; Duncan, 2004); (b) individual characteristics of the bully and the victim, such as physical size, sex, and age, (Atlas & Pepler, 1998; Farmer, et al., 2002; Underwood, Scott, Galperin, Bjornstand, & Sexton, 2004); (c) the presence and influence of peer groups (Pellegrini & Long, 2004; Rodkin, 2004); and (d) family related predictors such as parent/child relationships (Duncan, 2004; See Figure 1.1). In order to properly address these issues from a socio-ecological lens, this work is comprised of three studies, each of which represent a stand-alone manuscript, but that build upon one another. More specifically, Study 1 compares traditional bullying and victimization with online bullying/victimization by examining some of the individual and peer predictors of online vs. offline aggression. Study 2 uses a mixed-method approach to examine the construct of online aggression in a more in-depth way; specifically focusing on assessing some of the individual predictors unique to this form of bullying  3  (based on the findings which emerged from Study 1). This study also attempts to discern some of the motivational factors for aggressing online. Finally, Study 3 explores some of the contextual factors that may influence online aggression, including parenting behaviours at home, and an individual’s sense of school belonging and school climate. Figure 1.1. A Socio-ecological framework for examining bullying behaviours.  4  CHAPTER 2. THE CHANGING FACE OF BULLYING: AN EMPIRICAL COMPARISON BETWEEN TRADITIONAL AND INTERNET BULLYING AND VICTIMIZATION2 The primary purpose of this study is to compare online and offline bullying/victimization in order to determine the extent to which these two types of aggression are unique – whether they are predicted by distinct factors and are engaged in by different individuals. To accomplish this, the present study examines whether some of the individual, peer, and contextual factors that are related to offline bullying and victimization are also related to Cyberbullying.  2.1 BULLYING Bullying has been defined in a number of ways. In describing it from the victim’s perspective, Olweus states that, “A person is being bullied when he or she is exposed, repeatedly and over time, to negative actions on the part of one or more other persons” (1993a, p.9). Other researchers have described bullying as a form of extreme peer aggression where one or more students seek to gain status and power by harming and intimidating a more vulnerable peer through repeated physical, verbal or psychological aggression (Hoover, Oliver, & Hazler, 1992; Olweus, 1991, Pellegrini & Bartini, 2001; Smith & Boulton, 1990). It can take the form of overt, direct attacks on the victim, or can include social isolation or deliberate exclusion from a group (Olweus, 1994a). With these definitions in mind, it is clear that bullying is not limited to physical aggression, but that  2  A version of this chapter will be submitted for publication. Law, D. M., Hymel, S., & Shapka, J.D. (2009). The Changing Face of Bullying: An Empirical Comparison Between Traditional and Internet Bullying and Victimization  5  it also includes verbal attacks and that these attacks are intended to harm (Bosworth, Espelage, & Simmon, 1999; Underwood, Galen, & Paquette, 2001). Researchers also differentiate between physical and social/relational aggression. Physical aggression involves harming others through physical force, verbal threats, and/or instrumental intimidation (Atlas & Pepler, 1998; Craig et al., 2000; Crick & Grotpeter, 1995); whereas social aggression is a form of aggression designed to purposefully manipulate and damage peer relationships through gossiping, spreading rumours, and eliciting either verbal or non-verbal peer rejection (Crick, 1995; Crick & Grotpeter, 1995; Galen & Underwood, 1997; Underwood, et al., 2004; Xie, Swift, Cairns, & Cairns, 2002). Thus, bullying can be defined as intentional and repetitive physical, emotional, and/or social aggression, or exclusion of one or a group of individuals by another individual or group. 2.1.1. Cyberbullying Cyberbullying is a relatively new phenomenon that “involves the use of ICTs to support deliberate, repeated, and hostile behaviour by an individual or group that is intended to harm others” (Belsey, 2005, p. 8). Such technological tools used for this form of bullying include Instant Messaging (IM) and other chat tools, cell phone pictures, websites, text-messaging, blogs, and emails. In general, Cyberbullying incidents involve using ICTs in a variety of ways to send or post hateful and threatening messages over the Internet. To date, few empirically-based studies on Cyberbullying have been published in the academic literature. However, work is beginning to emerge at academic conferences, and the phenomenon has been well described by popular and news media, as well as in anecdotal reports written by principals, teachers, and school administrators. For 6  example, work conducted by Hinduja and Patchin (2008) identified a 12-year-old girl from Michigan who described how “one girl actually told me she would come and murder my parents and kill me personally. She made me cry so hard that I threw up. So, I know first hand what its like to be bullied beyond your imagination” (Hinduja & Patchin, Share Your Story section). Similarly, Internet privacy and security lawyer, Parry Aftab (n.d.), describes how: Sometimes the bullying takes place through the use of a website where the victims are ridiculed by being voted the ugliest, fattest or sluttiest student in the school. Sometimes cell phone cameras are used to take pictures of the victim in a locker room or dressing room or in an otherwise compromising situation. Their passwords may be stolen and they may find themselves locked out of their own accounts or blamed for hateful messages sent out by the bully posing as the victim. Their faces may be superimposed on pornographic images. Or their personal information may be posted online and used to provoke an offline attack by a hate or predator group. The methods used by bullies online are limited only by the limitless imagination of other preteens and teens (What is Netbullying or Cyberbullying section, para. 2). The consequences of Cyberbullying can be devastating and long lasting. A 13-year-old girl from Virginia describes how: This one guy asked me how I looked and wanted to know about my body and stuff and I just flat out told him leave me alone!!! …That was over four years ago and I still remember every thing he said to me. Every exact word. I felt awful. I hated it. I wanted to tell my parents but I was afraid that they would never let me  7  chat again and I know that's how a lot of other kids feel. It is a bad feeling knowing that people that don't know you are judging you (Hinduja & Patchin, Share Your Story section). As is evident from these examples, bullying is no longer limited to the playground or school setting. Children can become targets of Cyberbullying simply by carrying a cell phone or logging on to the Internet. Moreover, this aggression can be quite severe and extreme, and can lead to tremendous psychological distress that can transfer to offline physical harm. Benfer (2001) provides the example of vicious and offensive messages, about a high school girl who suffered from obesity and multiple sclerosis, that were posted anonymously online. These messages were designed to deliberately humiliate and demean her. Over time, the online hate entered the offline world when perpetrators wrote profanity on the sidewalk outside her home, vandalized her car, and threw acid-filled bottles at her front door. 2.1.2. Measuring Bullying One of the first goals of this study is to determine whether online and offline bullying are similar constructs. Traditional bullying has been measured in a variety of ways including self-report questionnaires (Espelage, Bosworth, & Simon, 2000; Hymel, Bowker, & Woody, 1993; Jeffrey, Miller, & Linn, 2000), longitudinally (Salmivalli, Lappalainen, & Lagerspetz, 1997), peer report techniques (Espelage & Holt, 2001), classroom and playground observations (Atlas & Pepler, 1998; Craig et al., 2000; Pepler, Craig, Roberts, 1998) and in experimental lab settings (Underwood & Bjornstad, 2001; Underwood, Hurley, Johanson, & Mosley, 1999).  8  To date, self-report questionnaires (administered either over the Internet, by telephone, or through paper and pencil methods) have been the primary method used to measure Cyberbullying in peer reviewed academic publications (Kowalski & Limber, 2007; Li, 2007; Patchin & Hinduja, 2006). The current study followed similar methods of data collection as previous researchers by using self-report information to compare online bullying/victimization with its traditional offline counterpart. Moreover, this work used Factor Analysis, Univariate Analysis of Variance, and Linear Regressions to make comparisons between online and offline bullying; a task yet to be undertaken in a single study. 2.1.3. Predictors of Bullying In comparing traditional forms of bullying with Cyberbullying, it is important to also compare the factors that influence bullying. In fact, several factors have consistently been identified as predictors of schoolyard bullying. These include individual characteristics, peer characteristics, family influences, as well as other contextual predictors, such as the classroom or playground (Espelage & Swearer, 2004). Unfortunately, we know little about how these factors are related to online aggression. As a starting place, this study compares how individual characteristics are related to both offline bullying and victimization with its online counterpart. Extant research has examined the contributions of a number of individual predictors of traditional aggression, such as sex, age, physical size, empathy and caring, and internalizing problem behaviours such as depression and anxiety (Espelage, Mebane, & Adams, 2004; Espelage, Mebane, & Swearer, 2004; Peplar et al., 1998; Swearer, Grills, Haye & Cary, 2004). Using the body of research on offline bullying as a foundation, this study compared offline bullying 9  with online bullying by examining their relationship with age, sex, self-esteem, perceptions of academic performance, perceptions of peer acceptance, and feelings of belonging and engagement in school – factors that have been well documented as having a relationship with offline bullying. Age. Research has found that the nature of aggressive acts changes with age. For example, regardless of sex, children as young as 3 years of age have been found to engage in both physical and social aggressions (Crick, Casas, & Mosher, 1997); by middle childhood (ages 6 to 12) social aggression seems to become more prominent, and boys seem to engage in physical aggression more than girls do (Coie & Dodge, 1998; Maccoby, 1998). During adolescence, physical aggression seems to decline for most (Coie & Dodge, 1998), while social aggression remains quite high (Sharp, 1995). The current study focused on adolescents between the ages of 13 and 18 because it is a developmental period that is not only characterized by physical, cognitive, emotional, and social changes (Brooks-Gunn; 1987, Harter, 1999), but also because both social aggression and Internet use increase dramatically during this time (MNet, 2005; Underwood, 2003). As such, it is necessary to investigate how online aggression might differ with age. Previous research on Internet use has found that adolescent Internet and computer access and use increase with age (MNet; Rideout, 2005). Other research has also found that with increased Internet use and access, adolescents are at increased risk for engaging in hateful, illegal, sexually provocative, bullying or harassing activities on the Internet (Young, 2008). With this in mind, for the current study, it was hypothesized that older adolescents would engage in Cyberbullying more than younger adolescents.  10  Sex. Previous research has identified sex differences in patterns of aggression. That is, research has found that boys tend to use considerably more physical aggression than girls (Coie & Dodge, 1998; Espelage, et al., 2000), while girls tend to be more socially aggressive (Crick et al., 1999). A possible explanation for the existence of sex differences in patterns of aggression was proposed by Crick (1995) and Maccoby (1998) who conceptualized a gender-sensitive model of aggression. This model postulates that children and adolescents choose the aggressive behavior that would be most detrimental to the social goals of their peers. For example, boys use physical aggression to impair the instrumental goals (i.e., establish dominance and social status in the class or school) of other boys (Pellegrini, 2001a; Pellegrini & Bartini, 2001; Pellegrini & Long, 2002). Olweus (1993b) and Atlas and Pepler (2001) studied the characteristics of the bullyvictim dyad and did find that bullies tended to be physically larger and stronger, with more aggressive personalities, while their victims were characterized as being smaller, weaker, more timid, and anxious. Meanwhile, research has found that girls are primarily engaged in social aggression because they place a high value on social relationships. As such, they tend to aggress in more indirect ways (such as by gossiping and spreading rumours) in order to damage or obstruct the social goals of other girls (Conway, 2005; Crick, 1995, 1997). Although some studies have supported this claim (Crick & Grotpeter, 1995; Lagerspetz, Bjorkqvist, & Peltonen, 1988), other studies have either found no significant sex differences (Rys & Bear, 1997) or that boys exhibit more socially aggressive behaviours than girls (Tomada & Schneider, 1997). Despite the inconsistency in research findings with respect to which sex aggresses socially more than the other, it is clear that boys are the dominant participants of physical aggression and  11  that physical size plays a large role in the aggressor/victim dyad (Coie & Dodge, 1998; Moffitt & Caspi, 2001). This said, research conducted by Cole, Zahn, Waxler, and Smith (1994) found that girls display two to three times as many positive emotions when faced with frustration or disappointment. This is compatible with research that suggests that girls are socialized against overtly expressing negative emotions, and instead to be nice at all times (Zahn-Waxler, 2000). In addition, research has also found that girls were more likely to express their irritation and anger through non-verbal gestures, for example, with hostile stares, gestures, and eye-rolling (Paquette & Underwood, 1999; Underwood, 2004). This form of social exclusion is, again, consistent with the literature that proposes that girls are socialized to refrain from blatantly displaying signs of hostility and aggression (Underwood et al., 2004; Zahn-Waxler). As evidence for this, in the Underwood et al. (2004) study, once the provoker was out of range, girls were more readily than boys to proceed to speak negatively about the provoker behind their back. In addition, research has proposed that girls are more likely to aggress socially because they are more aware of and have a more sophisticated understanding of people’s social ties than boys, which puts them in a better position to manipulate and hurt (Galen & Underwood, 1997). With the advent of Internet socializing it is possible that girls will feel even more comfortable being openly hostile online because the Internet offers a sense of protection from cultural norms and watchful eyes of parents. For example a 12year-old girl from mid-western Canada states, “In school ... you don't want anyone to think of you as a "gossip" or someone who says things about other people. Everyone wants to be "nice." You don't have to be nice [online] if you don't want to” (MNet, 2005).  12  In addition, research on adolescent online activities has found that children feel more comfortable communicating online and would say things they would not normally say offline (Peter et al., 2005; Ward & Tracey, 2004). With the various media for online socializing, it is possible that children and adolescents mask their emotions offline but not in an anonymous or protected online setting. For example, girls who engage in, or are victims of, social forms of bullying may be more cautious in displaying overt forms of hostility or retaliation offline, but may engage in social bullying online by expressing their hostility, textually, through rumour spreading and gossiping. Given that the Internet provides girls with an arena that facilitates more social forms of bullying, such as gossiping and rumour spreading, it is hypothesized that girls are more likely to engage in online bullying than boys. Self-Esteem: How an individual perceives him or herself is particularly heightened during adolescence because of the influx of physical, social and emotional changes that are simultaneously occurring during this developmental period (Arnett, 1999). During this time, adolescents undergo puberty and the related physical changes, experience an increase in abstract thinking, and are subjected to a host of differing and often conflicting social expectations (Brooks-Gunn; 1987, Harter, 1999). With these changes comes the increased risk for various externalizing (e.g., delinquency, substance abuse, aggression) or internalizing (e.g., depression, feelings of irritation, disruption to sleep) problems (Berryman, Smyth, Taylor, Lamont, & Joiner, 2002). Numerous studies have detailed the inverse relationship between self-esteem and problem behaviours – adolescents with higher self-esteem are less likely to experience both externalizing and  13  internalizing problem behaviours (Cole, etal., 2001; Haddock & Sporakowski, 1982; Labouvie-Vief, Chiodo, Goguen, & Diehl, 1995). In concordance with these views, researchers have found that lower self-esteem is linked to higher levels of aggression and victimization (Marsh, Parada, Yeung, & Healey, 2001; Moretti, Holland, & McKay, 2001; Olweus, 1994b). However, some researchers have found that high self-esteem can also be positively related to aggression and bullying (O’Moore & Kirkham, 2001; Orobio de Castro, Brendgen, Van Boxtel, Vitaro, and Schaepers, 2007). For example, Orobio de Castro and colleagues found that students between grades 3 and 7 who felt overly positive about themselves were more likely to engage in physical and proactive aggression. These findings are further supported by research which examined relational aggression and found that self-perceived popularity was positively related to relational aggression in grades 9 to 12 boys and girls (Mayeux & Cillessen, 2008). Although conflicting findings have been reported, it is clear that self-esteem plays a role in influencing aggressive behaviours. Recent research has found that adolescents tend to feel more comfortable online and, thus, potentially have positive self-esteem online (Peter et al., 2005). Given this, it is possible for those who do not normally aggress offline, due to lower self-esteem, to aggress online where they feel more comfortable. As such, it is hypothesized that adolescents who report lower levels of selfesteem are more likely to engage in Cyberbullying. Peer Acceptance: Research has identified the importance of peers and peer groups in human development. In fact, peer groups have played a significant role in fostering learning, kindness, and identity development (Harris, 1998; Hodges, Malone, & Perry, 14  1997; Hymel, Rubin, Rowden, & LeMare, 1990; Zimmer-Gembeck, Geiger, & Crick, 2005). Unfortunately, peer groups can also facilitate aggressive behaviours (McFarland, 2001). Research has established the importance of identifying with, and feeling accepted by a group (whether by race, academic achievement, sexual orientation), especially during adolescence (Brown, 2000; McPherson, Smith-Lovin, & Cook, 2001). Social identity theory proposes that the self-perceptions, attitudes, and behaviors individuals have toward group members (in-group) and non-group members (out-group) derive from their desire to belong to a group that they perceive to be superior. That is, in order to enhance their own self-esteem, adolescents strive to identify with a group that is perceived to be of a high status on the social hierarchy (Brown, 2000; Gini, 2006). As a result, group members demonstrate favoritism towards members of their own group and discriminate against out-group members. In fact, Ojala and Nesdale (2004) found that children tended to consider bullying behaviour more acceptable if it was consistent with the norms of their group and when it was directed toward an out-group member who posed a potential threat to the in-group. Therefore, when it comes to aggression, the interactions of aggressors and victims must be considered within the context of peer groups. In addition, other research on peer groups has determined that individuals tend to sort themselves across an array of similarities (i.e., homophily; Rodkin, 2004). Although individuals can selectively affiliate with others according to an infinite number of commonalities, they tend to aggregate according to three homophilic groups. First is the sharing of key beliefs and goals, as well as similar personality characteristics. For example, children are more likely to affiliate with those who are similarly motivated  15  academically, extraverted or introverted, aggressive or prosocial, among other characteristics (McPherson, Smith-Lovin, & Cook, 2001). The second factor involves demographic attributes such as sex, race, age, socioeconomic and social class (Benenson, Maisee, Dolenszky, Dolensky, Sinclair, & Simpson, 2002; Maccoby, 1998). The third grouping factor involves shared interests, such as participating in similar extracurricular activities, watching similar TV shows, enjoying similar pasttimes, or frequenting similar places such as malls, parks, or online venues (McPherson et al.). Adolescence is a time when the division of individuals across homophilic groups seems most prominent, as exemplified by the different cliques and crowds that exist in middle school and high school (Brown, Mory, & Kinney, 1994). Cliques are arranged hierarchically where leaders within a clique or group use their popularity as a method for controlling their group (Adler & Adler, 1998). Despite their popularity, research has found that cliques are unstable and members within a group may react aggressively against each other in order to maintain their position within the social hierarchy. More specifically, research has found that high network centrality and popularity are associated with more socially aggressive behaviours (Dodge, Coie, Pettit, & Price, 1990; Estell, Cairns, Farmer, & Cairns, 2002; Rodkin, Farmer, Pearl, & Van Acker, 2000; Vaillancourt, Hymel, & McDougall, 2003; Xie et al., 2002). Furthermore, recent findings indicate that having friends and being liked by peers decreases the likelihood of becoming victimized, especially if the friends are of high status (Pellegrini, Bartini, & Brooks, 1999; Pellegrini & Long, 2002). As such, it makes sense that the battle to be among the more popular peers is so important.  16  Group size also plays a large role in regulating bullying behaviour. In a study conducted by Benenson and colleagues (2002) children in groups were found to be more boastful and conflictual than children in dyads who were found to be more selfdeprecating and self-sacrificing. The authors further suggest that group interactions differ from dyadic interactions because groups afford greater anonymity and, compared with dyads, group members could potentially receive less serious consequences for their behaviour. In fact, many adolescents who are not aggressive themselves reinforce aggressive behaviours through applause, laughter, or supporting aggressive acts nonconfrontationally through gossip and rumour spreading and/or social exclusion, which can afford the aggressor more power (O’Connell, Pepler, & Craig, 1999; Salmivalli, Huttunen, & Lagerspetz, 1997). As noted earlier, adolescents use the Internet to interact and socialize with peers (Gross et al., 2002), to build and facilitate friendships and social relationships, and to overcome shyness (Maczewski, 2002; Valkenburg, Schouten, & Peter, 2005). Thus, the Internet is an ideal space for individuals to join a clique or feel they are members of a particular group. In fact, some children with certain temperament factors (e.g., shyness), may be more comfortable online and may be more successful in becoming part of a desired peer group through online socialization. Given this, it was hypothesized that there would be no significant differences in online aggression between adolescents who report higher or lower perceptions of peer acceptance. In other words, adolescents with high perceptions of peer acceptance would aggress online as a means for maintaining social status within their school, and adolescents with lower feelings of peer acceptance  17  would participate in online aggression as a means of interacting and fitting in with offline peers, peers they would otherwise never associate with. Perceptions of Academic Achievement: Research has demonstrated a significant relationship between academic achievement and bullying/victimization (Schwartz 2000; Beran, Hughes, & Lupart, 2008). Moreover, empirical evidence from a growing body of work suggests that bullies and victims may fair poorly in school, although for different reasons. For example, bullies perform poorly in school because they do not tend to be as invested in their school performance (Beran et al., 2008; Carlson & Cornell, 2008; Hoglund, 2007; Schwartz, Gorman, Nakamoto & Toblin, 2005; Song, Swearer, Eagle, & Cary, 2000). In contrast, victims tend to perform poorly academically due to the depressive symptoms (Hawker & Boulton, 2000), loneliness (Boivin et al, 1995), and lower self-esteem that are associated with being victimized (Boulton & Smith, 1994; Egan & Perry, 1998; Graham & Juvonen, 1998). In accordance with the findings from traditional bullying literature, it was hypothesized that lower perceptions of academic achievement would be associated with increased levels of online bullying. Belonging and engagement in school: The extent to which individuals feel a sense of belonging and engagement with their schools has been linked to various externalizing problem behaviours (e.g., substance abuse, school dropout; Hawkins, Catalano, & Miller, 1992), and to various positive outcomes (e.g., increased academic performance, increased social competence, good peer relationships; Luiselli, Putman, Handler, & Feinberg, 2005; Reid, Eddy, Fetrow, & Stoolmiller, 1999; Weissberg & Greenberg, 1997). Research has also revealed the importance for an individual to feel they belong and are engaged in their school and for the whole school to support  18  belongingness and student engagement in school (Luiselli et al., 2005). In fact, extant research has found that when teachers and staff create a school climate that promotes high achievement and socio-emotional well-being, students are more likely to feel that they belong in their school and are less likely to engage in aggressive behaviours (Ames, 1992; Griffith, 1995; Rutter, Maughan, Mortimore, Ouston, & Smith, 1979). More recent research conducted by Kasen, Berenson, Cohen, and Johnson (2004) found that students who attended well-organized and harmonious schools, which prioritize learning, experienced lower levels of aggressive and other problem behaviours. In contrast, students who attended schools which have an informal and casual atmosphere reported an increase in student problem behaviours. These findings are consistent with other work on aggression and bullying which stipulated that aggression, bullying and other problem behaviours were more likely to occur in areas with less supervision, such as the washroom or playground (Craig et al., 2000). Extending these findings, one might expect that, given the flexibility, anonymity and informal nature of the Internet, and the lack of surveillance and supervision in online activities (Young, 2007), the possibility for adolescents to engage in aggressive behaviours online is high. Since most online aggressive activities occur outside of school and in their homes, adolescents can feel comfortable initiating or joining in on online cruelty in order to feel they belong with their peer group, even if they do not feel engaged with or that they belong in school; as such, it was hypothesized that adolescents who have lower feelings of belonging and engagement in school would be more likely to engage in Cyberbullying. It is clear that as ICTs become more prevalent so do the possible venues for bullying. Moreover, ICTs remove many of the social and status barriers that are  19  associated with offline bullying, which leaves open the possibility that Cyberbullying may lead to a large increase of aggressive acts among adolescents. It may also lead to an increase in the number of bully-victims (individuals who bully and are also victims), as exemplified by work conducted by Ybarra and Mitchell (2004) which found that victims of offline bullying were significantly more likely to bully others online than non-victims. In a study they conducted with adolescents between the ages of 10 and 17, just over half of the online bullies reported being the targets of traditional bullying (Ybarra & Mitchell). With the widespread use of information technologies among adolescents, the potential exists for anyone to be a Cyberbully or victim. In order to intervene/prevent Cyberbullying from occurring, a first step is to determine how it is similar to or different from traditional forms of bullying. This is the primary goal of this study. Three main questions guide this work: (a) Are online and offline bullying separate constructs? (b) Are there age and Sex differences in offline and online bullying/victimization? and (c) Are there differences in the predictors of online and offline bullying/victimization?  2.2. METHOD 2.2.1. Procedures Data for this study were drawn from a larger survey, that is, the Vancouver Safe School and Social Responsibility Survey. This district-wide survey was developed in order to meet part of British Columbia’s Ministry of Education mandate for school accountability. Information from this survey provided a “baseline” for administrators and staff about the experiences of students in their high schools. Data were collected across  20  the Vancouver district in January/February 2006, was re-administered in the spring of 2008, and is expected to be administered again every 2-3 years for 10 years, in an effort to monitor changes in social responsibility, youth behaviours, and school aggression over time. Permission to use these data secondarily was granted by the school district and the UBC Behavioral Research Ethics Board, and has resulted in a number of professional conference presentations (e.g., Darwich & Hymel, 2008; Darwich, Hymel, Pedrini, Sippel & Waterhouse, 2008 a,b; Ott vandeKamp & Hymel, 2007; Pedrini, Sippel, Hymel & Waterhouse, 2007). The data for the current study include those from the first wave of data collection, which involved all of the high schools within the district. The questionnaire covered a broad spectrum of issues. Students were asked about their experiences with racism, sexual orientation discrimination, victimization (both on the Internet and at school), belongingness with peers, whether they had caring adults in their life, and whether they perceived their school climate to be accepting of individual differences. During questionnaire administration, students were instructed not to print their names on any part of the survey as a way to ensure that there would be no identifying information. The survey included a self-created “privacy code” to track students’ answers with a future administration of the survey while respecting privacy and confidentiality. The purpose of this was to ensure that students felt comfortable providing honest and thorough answers. The final sample for the current study included 19,551 participants (49% female) between grades 8 and 12. The number of students enrolled was almost equal across all grade levels, ranging from 19.1% to 20.8%. The population was ethnically diverse, with 54% of participants being East Asian, 19% being Caucasian, and 9% being of mixed ethnicity.  21  The remaining students – who were of Aboriginal, African/Caribbean, South Asian, Latin American, and Middle Eastern – comprised approximately 2% each of the sample. Descriptive crosstab analyses were performed to assess the relative associations among Traditional Bullies, Traditional Victims, Cyberbullies, and Cybervictims. To do this, items were dummy coded according to whether or not adolescents engaged in each of the bullying and victimization behaviours where “Never” was coded as 0, and responses of 1 through 5 were re-coded to equal 1 (See Table 2.1). Table 2.1. Crosstabs analyses for Bullying, Victimization, Cyberbullying and Cybervictimization. Traditional Victim Cyberbully Cybervictim  Traditional Bully 11% (n = 2079) 6.2% (n = 1033) 5% (n = 845)  Traditional Victim  Cyberbully  4.6% (n= 775) 6.4% (n = 1091)  5.5% (n=866)  2.2.2. Measures Outcome Variables In order to examine differences in the construct of online vs. offline bullying and to examine the differences in these types of bullying, items which measured victimization, bullying, and Cyberbullying were asked. Bullying and Victimization. In order to measure bullying and victimization, participants were asked to rate how often they had engaged in “bullying and harassment”, “physical bullying”, “verbal bullying”, “social bullying” and how often they had experienced being “bullied and harassed”, “physically bullied”, “verbally bullied”, and “socially bullied”. In addition they were also asked to rate how often they had been victims of Cyberbullying “at school” and “outside of school,” and how often “being 22  Cyberbullied caused problems at school.” Finally, they were also asked how often they had engaged in Cyberbullying “at school” or “outside of school” and how often their participation in this form of bullying “caused problems at school.” Following Vaillancourt and colleagues’ (2008) work, each form of bullying and victimization item was defined in an attempt to increase precision in measurement (See Appendix A, question 57). Participants responded to these items on a four-point Likert Scale ranging from “never” to “every week or more”. Covariate and Predictor Variables Self-Esteem: The measure for self-esteem was comprised of eight items from Marsh’s Self-Description Questionnaire. Examples of items include the degree to which participants, “do lots of important things” and “have a lot to be proud of.” Responses made ratings on a 5-point Likert scale ranging from “1 = strongly disagree” to “5 = strongly agree” (See Appendix A, question 18). A composite variable for self-esteem was created by taking the average across all eight of their responses. This measure had high internal consistency (α = .85). Perceptions of Academic Performance: Students were asked to describe their perceptions of their academic performance in elementary school, and in high school (their current school), where 1 = “Better than most students”, 2 = “About the same as most students”, and 3 = “Worse than most students” (See Appendix A, question 13). Responses were reverse coded so that higher responses reflected higher perceptions of academic performance. Since correlations were found between elementary and high school perceptions of academic achievement (r = .302, p < .001) a single composite  23  score was created by taking the mean of their responses to the elementary and the current school items3. Responses of “I’m not sure” were re-coded as missing. Perceptions of Peer Acceptance: Students were also asked to report how well they were liked in elementary school, and in their current school. Possible responses included “Better than most students” = 1, “About the same as most students” = 2, and “Worse than most students” = 3 (See Appendix A, question 14). Responses were reverse coded so that higher responses reflected higher perceptions of peer acceptance. As elementary and high school perceptions of peer acceptance were significantly correlated (r = .39, p < .001), a Perceptions of Peer Acceptance composite score was created by taking the mean of their elementary and high school responses1. Responses of, “I’m not sure” were recoded as missing. Student Belonging: Students were asked to respond to six questions related to their feelings of belonging in school. Students were asked to rate, on a 5-point Likert scale (ranging from “1 = strongly disagree” to “5 = strongly agree”), to what extent they feel that, “students are just looking out for themselves (reversed),” “students really care about each other,” and “others accept me as I am” (See Appendix A, questions 28, 31, 32, 33, 34, 37). The mean of participant responses was calculated to create the composite variable. Internal consistency for this measure was good (α = .76). Student Engagement: In order to measure students’ feeling of engagement with their school, they were asked to respond to five questions on a 5-point Likert scale (ranging from “1 = strongly disagree” to “5 = strongly agree”). Questions asked students to rate the degree to which they “work together to solve problems” and “have a say in  3  Analyses were also performed with elementary and high school perceptions inputted separately into the model. Doing so did not significantly change the results.  24  what’s going on” (See Appendix A, questions 26, 39, 42, 43, 44). Again, the mean of participant responses was calculated to create the composite variable. Internal consistency for this measure was also good (α = .70).  2.3. RESULTS 2.3.1. Question 1: Are Online and Offline Bullying Separate Constructs? In order to create the outcome variables and determine whether adolescents perceive online and offline bullying/victimization as separate constructs, exploratory factor analysis (EFA) was performed. The unweighted least squares method of extraction was chosen because it makes no assumption of multivariate normality, and the Varimax rotation strategy was chosen because it eases interpretability of the factor solution by maximizing the variance of the loadings and, consequently, clarifying which items are correlated with the factor (Tabachnick & Fidell, 2001). EFA was first run on the 14 items of the Bullying and Harassment portion of the Vancouver School Board Safe School and Social Responsibility Survey. The KaiserMeyer-Oklin value was .8, exceeding the recommended minimum value of .6 (Kaiser, 1970, 1974), and the Bartlett’s Test of Sphericity (Bartlett, 1954) reached statistical significance (p < .001). These results support the factorability of the correlation matrix. Three factors with eigenvalues greater than 1.0 emerged from the data. Eigenvalues for each of the factors were 6.8, 1.65, and 1.46, respectively. Upon inspection of the screeplot, a clear break after the third factor was revealed. Furthermore, upon assessment of the reproduced correlations only 18% of the residuals were nonredundant and had absolute values greater than .05. This means that when the observed  25  correlations are subtracted from the reproduced correlations, 82% are near 0, which provides further support for a three factor model being a good fit for the data. The rotated solution confirmed ‘very good’ to ‘excellent’ factor loadings for two factors: Traditional Bullying and Traditional Victimization (See Table 2.2). This decision is based on Comrey and Lee’s (1992) work which suggest that loadings above .71 are considered excellent, .63 are very good, .55 are good, .45 are fair, and .32 are poor. For Cyberbullying and Cybervictimization, however, there did not appear to be the same distinction between bullying and victimization. Most of the items loaded strongly on factor #1 with loadings in the very good to excellent range (Comrey & Lee, 1992). Interestingly, some cross-loadings were observed where the Cyberbullying items (but not Cybervictimization) also loaded on Traditional Bullying (cross-loadings were in the ‘good’ range between .56 and .57). These results demonstrate that there are some discrepancies among online bullying/victimization and offline bullying/victimization, with adolescents differentiating between traditional forms of bullying and victimization, but tending to view online bullying and victimization as a single construct (although still related to traditional forms of bullying, as indicated by the cross-loadings). Interestingly, physical, verbal, and social bullying and victimization did not load as separate constructs.  26  Table 2.2: Unweighted least squares factor analysis pattern matrix for bullying & victimization items. Item Cyberbullying and Cybervictimization: Cyberbullying at school to me Cyberbullying caused problems at school to me Cyberbullying outside of school to me Cyberbullying at school to others Cyberbullying caused problems at school to others Cyberbullying outside of school to others Traditional Bullying: Verbal bullying (name calling, teasing, threats) to others Bullying and harassment to others. Physical bullying (hitting, shoving, kicking, etc.) to others Social bullying (exclusion, rumours, gossip) to others Traditional Victimization: Verbal bullying (name calling, teasing, threats) to me Bullying and harassment to me Social bullying (exclusion, rumours, gossip) to me Physical bullying (hitting, shoving, kicking, etc.) to me  1 .76 .74 .70 .67 .65 .60  Factor 2  3 .37 .38 .43  .55 .53 .57 .75  .36  .68  .30  .67 .61 .80 .72 .66 .56  *Note: Only factor loadings of .32 or greater are reported here as is consistent with reporting procedures for Factor Analyses (Tabachnich & Fidell, 2001).  To explore these findings further, two additional EFAs were performed, one for the seven bullying items (including both the traditional and Cyberbullying items), and one for the seven victimization items. For bullying, only one factor emerged (eigenvalue = 4.1; See Table 2.3), and the loadings were in the very good to excellent range. For victimization, on the other hand, Traditional victimization and Cybervictimization items loaded as two separate factors (eigenvalues = 4.1, 1.1 respectively; See Table 2.4), suggesting that there is a distinction between online and offline victimization.  27  Table 2.3: Unweighted least squares factor analysis pattern matrix for bullying Items. Factor 1  Item Cyberbullying outside of school to others Cyberbullying at school to others Cyberbullying caused problems at school to others Physical bullying(hitting, shoving, kicking, etc.) to others Verbal bullying (name calling, teasing, threats, putdowns) to others Bullying and harassment to others Social bullying (exclusion, rumours, gossip, humiliation) to others.  .80 .80 .76 .74 .71 .71 .67  Table 2.4:Unweighted least squares factor analysis pattern matrix for Victimization Items. Factor Item  1  Verbal bullying (name calling, teasing, threats, putdowns) to me Bullying and harassment to me Social bullying (exclusion, rumours, gossip, humiliation) to me Physical bullying (hitting, shoving, kicking, etc.) to me Cyberbullying at school to me Cyberbullying caused problems at school to me Cyberbullying outside of school to me  .84 .74 .64 .57 .34  2  .33 .37 .82 .80 .80  One final EFA was run with just the Cyberbullying and Cybervictimization items to see if, on their own, two factors would emerge. As can be seen in Table 2.5, these items very strongly loaded together as one factor (eigenvalue = 4.1), with all items in the ‘excellent range’. Based on the pattern of results, it appears that adolescents do not distinguish as strongly between bullies and victims when the aggression is occurring online. Despite the crossloadings between Cyberbullying and Traditional Bullying, in general, Cyberbullying appears to be more closely related to the other cyber items.  28  Table 2.5: Unweighted least squares factor analysis pattern matrix for Cyberbullying and Cybervictimization items. Item Cyberbullying at school to others Cyberbullying at school to me Cyberbullying caused problems at school to others Cyberbullying outside of school to others Cyberbullying caused problems at school to me Cyberbullying outside of school to me  Factor 1 .81 .80 .79 .78 .77 .77  2.3.2. Question 2: Are There Grade and Sex Differences in Cyberbullying/victimization, Traditional Bullying, and Traditional Victimization? To develop the composite variables which were used for the remainder of this study, the outcomes from the initial EFA were utilized. Specifically, three composite variables were created: (1) Cyberbullying/Victimization (α = .90), (2) Traditional Bullying (α = .84), and (3) Traditional Victimization (α = .84)4. In order to assess grade and sex differences, three separate univariate analysis of variance were performed for each of the bullying and victimization types, with grade and sex entered as independent variables. As can be seen in Table 2.6, significant main effects for grade and for sex were found for all three models. More interesting, however, were the significant interaction effects found between grade and sex for all three of the outcomes. As shown in Figures 2.1, 2.2, and 2.3, boys and girls seem to report similar levels of bullying and victimization in the early grades, but boys tend to have higher levels than girls in the later grades. To examine grade differences more carefully, Tukey 4  It is important to note that all of the subsequent analyses were also run with the cyber items separated into Cyberbullying and Cybervictimization, and there were no differences in the pattern of results for these outcomes, which provides further evidence for a single construct.  29  HSD post hoc analyses were performed separately for boys and girls. For all three constructs, these analyses indicated that, for boys, the frequency of bullying and victimization increased until grade 10 and then decreased. Significant differences were found between grades 9 and 10 (p < .05) and grades 11 and 12 (p < .05) for Cyberbullying/victimization, grades 8 and 9 (p < .001) and grades 10 and 11 (p < .001) for Traditional Bullying and grades 10 and 11 (p < .001) for Traditional Victimization. Grade 8 girls behaved with similar frequencies as boys. However, these frequencies seemed to decrease over time. A significant difference was found between grade 9 and 10 girls for Traditional Victimization (p < .01). Table 2.6. Main and interaction effects for Cyberbullying/victimization, Traditional Bullying, and Traditional Victimization. df  Mean Square  F  Partial Eta Squared  Cyberbullying/victimization grade sex grade * sex Error Total  4 1 4 18436 18446  .72 12.01 1.64 .19  3.9** 64.8*** 8.8***  .001 .004 .002  Traditional Bullying grade sex grade * sex Error Total  4 1 4 18481 18491  7.07 54.40 1.93 .35  20.3*** 156.3*** 5.5***  .004 .008 .001  Traditional Victimization grade sex grade * sex Error Total  4 1 4 18631 18641  5.16 36.42 1.45 .38  13.7*** 96.6*** 3.8**  .003 .05 .001  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  30  Figure 2.1. Grade and sex interactions for Cyberbullying/victimization  Figure 2.2. Grade and sex interactions for Traditional Bullying  31  Figure 2.3. Grade and sex interactions for Traditional Victimization  2.3.3. Questions 3: Are There Differences in the Constructs Which Predict Traditional Bullying, Traditional Victimization, and Cyberbullying/victimization? In order to determine whether the predictors for Cyberbullying/victimization were different for Traditional Bullying and Traditional Victimization, hierarchical multiple regressions were performed separately for each of the three outcome variables. The grade and sex variables were entered in Block 1 as covariates, and Perceptions of Academic Performance, Peer Acceptance, Self-Esteem, Belonging, and Engagement in School were entered simultaneously as predictors in Block 2 of the regression. Grade did not appear to significantly correlate with Cyberbully/victimization or Traditional Bullying, and Peer Acceptance did not correlate with Cyberbully/victimization. These 32  non-correlating variables were kept in the models, however, in order to be consistent across analyses. All other variables were significantly correlated with the three dependent variables (p < .001), although these relationships were found to be low (r < 0.3; See Tables 2.7, 2.8, and 2.9). Tolerance levels for the models were between .65 and .91, thus assuring that the assumption for multicollinearity was not violated for any of the models. The assumption for linearity, normality, and outliers were assessed by examining the Normal Probability Plot and the Residuals Scatterplot. These graphs did not suggest any major violations of the assumptions. For Cyberbullying/victimization, 9.1% of the variance was explained by the final model (p < .001). All the independent variables, except for grade, made significant contributions to the model with Belonging in School and Engagement in School offering the largest unique contributions (β = -.15, p < .001 and β = -.16, p < .001 respectively). Inverse relationships were found for Sex, Perceptions of Academic Performance, SelfEsteem, Belonging, and Engagement in School. This negative relationship demonstrates that males are more likely to be involved in Cyberbullying/victimization than girls, and that as student perceptions of academic achievement, self-esteem, belonging, and engagement in school increased, Cyberbullying/victimization decreased (See Table 2.7). By contrast, a positive relationship was found for Cyberbullying/victimization and Perceptions of Peer Acceptance. In this case, as peer acceptance increases, so does Cyberbullying/victimization.  33  Table 2.7: Summary of hierarchical multiple regression model for Cyberbullying/victimization Cyberbullying/ Victimization Block 1 Grade Sex Block 2 Grade Sex Academic Performance Peer Acceptance Self-Esteem Belonging Engagement in School  B  SE  β  r  -.00 -.05  .00 .01  .01 -.06***  .01 -.06***  -.00 -.02 -.02 .05 -.05 -.10 -.11  .00 .01 .01 .01 .01 .01 .01  -.01 -.03*** -.02** .05*** -.07*** -.15*** -.16***  R2 .004  ∆R2 .004***  .091  .088***  .01 -.06*** -.07** .01 -.17*** -.25*** -.26***  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  For Traditional Bullying, 9% of the variance was explained by the final model (see Table 2.8), with school engagement making the largest unique contribution (β = -.24, p<.001). In the final model, both grade and sex were found to have a significant inverse relationship with Traditional bullying. That is, older males were more likely to engage in this form of aggression. As with Cyberbullying/victimization, all the independent variables, with the exception of perception of peer acceptance, were found to make significant inverse contributions to the model (See Table 2.8). In other words, as Perceptions of Academic Performance, Self-Esteem, Belonging, and Engagement in School increased, Traditional Bullying decreased. Conversely, for Perceptions of Peer Acceptance, an increase in Traditional Bullying was predicted by an increase in feelings of acceptance among peers.  34  Table 2.8: Summary of hierarchical multiple regression model for Traditional Bullying. Traditional Bullying Block 1 Grade Sex Block 2 Grade Sex Academic Performance Peer Acceptance Self-Esteem Belonging Engagement in School  B  SE  β  r  -.00 -.11  .00 .01  -.00 -.09***  -.01 -.09***  -.01 -.73 .03 .13 -.05 -.04 -.21  .00 .01 .01 .01 .01 .01 .01  **  -.02 -.06*** -.03*** .10*** -.05*** -.05*** -.24***  R2 .008  ∆R2 .008***  .092  .084***  -.01 -.09*** -.07*** .06*** -.13*** -.17*** -.28***  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  The model for Traditional Victimization differed from that of Traditional Bullying and Cyberbullying/victimization. For this model, 12.7% of the variance could be explained by the final model, with belonging in school offering the largest unique contribution (β = -.25, p<.001). Grade and sex were found to be significant, with older boys being traditionally victimized more than girls. Unlike Traditional Bullying and Cyberbullying/victimization, Perceptions of Academic Performance did not significantly contribute to the model and Peer Acceptance was inversely related to Traditional Victimization. The other predictors (i.e., Self-Esteem, Belonging in school, and Engagement in School) were also found to have inverse relationships to Traditional Victimization (See Table 2.9). In other words, as Self-Esteem, Belonging in school, Perceptions of Peer Acceptance, and Engagement in school increased, Traditional Victimization decreased.  35  Table 2.9: Summary of hierarchical multiple regression model for Traditional Victimization. Traditional Victimization Block 1 Grade Sex Block 2 Grade Sex Academic Performance Peer Acceptance Self-Esteem Belonging Engagement in School  B  SE  -.02 -.09  .00 .01  -.04*** -.07***  -.04*** -.07***  .00 .01 .01 .01 .01 .01 .01  ***  ***  -.03 -.05 -.00 -.07 -.03 -.23 -.11  β  r  -.07 -.04*** -.01 -.05*** -.03*** -.25*** -.12***  R2 .007  ∆R2 .007***  .127  .121***  -.04 -.07*** -.06*** -.12*** -.19*** -.33*** -.26***  Note. *p < 0.05 **p < 0.01 ***p < 0.001.  After running the regression analyses, the independent variables that emerged as significant for each of the three models were assessed for their relative importance using the Relative Pratt Index (RPI; Thomas, Hughes & Zumbo, 1998). The RPI is a statistical measure that assesses the relative contribution of the explanatory variables to the total variance explained by the regression model using the following formula: Dj = (βj*rj)/R2 where Dj represents the relative importance of the item (RPI), βj symbolizes the standardized Beta, rj is the simple correlation between the explanatory variable and the dependent variable. Thomas, Hughes and Zumbo suggest that if Dj < ½(p), where p is the number of items in the model, then the item does not make an important contribution to the model even if it does emerge as statistically significant. Upon employing the Pratt Index on each of the significant items, Engagement in School, and Belonging in school made the most important contributions for Cyberbullying/victimization, Traditional 36  Bullying, and Traditional Victimization (See Table 2.10). For both Cyberbullying/victimization and Traditional bullying, Engagement in School emerged as most important, while Belonging emerged as second most important. For Traditional Victimization, it was reversed, with Belonging being the most important and Engagement being the second most important. Regardless of rank, however, the more adolescents felt that they belonged in school and that they were engaged in school, the less they reported being engaged in online or offline bullying/victimization. Table 2.10: Summary of important items according to the Pratt index. Item DV = Cyberbullying/victimization Engagement in School Belonging in School Self-Esteem Perceptions of Peer Acceptance Perceptions of Academic Achievement Grade  Dj .31* .27* .09* .01 .01 .00  DV = Traditional Bullying Engagement in School Belonging in School Self-Esteem Perceptions of Peer Acceptance Academic Achievement Sex Grade  .63* .08* .07* .08 .02 .01 .00  DV = Traditional Victimization Belonging in School Engagement in School Perceptions of Peer Acceptance Self-Esteem Sex Grade  .60* .22* .07* .04 .03 .01  Note. Dj < ½(p) = .07. * Important Variable.  37  2.4. DISCUSSION Traditional forms of bullying have long been studied. However, work on Cyberbullying has only recently begun to emerge (e.g., Agatston, Kowalski & Limber, 2007; Patchin & Hinduja 2006; Li, 2007). Most of this early work has examined frequencies of Cyberbullying compared to traditional bullying, but has failed to examine the nature of these two forms of bullying in relation to each other, or even whether they are unique constructs. This study used a series of factor analyses to show that there is some indication that the construct of Cyberbullying is different from traditional bullying and must be examined more thoroughly in its own right. The four EFA models revealed that while face-to-face forms of bullying and victimization are considered distinct constructs, this distinction is not as clear when the aggression is occurring online. More specifically, it appears that adolescents view Cyberbullying and Cybervictimization as a single construct, which is in contrast to how they view Traditional Bullying and Traditional Victimization. The bullying aspects of online aggression do appear to overlap with offline bullying (these items cross-loaded with the online aggression factor and the offline bullying factor), but there did not appear to be a strong relationship between being a victim online and being a victim offline. Overall, it appeared that adolescents were less certain in distinguishing between bullying and victimization when it was occurring online. One possible explanation is that in an online venue, victims were much more comfortable and capable of retaliating to aggressive acts. For example, if an individual said something mean to another online, and the initial “target” responded aggressively in return, both individuals have essentially engaged in bullying behaviour and have also 38  been the victim of such behaviour. Moreover, it is possible that the initial event and response could expand quite rapidly into a series of bullying/victimization incidents between the two individuals, which may make it difficult to clearly differentiate who is the victim and who is the bully. In addition, as the online retaliatory interactions continue, it is likely to draw in friends on both sides, who might also contribute to both types of behaviour. Further work must be conducted to address these hypotheses. A similar situation would be far less likely to occur offline due to the nature of typical face-to-face bullying situations. The power imbalance that is normally present in traditional physical bullying situations makes it unlikely that victims would immediately respond in kind to the perpetrator, due to personality and physical characteristics (e.g. shyness, introversion, small stature). If there was any retaliation, it would likely occur at a different time when the original target was more prepared, and be thought of as a separate incident. This makes it easier to distinguish between the bully and the victim in offline situations. This possibility is compatible with research showing that individuals are more comfortable saying things online than offline due to the protectiveness of the screen (Peter et al., 2005; Ward & Tracey, 2004). Additional work is necessary to elucidate more fully adolescents’ perceptions of online aggression/victimization and their motivations for engaging in these aggressive acts. Some of this will be addressed in Study II of this work. Univariate Analyses of Variance found both similarities and differences among Cyberbullying/victimization, Traditional bullying, and Traditional victimization. Specifically, significant grade and sex interactions occurred across all forms of bullying and victimization. That is, boys reported being more likely to participate in  39  Cyberbullying/victimization, Traditional Bullying, and Traditional Victimization than girls as grades increased. For girls, however, this trend decreased slightly as they got older, but this pattern was largely non-significant. For boys, all three bullying and victimization outcomes occurred most frequently when students were in grades 9 or 10, as compared to when they were in grades 8, 11, or 12. Conversely, girls reported being bullied or victimized more in grades 8 and 9 than in grades 10, 11, or 12. Although the effect size for these grade differences was found to be small, there are several explanations for these discrepancies. One reason for the increase in bullying/victimization for boys and girls between grades 8 and 10 is the hierarchical position of these adolescents in their school. Research has shown that adolescence is a developmental period when the division of individuals across homophilic groups seems most prominent, as exemplified by the different cliques and crowds that exist in middle and high school (Brown et al., 1994). Specifically, research has found that high network centrality and popularity are associated with socially aggressive behaviours as group leaders strive to control their group and maintain their position within the social hierarchy (Adler & Adler, 1998; Dodge et al., 1990; Estell et al., 2002; Rodkin et al., 2000; Vaillancourt et al., 2003; Xie et al., 2002). With this in mind, it makes sense that the battle to be among the more popular peers is especially so for students between grades 8 and 10, who are likely to attempt to exercise authority over one another. As adolescents enter grades 11 and 12 and their peer groups become established, this struggle to climb the social ladder may become less important as their focus shifts toward the formation of romantic relationships with a significant other (Pellegrini, 2001a).  40  Despite these arguments, the current results did run somewhat contrary to previous work on traditional bullying and victimization. Previous studies have shown that as adolescents transition into high school, a marked increase in bullying and victimization occurs as they attempt to establish their peer group in their new social environment (Kochenderfer & Ladd, 1996), but that once these new groups have been formed, and have stabilized in the new school, bullying and victimization decreased (Asidao, Vion, & Espelage, 1999; Kochenderfer & Ladd, 1996; Pellegrini & Bartini, 1999). Although this pattern was shown for girls in this study, the spike in bullying did not occur until the second year of high school for boys. Of importance to note is the fact that participants in this study started high school in grade 8, which is earlier than usual. Given this, it is possible that both developmental and school transition effects are at play in this sample. More specifically, it is possible that the purpose of the bullying is – as argued above – to establish peer hierarchies, but the timing of it may be developmentally driven, and not entirely due to the school transition. Future work needs to tease apart these possibilities. Another factor that might explain the increase in online bullying/victimization for students in grades 9 and 10 (specifically for boys in this study) is the amount of time spent online by adolescents. Research has found that older adolescents tend to use the Internet more and tend to use the Internet with less supervision than younger adolescents (MNet, 2005). This may be affording them more opportunity to say mean things online should they choose to do so. Assuming that the grade-related differences are robust, additional research to identify the reasons why adolescents are more likely to engage in bullying and Cyberbullying in grades 9 and 10 will be important to facilitate the  41  development and implementation of intervention and prevention programs. However, as a first step, these results need to be replicated longitudinally to ensure that these findings were not simply an artifact of using cross-sectional data. One of the vulnerabilities of any cross-sectional study is that there could be some other characteristic – beyond maturation – that is unique to the specific cohorts of students in the sample that could explain the findings. Results from this study also showed that older boys seem to engage in bullying/victimization behaviours more often than girls. Although inconsistent with the original hypothesis that girls would engage in online aggression more than boys the findings from this study are supported by literature showing that boys have typically reported higher aggression levels than girls (Olweus, 1991; Pellegrini & Long, 2002; Ross, 1996). That said, research also stipulates that girlhood aggression is often under reported because girls are more covert and indirect in their aggressive behaviour (i.e. gossiping, dismissive glances, social exclusion), and that girls are more likely to make strong attempts to appear nice all the time (Underwood et al., 2004; Zahn-Waxler, 2000). With this in mind, a possible reason why boys may have reported being more involved in aggressive activities than girls in this study is because girls are less likely to view or admit that their behaviour is bullying behaviour. Future work needs to explore more fully the existence of sex differences for online bullying and victimization. One important variable that will need to be considered is the amount of time an individual is spending online. It will also be important to know the kinds of activities adolescents are engaging in online (for example, socializing online is likely to be associated with increased bullying opportunities and incidences in comparison with web browsing). For example,  42  if specific questions were asked on how adolescents are bullying online (i.e. spreading rumours over Instant Messaging vs. creating websites that are intended to harm), then more girls may have reported to engaging in online aggression more frequently. This study also assessed differences in predictors of traditional vs. online bullying and victimization. These findings revealed similar outcomes for each of the predictors. As hypothesized, Cyberbullying/victimization and Traditional Bullying were both predicted by perceptions of academic achievement, and this relationship was found to be negative. That is, reported levels of Cyberbullying/victimization and traditional bullying decreased as perceptions of academic achievement increased. These findings make sense given the extant literature that has found inverse relationships between traditional bullying and academic performance (Beran et al., 2008; Schwartz et al., 2005; Song et al., 2000). In contrast, perception of academic achievement did not significantly predict Traditional Victimization. This finding confirms the outcomes of the EFA which demonstrated that bullies and victims are two separate groups of individuals with different characteristics. These findings also support the need to examine these three constructs more closely to determine exactly how Cyberbullies/victims are different from Traditional Bullies and Victims. Adolescent perceptions of peer acceptance made significant contributions to all three models of aggressive/victim behavior; however, the directionality of this relationship varied. That is, adolescents with higher perceptions of peer acceptance were more likely to engage in Cyberbullying/victimization and Traditional bullying. This finding is consistent with research that stipulates that adolescents who feel accepted by their peers are more likely to bully in order to maintain their position in the social  43  hierarchy (Dodge et al., 1990; Vaillancourt et al., 2003; Xie et al., 2002). By contrast, an inverse relationship was found for perceptions of peer acceptance and Traditional Victimization. In other words, the more adolescents feel they are accepted by their peers, the less likely they are to be victimized. This finding supports those of previous research which found the same outcome (Pellegrini et al., 1999; Pellegrini & Long, 2002). Finally, for all three constructs, self-esteem, belonging in school, and engagement in school all significantly predicted bullying/victimization. Specifically, no matter whether adolescents were bullied/victimized online or offline, those who reported higher self-esteem, and who reported greater school belonging and engagement in school were less likely to be involved in bullying/victimization, as bullies or as victims. These findings are consistent with previous research which has found that individuals with higher self-esteem, and who feel a higher sense of school community and belonging are less likely to be involved in schoolyard aggressive behaviours (Ames, 1992; Luiselli et al., 2005). Moreover, the outcome of the Pratt Index indicated that for each of the three bullying and victimization constructs, Belonging and Engagement in school were the only variables to emerge as important predictors across all three constructs. That is, although the other variables may have made statistically significant contributions to the prediction of the constructs, only Belonging and Engagement made important contributions to explaining the constructs. These results are supported by research findings that emphasize the importance of developing a school climate that fosters school belonging and engagement, and that schools which emphasize the importance of such environments are more likely to have students who are less likely to engage in aggressive  44  behaviours (Luiselli, et al., 2005). This also suggests that although this study has provided evidence that these constructs are unique, it also appears that there are commonalities in the factors that predict aggression, whether it is online or offline. 2.4.1. Limitations There are several limitations to this study that should to be discussed. First, the sample size for this study was very large (N=19551), thus inflating the statistical power of the analyses. As it is easy to achieve statistical significance with such a large sample, results for this study must be examined closely and must consider how well they fit with theory. Research on Cyberbullying is in its infancy; as such, it is also critical that research be conducted to replicate these findings. The cross-sectional nature of this study is also a limitation in that we are unable to establish developmental differences within individuals over time. Future work in this area needs to collect data on Cyberbullying longitudinally in order to parse out betweenperson effects from within-person effects. It would also be meaningful to employ observational techniques of Cyberbullying in order to assess the true nature of this form of aggression. Another limitation of this study is that the questionnaire items related to Cyberbullying were quite general. That is, the items asked participants to rate to what degree they have been a Cyberbully or Cybervictim at school or outside of school. This type of question yields a general response that allows for a preliminary look at this form of aggression; however, given that Cyberbullying is a new phenomenon, it is imperative for future research to find out what the term Cyberbullying even means to adolescents. Further work in this area should also include items on computer and Internet usage and 45  access, in order to control for these potential confounds. In addition, data about Cyberbullying should be collected through means other than self-report questionnaires. For example, interview data would be a valuable method for gaining a deeper understanding of this phenomenon. In an attempt to address many of these concerns, Study II takes a closer look at the construct of Internet aggression, and utilizes a mixedmethod approach to do so. 2.4.2. Conclusion Despite these limitations, the results of this study indicate that adolescents perceive aggression that is occurring online to be distinct from other forms of aggression. Adolescents don’t distinguish between Cyberbullying and Cybervictimization in the same way that they distinguish between traditional bullying and traditional victimization. However, there are many similarities in the factors that predict all three constructs. There were similar sex and grade differences, as well as similarities in the individual and contextual factors that predict these constructs. Overall though, this study confirms the need to further investigate Cyberbullying and Cybervictimization in order to accurately understand these phenomena and how they manifest in the lives of adolescents. If we want to be committed to enhancing social responsibility among children, we need to make sure our intervention and prevention attempts go beyond the schoolyard. This study provides an initial framework for examining online aggression more closely. It is clear from this work that the construct of Cyberbullying must be deconstructed, and that future research is required to examine more closely the predictors of online aggression from a socio-ecological perspective. Study II and III are first steps in this direction.  46  2.5. REFERENCES Adler, P.A., & Adler, P. (1998). Peer power: Preadolescent culture and identity. New Brunswick, NJ: Rutgers University Press. Aftab, P. (n.d.). Net Bullies. Retrieved June 16, 2006, from http://www.netbullies.com/pages/1/index.htm Agatston, P.W., Kowalski, R. & Limber, S. (2007). Students’ perspectives on Cyberbullying. Journal of Adolescent Health, 41, 59-60. Ames, C, 1992). Classrooms: Goals, structures, and student motivation. Journal of Educational Psychology, 84, 261-271. Arnett, J.J. (1999). Adolescent storm and stress, reconsidered. American Psychologist, 54, 317-326. Atlas, R.S. & Pepler, D.J. (1998). Observations of bullying in the classroom. The Journal of Educational Research, 92, 86-99. Bartlett, M.S. (1954). A note on the multiplying factors for various chi square approximations. Journal of the Royal Statistical Society, 16 (Series B), 296-298. Belsey, B. (2005). Cyberbullying: An emerging threat to the “Always On” generation. Bullying.org Canada. Retrieved June 17, 2006, from http://www.cyberbullying.ca/pdf/feature_dec2005.pdf Benfer, A. (2001). Cyber slammed. Retrieved July 7, 2001, from http//www.dir.salon.com/mwt/feature/2001/07/03/cyber_bullies/index.html. Benenson, L.F., Maisese, R., Dolenszky, E., Dolensky, N., Sinclair, N., & Simpson, A. (2002). Group size regulate self-assertive versus self-deprecating responses to internal competition. Child Development, 73, 1818-1829. 47  Beran, T.N., Hughes, G., & Lupart, J. (2008). A model of achievement and bullying: analyses of the Canadian National Longitudinal Survey of Children and Youth data. Educational Research, 50, 25-40. Beran, T. & Li, Q. (2005). Cyber-Harassment: A study of a new method for an old behavior. Journal of Educational Computing Research, 32, 265-277. Berryman, J. C., Smythe, P. K., Taylor, A., Lamont, A., & Joiner, R. (2002). Developmental psychology and you (2nd ed.). Oxford, UK: The British Psychological Society and Blackwell Publishers. Boivin, M., Hymel, S., & Bukowski, W.M. (1995). The roles of social withdrawal, peer rejection, and victimization by peers in predicting loneliness and depressed mood in children. Development and Psychopathology, 7, 765-786. Bosworth, K., Espelage, D.L., & Simon, T.R. (1999). Factors associated with bullying behavior in middle school students. Journal of Early Adolescence, 19, 341-362. Boulton, M.J., & Smith, P.K. (1994). Bully/victim problems in middle-school children: Stability, self-perceived competence, peer perceptions and peer acceptance. British Journal of Developmental Psychology, 12, 315-329. Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design. Cambridge, MA: Harvard University Press. Brooks-Gunn, J. (1987). Pubertal processes: Their relevance for developmental research. In V. B. Van Hasselt & M. Hersen (Eds.), Handbook of adolescent psychology (pp.111-130). New York: Pergamon. Brown, R. (2000). Social identity theory: Past achievements, current problems and future challenges. European Journal of Social Psychology, 30, 745-778.  48  Brown, B.B., Mory, M.S., & Kinney, D. (1994). The complex world of peer crowds: Casting crowds in a relational perspective: Caricature, channel, and context. In R. Montemayor, G.R. Adams, & T.P. Gullotta (Eds.), Advances in adolescent development: Vol. 6. Personal relationships during adolescence (pp.123-167). Carlson, L.W., & Cornell, D.G. (2008). Differences between persistent and desistent Middle School Bullies. School Psychology International, 29, 442-451. Coie, J.D. & Dodge, K.A. (1998). Aggression and antisocial behavior. In N. Eisenberg (Ed.), Handbook of Child Psychology, Vol. 3 (pp.779-862). New York: Wiley. Coie, J.D. & Jacobs, M.R. (1993). The role of social context in the prevention of conduct disorder. Development and Psychopathology, 5, 263-275. Cole, D.A., Scott, M.E., Martin, J.M., Lachlan, P.G., Seroczynski, A.D., Tram, J.M., Hoffman, K.B., Ruiz, M.D., Jacquez, F., & Maschman, T. (2001). The development of multiple domains of child and adolescent self-concept: a cohort sequential longitudinal design. Child Development, 72 (6), 1723-1746. Cole, P.M., Zahn-Waxler, C., & Smith, K.D. (1994). Expressive control during a disappointment: Variations related to preschooler’s behavior problems. Developmental Psychology, 30, 835-845. Comrey, A.L., and Lee, H.B. (1992). A First Course in Factor Analysis. (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers. Conway, A.M. (2005). Girls, aggression, and emotion regulation. American Journal of Orthopsychiatry 75, 334 – 339. Craig, W.M., Pepler, D., & Atlas, R. (2000). Observations of bullying in the playground and in the classroom. School Psychology International, 21, 22-36.  49  Crick, N.R. (1995). Relational aggression: The role of intent attributions, feelings of distress, and provocation type. Development and Psychopathology, 7, 313-322. Crick, N.R. (1997). Engagement in gender normative versus non-normative forms of aggression: Links to social-psychological adjustment. Developmental Psychology, 33, 310-317. Crick, N.R., Casas, J.f., & Mosher, M. (1997). Relational and overt aggression in preschool. Developmental Psychology, 33(4), 589-600. Crick, N.R., & Grotpeter, J.K., (1995). Relational aggression, gender, and socialpsychological adjustment. Child Development, 66, 710 – 722. Crick, N.R., Grotpeter, J.K., & Rockhill, C.M. (1999). A social information-processing approach to children’s loneliness. In K.J. Rotenberg & S. Hymel (Eds.), Loneliness in childhood and adolescence. New York: Cambridge University Press. Darwich, L. & Hymel, S. (2008). Lesbian, gay, bisexual, and questioning adolescents: victimization, belonging, safety and the role of supportive adults in high school. Education.com, Special Issue on Bullying (www.education.com ) Darwich, L., Hymel, S., Pedrini, L., Sippel, J., & Waterhoouse, T. (2008a). Lesbian, gay, bisexual, and questioning adolescents: Their social experiences and the role of supportive adults in high school. Invited presentation at the Pride in Education Conference: Addressing Homophobia in BC Schools, Vancouver, BC (24 October, 2008).  50  Darwich, L., Hymel, S., Pedrini, L., Sippel, J. & Waterhouse, T. (2008b). Lesbian, gay, bisexual and questionning adolescents:Their social expereinces and the role of supportive adults in high school. Paper presented at the biennial meeting of the Society for Reseearch in Adolescence, Chicago. IL, USA (8 March 2008). Dodge, K.A., Coie, J.D., Pettit, G., & Price, J. (1990). Peer status and aggression in boys’ groups: Developmental and contextual analyses. Child Development, 61, 1289-1309. Doll, B., Song, S., & Siemers, E. (2004). Classroom ecologies that support or discourage bullying. Espelage & S. Swearer (Eds.), Bullying in American schools: A socialecological perspective of prevention and intervention (pp 161 – 183). Mahwah, NJ: Lawrence Erlbaum Associates. Duncan, R.D. (2004). The impact of family relationships on school bullies and victims. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 227 – 244).  Mahwah, NJ:  Lawrence Erlbaum Associates. Egan, S. K., & Perry, D.G. (1998). Does low self-regard invite victimization? Developmental Psychology, 34, 299-309. Espelage, D.L., Bosworth, K., & Simon, T.R. (2000). Examining the social context of bullying behaviours in early adolescence. Journal of Counseling & Development, 78, 326-333. Espelage, D. L., & Holt, M.K. (2001). Bullying and victimization during early adolescence: Peer influences and psychosocial correlates. Journal of Emotional Abuse, 2, 123-142.  51  Espelage, D.L., Mebane, S.E., & Adams, R.S. (2004). Empathy, caring, and bullying: Toward an understanding of complex associations. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 37-61). Mahwah, NJ: Lawrence Erlbaum Associates. Espelage, D.L., Mebane, S.E., & Swearer, S. M. (2004). Gender differences in bullying: Moving beyond mean level differences. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 15-35). Mahwah, NJ: Lawrence Erlbaum Associates. Espelage, D.L. & Swearer, S.M. (Eds.). (2004). Bullying in American schools: A Socioecological perspective on prevention and intervention. Mahwah, NJ: Lawrence Erlbaum Associates, Inc., Publishers. Estell, D.B., Cairns, R.B., Farmer, T.W., & Cairns, B.D. (2002). Aggression in inner-city early elementary classrooms: Individual and peer-group configurations. MerrillPalmer Quarterly, 48, 52-76. Farmer, T.W., Leung, M-C., Pearl, R., Rodkin, P.C., Cadwallader, T.W., & Van Acker, R. (2002). Deviant or diverse peer groups? The peer affiliations of aggressive elementary students. Journal of Educational Psychology 94, 611-620. Fraser, M.W. (1996). Aggressive behavior in childhood and early adolescence: An ecological-developmental perspective on youth violence. Social Work, 41, 347361. Galen, B.R., & Underwood, M.K. (1997). A developmental investigation of social aggression among children. Developmental Psychology, 33, 589-600.  52  Garbarino, J. (2001). An ecological perspective on the effects on violence on children. Journal of Community Psychology, 29, 361-378. Garbarino, J. & deLara, E. (2002). And words can hurt forever: How to protect adolescents from bullying, harassment, and emotional violence. New York: The Free Press. Gini, G., 2006. Bullying as a social process: The role of group membership in students’ perception of inter-group aggression at school. Journal of School Psychology, 44, 51-65. Graham, S., & Juvonen, J. (1998). Self-blame and peer victimization in middle school: An attributional analysis. Developmental Psychology, 34, 587-599. Griffith, J. (1995). An empirical examination of a model of social climate in elementary schools. Basic and Applied Social Psychology, 17, 97-117. Gross, E.F., Juvonen, J., & Gable, S.L. (2002). Internet use and well-being in adolescence. Journal of Social Issues 58, 75-90. Haddock, B.L., & Sporakowski, M.J. (1982). Self-concept and family communication: A comparison of status and criminal offenders and non-offenders. Journal of Offender Counseling, Services and Rehabilitation, 7, 61-74. Harris, J.R. (1998). The Nurture Assumption: Why children turn out the way they do. New York, NY: Touchstone. Harter, S. (1999). The construction of self: A developmental perspective. New York, NY: The Guilford Press.  53  Hawker, D.S.J., & Boulton, M.J. (2000). Twenty years’ research on peer victimization and psychosocial adjustment: A meta-analytic review of cross-sectional studies. Journal of Child Psychology and Psychiatry and Allied Disciplines, 41, 441-455. Hawkins, J.D., Catalano, R.F., & Miller, J.Y. (1992). Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood. Implications for substance abuse prevention. Psychological Bulletin, 112, 64-105. Hinduja, S., & Patchin, J.W. (2008). Cyberbullying.us. Retrieved November 2008, from http://www.cyberbullying.us/shareyourstory.php. Hodges, E.V.E., Malone, M.J. & Perry, D.G., 1997. Individual risk and social risk as interacting determinants of victimization in the peer group. Developmental Psychology, 33, 1032-1039. Hoglund, W.L.G. (2007). Social functioning in early adolescence: Gender linked responses to peer victimization. Journal of Educational Psychology, 99, 683-699. Hoover, J.H., Oliver, R., & Hazler, R.J. (1992). Bullying: Perceptions of adolescent victims in the Midwestern U.S.A. School Psychology International, 13, 5-16. Howell, D.C. (1997). Statistical Methods for Psychology (4th ed.). Toronto: Wadsworth Publishing Company. Hymel, S., Bowker, A., & Woody, E. (1993). Aggressive versus withdrawn unpopular children: Variations in peer and self-perceptions in multiple domains. Child Development, 64, 879-896. Hymel, S., Rubin, K.H., Rowden, L., & LeMare, L. (1990). Children’s peer relationships: Longitudinal prediction of internalizing and externalizing problems from middle to late childhood. Child Development, 61, 2004-2021.  54  Kasen, S., Berenson, K., Cohen, P., & Johnson, J.G. (2004). The effects of school climate on changes in aggressive and other behaviors related to bullying. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 107 – 117). Mahwah, NJ: Lawrence Erlbaum Associates. Kowalski, R.M., & Limber, S.P. (2007). Electronic bullying among middle school students. Journal of Adolescent Health, 41, 22-30. Labouvie-Vief, G., Chiodo, L.M., Goguen, L.A., Diehl, M., & Orwoll, L. (1995). Representations of self across the life span. Psychology of aging, 10 (3), 404-415. Lagerspetz, K.M.J., Bjorkqvist, K., & Peltonen, T. (1988). Is indirect aggression typical of females? Gender differences in aggressiveness in 11- to 12-year-old children. Aggressive Behaviour, 14, 403-414. Lei, J. & Zhao, Y. (2007). Technology uses and student achievement: A longitudinal study. Computers & Education, 49, 284-296. Li, Q. (2007). New bottle but old wine: A research of cyberbullying in schools. Computers in Human Behavior, 23(4), 1777-1791. Luiselli, J.K., Putnam, R.F., Hndler, M.W., & Feinberg, A.B. (2005). Whole-school positive behaviour support: Effects on student discipline problems and academic performance. Educational Psychology, 25, 183-198. Maccoby, E.E. (1998). The two sexes: Growing up apart, coming together. Cambridge, MA: Harvard University Press. Macklem, G.L. (2003). Bullying and Teasing: Social power in children’s groups. New York, Kluwer Academic/Plenum Publishers.  55  Maczewski, M. (2002). Exploring identities through the Internet: Youth experiences online. Child & Youth Care Forum, 31, 111-129. Marsh, H.W., Parada, R.H., Yeung, A.S., & Healey, J. (2001). Aggressive troublemakers and victims: A longitudinal model examining the pivotal role of self-concept. Journal of Educational Psychology, 93(2), 411-419. Mayeux, L. & Antonius, H.N., (2008). It’s not just being popular, it’s knowing it, too: The role of self-perceptions of status in the associations between peer status and aggression. Social Development, 4, 871-888. McFarland, D.A. (2001). Student resistance: How the formal and informal organization of classrooms facilitate everyday forms of student defiance. American Journal of Sociology, 107, 612-678. McPherson, M., Smith-Lovin, L., & Cook, J.M. (2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415-444. Media Awareness Network (2005). Young Canadians in a Wired World: Phase II. Retrieved 7 January, 2007 from http://www.mediaawareness.ca/english/research/YCWW/phaseII/ Moffitt, T.E., & Caspi, A. (2001). Childhood predictors differntiate life-course persistent and adolecscent-limited antisocial pathways among males and females. Development and Psychopathology, 13, 355-375. Morettie, M.M., Holland, R., & McKay, S. (2001). Self-other representations and relational and overt aggressionin adolescent girls and boys. Behavioural Schience and the Law, 19, 109-126.  56  O’Connell, P., Pepler, D., & Craig, W. (1999). Peer involvement in bullying: Insights and challenges for intervention. Journal of Adolescence, 22, 437-452. O’Moore, M., & Kirkham, C. (2001). Self-esteem and its relationship to bullying behavior. Aggressive Behavior, 27(4), 269-283. Ojala, K., & Nesdale, D. (2004). Bullying and social identity: The effects of group norms and distinctiveness threat on attitudes towards bullying. British Journal of Developmental Psychology, 22, 19-35. Olweus, D. (1991). Bully/victim problems among school children: Basic facts and effects of a school based intervention program. In Rubin & D. Pepler (Eds.), The Development and Treatment of Childhood Aggression. Hillsdale, NJ: Erlbaum. Olweus, D. (1993a). Bulling at school. Cambridge, MA: Blackwell. Olweus, D. (1993b). Bullies on the playground: The role of victimization. In C.Hart (Ed.). Children on the playground: Research perspectives and applications (pp.85-128). New York: SUNY Press. Olweus, D. (1994a). Annotation: Bullying at school: Basic facts and effects of a schoolbased intervention program. Journal of Child Psychology and Psychiatry, 35, 1171-1190. Olweus, D. (1994b). Bullying at school: Long-term outcomes for the victims and an effective school-based intervention program. In L.R. Huesmann (Ed.), Aggressive behaviour: Current perspectives (pp. 97-130). New York: Plenum Press.  57  Orobio de Castro, B., Brendgen, M., Van Boxtel, H., Vitaro, F., & Schaepers, L. (2007). “Accept me , or else…”: Disputed overestimation of social competence predicts increases in Proactive Aggression. Journal of Abnormal Child Psychology, 35, 165-178. Ott VandeKamp, K., & Hymel, S. (2007). School Safety: Perceptions of secondary students with low incidence disabilities. Paper presented at the 39th Annual Meeting of the National Association of School Psychologists, New York, NY. (29 March 2007). Paquette, J.A., & Underwood, M.K. (1999). Young adolescents’ experiences of peer victimization: Gender differences in accounts of social and physical aggression. Merrill-Palmer Quarterly, 45, 233-258. Patchin, J.W., & Hinduja, S. (2006). Bullies move beyond the schoolyard: A preliminary look at Cyberbullying. Youth Violence and Juvenile Justice, 4, 148-169. Pedrini, L., Sippel, J., Hymel, S., & Waterhouse,T. (March, 2007). Civic education, moral education, and Democracy in a Global Society. Paper presented at the 33rd Annual meeting of the Association for Moral Education, New York, NY. Pelligrini, A.D. (2001a). The roles of dominance and bullying in the development of early heterosexual relationships. In R.A. Geffner, M. Loring, & C. Young (Eds.), Bullying behaviour: Current issues, research, and interventions (pp. 63-73). New York: The Haworth Press, Inc. Pellegrini, A.D., & Bartini, M. (2001). Dominance in early adolescent boys: Affiliative and aggressive dimensions and possible functions. Merrill-Palmer Quarterly, 47, 142-163.  58  Pellegrini, A.D., Bartini, M., & Brooks, F. (1999). School bullies, victims, and aggressive victims: Factors relating to group affiliation and victimization in early adolescence. Journal of Educational Psychology, 91, 216-224. Pellegrini, A.D., & Long, J. (2002). A longitudinal study of bullying, dominance, and victimization during the transition from primary to secondary school. British Journal of Developmental Psychology, 20, 259-280. Pepler, D.J., Craig, W.M., & Roberts, W.L. (1998). Observations of aggressive and nonaggressive children on the school playground. Merrill-Palmer Quarterly, 44, 55-76. Peter, J., Valkenburg, P.M., & Schouten, A.P. (2005). Developing a model of adolescent friendship formation on the Internet. CyberPsychology & Behavior, 8, 423-430. Prinstein, M.J., Boergers, J., & Vernberg, E.M. (2001). Overt and relational aggression in adolescents: Social-psychological adjustment of aggressors and victims. Journal of Clinical Child Psychology, 30, 479-491. Reid, J.B., Eddy, J.M., Fetrow, R.A., & Stoolmiller, M. (1999). Description and immediate impacts of a preventive ntervention for conduct problems. American Journal of Community Psychology, 27, 483-517. Ridout, V., Roberts, D.F., & Foehr, U.G., (2005). Generation M: Media in the Lives of 8-18 year-olds. A Kaiser Family Foundation Study. Retrieved March, 2007 from http://www.kff.org/entmedia/upload/Executive-Summary-Generation-M-Mediain-the-Lives-of-8-18-Year-olds.pdf  59  Rodkin, P.C. (2004). Peer ecologies of aggression and bullying. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 87 – 106). Mahwah, NJ: Lawrence Erlbaum Associates. Rodkin, P.C., Farmer, T.W., Pearl, R., & Van Acker, R. (2002). Heterogeneity of popular boys: Antisocial and prosocial configurations. Developmental Psychology, 36, 14-24. Ross, D. (1996). Childhood bully and teasing: What school personnel, other professionals and parents can do. Alexandria, Virginia: American Counseling Association. Rutter, M., Maughan, B., Mortimore, P., Ouston, J., & Smith, A. (1979). Fifteen thousand hours: Secondary schools and their effects on children. Cambridge, MA: Harvard University Press. Rys, G.S., & Bear, G.G. (1997). Relational aggression and peer relations: Gender and developmental issues. Merrill-Palmer Quarterly, 43, 87-106. Salmivalli, C., Huttunen, A., & Lagerspetz, K.M.J. (1997). Peer networks and bullying in schools. Scandinavian Journal of Psychology, 38, 305-312. Salmivalli, C., Lappalainen, M., & Lagerspetz, M.J. (1997). Stability and change of behaviour in connection with bullying in schools: A two-year follow-up. Aggressive Behaviour, 24, 205-218. Schwartz, D. (2000). Subtypes of victims and aggressors in children’s peer groups. Jouranl of Abnormal Child Psychology, 28, 181-192.  60  Schwartz, D., Gorman, A.H., Nakamoto, J., & Toblin, R.L. (2005). Victimization in the peer group and children’s academic functioning. Journal of Educational Psychology, 97, 425-435. Sharp, S. (1995). How much does bullying hurt?: The effects of bullying on the personal well-being and educational progress of secondary aged students. Educational and Child Psychology, 12, 81-88. Smith, P.K., & Boulton, M. (1990). Rough and tumble play, aggression, and dominance: Perception and behavior in children’s encounters. Human Development, 33, 271282. Song, S., Swearer, S., Eagle, J., & Cary, P. (2000, April). Bullying and school achievement: Understanding a complex relationship. Poster presented at the annual convention of the National Association of School Psychologists, New Orleans, LA. Swearer, S.M., & Doll, B. (2001). Bullying in schools: An ecological framework. Journal of Emotional Abuse, 2, 7-23. Swearer, S.M., & Espelage, D.L. (2004). Introduction: A social-ecological framework of bullying among youth. D. L. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 1 – 12). Mahwah, NJ: Lawrence Erlbaum Associates.  61  Swearer, S.M., Grills, A.E., Haye, K.M. & Cary, P.T. (2004). Internalizing problems in students involved in bullying and victimization: Implications for intervention. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 63-83). Mahwah, NJ: Lawrence Erlbaum Associates. Swearer, S.M., Song, S.Y., Cary, P.T., Eagle, J.W., & Mickelson, W.T. (2001). Psychosocial correlates in bullying and victimization: The relationship between depression, anxiety, and bully/victim status. Journal of Emotional Abuse, 2, 95121. Tabachnick, B. G., & Fidell, L.S. (2001). Using Multivariate Statistics (4th Ed). MA: Allyn & Bacon. Thomas, D. R., Hughes, E., & Zumbo, B. D. (1998). On variable importance in linear regression. Social Indicators Research: An International and Interdisciplinary Journal for Quality-of-Life Measurement, 45, 253-275. Tomada, G., & Schneider, B.H. (1997). Relational aggression, gender, and peer acceptance: Invariance across culture, stability over time, and concordance among informants. Developmental Psychology, 33, 601-609. Underwood, M.K. (2003). Social Aggression Among Girls. New York: Guildford Press. Underwood, M.K. (2004). Glares of contempt, eye rolls of disgust and turning away to exclude: Non-verbal forms of social aggression among girls. Feminism & Psychology, 14, 371-375.  62  Underwood, M.K., & Bjornstad, G.J. (2001). Children’s emotional experience of peer provocation: The relation between observed behaviour and self-reports of emotions, expressions, and social goals. International Journal of Behavioural Development, 25, 320-330. Underwood, M.K., Galen, B.R., & Paquette, J.A. (2001). Top ten challenges for understanding gender and aggression in children: Why can’t we all just get along? Social Development, 10, 248-266. Underwood, M.K., Hurley, J.C., Johanson, C.A., & Mosley, J.E. (1999). An experimental, observational investigation of children’s responses to peer provocation: Developmental and gender differences in middle childhood. Child Development, 70, 1428-1446). Underwood, M.K., Scott, B.L., Galperin, M.B., Bjornstad, G.J., & Sexton, A.M. (2004). An observational study of social exclusion under varied conditions: Gender and developmental differences. Child Development, 75, 1538-1555. Vaillancourt, T., Hymel, S., & McDougall, P. (2003). Bullying is power: Implications for school intervention programs. Journal of Applied School Psychology. Vaillancourt, T., McDougall, P., Hymel, S., Krygsman, A., Miller, J., Stiver, K., & Davis, C. (2008). Bullying: Are researchers and children/youth talking about the same thing? International Journal of Behavioural Development, 32 I (6), 486-495. Valkenburg, P.M., Schouten, A.P., & Peter, J. (2005). Adolescents’ identity experiments on the internet. New Media & Society, 7, 383-402. Wallace, P. (1999). The psychology of the Internet. New York: Cambridge University Press.  63  Ward, C. & Tracey, T.J.G. (2004). Relation of shyness with aspects of online relationship involvement. Journal of Social and Personal Relationships, 21, 611623. Weissberg, R.P., & Greenberg, M.T. (1997). School and community enhancement and prevention programs. In W. Damon, I.E. Sigel, & K.A. Renninger (Eds.), Handbook of child psychology (Vol.5). New York: John Wiley and Sons. 255280. Xie, H, Swift, D.J., Cairns, B.D., & Cairns, R.B. (2002). Aggressive behaviors in social interaction and developmental adaptation: A narrative analysis of interpersonal conflicts during early adolescence. Social Development, 11, 205-224. Ybarra, M.L. & Mitchell, K.J. (2004). Online aggressor/targets, aggressors, and targets: A comparison of associated youth characteristics. Journal of Child Psychology and Psychiatry, 45, 1308-1316. Young, L. (2008). Understanding a Child’s Virtual World: A handbook for parents and guardians. Available from http://www.incredibleinternet.com/online-safety/drlinda-young/parent-guide. Zahn-Waxler, C. (2000). The development of empathy, guilt, and internalization of distress. In R. Davidson (Ed,), Anxiety, depression, and emotion: Wisconsin Symposium on Emotion, Volume II (pp. 222-265). New York: Oxford University Press. Zimmer-Gembeck, M.J., Geiger, T.C., & Crick, N.R. (2005). Relational and physical aggression, prosocial behavior, and peer relations: Gender moderation and bidirectional associations. Journal of Early Adolescence 25, 421- 452.  64  CHAPTER 3. CYBERBULLYING VERSUS SCHOOLYARD BULLYING: SAME OR DIFFERENT?5 The purpose of study I was to investigate the construct of online vs. offline bullying/victimization and to examine some of the factors that predict online bullying/victimization among adolescents. The results of Study I revealed that adolescents perceive online bullying differently from offline bullying. This highlights the need to investigate this form of aggression more closely in its own right. Study II is designed to accomplish this, to further elucidate the construct of online aggression, and to assess some of the individual predictors and motivations for engaging in it.  3.1. AGGRESSIVE BEHAVIOURS DURING ADOLESCENCE Research on aggressive behaviours in children and adolescents has long been studied (see Rubin, Bukowski, & Parker, 2006; Tremblay, Hartup, & Archer, 2005 for reviews). Most researchers agree that aggression includes behaviours that are conducted with the intention to hurt or harm others (Berkowitz, 1993; Crick & Grotpeter, 1995; Underwood, Scott, Galperin, Bjornstad, & Sexton, 2004). However, the motivations underlying different aggressive acts and the ways they are exerted appear to differ, leading to distinctions between several subtypes of aggression including: physical versus social aggression (Galen & Underwood 1997; Paquette & Underwood, 1999), direct versus indirect (Bjorkqvist, Lagerspetz, & Kaukiainen, 1992), overt versus relational (Crick, Bigbee, & Howes, 1996; Crick & Grotpeter, 1995), reactive versus proactive (Dodge & Coie, 1987), relational versus social versus indirect (Crick & Grotpreter, 1995;  5  A version of this chapter will be submitted for publication. Law, D. M., Shapka, J.D. & Domené, J.F. (2009). The Changing Face of Bullying: An Empirical Comparison Between Traditional and Internet Bullying and Victimization  65  Oesterman, Bjoerkqvist, Lagerspetz, Kaukiainen, Landau, Fraczek, Caprara, 1998; Underwood, 2003; Underwood, Galen, & Paquette, 2001), and confrontational versus non-confrontational (Xie, Cairns & Cairns, 2004). Several of these sub-types of aggression are relevant to the study of online bullying. Specifically, the current study differentiates between proactive and reactive aggression, confrontational and non-confrontational, and physical, verbal, and relational/social aggression. Each of these types of aggression is described below. 3.1.1. Proactive vs. Reactive Aggression Proactive aggression refers to aggression that is used to intentionally obtain a resource or a goal (Crick & Dodge, 1996; Dodge & Coie, 1987). Arguably, bullying behaviours fall under this domain because people who bully do so in an attempt to demonstrate power and authority over the target (Atlas & Peplar, 2001; Olweus, 1991). Bullying behaviours can involve either physical or direct forms of aggression, such as kicking and punching (Olweus, 1991; Pellegrini & Bartini, 2001; Smith & Boulton, 1990) or relational aggression, such as rumour spreading, gossiping, and social isolation (Bosworth, Espelage, & Simon, 1999; Underwood et al., 2001). The key differences between bullying and other forms of aggression are intentionality, repetition of the aggressive act, and the power differential between bully and victim. That is, in a bullying situation, the individual intentionally chooses to inflict harm on another, and does so repeatedly over time in order to dominate or impress power over the victim (Olweus, 1991). With this in mind, it is important to note that there are many acts of Proactive Aggression that do not include all three of these components, and are therefore not considered bullying. For example, a proactive aggressive incident might include pushing 66  someone in order to be first in line for lunch. This is not considered bullying because it is a onetime occurrence. Reactive aggression is aggression that is exerted as a reaction to provocation (Crick & Dodge, 1996; Dodge & Coie, 1987). For example, an individual may interpret the intentions of another as being deliberately hurtful, and so may retaliate in anger (they are reacting to a perceived threat). As with proactive aggression, not all instances of reactive aggression are considered bullying. Keeping the three components of bullying in mind, reactive aggression would only be considered bullying if it occurs repeatedly, with intention to harm, and with the purpose to impress power on the victim. For example, if an individual interprets someone bumping into him/her as intentional, and thus, reacts aggressively by pushing the “perpetrator” in return, this instance would not be considered bullying if it only occurs one time. However, if the “victim” continues to push and exert aggression on the “perpetrator” with the idea that he/she must demonstrate power and strength over him/her, then this may be considered bullying. 3.1.2. Confrontational vs. Non-Confrontational Aggression Beyond the distinctions between proactive and reactive aggression, researchers have also differentiated between confrontational and non-confrontational aggressive behaviours (Xie, Cairns, & Cairns, 2004). Confrontational aggression is a form of aggression which occurs directly between one person and another, where the perpetrator’s identity is known. Acts include such behaviours as telling another person that “I don’t like you anymore” or “You can’t play with us because you’re weird”. Nonconfrontational aggression, on the other hand, refers to aggressive acts where the perpetrator’s identity is not necessarily known, for example, saying something behind the 67  victim’s back, telling peers not to talk to a certain person, or spreading rumours about another person. Research conducted by Xie et al. (2004) found that children who engaged in confrontational aggression were more likely to do poorly in school, to be unpopular among their peers, and to engage in various problem behaviours. None of these difficulties were found among children who engaged in non-confrontational forms of aggression. 3.1.3. Relational Aggression Whether they are proactive or reactive, confrontational or non-confrontational, aggressive acts can be exerted in different ways. Researchers have distinguished between three main forms of aggression: (1) Physical Aggression, that is, harming others through physical force, and/or instrumental intimidation (Atlas & Pepler, 1998; Craig, et al., 2000; Crick & Grotpeter, 1995), (2) Verbal Aggression, that is, yelling, insulting, namecalling, and threatening (Salmivalli, Kaukiainen, & Lagerspetz, 2000), and (3) Relational Aggression, that is, gossiping, spreading rumours, and social exclusion (Crick & Grotpreter, 1995; Lagerspetz, & Bjorkqvist, 1994; Lagerspetz, Bjorkqvist, Peltonen, 1988; Oesterman et al., 1998; Underwood, 2003; Underwood et al., 2001). Addressing all of these forms of aggression is beyond the scope of this study. Therefore, I will be focusing specifically on Relational Aggression, as this type of aggression is likely the most related to online aggression. Also referred to as Indirect or Social aggression (Crick & Grotpreter, 1995; Oesterman et al., 1998; Underwood et al., 2001), Relational Aggression is comprised of four main factors: (1) it is purposefully used to manipulate and damage peer relationships through gossiping, spreading rumours, or eliciting either verbal or non-verbal peer 68  rejection; (2) it can involve confrontational tactics to inflict harm (e.g., telling someone you don’t like them and they are not welcome to join the group), (3) it can also include non-confrontational methods for inflicting harm, thus, making it difficult for the victim to know who initiated or who is involved in the attack (e.g., spreading rumours anonymously), and (4) it can include both verbal and non-verbal behaviours such as calling someone names or ignoring someone (Putallaz, Kupersmidt, Coie, McKnight, & Grimes, 2004). Although some researchers have made differentiations among Relational, Indirect, and Social aggression (Crick & Grotpeter, 1995; Oesterman et al., 1998; Underwood et al., 2001) Relational aggression will be defined as any form of aggression that is exerted through rumour-spreading, gossiping, social exclusion/manipulation, or rejection. 3.1.5. Online Aggression As Cyberbullying, Cyber-aggression, or Cyber-harassment become more prominent in the news and popular media, as well as in academic works (e.g., Hinduja & Patchin, 2006; Li, 2006; Ybarra & Mitchell, 2004), it becomes important to determine exactly what this form of aggression is and how it manifests. The results of Study I showed that adolescents tend to view Cyberbullying and Cybervictimization as a single construct, and that, while related, Cyberbullying appears to be distinct from Traditional Bullying and Traditional Victimization. With this in mind, this study will refer to aggressive acts that occur over the Internet as online aggression, as the terms Cyberbullying or Cybervictimization may not be appropriate for this type of aggression. Previous work on bullying and aggression has stipulated that a critical difference between bullying and aggression is that bullying involves, (a) intention to harm, (b) repetition, and 69  (c) a power differential (Hoover, Oliver, & Hazler, 1992; Olweus, 1991, Pelligrini & Bartini, 2001; Smith & Boulton, 1990). As yet, research on “Cyberbullying” has yet to determine whether the construct measured includes the intentionality, repetition, and power differences (whether physical or social) that define bullying per se. The current study examined several questions to further understand how and why aggression is occurring online. This work also explored whether adolescents were choosing to aggress online due to proactive or reactive factors. It is important to determine whether adolescents are aggressing online because they feel they have been wronged in some way or because they are attempting to gain power, in order to ensure that appropriate intervention and prevention strategies are developed. This study also examined whether adolescents were more likely to be confrontational or nonconfrontational in their online aggression. Perhaps the use of the Internet and other forms of technological communications enables users to be simultaneously confrontational and non-confrontational. For example, although one’s name is attached to an email, the physical absence between sender and receiver can give the illusion of being nonconfrontational. With the advent of the Internet, social networks can be broadened to include up to hundreds of people, which may prove to have a profound influence on the impact of a bullying incident. The current study explores how Information Communication Technologies (ICTs) are linked to online aggression.  3.2. ADOLESCENT INTERNET USE AND ONLINE AGGRESSION Internet use is growing at an exponential rate and has become a ubiquitous presence in the lives of Canadian adolescents. In 2005, Media Awareness Network (MNet) conducted a national survey among 5,272 Canadian students in grades 4 to 11 70  and found that 94% of children have Internet access at home, 61% of whom have high – speed connection. Moreover, by grade 11, 51% have access to their own Internetconnected computer. Furthermore, communication via ICTs is no longer limited to the computer. Sixty-eight percent of children have access to cell phones, 44% of whom can use these phones to surf the Internet, while 56% use them to text message their friends (MNet, 2005). Similar trends have been found in the United States, Europe, and Asia (Ridout, Roberts, & Foehr, 2005; Kraut, Brynin, & Kiesler, 2006). Despite the prevalent use of ICTs in the lives of adolescents, we know very little about how the Internet or cell phones are influencing adolescents’ communication skills and social relationships. In examining how ICTs are related to aggressive communication, the current study will focus on four popular methods of online communication: Instant Messaging, cell phones, social networking sites, and YouTube. 3.2.1. Instant Messaging Instant Messaging (IM) allows individuals to textually communicate, in real-time, with others through a software application (e.g., MSN, Yahoo Chat, ICQ). Generally included in the IM software is a feature that allows users to easily see whether a chosen friend or “buddy” is online. Chosen individuals can be added to a user’s “buddy list” by entering their email address and requesting acceptance as a friend. IM has been found to be a very common method of communication among adolescents during non-school hours (Boneva, Quinn, Kraut, Kiesler, & Shklovski, 2006). In fact, 80% of adolescents use IM on a daily basis, and research on IM use has found that by grade 6, Canadian girls prefer IM’ing over any other online activity. Similarly, other data has shown that from  71  grade 9 onwards, IM is the favored online activity for 54-61% of boys and 80-83% of girls (MNet, 2005). To determine who adolescents are chatting with over IM, recent research has shown that adolescents use IM to communicate with people they see at school, or to keep in touch with friends and relatives who do not live in their immediate community (i.e., outside of the city or country). As such, most individuals on a given “buddy list” are known to the users in “real life” (Greenfield, Gross, Subrahmanyam, Suzuki, & Tynes, 2006; Gross, et al., 2001; Pew Internet & American Life Project, 2003; Ramirez, Dimmick, & Lin, 2004). Most exchanges are text-only, although, popular services now allow voice messaging, file sharing, and even video chat when both users have Webcams. IM use has not replaced traditional after-school activities for adolescents (i.e., clubs, sports, or meeting face-to-face with friends). In fact, recent research among college students has found that face-to-face communication was significantly more gratifying and useful than communicating with any form of technology, including the landline phone (Flanagin, 2005). Nevertheless, IM does provide adolescents with another avenue for communicating with friends, and for creating closer relationships (Flanagin, 2005; Gross et al., 2002). Moreover, adolescents report IM as being a useful aid for reducing shyness (Boneva et al., 2006), and as being socially gratifying (Flanagin, 2005; Ramirez et al., 2004). Although IM is one of the most widely used communication devices among adolescents, little research has been conducted on how it may be facilitating aggression and bullying. The possibility for IM to be used as an avenue for saying mean things, spreading rumours and gossip is quite evident. This is exemplified by one 15-year-old’s  72  explanation of how “One of my friends started hassling me on MSN messenger; she was sending me nasty messages and text messages and this carried on at school” (Hinduja & Patchin, 2008, Share Your Story section). Furthermore, communicating via IM allows perpetrators to be simultaneously confrontational and non-confrontational in their aggressive acts since their names are attached to the messages they send out, but they are protected by the large physical distance that lies between the perpetrator and the victim. Alternatively, if a perpetrator acquires someone’s IM account information, it allows the perpetrator to say mean things or spread rumours about others, without the person’s knowledge and without taking any blame. The current research is a first step at examining how these factors might change the nature of aggression, as well as who is doing the aggressing. 3.2.2. Cell Phones An increasing number of adolescents are using cell phones. As previously mentioned, almost 70% of Canadian children have access to a cell phone (MNet, 2005). Moreover, in 2005, 45% of American adolescents owned a cell phone (Pew Internet and American Life Project, 2005). In Norway, between 1997 and 2001, the use of cell phones among adolescents increased from approximately 15% to over 90% (Ling & Yttri, 2006). With this increase in cell phone use comes an increase in cell phone text messaging (also called Short Message Service or SMS). Almost 60% of Canadian adolescents and 33% of American adolescents use their cell phones to text message their friends (MNet; Pew Internet and American Life Project). SMS is a service available on most digital cell phones and allows users to send short alpha-numeric messages to each other. Popular IM  73  programs, such as MSN, also allow their users to send text messages to the cell phones of other members of the MSN community. Research conducted in Europe has found that adolescents are the primary users of SMS text messaging (Ling & Yttri, 2006). The relatively low cost, inclusion in a number of cell phone plans (e.g., Telus, Rogers, Bell in Canada), and ability to allow young people to silently contact friends at times where verbal telephone communication may be inappropriate (e.g., during class, or in the middle of the night) has contributed to the growing popularity of SMS (Ling & Yttri). Similar to IM, texting via cell phones has provided new opportunities for peer group interaction. In fact, 20% of Norwegian adolescents reported sending cell phone text messages at least once a week between midnight and 6am. Adolescents are no longer hiding beneath their covers reading with a flashlight; rather, they are sending and receiving text messages with their friends, as substantiated by the following focus group data from Ling and Yttri’s study (p. 219): Kai (15): You can’t call your friends at one o’clock at night you know, that’d really piss off their parents. Moderator: One o’clock at night? By then I’m already in bed and asleep for a couple of hours. Harald (15): But it’s way better at night, being on your mobile phone under the covers, instead of sitting in the middle of the living room. Ola (14): My cell phone’s always on. I just turn off the sound… In addition to using cell phones for talking to friends and sending text messages, 17% of Canadian adolescents have cell phone cameras (MNet, 2005). Moreover, a growing number of cell phones are equipped with the capacity for users to take short  74  video clips of themselves and others. Although no known empirical research has been conducted on how adolescents are using video cell phone features, MNet has found that it has become common practice to post pictures taken from cell phones on MSN profiles and social networking sites (described below), and video clips on websites such as YouTube (described below). From a bullying perspective, the ease with which an individual could take an inappropriate or unflattering photo of someone and post it publicly is quite concerning. 3.2.3. Social Networking Sites No formal definition for social networking sites currently exists. However, researchers generally define these websites as sites where users can create a personal profile of themselves and connect that profile to the profiles of others, for the purposes of establishing or maintaining connections with friends and for creating an explicit personal social network (Ellison, Steinfield, & Lampe, 2006; Pew Internet & American Life Project, 2007, Valkenburg, Peter, & Schouten, 2006). Recent research, conducted by Pew Internet and American Life Project (2007) has found that 55% of American adolescents have used and created personal profiles on social networking sites, such as MySpace or Facebook, with almost 50% of participants visiting these sites more than once a day. Social Networking Profiles typically allow users the option of posting a picture of themselves and demographic information, such as age, and education, as well as their interests and hobbies. In addition, users are able to post messages to people within their network, upload online photo albums for others to see, and post comments about the photos of others. A recent study conducted in the Netherlands revealed an association between the feedback adolescents receive from others about their pictures and 75  postings and their self-esteem/well-being; positive feedback from others was related to increased self-esteem, whereas negative feedback decreased their self-esteem (Valkenburg et al., 2006). This finding is relevant to the current study as it suggests that aggressive online acts may have a negative effect on adolescent’s developing sense of self. 3.2.4. YouTube YouTube is a website that allows users to upload, view, and share up to 10-minute video clips with others. The website not only informs viewers of how many times a clip has been watched, but also allows users to rate and post comments about the clip. Empirical research on how YouTube is being used among adolescents is non-existent, although stories in the popular media indicate that it is common for adolescents to film school yard fights or students changing in locker rooms and uploading them to YouTube (Smith, 2007). This raises concerns about the use of this online venue for engaging in online aggression.  3.3. PREDICTORS OF BULLYING BEHAVIOURS AND VICTIMIZATION Research has identified a number of predictors of traditional bullying and victimization. A useful framework for thinking about the myriad of influences on bullying is a socio-ecological approach that simultaneously encompasses the role of individual characteristics, peer groups, and school factors (Espelage & Swearer, 2004). This current study examined some of the individual factors that have been linked to traditional bullying, including age, sex, and self-esteem differences. This study also  76  assessed whether adolescents are proactive or reactive in their online hostility and whether their aggression is confrontational or non-confrontational in nature. 3.3.1. Age and Sex Differences. As described in Study I, there are significant age and sex differences when it comes to bullying and aggression. To recap, it appeared that aggressive behaviours tend to become less physical and more social over time (Coie & Dodge, 1998; Sharp, 1995) and that girls and boys tend to aggress differently according to their social priorities, with boys aggressing in more physical ways and girls aggressing in more social ways (Conway, 2005; Crick, 1995, 1997; Pellegrini, 2001a; Pellegrini & Bartini, 2001; Pellegrini & Long, 2002). The current study explored sex and grade variables to see if these are significant factors in the different facets of online aggression. 3.3.2. Self-Esteem. Several different terms have been used to refer to self-perceptions including selfconcept, self-esteem, self-evaluations, self-image, self-representations, and self-worth (Harter, 1999). The primary distinction is between terms that reflect a person’s evaluation of his or her worth or value as a person (e.g., self esteem; self worth, Harter, 1989) and terms which refer to a person’s perception of how well they function in different domains of their life (e.g., self-concept, Rosenberg, 1986). From this perspective, self-esteem or self-worth can be considered evaluative terms and self-concept a descriptive term. This study utilizes self-esteem instead of self-concept, as it has been the primary construct used in previous research looking at the relationship between self-perceptions and  77  aggression (Boivin, Thomassin, & Alain, 1989; Hymel et al., 1993; Kinard, 1978; Leary, Schreindorfer, & Haupt, 1995; Stucke & Sporer, 2002). Study I contained a detailed explanation of how self-esteem was related to traditional bullying. In summary, individuals’ self-esteem plays a role in predicting bullying and victimization, although the research in this area is mixed. For example, studies have found that adolescents who feel poorly about themselves are more likely to engage in aggressive behaviours (Marsh et al., 2001; Moretti et al., 2001), whereas other research has found that adolescents who hold themselves in high regard are more likely to engage in this sort of behaviour (Mayeux & Cillessen, 2008; Orobio de Castro, et al., 2007). These findings are likely a result of the different sub-groups of aggressors and their relationships with their peers. For example, research has found that aggressors who are generally viewed by peers as popular and powerful have higher self-perceptions (Vaillancourt, Hymel, & McDougall, 2003), perhaps for the purpose of maintaining status along the social hierarchy (Dodge, Coie, Pettit, & Price, 1990; Estell, Cairns, Farmer, & Cairns, 2002; Rodkin, Farmer, Pearl, & Van Acker, 2000; Vaillancourt et al.; Xie et al., 2002). On the other hand, those with lower self-esteem and those who perceived themselves as disliked among their peers were also found to demonstrate an increase in aggression over time (Mayeux et al.), possibly because they are striving to identify with their peers and feel a sense of belonging (Brown, 2000; Gini, 2006). The Internet provides another venue for adolescents with higher or lower self-esteem to engage in aggression. This study examines this relationship more closely.  78  3.3.3. Summary Previous research has revealed that aggression is multidimensional, multifaceted, and complex. Moreover, aggressive behaviours are an interaction of multiple variables, such as individual and contextual factors, and needs to be considered in the prediction of aggressive behaviours. The purpose of this study was to empirically examine the nature of online aggression, by addressing four specific research questions: 1. What are the sex and grade differences associated with online aggression? 2. How does self-esteem predict online aggression? 3. Is online aggression primarily Proactive or Reactive? 4. Are adolescents using confrontational or non-confrontational means for engaging in online aggression? The overarching hypothesis for this research, based on the results of Study 1 and previous literature, was that online aggression would be distinct from traditional forms of bullying, and that this was likely related to the medium being used for aggression. Regarding the intention underlying the aggressive acts, given that the Internet provides ample opportunity for individuals to engage in relationally aggressive acts, such as gossiping and spreading rumours (e.g., over IM, social networking sites, and/or SMS text messaging), it was postulated that, online, adolescents would engage in both confrontational and non-confrontational aggression (e.g., saying mean things directly to someone over IM, or spreading rumours covertly among peers through private messages). Furthermore, it was hypothesized that older girls would engage in this form of aggression more than younger adolescents and boys. This stipulation is based on previous work on relational aggression and Internet use that has found that girls are more likely to engage  79  in this form of aggression than boys and that girls tend to use venues such as IM and social networking sites to facilitate communication and relationship among peers more than boys (Flanagin, 2005; Gross et al. 2002; MNet, 2005). With respect to self-esteem, based on the literature, and the results of Study 1, it was hypothesized that those who engaged in online aggression would have lower self-esteem (Mayeux & Cillessen, 2008). In order to answer these questions and confirm these hypotheses, this study employed a positivist, concurrent mixed-methods research design for examining online aggression. Specifically, this research consisted of quantitative, self-report survey data that examined the statistical trends in online aggression, which were then supplemented by data from semi-structured interviews, to obtain a deeper and richer understanding of this phenomenon.  3.4. METHOD Data for this study were drawn from a self-report questionnaire, as well as semistructured interviews conducted with a subset of participants. The study was conducted using a Concurrent Nested mixed-method research strategy (Creswell, 2009). That is, the purpose of the questionnaire was to gain a broad perspective of online aggression and some of its predictors, whereas the purpose of the interviews was to supplement the quantitative findings by exploring the motivations and means by which adolescents are engaging in online aggression. Thus, the qualitative component is nested within the larger context of a primarily quantitative study. Due to the fact that the quantitative component is the primary part of this nested mixed-method design, the entire study resides within the positivist research paradigm. Thus, positivist approaches to interviewing, analysis of interview content, and validation of findings were employed. Survey and interview data 80  were collected concurrently, and findings from each method of data collection were merged and examined together to develop a clearer sense of the predictors and motivations underlining online aggression. 3.4.1. Participants Elementary and high school students between the ages of 10 and 18 from the Lower Mainland of British Columbia were invited to participate in this study. Upon obtaining ethics approval from the UBC Behavioural Research Ethics Board and four school boards in the Lower Mainland of British Columbia (Appendix B), principals and teachers were contacted, and a request to visit their classrooms was made. Prior to collecting the data, the researchers visited the classrooms of consenting teachers and explained the study to the students. At this time, letters and accompanying consent forms were distributed to students, to be passed to their parents for parental consent (Appendix B). Students who were given parental consent and who assented themselves were asked to complete the questionnaire. In total, 1487 consent forms were distributed with a response rate of 49%. The questionnaire, Social Responsibility on the Internet (Appendix C) was developed specifically for the purpose of this study. In developing this questionnaire, various researchers with expertise in developmental psychology, socioemotional learning, technology use, and measurement were consulted. Prior to completing the questionnaire, participants were asked whether they wanted to be considered for a future interview, and if so they were asked to provide their email address and/or phone number so that they could be contacted in the future. The final sample for the questionnaire portion of the study included 733 (454 females; 255 males) elementary and high school students between the ages of 10 and 18. 81  Grade 8 and Grade 10 students comprised 20% of the sample each, whereas grades 6, 7, 9, 11 and 12 students comprised a total of 53% of the sample; approximately 10% per grade. Only 4% of the sample was in grade 5. It should be noted that although only a small number of participants were in grade 5, given the large size of the overall sample, and the robustness of regression for dealing with unbalanced designs (Howell, 1997), they were not collapsed with the next grade. All analyses were also run with a collapsed grade 5 and 6 group to ensure that there were no differences in the outcomes. In terms of ethnicity, 45% of the participants were of East Asian descent (e.g. Cambodian, Chinese, Japanese, Korean, Taiwanese, Vietnamese, Filipino), whereas 34% were of European descent. The remaining ethnic groups included Aboriginal, African/Caribbean, South Asian, Latin American, Middle Eastern, and mixed background; all of which comprised approximately 2% of the sample. In addition to this, information regarding a series of Yes/No questions about whether participants had Internet access, or owned a cell phone were asked to ascertain how accessible these technologies were to the participants. A large majority of the students (81%) had high speed Internet, and approximately 60% of participants had their own cell phone. 3.4.2. Questionnaire Data – Outcome Variables Online Aggression. To measure online aggression, 16 questions were included in this study, all aimed at tapping into different aspects of being aggressive online, being aggressed upon online, and witnessing aggressive acts online (See Appendix C questions 28 and 29). These questions were derived from traditional research which has identified three key players in bullying situations; the bully, the victim, and the witness (Craig & Peplar, 1995; 2000). In order to examine whether online aggression follows the same 82  patterns as offline bullying, questions on the roles adolescents might play in a bullying situation were asked. In addition to this, adolescents were asked to explain whether they engaged in online aggressive acts “with friends” or “alone”. These questions were posed to examine the influence peer groups might have when it comes to online aggressive situations. Traditional bullying literature has described how group size plays a role in regulating bullying behaviours, and that having peers around can reinforce aggressive behaviours through approval or rumour spreading (O’Connell, Pepler, & Craid, 1999; Salmivalli, Huttunen, & Lagersptz, 1997). Given that most “Cyberbullying” incidents occur in the home (Young, 2008), it is important to assess whether adolescents are more or less likely to aggress online with peers or alone. The purpose of having multiple online aggression items was to measure various aspects of online bullying and victimization. As such, the first level of analyses was to conduct an unweighted least squares exploratory factor analyses to determine the factor structure of these items. These analyses are described in the Results section. 3.4.2. Questionnaire Data – Predictor Variables Self-Esteem. To measure adolescent self-esteem, Marsh’s (1990) Self Description Questionnaire was used (See Appendix C, question 26.2, 26.9. 26.15, 26.21, 26.27, 26.37, 26.43, 26.48, 26.51). This study only included the nine items of Marsh’s ‘general’ selfesteem subscale. Items for this measure were structured on a 6-point Likert scale format and respondents were asked to rate whether each statement was False, Mostly false, More False than True, More True than False, Mostly True, or True. The composite for this sub-scale was calculated by taking the mean of the items for this domain. Higher scores  83  indicated higher self-esteem. In terms of reliability, high internal consistency scores for the Self-Esteem score was found to be 0.946. Proactive/Reactive Aggression. In order to evaluate proactive and reactive aggression, Raine and colleagues’ (2006) Reactive-Proactive Aggression Questionnaire was modified so that it specifically focused on online aggression (See Appendix C question 32). Items were structured using a 3-point scale where 0 = never, 1 = sometimes, and 2 = often. Similar to the online aggression items, a factor analyses was run to determine the factor structure of these adapted items. 3.4.3. Questionnaire Data – Covariates Computer in the Bedroom: Having a computer in the bedroom has been linked to increased misbehavior on the Internet (Ridout, Roberts, & Foehr, 2005); as such, it was considered important to control for this. Adolescents were asked to respond yes or no to the question “Do you have a computer in your bedroom?” where 0 = No and 1 = Yes. Lone Parenting. Research has shown an inverse relationship between socioeconomic status and lone parenting (Rahkonen, Laaksonen, & Karvonen, 2004). Given this, whether adolescents lived with one parent or both was used as a proxy for determining socio-economic status. Previous research has found that adolescents who come from lower income families might be more prone to engaging in inappropriate behaviours both offline and online (Ridout et al., 2005); as such, it was deemed important to control for this effect. The Lone Parenting variable was constructed by recoding question 13, “Which of these adults do you live with MOST OF THE TIME? Check all  6  It should be noted that Marsh’s model of self-esteem was chosen over Harter’s scale (1988) because it was suitable for the entire age-range of participants in this study, whereas the Harter scale uses separate child and adolescent versions.  84  the adults you live with” (See Appendix C, question 13), into dummy variables where 0 = not a lone parent and 1 = lone parent. Interview Data Data collection. To address questions 3 and 4 of the study, semi-structured interviews were conducted with 15 adolescents (10 females and 5 males), in order to gain a more thorough understanding of the relationships between online aggression and confrontational, non-confrontational, reactive, and proactive forms of aggression. Interview participants were drawn from the pool of students who participated in the questionnaire portion of the study, and who had expressed interest in participating in an interview. Participants were given the choice of being interviewed in either a face-toface setting or via IM, with 4 participants choosing a face-to-face interview, and 11 participants choosing to be interviewed over IM. As part of the interview protocol, participants were asked to recall and describe times that they had been hurt online (e.g., been a target of an aggressive act), that they had engaged in hurting others online (e.g., perpetrated an aggressive act), and that they had seen others being hurt online (e.g., witnessed an aggressive act). Participants were asked several questions for each of these experiences, and follow-up probes were used to gain a deeper and richer sense of what had occurred and why (see Appendix D for specific interview questions). Interview times ranged from 15 minutes to 1 hour for the face-to-face interviews, and 30 minutes to 1.5 hours for the MSN interviews. Data analysis. Prior to analyzing the data, voice recorded face-to-face interview data were transcribed into textual format and MSN log histories were imported into a word processor and formatted to match the structure of the face-to-face interview  85  transcripts. Transcription was initially conducted by a professional transcriber, but the interview content was subsequently reviewed and corrected by the original interviewer, to improve the accuracy of the transcription process. All interviews were then coded according to four central themes: (a) Proactive Aggression, (b) Reactive Aggression, (c) Confrontational Aggression, and (d) NonConfrontational Aggression. Consistent with the positivist nested research design that was used, these categories were formed on an a priori basis, based on the existing literature and the categories used in the quantitative analyses. Coding was accomplished by attending to the manifest content of participants’ statements, to identify sections related to the four themes. These statements were then categorized and extracted to formulate a description of participants’ experiences that, as much as possible, preserved participants’ own words. Data from the Proactive/Reactive themes were used to elaborate and clarify the findings derived from the questionnaire on this form of aggression. Data from the Confrontational/Non-Confrontational themes were used to obtain a deeper understanding of the motivations behind online aggression and whether the security of being behind a screen facilitates how participants choose to react to this form of aggression. To improve the integrity of the analysis, a credibility checking process was conducted as part of the analysis. The credibility process was adapted from the Critical Incident Technique (CIT), a widely used qualitative research method that is also located within the positivist paradigm (Butterfield, Borgen, Amundson, & Maglio, 2005; Chell, 1998; Woolsey, 1986). First, the integrity of the coding process was evaluated by having a research assistant independently code 20% of the interviews, to assess for concordance  86  between the codes generated by the primary investigator, and those generated by the independent coder. This initial credibility check revealed some variability between the two coders’ thematic placement. Specifically, only about 35% of incidences were similarly coded. The primary source of disagreement between the coders was in the categorization of participants’ described experiences as Confrontational versus NonConfrontational Aggression. Consequently, a Theoretical Validity Check (Maxwell, 1992) was performed, with the result that the original Confrontational versus Non-Confrontational category was modified to increase specificity. Review of relevant theory revealed that, although aggression literature utilizes the Confrontational and Non-Confrontational categories (Xie et al., 2004), this may apply primarily to face-to-face aggression. In contrast, in an online setting, it is possible for individuals to be simultaneously confrontational and nonconfrontational (e.g., sending a mean message over MSN, but avoiding saying those same mean words in person). Moreover, the results of Study 1 suggested that online aggression is a distinct phenomenon from face-to-face aggression, suggesting that it may be important to distinguish between the setting (i.e., online vs. offline) in addition to the style (Confrontational vs. Non-confrontational) of aggression. In the end, the original Confrontational and Non-Confrontational categories were divided into four sub-categories: (a) Non-Confrontational online (i.e., gossip online), (b) Non-Confrontational offline (i.e., socially excluding someone offline, gossiping and rumour spreading offline), (c) Confrontational online (i.e., saying mean things directly to someone over MSN), and (d) Confrontational offline (i.e., saying mean things to  87  somebody’s face). The three interviews used in the initial credibility check were then independently re-coded by both coders, according to the new categorization system. This process yielded a much higher degree of correspondence, with both coders being in agreement 95% of the time. Subsequently, the entire data set was recoded using the four separate categories of confrontation. The original Reactive and Proactive themes were retained, because those categories received 98% concordance in the first credibility check. The remaining discrepancies in coding were discussed by the primary researcher and the additional coder, working towards a final consensus as to what the appropriate code should be.  3.5. RESULTS 3.5.1. Determining the Factor Structures of Online Aggression and Reactive/Proactive Aggression. Online Aggression. Prior to creating the online aggression composite variables, an unweighted least squares exploratory factor analysis was run in order to determine the latent structure of the items. Seeing as this measure was a newly developed for the current study, it is imperative that the underlying relationship among items be explored before composite variables are created (Tabachnick & Fidell, 1996). The unweighted least squares extraction method was chosen because it makes no assumption of multivariate normality. Upon inspection of the correlation matrix (See Table 3.1), each item significantly correlated with each other (p < .001). In order to allow the underlying factors to be correlated, the most commonly used oblique rotation method was chosen; namely, direct oblimin (Tabachnick & Fidell).  88  A preliminary factor analysis revealed that two of the items that did not fit well due to cross loadings and/or low factor loadings, should be removed7. These items were: “How often have you had experience with seeing and hearing about mean things…said through the Internet…” and “How often have you had experience with mean things…said or done to you through the Internet…”. After removing the items the analysis was re-run. The Kaiser-Meyer-Oklin value was .9, exceeding the recommended minimum value of .6 (Kaiser, 1970, 1974), and the Bartlett’s Test of Sphericity (Bartlett, 1954) reached statistical significance (p < .001). These results support the factorability of the items. Three factors with eigenvalues greater than 1.0 emerged from the data. Eigenvalues for each of the factors were 5.3, 1.6, and 1.1 respectively. Upon inspection of the screeplot, a clear break after the third factor was revealed, which indicated that three online aggression factors emerged: (1) Aggressive Messaging, (2) Developing Hostile Websites, and (3) Posting/Commenting about Embarrassing Photos or Videos (See Table 3.2). As the strengths of the correlations were found to vary (see Table 3.1)8, Varimax rotation (which ensures that the underlying constructs are not allowed to correlate; Tabachnick & Fidell) was employed as a precautionary measure (although it should be noted that the outcome was the same as that of the oblimin rotation). The three factor model that emerged from the factor analysis was deemed a good fit to the data.  7  These items were removed in accordance to literature which stipulates that items should be thoughtfully removed if they do not contribute to the model or diminish reliability (Pett, Lackey, & Sullivan, 2003). 8 Relationship strength was derived according to Cohen’s (1988) guidelines which indicate that r = +/- .10 to .29 is small, r = +/- .30 to .49 is medium, and r = +/- .50 to 1.0 is large.  89  Table 3.1. Correlation matrix for online aggression items. EP to me Saw EP of others Sent AM with friends Receiving AM with friends Receiving AM with alone Replying AM with friends Replying AM with alone Create HW with friends Create HW alone  Saw EP of others  Sent AM with friends  Receiving AM with friends  Receiving AM with alone  Replying AM with alone  Create HW with friends  .50*** .30***  .31***  .25***  .36***  .47***  .22***  .27***  .37***  .58***  .26***  .35***  .55***  .62***  .48***  .23***  .26***  .37***  .45***  .56***  .62***  .31***  .26***  .31***  .25***  .21***  .31***  .31***  .19***  .20***  .25***  .19***  .25***  .22***  .30***  Note: EP = Embarrassing pictures/videos; AM = Aggressive Messaging; HW = Hostile Websites *p < .05. **p < .01. ***p < .001  90  Replying AM with friends  .60***  Table 3.2: Unweighted least squares factor analysis pattern matrix for all online aggression items. Item Aggressive Messaging Replying to mean messages about others online with friends Receiving mean messages online alone Replying to mean messages about others online alone Receiving mean messages online with friends Sending mean messages online with friends Took part in mean things to others online Replying to mean messages about you online with friends Replying to mean messages about you online alone Sending mean messages online alone Hostile Website Development Creating websites to embarrass or make fun of people online Creating websites to embarrass or make fun of people with friends Posting/Commenting on Embarrassing Photos or Videos Had experiences with embarrassing pictures or clips online about me Saw embarrassing pictures or clips to others online Took part in embarrassing pictures or clips to others online  1  Factor 2  3  .774 .771 .748 .729 .571 .497 .480 .459 .450 .901 .585 .734 .630 .628  *Note: Only factor loadings of .32 or greater are reported here as is consistent with reporting procedures for Factor Analyses (Tabachnich & Fidell, 2001).  It is interesting to note that the factor structure did not differentiate between bully, victim, and witness, as is typical of traditional forms of bullying (Craig & Peplar, 1995; 2000). In contrast, the factor analyses clearly revealed that for online forms of bullying, adolescents differentiate their experiences according to the method used for the aggressive act, rather than whether one is committing, witnessing, or being a victim of the aggression. Based on these analyses, three composite variables were created by taking the average of the items for each factor: (1) Aggressive Messaging (α = .87), (2) Hostile Websites (α = .74), and (3) Posting/Commenting on Embarrassing Photos (α = .70). These composites were used in all subsequent analyses. 91  Reactive vs. Proactive Aggression. In an attempt to create proactive and reactive composite variables (See Appendix C for proactive items 33 b, d, f, i, k, m and reactive items 33 a, c, e, g, h, j, l), a second exploratory factor analysis using unweighted least squares and oblimin rotation was employed. The correlation matrix revealed significant correlations among all items but one (“Threatened or posted things so others would do things for you”; See Table 3.3). According to the recommendations of Pett and colleagues (2003), the one uncorrelated item was removed and the factor analysis was run. The Kaiser-Meyer-Oklin value was .9, exceeding the recommended minimum value of .6 (Kaiser, 1970, 1974), and the Bartlett’s Test of Sphericity (Bartlett, 1954) reached statistical significance (p < .001), thus supporting the factorability of the items. Two factors with eigenvalues greater than 1.0 emerged from the data (eigenvalues = 4.5 and 1.2, respectively). The screeplot revealed a clear break after the first factor, which seemed to indicate that all items load on one factor (See Figure 3.1). For a two-factor solution (See Table 3.4) the last Proactive items either crossloaded (Said mean things just for fun) or were the only items to load more strongly on factor 2 (Said mean things to be cool). For a one-factor solution, all the items loaded strongly on one factor (See Table 3.5). Based on this outcome, a one-factor model was deemed to be the best fit for the data. Therefore, it was concluded that Proactive and Reactive aggression appear to be the same construct online. In order to tease this out more thoroughly, a hierarchical regression analysis was performed, and the Relative Pratt Index (RPI; Thomas, Hughes & Zumbo, 1998) was calculated for each of the online aggression constructs: (a) Aggressive Messaging, (b) Hostile Website Development, and (c) Commenting/Posting Embarrassing  92  Photos/Videos. This procedure was performed to assess the relationships that the Proactive and Reactive items have with each of the dependent variables, and to determine which items make important contributions to predicting the three outcomes. Specifically, for each of the three regression models, sex and age were entered as covariates, and then all 13 Proactive/Reactive items were entered into the regression in a separate block. After running the analyses, all Proactive/Reactive items that significantly predicted the outcome variable were assessed for their relative importance using the RPI9. The items that were found to be important were then used for further analyses (See Table 3.6). It is important to note that for Hostile Website Development only the Proactive items emerged as important, while for Commenting/Posting Embarrassing Photos/Videos only the Reactive items emerged as important. For Aggressive Messaging it appears that both the Reactive and Proactive items were important, however, upon closer examination it seems apparent that the two Proactive items, Threatened/bullied someone online and Gotten others to defend you or someone else, could also be interpreted by the participants as being Reactive rather than Proactive. For example, if a participant is asked to what extent they have threatened/bullied someone online they may respond Sometimes, however the motivation for this could have been reactive in nature. Given this, these two proactive items were re-labeled as Reactive.  9  Please refer to Study I for a detailed description of the RPI procedures  93  Table 3.3. Correlation matrix for Proactive and Reactive items.  P:V32b R:V32c P:V32d R:V32e P:V32f R:V32g R:V32h P:V32i R:V32j P:V32k R:V32l P:V32m  R: V32a .43*** .46*** .33*** .42*** .18*** .25*** .33*** .32*** .44*** .33*** .37*** -.02  P: V32b  R: V32c  P: V32d  R: V32e  P: V32f  R: V32g  R: V32h  P: V32i  R: V32j  P: V32k  R: V32l  .40*** .37*** .31*** .35*** .32*** .29*** .34*** .32*** .29*** .31*** .01  .23*** .43*** .15*** .26*** .31*** .26*** .47*** .29*** .36*** -.02  .28*** .39*** .25*** .21*** .35*** .27*** .29*** .29*** .00  .26*** .33*** .31*** .24*** .38*** .33*** .40*** .00  .35*** .24*** .33*** .21*** .26*** .29*** .02  .28*** .26*** .26*** .28*** .28*** .02  .28*** .48*** .28*** .43*** .01  .35*** .25*** .29*** .02  .30*** .48*** -.01  .34*** .02  .01  Note:R = Reactive Item P = Proactive Item *p < .05. **p < .01. ***p < .001  94  Figure 3.1. Screeplot of factor loadings for Proactive and Reactive Aggression.  5  Eigenvalue  4  3  2  1  0 1  2  3  4  5  6  7  8  9  10  11  12  Factor Number  Table 3.4: Factor loadings from unweighted least squares factor analysis of Proactive and Reactive items forced to two factors. Factor Item Posted mean things to defend yourself or someone else (R) Reacted angrily when provoked by others (R) Posted mean things when they annoyed you (R) Said or posted mean things when you are teased (R) Said mean things because felt mad (R) Posted mean things because have been threatened (R) Posted mean things to show who was on top (P) Gotten others to say or post mean things about someone else (P) Threatened/bullied someone online (P) Became angry when didn’t get own way so take it out online (R) Said mean things just for fun (P) Posted mean things to be cool (P)  1 .74 .69 .68 .63 .61 .58 .51 .47 .42 .40 .38 .25  2  .36 .61  *Note: Only factor loadings of .32 or greater are reported here as is consistent with reporting procedures for Factor Analyses (Tabachnich & Fidell).  95  Table 3.5: Factor loadings from unweighted least squares factor analysis of Proactive and Reactive items forced to one factor. Item Posted mean things to defend yourself or someone else (R) Posted mean things when they annoyed you (R) Said or posted mean things when you were teased (R) Said mean thing because felt mad (R) Posted mean thing to show who was on top (P) Reacted angrily when provoked by others (R) Post mean things because you have been threatened (R) Threatened/bullied someone online (P) Gotten others to say or post mean things about someone else (P) Said mean things just for fun (P) Become angry when don't get your way so take it out online (R) Posted mean things to be cool (P)  Factor 1 .66 .63 .63 .60 .60 .59 .56 .52 .52 .51 .49 .46  Table 3.6: Summary of important proactive and reactive items according to the Pratt index. Item  Dj  DV = Aggressive Messaging Reactive: Posted mean things when they annoyed you Reactive: Posted mean things to defend yourself or someone else Reactive: Said or posted mean things when you were teased Proactive: Threatened/bullied someone online Proactive: Gotten others to say or post mean things about someone else Reactive: Become angry when don’t get your way to take it out online  .262* .173* .089* .081* .080* .062*  DV = Hostile Websites Proactive: Posted mean things to show who was on top Proactive: Said mean things just for fun Proactive: gotten others to say or post mean things about someone else  .195* .183* .157*  DV = Embarrassing Photos Reactive: Posted mean things to defend yourself or someone else Reactive: Posted mean things when they annoyed you  .133* .120*  Note. Dj < ½(p) = .03 * Important Variable.  96  3.5.2. Research Question 1: What are the Sex and Grade Differences Associated with Online Aggression? A preliminary examination of the data revealed that 27% of adolescents reported being aggressive via Email or Instant Messaging, 12% of adolescents reported being involved in creating Hostile Websites, and 34% of adolescents reported being involved in posting and making inappropriate comments about others through pictures and videos. To examine the gender and grade differences for each of the different forms of online aggression that emerged from the factor analyses, three separate univariate analyses of variance (ANOVA) were run for Aggressive Messaging, Hostile Websites and Embarrassing Photos. For Aggressive Messaging, there were significant main effects for grade (F(7, 715) = 5.8, p = < .001) and sex (F(1, 715) = 4.3, p < .05) but no significant interaction effects, as can be seen in Figure 3.1. The effect sizes were found to be small to medium10 for grade (eta square = .05) and small for gender (eta square = .006). Tukey HSD post hoc analyses revealed no significant differences in Aggressive Messaging among high school students (grades 7 to 12); however, grade 5 students were significantly less likely to send Aggressive Messages than students in grades 10 to 12. Similarly, grade 6 students were significantly less likely to be aggressive in this way than participants in grades 9 to 12. With the exception of grades 9, 11 and 12, girls tend to engage in Aggressive Messaging more than boys.  10  These effect sizes are interpreted according to Cohen’s (1988) commonly used guidelines which stipulate that .01=small, .06=moderate, and .14=large effect.  97  Figure 3.2. Grade and sex differences for Aggressive Messaging.  1.80  gender boy  Estimated Marginal Means  girl  1.60  1.40  1.20  1.00  grade 5 grade 6 grade 7 grade 8 grade 9  grade 10  grade 11  grade 12  For Hostile Websites, significant main effects were found for grade (F(7,713) = 4.1, p < .001 eta sqare = .04, with older adolescents reporting higher incidences of developing Hostile Websites. There were also a significant grade-by-sex interaction (F(7,713) = 3.1, p < .01, eta square = .03). As can be seen from the estimated means presented in Figure 3.3, grade 12 boys appeared to be considerably different from students at all grades; as such, the interaction that was found could be misleading due to this single discrepancy. To test for this, an ANOVA was conducted with the grade 12s removed from the analysis. This analysis demonstrated no significant main or interaction effects across grade or sex for students in grade 5 through grade 11. Thus, the interaction effect that was shown in the original ANOVA was entirely due to the differences observed at grade 12. Tukey HSD post hoc analysis for the gender effect for grade 12s 98  was statistically significant (p < .001). In other words, grade 12 boys created significantly more hostile websites than any of the other students. Figure 3.3. Grade and sex differences for Creating Hostile Websites.  1.80  gender  Estimated Marginal Means  boy 1.60  girl  1.40  1.20  1.00  0.80 grade 5 grade 6 grade 7 grade 8 grade 9  grade 10  grade 11  grade 12  For Embarrassing Photos, there were significant main effects for grade, with older students appearing to engage in this kind of activity; F(7,702) = 13.2; p < .001, eta-sq .03. There was also a significant interaction effect, (F(7,702) = 2.6, p = .01, eta-sq .03). Tukey HSD post hoc analyses indicated no differences in commenting/posting embarrassing photos/videos between boys and girls at grade 5 through 8. However, one significant difference was found between girls and boys at grade 10 (p = .001). In general, both boys and girls showed no differences in commenting/posting embarrassing photos/videos in grades 9, 11, and 12, although these students engaged in this behaviour more frequently than students in grade 5 though 8. 99  Figure 3.4. Grade and sex differences for Commenting/Posting Embarrassing Photos/Videos  2.00  gender  Estimated Marginal Means  boy 1.80  girl  1.60  1.40  1.20  1.00 grade 5 grade 6 grade 7 grade 8 grade 9  grade 10  grade 11  grade 12  3.5.3. Research Questions 2 and 3: How Does Self-Esteem Predict Online Aggression? Is Online Aggression Primarily Proactive or Reactive? In order to answer research questions 2 and 3, three hierarchical multiple regressions were performed separately for each type of online aggression. Having a computer in their bedroom, coming from a Lone Parent family, sex, and grade were entered into Block 1 as covariates. Self-Esteem was entered into Block 2 and the Proactive/Reactive items, as determined via the Pratt Index (see the first section of the Results), were entered in Block 3. The assumptions for each of the three models were largely met. However, there were some minor violations. Most of the independent variables were significantly 100  correlated with all three dependent variables (p < .001), with the exception of Lone Parenting and Self-Esteem (See Tables 3.7, 3.8, 3.9). Also, Hostile Websites and Embarrassing Photos and/or Videos were not significantly correlated with Sex (See Table 3.7, 3.8). Tolerance levels for all three models were between .69 and .99, thus ensuring that the assumption for multicollinearity was met. The assumption for linearity, normality, and outliers were assessed by examining the Normal Probability Plot and the Residuals Scatterplot. For Aggressive Messaging and Commenting/Posting Embarrassing Photos/Videos the points lay roughly in a straight diagonal line and the residuals were rectangular and centering along 0 thus, meeting these assumptions. For the Creating Hostile Websites variable, however, the points on the plots failed to lie within the acceptable range, thus violating the assumption of normality. A square root transformation was applied to the data to account for this deviation (Tabachnick & Fidell, 2001) but no change was observed; as such, the data were analyzed using the original distribution. Despite these minor violations in correlations and normality, it was considered safe to proceed with the analyses since regression analyses are, for the most part, robust to these assumption violations (Howell, 1997) and since the sample size was large. Aggressive Messaging. Upon examination of the covariates, having a computer in the bedroom, grade, and sex significantly predicted whether an individual sent aggressive messages online. Lone Parenting did not predict this behaviour (See Table 3.7). It should be noted that all of these variables, except for computer in the bedroom, remained significant when the rest of the variables were entered in Block 2 and 3. Self-esteem was not initially significant when it was entered into the model, but in the presence of the  101  Proactive/Reactive variables (in Block 3), it emerged as significant. In the final main effects model (Block 3), grade, sex, self-esteem, and all the Proactive/Reactive items were found to make significant contributions to the model (R2 =.482, ∆R2 = 41, p < .001). Upon closer examination of the coefficients, sex, grade, and self-esteem had a positive association with sending aggressive messages online. Specifically, girls were more likely to engage in this activity than boys (β=.12, p < .001), older students were more likely to send aggressive messages online than younger students (β=.10, p = .001), and individuals with higher self-esteem were more likely to send hurtful and aggressive messages than those with lower self-esteem (β=.07, p < .05). Interactions of sex-bygrade, sex-by-self-esteem, and grade-by-self-esteem were entered into a fourth block in the regression, but none of them were found to be significant, so they were not included in the final model.  102  Table 3.7: Summary of hierarchical multiple regression model for Aggressive Messaging. Aggressive Messaging Block 1 Computer in bedroom Grade Sex Lone Parent Block 2 Computer in bedroom Grade Sex Lone Parent Self-Esteem Block 3 Computer in bedroom Grade Sex Lone Parent Self-Esteem Reactive: Posted mean things when they have annoyed you Reactive: Become angry when don’t get your way, so take it out online Reactive: Threatened/bullied someone online Reactive: Posted mean things to defend yourself or someone else Reactive: Gotten others to defend you or someone else Reactive: Said or posted mean things when you were teased  B .09 .04 .08 .03 .09 .04 .08 .03 -.01  SE .03 .01 .03 .03 .03 .01 .03 .03 .02  β  R2  r  .11** .20*** .12** .04 **  .11 .20*** .11** .04 -.03  .13*** .20*** .10*** .05 ***  UR2  .069  .069***  .069  .001  .482  .413***  .13 .20*** .10*** .05 -.03  .02 .02 .09 .02 .03  .02 .01 .02 .02 .01  .03 .10** .12*** .03 .06*  .13*** .20*** .10*** .05 -.03  .17  .02  .27***  .54***  .14  .03  .13***  .35***  .15  .04  .13***  .39***  .13  .02  .21***  .51***  .13  .03  .12***  .39***  .11  .03  .13***  .44***  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  103  Hostile Websites. As with Aggressive Messaging, the overall model was found to be significant. For the first block, having a computer in the bedroom significantly increased the likelihood that an individual would engage in creating Hostile Websites (β = .10, p < .01). However, this did not hold when the Proactive/Reactive items were added in Block 3. Significant grade differences were found in the final model, with older adolescents engaging in this type of aggressive behaviour more than younger adolescents (β = .11, p < .001). All of the Proactive/Reactive items were significant and their addition to the model explained an additional 13% of the variance. The interaction terms, sex by grade, sex by self-esteem, and grade by self-esteem were entered into a fourth block, but again were not found to be significant, so were removed from the model (See Table 3.8). Embarrassing Photos. For this outcome variable, grade and sex were significant predictors in the model, with students in higher grades (β = .09, p < .001) and girls (β = .09, p < .05) participating in this form of aggression more than boys and students in younger grades, respectively. The two reactive items (Reactive: “Posted mean things when they annoyed you” and Reactive: “Posted mean things to defend yourself or someone else”) also significantly predicted whether an individual would engage in posting and/or making comments about pictures posted online (β = .19, p < .001 and β = .19, p < .001, respectively). In the fourth block, the interactions of sex by grade, sex by self-esteem, and grade by self-esteem were entered. As with the previous models, these were not found to be significant and were removed from the model. Please see Table 3.9 for the final regression model.  104  Table 3.8: Summary of hierarchical multiple regression model for Hostile Websites. Hostile Websites Development Block 1 Computer in bedroom Grade Sex Lone Parent Block 2 Computer in bedroom Grade Sex Lone Parent Self-Esteem Block 3 Computer in bedroom Grade Sex Lone Parent Self-Esteem Proactive: Posted mean things to show who was on top Proactive: Said mean things just for fun Proactive: Had others say or post mean things about someone else  B .06 .02 -.03 .02 .06 .02 -.03 .02 -.01  SE .02 .01 .02 .02 .02 .01 .02 .02 .01  β .10** .15*** -.05 .04 **  .10 .15*** -.05 .04 -.02  .04  .04***  .04  .00  .17  .13***  .12 .15 -.05 .05 -.01  .02 .01 .02 .02 .01  .05 .11*** -.03 .02 .01  .12 .15 -.05 .05 -.01  .13  .03  .18***  .04  .12  .03  .16***  .29  .16***  .26  .03  UR2  .12 .15 -.05 .05  .03 .02 -.02 .01 .00  .13  R2  r  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  105  Table 3.9. Summary of results from the hierarchical multiple regression model for Embarrassing Photos and/or Videos. Embarrassing Photos Block 1 Computer in bedroom Grade Sex Lone Parent Block 2 Computer in bedroom Grade Sex Lone Parent Self-Esteem Block 3 Computer in bedroom Grade Sex Lone Parent Self-Esteem Reactive: Posted mean things when they annoyed you Reactive: Posted mean things to defend yourself or someone else  B .04 .09 .09 .05 .05 .09 .08 .05 .01  SE .04 .01 .04 .05 .04 .01 .04 .05 .03  β .04 .31*** .07 .04 .04 .31*** .07 .04 .01  R2  r  UR2  .11  .11***  .11  .00  .20  .10***  .06 .30 .08 .05 .06 .30 .08 .05 .03  .01 .07 .09 .06 .04  .04 .01 .04 .04 .02  .01 .26*** .08* .05 .05  .06 .30 .08 .05 .03  .19  .04  .18***  .31  .19  .04  .19***  .30  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  Interview Findings In addition to these analyses, interviews were conducted with a subset of 15 adolescents (10 females and 5 males) to shed further light on the issue of Proactive versus Reactive aggression. During these interviews 21 separate incidences of online aggression were discussed. As can be seen from Table 3.10, six of the online aggression incidences were solely Proactive in nature, one was solely Reactive, while the majority (14) included 106  a combination of proactive and reactive aggression. It would appear, therefore, that aggressive interactions in an online setting are primarily proactive and reactive at the same time, rather than being one or the other.  107  Table 3.9: Summary of interview themes Participant 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15  108  Aggressive Incident  Proactive Aggression  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Total  √ √ √ √ √ √ √ √ √ √ √  Reactive Aggression √ √ √  √ √ √ √ √ √ √  √ √ √ √ √ √ √ √ √ 20  NonConfrontational Online √  √ √ 14  √ √  √ √ √ √ √ √  √ √ √ √ √  √  √  √  √  √ √ √  NonConfrontational Offline √  Confrontational Online  Confrontational Offline  √ √ √ √ √ √ √ √ √ √  √  √ √ √ √  √ √ √ √  √ √  √ 10  10  √ √ 16  8  In most of the cases where participants described online aggression experiences that were both Proactive and Reactive in nature, targets of the aggression perceived the perpetrator to be mean to them for proactive reasons, whereas their own behaviour was retaliatory and, therefore, justified. For example, a grade 6 girl explained how she was a target of online aggression when she had to suddenly log off MSN because her parents were yelling at her. The friend she was chatting with took this as a slight and started reacting aggressively in the days to follow: Participant: I had some threats, like, 'I'm gonna do something to you today' or anything, and nothing happened, basically. She basically said, 'Oh, something's gonna- bad's gonna happen to you at recess or lunch'. And...I basically stayed with my friends and nothing happened. And one time she's like, 'I'm gonna bring a...lot of people, a lot of my friends, and we're gonna yell- talk to you'. And I'm like, 'Go ahead, I'm not scared'. (laughs). And um she gave me people's names and then I'm like, it's OK. 'Cause they're all my- they were my friends, too. And I sit next to one of them. I'm like, 'Did- did you agree to come and talk to me?' and she's like, 'No, [name] just made that up'. So, kind of got that info from my other friends that were also her friends, so. Yeah, that's-…. after that incident, she's like - ah, she e-mailed me and she's like, 'Something bad's gonna happen to you tomorrow'. Nothing happened… It was kind of awkward, because, like, a lot of people knew that we weren't talking. And it was awkward to, like, be next to each other at times. And we would be on the same team and we'd talk when we were fighting. Like, if we had to, for class projects or anything… Interviewer: So … when all of this is happening… it seemed like, on MSN she was saying those mean things.  109  Participant: Yeah, MSN. That's it. Like, MSN and e-mail. Nothing in person. And she's got- she got her friends to say stuff to me, too, and stuff like that, but not in person. All of them were too scared to say anything in person. In this example, the aggressive acts started in reaction to a misinterpretation. The person who started making threats did so because she was hurt or insulted that her friend stopped talking to her on MSN and because she did not know the reason behind this abandonment. In reaction to this, she started proactively exerting her power over the “target” by making threats, spreading rumours, and attempting to convince others to “side” with her. In turn, the “target” retaliated by exerting her power, as exemplified when she says “go ahead, I’m not scared…it’s OK, 'Cause they're all my- they were my friends, too”. Interestingly, all the aggressive behaviours occurred confrontationally online in that the participants would make threats and speak with hostility openly and directly with each other, but would remain civil in person. Another example of individuals engaging in both proactive and reactive aggression involves grade 6 students: Participant: Like, um, I know my friend was going out with this guy. And she kinda has glasses and she has braces. And this guy's really, really popular. …. a lot of people were, like, talking behind their back and spreading rumours about them, because she kinda, they called her nerdy and all that 'cause she had braces and glasses. And...so she was kind of....sad about it, upset, 'cause he was more popular than- they were, like, saying that he's gonna ruin his reputation because of her and stuff like that. Interviewer: Oh, wow. So was this, all of this talk, was that happening online or offline or both? Participant: Both. … Like, people had their names and personal messages as '[friend]'s a nerd', or ah '[friend and boy] don't go together', stuff like that…Awkward and since she's  110  my closest- one of my closest friends, it was kind of sad, too. And this guy, I'm- I was surprised. But because of his so-called reputation, he dumped her on Facebook. And a fifth-grader's, like, what kind of person would dump someone on the internet, where, like, over five hundred people can see. And...the next day after he dumped her, most girls here, um, everybody knew, right. And my friend was, like, all crying about it still. And everybody was asking her, some people were making fun of her, and some people were even mad at the guy. So yeah, it was kind of awkward, 'cause we were near her. And then people would be staring at us and stuff, and be like, 'Why are you hanging out with six-graders, 'cause you're'- she's seventh-grade. And she, she comes over to my house on Fridays and Thursdays. …. I told [boy] that it wasn't nice of him to go dump her where everyone, like, people that don't even know you, sometimes, that [unclear] can even see. And most of the school knows now. And you kind of ruined her life sort of, because people are making fun of her, people are talking to her and saying rude things to her. Some people are soothing her, 'cause we're her friends, but some people are really rude about it. And his friends are now disliking him, too, now. So he's, like...really hated at this point. So, yeah. And the way he said it was kind of rude; the way he put his words...was really rude. Interviewer: Why do you think he did it that way? Participant: He says that his reason for breaking up with her was um because he never liked her from the beginning. And we asked him why he even started to date her - she just asked him, because - this girl I know, she was like, 'This person's gonna ask you out'. And then, she went and asked him instead, and he said yes. And then kinda said- he said that he was just pretending this whole time…Like, she- sometimes she starts  111  crying. And now he's, like, even - she's like, she wants to just be friends - and now he won't even let her say hi to him. He's like, 'Go tell her not to say hi to me anymore. Don't- tell her not to touch me. Tell her not to even talk to me'. And it's just...weird. So yeah… If no one had started the rumour of her being...a nerd, as they say, then she w- this wouldn't have happened, I don't think. At the beginning of this experience, it seems that the more popular boy had no qualms for dating the less popular girl. However, as his peers proactively informed him online and in person that his reputation would be compromised, he reacted aggressively by publicly breaking up with her on Facebook. His own aggression became proactive when he attempted to regain his status on the social totem pole by deliberately avoiding and ignoring her. Moreover, the menacing behaviour of his peers as they continued to gossip and spread rumours about the girl contributed to this Proactive aggression. It must be recognized that, although online aggression usually involves both proactive and reactive elements, this is not always the case. For example, the following description from a Grade 8 female who was dealing with school rumours about her engaging in self-harm (cutting) behaviour and being “emo” demonstrates that, sometimes, online aggression is only proactive. In this case, the participant perceived herself to be only the recipient, not a contributor, to online aggression: Participant: It was mostly MSN.. but people posted some things on Facebook, too. … one kid would just say 'hello emo kid' to me all the time. …. Participant: Those kids were always calling me emo, and I was telling them to stop… so we got in some arguments.  112  Interviewer :...what did these arguments look like? Participant: Them: Hello, emo kid! Why are you so sad? Me: I'm not emo, and I don't appreciate you saying that to me. Do you know how that makes me feel? Them: Why don't you go cut yourself, then? In the one reported incident involving only Reactive Aggression, a grade 10 female explained how she posted aggressive comments and notes about a teacher she did not like at school, explaining that she thought he was racist and unfair. She described how her post online was more of a general post for the purpose of venting her frustrations, and not one to genuinely inflict harm on her teacher. Thus, although she was the one engaging in the online aggression, she perceived what she had done so as a response to the teacher’s provocation of treating her unfairly. Overall, the findings from the questionnaire and interview data were consistent with each other. Specifically, both the survey and the interview data revealed that adolescents perceived their own behaviour as Reactive. However, the interview data supplemented these findings by demonstrating that adolescents tend to see other people’s aggression as Proactive. 3.5.4. Research Question 4: Are Adolescents Using Confrontational or NonConfrontational Means for Engaging in Online Aggression? Research question 4 was also addressed using data from the 15 interviews. As can be seen from Table 3.10, most of the incidents reported in this study involved Confrontational Aggression Online, usually in combination with at least one other category of behaviour (Nonconfrontational Aggression Online, Confrontational Aggression Offline, and/or Nonconfrontational Aggression Offline). Furthermore, all but one of these experiences of Confrontational Aggression Online involved perpetrators saying aggressive things directly to the 113  target over the Internet. The following example, from a girl in Grade 10, presents an interesting illustration of how verbal Confrontational Aggression in an online setting can be generated following physical aggression (Confrontational Aggression Offline) at school: Interviewer: have you ever said or posted anything mean about other people? Participant: just once, after i had a fight with a girl Interviewer: so you had the fight with the girl offline so then kind of "retaliated" online? Participant: yea, she was sending people some anti-me propaganda so i sent her a public message, and that was the only time Interviewer:: ohh...was it through msn or facebook or nex..email? Participant: nexopia. Interviewer::oh ok...and was she someone you knew offline too? Participant: yes. i see her around a lot Interviewer:: oh ok...if you think back to this time, do you think your actions were justified? Participant: i think it was more defence on my part, she was telling poeple i started the fight. she knew i was a vegetarian and took some pork and rubbed it in my face so i punched her. so i think i was fine in calling her for it. i dont even remember what i said… Interviewer: so it started offline...and then she started posting "anti-you" things on nexopia...and then got in your face with the pork and so you punched her? Participant: yea, the pork and punch was before we were online Interviewer: so your retaliation was offline and this whole thing was a combination of offline and online? Participant: yea, offline first, then online  114  Only four of the incidents described by participants did not include a Confrontational Aggression Online component. For example, a grade 8 girl described the following situation in her MSN interview, illustrating not only both online and offline Non-confrontational aggression, but also how easy it is to spread information widely over the internet, in an anonymous way: Participant: well it was in gr 7 and i liked this kid and someone wrote that i liked that kid on msn … everyone found out about it and it was pretty hard in class to concentrate Interviewer: do you know the person who said, on MSN, that you liked that kid? Participant: no it was passed around through people Interviewer: so you don't know who started it? Participant: nope Interviewer: do you know approx how many people got the message? Participant: umm about 90 people Interviewer: what did you do when you found out that everyone knew? Participant: i didnt talk to anyone for a while until it stopped Interviewer: how long did it take before it stopped? Participant: umm about two months or so Interviewer: so...for two months...what was everyone doing? They just kept talking about it on MSN? or were they doing other things too? Participant: they would laugh quietly behind my back and also whisper Interviewer: ohhh...so this whispering was happening at school Participant: yaa Interviewer: but it started on MSN? Participant: yupp  115  Another grade 8 female, who experienced name calling and rumour spreading over Facebook and MSN, explains how “some people use the internet rather than face to face, because they feel more powerful and they can say meaner things.” A common thread among these four cases is that they involved both Non-Confrontational Aggression online and offline. Therefore it appears that some individuals are vigilant in keeping their aggressive acts discrete and that they use both online communication and traditional means of communication to gain power through social manipulation rather than through overt aggression that requires confronting the target personally. Four incidents involved only a single form of aggression, and all of these incidences were Confrontational Online. Interestingly, all of these incidences were described by those who had either witnessed the aggressive incidences online but had not directly participated in them, or who were part of a gaming venue or forum where they did not know the other users in “real life”. Overall, it is clear that adolescents are using both confrontational and non-confrontational aggression online. Moreover there are instances when one aggressive experience can involve both confrontational and non-confrontational aggression simultaneously. It is apparent that ICT’s can be used to facilitate these forms of aggression among adolescents.  3.6. DISCUSSION Adolescents who choose to engage in some form of online aggression primarily take part in sending mean messages online (including gossiping and rumour spreading) and posting/making comments about pictures that are posted online (27% and 34%, respectively). A lower percentage of adolescents engaged in creating hostile websites (12%). Analyses of the questionnaire data revealed that individuals who participate in sending aggressive messages or posting/commenting on embarrassing photos believe they are doing so for reactive reasons, rather than proactive reasons. Moreover, analyses of the interview data confirmed that 116  adolescents tended to respond aggressively for retaliatory reasons. The interview data extrapolated upon this finding by demonstrating that adolescents often perceived other people’s posts as being proactive while perceiving their own actions as reactive. The idea that one’s own aggressions are reactive while others’ reactions are proactive was particularly evident in the interview data where participants framed themselves as the target of aggression, with seemingly no choice but to retaliate, while the perpetrator was viewed as aggressing proactively towards them. The following sequence from a Grade 10 male demonstrates the way that adolescents frame the situation, despite the fact that alternative interpretations are clearly possible: Participant: well... it was about these girls that my friend has been having trouble with and her nickname is “Cosplay". (inside joke) and they've been talking behind our backs and so we started to talk about how she dresses and ya… Interviewer: so just to get this straight, your friend has been having trouble with these girls who talk about you and your friend behind your backs...so then you started talking about how they dress? Participant: first things first... my friends and i are into ANIME... Interviewer: ok... Participant: one of the girls had hair that was perfect for Instant Cosplay [Anime character’s name] and so we call her Cosplay. So ... she started talking about my friend behind her back and my friend got really mad... so she confronted her and Cosplay answered sarcastically that she won’t say things behind her back anymore. The next day... some of our other friends noticed that she was talking about my friend and they told us.... and during that time we had gossip-like discussions about her Interviewer: and all of this gossiping and stuff was happening online or offline or both?  117  Participant: well both... more online though Interviewer: so would you say that you gossiping about her started because she was gossiping and saying stuff about your friend? Participant: yep.. pretty much Despite the participant’s framing of the situation, it could be that the participant and his friends initiated the hostile behaviour by referring to the “perpetrator” as an Anime character and calling her names. Annoyed, the “perpetrator” may have retaliated by spreading rumours about his friend. However, it is clear that the “target” and her friends viewed their mean actions as retaliatory, while remaining ambiguous about their own aggressive behaviour toward the “perpetrator”. The conclusion that adolescents view their own online aggression as reactive and others’ online aggression as proactive is consistent with research which indicates that individuals tend to aggress online because they want revenge (Hinduja & Patchin, 2006), and because they possess external loci of control. Specifically, research on traditional aggression and bullying has found a positive association between aggression and individuals’ beliefs about the causes of that aggression (Halloran, Doumas, John, & Margolin, 1999; Oesterman et al., 1999; Slee, 1993). For example, individuals who believe they are in control of the events in their lives and recognize the role they have played in a given situation are said to have an internal locus of control. These individuals have been found to react differently to a given circumstance when compared to individuals who have an external locus of control (i.e., believing that other factors besides themselves control events in their life). The idea that adolescents are sending aggressive messages and posting/commenting on pictures for reactive reasons supports locus of control theories in that they feel justified in retaliating. That is, rather than examining their own  118  behaviours and how their own behaviours may have contributed to the “perpetrator’s” aggressive acts, they focus only on the aggressive acts of the “perpetrator” and react aggressively in return. Moreover, the online environment provides more opportunities for retaliation. Individuals who might never retaliate offline due to small physical stature, shyness, or self-esteem, might feel quite comfortable doing so online. A significant main effect for both grade and sex was found for aggressive messaging. Specifically girls send more aggressive messages than boys do, and individuals in older grades engage in these behaviours more than those in younger grades. This result is not surprising as it supports the research that finds that older adolescents tend to use technology and computers for socializing more frequently than younger adolescents (MNet, 2005). Posting aggressive messages online was the only form of online aggression to exhibit a significant sex difference; that is, older girls were more likely to make belligerent comments online when compared to boys. These findings are compatible with research on the social nature of girlhood aggression. Work in this area has found that as girls mature, their social goal of maintaining popularity among their peers increases (Block, 1983; Crick & Grotpeter, 1995; Pellegrini, 2002). Moreover, relational aggression and the ability to manipulate one’s social environment, as exhibited by girls, requires enhanced cognitive abilities that come with age (Crick & Rose, 2001). Based on these findings from traditional literature on aggression, it makes sense that older adolescent girls might use the Internet for starting or continuing their relational aggression and social manipulations. For Aggressive Messages, a positive relationship was found between self-esteem and this form of aggression. That is, individuals with higher self-esteem were found to aggress more than those with lower self-esteem. These findings are consistent with the findings of traditional  119  offline aggression and self-esteem which show that individuals with higher self-esteem are more likely to be engaged in proactive and relational aggression (Mayeux & Cillessen, 2008; Orobio de Castro et al., 2007; Ybrandt, 2007). In addition, there is the potential that individuals’ perceptions of themselves change when it comes to online interactions, due to the anonymity afforded by the computer screen, as is indicated by this conversation with a Grade 10 male: Interviewer: I'm just wondering how you think this game playing has shaped your life? Like, do you feel like you can be more "yourself" online vs. offline? Do you feel more comfortable online than offline? Participant: again, this sort of ..varies =P when you're online it's usually alot easier to express yourself and show who you really are, simply because of the anonymity provided online..like if you wanted, you could be a pink-wearing cowboy of a mage, simply because you can..(not that I am =P) but offline you sort of have to fit into cliques, imo [In My Opinion]. Like, in real life, could you ever see anyone expressing themself like that? … Interviewer: so...you don't really feel like you have an online self vs. and offline self? Participant: it sort of depends on the way you mean this. online I'm pretty different.. when it calls for it i can be crude and vicious (if say, someone is annoying my friends etc or bringing the party down). Whereas offline i'd sort of let it slide and let somebody else do it- likely due to anonymity online. otherwise, I'm about the same. From this account, it is clear that online venues give individuals an opportunity to explore different aspects of themselves and that individuals behave differently and have different perceptions of themselves online than they would offline. This finding is consistent with other research in this area (Gross et al., 2002).  120  Having a computer in the bedroom significantly predicted the likelihood of adolescents sending aggressive messages and creating hostile websites. But this variable became nonsignificant when the Proactive/Reactive items were included in the model. This suggests that when there are motives for aggressing online, individuals are likely to engage in these forms of aggression despite the risk of being caught by an adult (e.g., the location of the computer becomes irrelevant). They may feel more justified in their actions because they are defending themselves or someone else online. They do not feel they are doing anything wrong and, consequently, feel comfortable posting these sorts of things online even when the computer is in a public area of the home. Overall, fewer adolescents engage in creating hostile websites for the purpose of hurting others (12%), and the adolescents who did so appeared to be a unique group. In particular, adolescents who engaged in this form of online aggression appeared to do so for Proactive reasons. This is in contrast to adolescents who engage in sending aggressive messages or posting/commenting on embarrassing photos, who do so for largely Reactive reasons. It is very important to explore this finding further as this particular group of individuals, namely grade 12 boys, appear to be intentional in their malicious online behaviour and are willing to invest much time and energy in creating a special website for hurting others. Furthermore, the aggressiveness associated with this kind of bullying is very close to that of traditional bullying, in that it is done with the intention to harm another individual, and the permanence of a website parallels the repeated aggressive behaviour generally associated with bullying behaviour. Not surprisingly, as with aggressive messages and posting/commenting on embarrassing photos, older students were more likely to create hostile websites than younger  121  students. This finding makes sense given that older students likely have more online access, as well as increased technological sophistication (Lei & Zhao, 2007). The fact that self-esteem, having a computer in the bedroom, and coming from a single parent family (all variables that one would hypothesize as being linked to irresponsible Internet use) did not emerge as significant factors with respect to creating Hostile Websites indicates the need for more research in this area. It is possible that self-esteem did not emerge as significant because it was examined too generally. Previous research on self-esteem, as it applies to aggression, has looked specifically at one’s perceptions of social competence and peer acceptance (Mayeux & Cillessen, 2008; Orobio de Castro et al., 2008). Alternatively, website creation is a sufficiently technologically sophisticated task that, unlike the use of IM or tools to post material to existing websites, the variance associated with skill at creating websites may be the primary characteristic that this sub-group of adolescents have in common. Future research should examine how this unique group of individuals perceives their friendships and social competence, and the nature of their defining characteristics. With respect to confrontational and non-confrontational aggression, the interview data revealed that most aggressive experiences involved more than one form of aggression; online and offline; confrontational and non-confrontational. From these findings it seemed that individuals were engaging in online aggression because it was easier to continue or start gossiping and spreading rumours online, and there was less likelihood of being caught by an adult. As such, adolescents were not afraid to post mean comments online with their name attached to it. This Grade 7 Female explains how: Participant: its really bad that ppl get all hardcore on the keyboard but in real life they don’t say anything  122  Interviewer: why do you think they get all hardcore on the keyboard but not in real life? Participant: well they think no 1 can hurt them and they can’t get into trouble Interviewer: do you think that's true? Participant: yea Interviewer: can you expand a bit on why you think that? Participant: well they think that its okay to be all mean online because they can get away with it. There does seem to be a distinction between bullying and aggression online. For the most part, the descriptions of online aggression provided by the sample suggest that adolescents seem to be engaging in reciprocal banter whereby each participant becomes both the target and the perpetrator. This finding is consistent with the factor structure that emerged from this study (i.e., when online, individuals are no longer categorized as perpetrators, targets, and witnesses), as well as results from Study I (where individuals are unclear as to whether they are Cyberbullies or Cybervictims). The questionnaire data indicated that most individuals’ aggressive acts were retaliatory, whereas the interview data revealed that individuals were both proactive and reactive in their aggressive acts. As mentioned, it is believed that adolescents are able to justify their mean behaviour and, for the most part, felt they were reacting to an incident, rather than because they just wanted power. If we re-examine the definition of bullying, it is characterized by aggressive behaviour that is done to another person with the intention to harm, it is repeated, and it is done in order to gain power (Olweus, 1993). A grade 10 student also described bullying as “1-sided or, you know, lots of people picking on 1 person.” In contrast, in the online aggression incidences described by participants in this study, actual one-sided bullying incidents occurred in only 5 of 21 incidences, while the remaining incidences (14) were vicious reciprocal arguments  123  that were exacerbated online. Thus, online aggression appeared to be distinct from traditional bullying in that it appeared to be more reciprocal, with the same individuals being victim and bully, than has been found in the traditional bullying literature. 3.6.1. Conclusions Overall, it appeared that online aggression differed according to the type of activity adolescents were using as the instrument of their aggression. When it came to sending aggressive messages or comment/posting embarrassing pictures, adolescents reported engaging in these forms of aggression for reactive reasons. However, it is important to note that although the questionnaire data suggested that most adolescents participated in reactive aggression, interview data also revealed that aggressive situations were a combination of proactive and reactive aggression; many individuals felt justified in their own aggressive acts as a form of retaliation, while interpreting the “perpetrator’s” aggression as proactive, with intention to harm. By contrast, when it came to creating Hostile Websites, questionnaire data showed that adolescents did so for solely proactive reasons, thus demonstrating that this is a unique group of individuals who are intentional in their online harm. Unfortunately, of the 21 incidences reported in the questionnaire data, none involved the creation of hostile websites; as such, a more in-depth understanding for this group of individuals could not be examined. With regard to online aggression, adolescents typically used a combination of nonconfrontational and confrontational online aggression. The primary reason for these decisions seemed to lie in the feelings of anonymity when aggressing online and also the feeling that they would not be caught and/or reprimanded for their mean behaviour. These findings are consistent with traditional bullying patterns, where, due to limited adult supervision, aggression and bullying are more likely to occur on the playground (Atlas & Peplar, 2001; Craig et al., 2000). 124  Other important findings from this work seem to lie in the area of self-esteem and whether or not the computer is in the adolescent’s bedroom. When the proactive/reactive items were removed from the regression model, self-esteem predicted whether individuals would engage in sending aggressive messages, while having a computer in the bedroom was related to both the sending of aggressive messages and developing hostile websites. Future work in this area should examine these inconsistencies more closely, perhaps by looking at the domain specific attributes of self-esteem. 3.6.2. Limitations This study had several limitations that need to be acknowledged. First, although the overall sample size was relatively large (N = 733), participants were not evenly distributed across grade, sex, and the form of online aggression that occurred. As a result, the power of the analytical procedures for detecting significant differences was probably weakened by the relatively few cases of some of the online venues (specifically creating Hostile Websites). Future research should examine differences across these online venues with more evenly distributed sample sizes. Second, a general self-esteem measure was used to assess whether this construct predicts online aggression. Future work should examine the predictive value of specific domains of selfperceptions, namely self-conceptions of popularity and social competence; two constructs which have been shown to be predictive for traditional aggression (Mayeux & Cellessen, 2007). Another limitation of the study is that we were unable to find interview participants who engaged in creating Hostile Websites; as such we were not able to tease out the motivations behind this form of online aggression. Future work should consider examining this form of aggression in its own right, as it does appear to be unique. Also, although 15 interviews are 125  within the range of acceptable sample sizes for qualitative research, the positivist content analytic strategy that was used would have benefited from a larger data set and a greater number of reported incidences. This may be obtained by either spending more time with the existing participants, or interviewing others. Part of the difficulty was that many adolescents who were willing to complete a survey were much more hesitant about being interviewed. Finally, the cross sectional nature of this work means we cannot be confident that our age effects represent developmental changes that we would expect to find. It would be beneficial to conduct this study longitudinally in order to determine within-person change. Despite these limitations, this work is among the first to examine the construct of online aggression and to tease out the constructs which predict it. In particular, this study is the first to distinguish differences among the methods adolescents choose to use when aggressing online, and whether they are doing so for proactive or reactive reasons. Furthermore, this is the first study to employ a mixed-method design for examining online aggression and adolescents’ motivations for engaging in it. This work is a first step at looking at the individual contributors to online aggression. Future work should examine the contextual factors that might influence mean behaviour online. In keeping with the socio-ecological approach, Study III focused on Internet aggression and its relationship to parenting behaviours and school belonging.  126  3.7. REFERENCES Atlas, R.S. & Pepler, D.J. (1998). Observations of bullying in the classroom. The Journal of Educational Research, 92, 86-99. Bartlett, M.S. (1954). A note on the multiplying factors for various chi square approximations. Journal of the Royal Statistical Society, 16 (Series B), 296-298. Berkowitz, L. (1993). Aggression: Its causes, consequences, and control. New York: Academic Press. Bjorkqvist, K., Lagerspetz, K., & Kaukiainen, A. (1992). Do girls manipulate and do boys fight? Developmental trends in regard to direct and indirect aggression. Aggressive Behavior, 18, 117-127. Block, J.H. (1983). Differential premises arising from differential socialization of the sexes: Some conjectures. Child Development, 54, 1335-1354. Boivin, M., Thomassin, L., & Alain, M. (1989). Peer rejection and self-perceptions among early elementary school children: Aggressive rejectees versus withdrawn rejectees. In B.H. Schneider, G. Attili, J. Nadel, & R.P. Weissburg (Eds.), Social competence in developmental perspective (pp. 392-393). Boston: Kluwer. Boneva, B.S., Quinn, A., Kraut, R.E., Kiesler, S., & Shklovski, I. (2006). Teenage Communication in the Instant Messaging Era. In R. Kraut, M. Brynin, & S. Kiesler (Eds.), Computers, Phones, and the Internet: Domesticating information technology (pp 201-218). New York, NY: Cambridge University Press.  127  Bosworth, K., Espelage, D.L., & Simon, T.R. (1999). Factors associated with bullying behavior in middle school students. Journal of Early Adolescence, 19, 341-362. Butterfield, L.D., Borgen, W.A., Amundson, N.E., & Maglio, A-S. T. (2005). Fifty years of the critical incident technique: 1954-2004 and beyond. Qualitative Research, 5, 475-497. Brown, R. (2000). Social identity theory: Past achievements, current problems and future challenges. European Journal of Social Psychology, 30, 745-778. Coie, J.D. & Dodge, K.A. (1998). Aggression and antisocial behavior. In N. Eisenberg (Ed.), Handbook of Child Psychology, Vol. 3 (pp.779-862). New York: Wiley. Conway, A.M. (2005). Girls, aggression, and emotion regulation. American Journal of Orthopsychiatry 75, 334 – 339. Craig, W.M., Pepler, D., & Atlas, R. (2000). Observations of bullying in the playground and in the classroom. School Psychology International, 21, 22-36. Creswell, J.W. (2009). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches (3rd ed.). Los Angeles, Sage Publications Ltd. Crick, N.R. (1995). Relational aggression: The role of intent attributions, feelings of distress, and provocation type. Development and Psychopathology, 7, 313-322. Crick, N.R., Bigbee, M.a., & Howes, C. (1996). Gender differences in children’s normative beliefs about aggression: How do I hurt thee? Let me count the ways. Child Development, 67, 1003-1014. Crick, N.R. & Dodge, K.A. (1996). Social information-processing mechanisms in reactive and proactive aggression. Child Development, 67, 993-1002. Crick, N.R., & Grotpeter, J.K., (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710 – 722.  128  Crick, N.R., & Rose, A.J. (2001). Toward a gender-balanced approach to the study of socialemotional development: A look at relational aggression. In P.H. Miller & E.K. Scholnick (Eds.), Toward a feminist developmental psychology (pp.153-168). New York: Routledge. Dodge, K.A., & Coie, J.D. (1987). Social information processing factors in reactive and proactive aggression in children’s peer groups. Journal of Personality and Social Psychology, 53, 1146-1158. Dodge, K.A., Coie, J.D., Pettit, G., & Price, J. (1990). Peer status and aggression in boys’ groups: Developmental and contextual analyses. Child Development, 61, 1289-1309. Ellison, N., Steinfield, C., & Lampe, C. (June, 2006). Spatially bounded online social networks and social capital: The role of Facebook. Paper presented at the Annual Conference of the International Communication Association, Dresden, Germany. Espelage, D.L. & Swearer, S.M. (Eds.). (2004). Bullying in American Schools: A SocioEcological Perspective on Prevention and Intervention. Mahwah, NJ: Lawrence Erlbaum Associates, Inc., Publishers. Estell, D.B., Cairns, R.B., Farmer, T.W., & Cairns, B.D. (2002). Aggression in inner-city early elementary classrooms: Individual and peer-group configurations. Merrill-Palmer Quarterly, 48, 52-76. Flanagin, A.J., (2005). IM Online: Instant Messaging use among college students. Communication Research Reports, 22, 175-187. Galen, B.R., & Underwood, M.K. (1997). A developmental investigation of social aggression among children. Developmental Psychology, 33, 589-600.  129  Gini, G., 2006. Bullying as a social process: The role of group membership in students’ perception of inter-group aggression at school. Journal of School Psychology, 44, 51-65. Greenfield, P.M., Gross, E.F., Subrahmanyam, K., Suzuki, L.K., & Tynes, B. (2006). Teens on the Internet: Interpersonal connection, identity, and information. In R. Kraut, M. Brynin, & S. Kiesler (Eds.), Computers, Phones, and the Internet: Domesticating information technology (pp 185-200). New York, NY: Cambridge University Press. Gross, E.F., Juvonen, J., & Gable, S.L. (2002). Internet use and well-being in adolescence. Journal of Social Issues 58, 75-90. Halloran, E.C., Doumas, D.M., John, R.S., & Margolin, G. (1999). The relationship between aggression in children and locus of control beliefs. Journal of Genetic Psychology, 160, 5-21. Harter, S. (1989). Causes, correlates, and the functional role of global self-worth: A life span perspective. In J. Kolligian and R. Stennberg (Eds.), Perceptions of competence and incompetence across the life span (pp. 67-97). New Haven, CT: Yale University Press. Harter, S. (1999). The construction of self: A developmental perspective. New York, NY: The Guilford Press. Hinduja, S., & Patchin, J.W. (2008). Cyberbullying.us. Retrieved November 2008, from http://www.cyberbullying.us/shareyourstory.php. Hoover, J.H., Oliver, R., & Hazler, R.J. (1992). Bullying: Perceptions of adolescent victims in the Midwestern U.S.A. School Psychology International, 13, 5-16. Howell, D.C. (1997). Statistical Methods for Psychology (4th ed.). Toronto: Wadsworth Publishing Company.  130  Hymel, S., Bowker, A., & Woody, E. (1993). Aggressive versus withdrawn unpopular children: Variations in peer and self-perceptions in multiple domains. Child Development, 64, 879-896. Kaiser, H. (1970). A second generation Little Jiffy. Psychometrika, 35, 401-415. Kaiser, H (1974). An index of factorial simplicity. Psychometrika, 39, 31-36. Kinard, E.M. (1978). Emotional development in physically abused children: A study of selfconcept and aggression. California: R & E Research Associates, Inc. Kraut, R., Brynin, M., & Kiesler, S. (2006). Computers, Phones, and the Internet: Domesticating information technology . New York, NY: Cambridge University Press. Lagerspetz, K.M.J. & Bjorkqvist, K. (1994). Indirect aggression in girls and boys. In L.R. Huesmann (Ed.), Aggressive behavior: Current perspectives (pp. 131-150). New York: Plenum. Lagerspetz, K.M.J., Bjorkqvist, K., & Peltonen, T. (1988). Is indirect aggression typical of females? Gender differences in aggressiveness in 11- to 12-year-old children. Aggressive Behaviour, 14, 403-414. Leary, M.R., Schreindorfer, L.S., & Haupt, A.L. (1995). The role of self-esteem in emotional and behavioural problems: Why is low self-esteem dysfunctional? Journal of Social and Clinical Psychology, 14, 297-314. Lei, J. & Zhao, Y. (2007). Technology uses and student achievement: A longitudinal study. Computers & Education, 49, 284-296. Li, Q. (2007). New bottle but old wine: A research of cyberbullying in schools. Computers in Human Behavior, 23(4), 1777-1791.  131  Ling, R., & Yttri, B. (2006). Control, emancipation, and status: The mobile telephone in teens’ parental and peer relationships. In R. Kraut, M. Brynin, & S. Kiesler (Eds.), Computers, Phones, and the Internet: Domesticating information technology (pp 219-234). New York, NY: Cambridge University Press. Marsh, H.W. (1986). Global self-esteem: Its relation to specific facets of self-concept and their importance. Journal of Personality and Social Psychology, 51, 1224-1236. Marsh, H.W. (1987). The hierarchical structure of self-concept and the application of hierarchical confirmatory factor analysis. Journal of Educational Measurement, 24, 1719. Marsh, H.W., Parada, R.H., Yeung, A.S., & Healey, J. (2001). Aggressive troublemakers and victims: A longitudinal model examining the pivotal role of self-concept. Journal of Educational Psychology, 93(2), 411-419. Media Awareness Network (2005). Young Canadians in a Wired World: Phase II. Retrieved 7 January, 2007 from http://www.media-awareness.ca/english/research/YCWW/phaseII/ Morettie, M.M., Holland, R., & McKay, S. (2001). Self-other representations and relational and overt aggressionin adolescent girls and boys. Behavioural Schience and the Law, 19, 109-126. Mayeux, L. & Cillessen, A.H.N. (2008). It’s not just being popular, it’s knowing it, too: The role of self-perceptions of status in the associations between peer status and aggression. Social Development, 17¸871-888. O’Connell, P., Pepler, D., & Craig, W. (1999). Peer involvement in bullying: Insights and challenges for intervention. Journal of Adolescence, 22, 437-452.  132  Olweus, D. (1991). Bully/victim problems among school children: Basic facts and effects of a school based intervention program. In Rubin & D. Pepler (Eds.), The Development and Treatment of Childhood Aggression. Hillsdale, NJ: Erlbaum. Orobio de Castro, B., Brendgen, M., Van Boxtel, H., Vitaro, F., & Schaepers, L. (2007). “Accept me , or else…”: Disputed overestimation of social competence predicts increases in Proactive Aggression. Journal of Abnormal Child Psychology, 35, 165-178. Osterman, K., Bjoerkqvist, K., Lagerspetz, K.M. J., Kaukiainen, A., Landau, S.F., Fraczek, A., Caprara, G.V. (1998). Cross-cultural evidence of female indirect aggression. Aggressive Behavior, 24, 1-8. Paquette, J.A., & Underwood, M.K. (1999). Young adolescents’ experiences of peer victimization: Gender differences in accounts of social and physical aggression. MerrillPalmer Quarterly, 45, 233-258. Pellegini, A.D. (2002). Bullying and victimization in middle school: A dominance relations perspective. Educational Psychologist, 37, 151-163. Pelligrini, A.D. (2001a). The roles of dominance and bullying in the development of early heterosexual relationships. In R.A. Geffner, M. Loring, & C. Young (Eds.), Bullying behaviour: Current issues, research, and interventions (pp. 63-73). New York: The Haworth Press, Inc. Pellegrini, A.D., & Bartini, M. (2001). Dominance in early adolescent boys: Affiliative and aggressive dimensions and possible functions. Merrill-Palmer Quarterly, 47, 142-163. Pellegrini, A.D., & Long, J. (2002). A longitudinal study of bullying, dominance, and victimization during the transition from primary to secondary school. British Journal of Developmental Psychology, 20, 259-280.  133  Pett, M.A., Lackey, N.R., & Sullivan, J.J. (2003). Making Sense of Factor Analysis: The use of factor analysis for instrument development in health care research. California: Sage Publications, Inc. Pew Internet and American Life Project (2003). American’s online pursuits: The changing picture of who’s online and what they do. Retrieved 03 March 2007 from: http://www.pewinternet.org/pdfs/PIP_Online_Pursuits_Final.PDF. Pew Internet and American Life Project (2005). Teens and Technology: Teens are leading the transition to a fully wired and mobile nation. Retrieved 03 March 2007 from http://www.pewinternet.org/pdfs/PIP_Teens_Tech_July2005web.pdf. Pew Internet and American Life Project (2007). Social Networking Websites and Teens Overview. Retrieved 03 March 2007 from http://www.pewinternet.org/pdfs/PIP_Social_Network.pdf Putallaz, M, Kupersmidt, J.B., Coie, J.D., McKnight, K., & Grimes, C.L. (2004). A behavioral analysis of girls’ aggression and victimization. In M. Putallaz & K.L. Bierman (Eds.), Aggression, Antisocial Behaviour, and Violence Among Girls (pp.110-134). New York, NY: Guilford Press. Ramirez, A.R., Dimmick, J., & Lin, S. (May, 2004). Revisiting media competition: The gratification niches of instant messaging, e-mail, and telephone. Paper presented at the Annual Meetings of the International Communication Association, New Orleans, LA. Rahkonen, O., Laaksonen, M., & Karvonen, S. (2004). The contribution of lone parenthood and economic difficulties to smoking. Social Science & Medicine, 61, 211-216.  134  Ridout, V., Roberts, D.F., & Foehr, U.G., (2005). Generation M: Media in the Lives of 8-18 year-olds. A Kaiser Family Foundation Study. Retrieved March, 2007 from http://www.kff.org/entmedia/upload/Executive-Summary-Generation-M-Media-in-theLives-of-8-18-Year-olds.pdf. Rodkin, P.C., Farmer, T.W., Pearl, R., & Van Acker, R. (2002). Heterogeneity of popular boys: Antisocial and prosocial configurations. Developmental Psychology, 36, 14-24. Rosenberg, M. (1986). Self-concept from middle childhood through adolescence. In J. Suls (Ed.), Psychological perspectives on the self (Vol. 3, pp. 107-136). Hillsdale, NJ: Erlbaum. Rubin, K.H., Bukowski, W. M., & Parker, J.G. (2006). Peer Interactions, Relationships, and Groups. In W. Damon & R.M. Lerner (Series Eds.) & N. Eisenberg (Vol. Ed.), Handbook of Child Psychology: Vo.l 3. Social, Emotional, and Personality Development (6th ed., pp. 571-645). John Wiley & Sons, Inc. Salmivalli, C., Kaukiainen, A., & Lagerspetz, K. (2001). Aggression in the social relations of school-aged girls and boys. In P.T. Slee & K. Rigby (Eds.), Children’s peer relations (pp.60-75). New York: Routledge. Salmivalli, C., Huttunen, A., & Lagerspetz, K.M.J. (1997). Peer networks and bullying in schools. Scandinavian Journal of Psychology, 38, 305-312. Sharp, S. (1995). How much does bullying hurt?: The effects of bullying on the personal wellbeing and educational progress of secondary aged students. Educational and Child Psychology, 12, 81-88. Slee, P.T. (1993). Bullying: A preliminary investigation of its nature and the effects of social cognition. Early Child Development and Care, 87, 47-57.  135  Smith, P.K., & Boulton, M. (1990). Rough and tumble play, aggression, and dominance: Perception and behavior in children’s encounters. Human Development, 33, 271-282. Smith, B. (2007). Schools ban YouTube site in cyber-bully fight. Retrieved 23, July, 2007 from: http://www.theage.com.au/news/national/schools-ban-youtube-sites-in-cyberbullyfight/2007/03/01/1172338796092.html# Stucke, T.S. & Sporer, S.L. (2002). When a grandiose self-image is threatened: Narcissism and self-concept clarity as predictors of negative emotions and aggression following egothreat. Journal of Personality, 70(4), 509-532. Tabachnick, B. G., & Fidell, L.S. (2001). Using Multivariate Statistics (4th Ed). MA: Allyn & Bacon. Thomas, D. R., Hughes, E., & Zumbo, B. D. (1998). On variable importance in linear regression. Social Indicators Research: An International and Interdisciplinary Journal for Quality-ofLife Measurement, 45, 253-275. Tremblay, R.E., Hartup, W.W., & Archer, J. (Eds.). (2005). Developmental Origins of Aggression. Guilford Press. Underwood, M.K. (2003). Social Aggression Among Girls. New York: Guildford Press. Underwood, M.K., Galen, B.R., & Paquette, J.A. (2001). Top ten challenges for understanding gender and aggression in children: Why can’t we all just get along? Social Development, 10, 248-266. Underwood, M.K., Scott, B.L., Galperin, M.B., Bjornstad, G.J., & Sexton, A.M. (2004). An observational study of social exclusion under varied conditions: Gender and developmental differences. Child Development, 75, 1538-1555.  136  Valkenburg, P.M., Peter, J., & Schouten, A.P. (2006). Friend networking sites and their relationship to adolescents’ well-being and social self-esteem. CyberPsychology, 9, 584590. Vaillancourt, T., Hymel, S., & McDougall, P. (2003). Bullying is power: Implications for school intervention programs. Journal of Applied School Psychology, 19, 157-176. Woolsey, L.K. (1986). The Critical Incident Technique: An innovative qualitative method of research. Canadian Journal of Counselling, 20, 242-254. Xie, H., Cairns, B.D., &Cairns, R.B. (2004). The development of aggressive behaviors among girls: Measurement issues, social functions and differential trajectories. In D.J. Pepler, K. Madsen, C. Webster, & K. Levene (Eds.), The development and treatment of girlhood aggression (pp. 103-134). Mahwah, NJ: Erlbaum. Xie, H, Swift, D.J., Cairns, B.D., & Cairns, R.B. (2002). Aggressive behaviors in social interaction and developmental adaptation: A narrative analysis of interpersonal conflicts during early adolescence. Social Development, 11, 205-224. Ybarra, M.L. & Mitchell, K.J. (2004). Online aggressor/targets, aggressors, and targets: A comparison of associated youth characteristics. Journal of Child Psychology and Psychiatry, 45, 1308-1316. Ybrant, H. (2007). The relation between self-concept and social functioning in adolescence. Journal of Adolescence, 31, 1-16. Young, L. (2008). Understanding a Child’s Virtual World: A handbook for parents and guardians. Available from http://www.incredibleinternet.com/online-safety/dr-lindayoung/parent-guide.  137  CHAPTER 4. IS MONITORING ADOLESCENT INTERNET USE REALLY THE ANSWER? THE ASSOCIATION AMONG PARENTING, SCHOOL, AND PEER FACTORS ON ONLINE AGGRESSION11 This dissertation is a series of studies looking at online aggression. In Study I offline bullying/victimization was compared with online bullying/victimization; it was determined that the construct of online bullying was unique from traditional forms and warranted further examination. Study II attempted to address this issue by using a mixed-method approach for examining the construct of online bullying more closely, and to examine some of the factors that predict it (i.e., age, sex, self-esteem). Study II also explored some of the motivating factors adolescents have for engaging in this type of aggression. In keeping with a socio-ecological approach, Study II focused on some of the individual characteristics deemed to be related to online aggression. Study III extends these findings by exploring the relationship between online aggression and family and school contexts.  4.1.1. FAMILY PREDICTORS OF BULLYING Previous research has identified a number of parenting factors that contribute to antisocial behaviours among adolescents, including parental characteristics and parental monitoring (Bowers, Smith, & Binney, 1994; Duncan, 2004; Olweus, 1992; Rigby, 1994; Stattin & Kerr, 2000). With respect to parental characteristics, research has shown that mothers of male bully victims tend to be overly intense, over protective, controlling, and restrictive (Bowers et al., 1994; Olweus, 1992, 1993a). Fathers of male victims, however, tend to be distant, critical, or absent (Fosse & Holen, 2002; Olweus, 1993a). It has been theorized that boys learn how to interact with other males and keep themselves safe from aggressors based on their interactions 11  A version of this chapter will be submitted for publication. Law, D.M., & Shapka, J.D. (2009). Is monitoring adolescent Internet use really the answer? The association among parenting, school, and peer factors and Online Aggression  138  with their fathers. Thus, boys with critical or disengaged fathers may not develop these skills to the same extent as boys with close relationships with their fathers (Fosse & Holen, 2002). The opposite pattern of findings has been proposed for female victims. Female victims have described their mothers as hostile, rejecting, likely to withdraw love and threaten abandonment when they misbehaved (Finnegan, Hodges, & Perry, 1998). In addition, female victims tend to describe their home life as poor functioning with inadequate family communication (Rigby, 1994). It has been suggested that, because the mothers of female victims fail to model the healthy social skills necessary for developing relationships with peers, female victims have difficulty regulating their emotions and communicating effectively, thus making them easy targets for bullies. Not surprisingly, bullying aggressors of both sexes have been found to come from families that exhibit physical and emotional aggression in the home (Olweus, 1994; Strassberg, Dodge, Pettit, & Bates, 1994). In addition to this, it appears that parents of aggressors are less likely to place limits on their children’s activities and behaviours (Olweus, 1994; Pettit, Laird, Dodge, Bates, & Criss, 2001). Parental monitoring of adolescent behaviour has long been touted as an integral method for minimizing children’s aggressive or delinquent behaviour (Reid & Patterson, 1989; Snyder & Patterson, 1987). In fact, much research has shown that poorly monitored adolescents are more likely to engage in substance abuse (Biglan, Duncan, Ary, & Smolkowski, 1995; Flannery, Vazsonyi, Torquati, & Fridrich, 1994), sexual risk taking (Metzler, Noell, Biglan, Ary, & Smolkowski, 1994), and are more likely to hang out with deviant peers (Dishion, Capaldi, Spracklen, & Li, 1995).  139  Despite these research findings, Stattin and Kerr (2000) assert that much of the work in this area has not measured the construct of monitoring, but rather parental knowledge. Parental monitoring, according to Dishion and McMahon (1998) is “a set of correlated parenting behaviours involving attention to and tracking of the child’s whereabouts, activities, and adaptations” (p. 61). By this definition, parental monitoring involves actively controlling and keeping watch over their children. In practice, however, the measures used to examine parental monitoring focus on what parents know about their child’s behaviour (e.g., How much do your parents REALLY know…where you go at night?) rather than how they acquired this knowledge (Fletcher, Darling, & Steinberg, 1995). To address this discrepancy, Stattin and Kerr (2000) conducted a study to determine the impact of whether children spontaneously disclosed to parents without prompting (child disclosure), whether parents solicited information from their children, or their children’s friends, by asking questions (parent solicitation), or whether parents impressed rules and restrictions upon their children (parental control). The authors found that parent solicitation was more highly linked to norm-breaking behaviour. That is, the more parents asked about what their children were doing, the more their child was likely to engage in antisocial or deviant behaviour. Parental control was only marginally associated with lower deviant behaviour, and only under some circumstances. Setting rules and demanding to know where their children were going, and how they were spending their time and money was only somewhat associated with lower normbreaking behaviour. In contrast, child self-disclosure was consistently linked to lower engagement in antisocial or deviant behaviour. Specifically, it was found that it is not parents’ constant questioning or rule setting that most protects children from antisocial behaviour, but it is the child’s willingness to self-disclose information of their own volition.  140  From this study it seems that, although setting rules and boundaries are important, it is even more critical for parents to create an environment and a relationship with their children that encourage self-disclosure. Essentially, it looks as if it is the child’s emotional bond or attachment to their parents that most predicts their engagement in risky behaviours. This idea is consistent with research that confirms that attachment to parents is associated with lower incidences of delinquent behaviour (Benda & Whiteside, 1995; Sokol-Katz, Dunham, & Zimmerman, 1997). Moreover, it appears that although parents are well-intentioned in asking about what their children are doing, these well-meaning attempts must correspond with their child’s willingness to disclose (Crouter, MacDermid, McHale, & Perry-Jenkins, 1990; Weintraub & Gold, 1991). In short, it seems that “Parents who are good monitors have made the effort to establish channels of communication with their child, and as a result of their relationship with the child, they are knowledgeable about the child’s daily experiences” (Pulkkinen, 1982, p. 656). This information is critical when we think about parental monitoring and rule setting with respect to Internet use. From the literature, it is evident that the relationships children have with their parents influences whether they are more prone to victimization, aggressive behaviour, and other anti-social behaviours. To date, very little research on parental monitoring and computer/Internet use has been conducted. Despite the paucity of research in this area, numerous people have espoused the importance of monitoring children’s online and computer activities through rule setting and monitoring software. These ideas are exemplified through the media and popular Internet safety books and websites. Such resources advise parents to keep track of what their children are doing online by installing monitoring software such as NetNanny, Razzul, or BlogBeware, and by regularly examining the computer’s browsing and messaging histories (see Nickel, 2006 for example). According to recent research on parental  141  control, and the findings of Stattin and Kerr (2000), it seems that installing monitoring software, checking message histories, and controlling adolescent online behaviours might be counterproductive and may actually play a part in increasing poor behaviour on the Internet. It is hypothesized that the relationship adolescents have with their parents, and the willingness of the adolescent to discuss their online activities with their parents, is more important than monitoring, tracking, and soliciting information from adolescents about their Internet use is. This hypothesis is supported by the work of Ybarra and Mitchell (2004) who found that poor caregiver-child relationships were significantly related to online aggression. In addition, similar to the parent-child dyad for offline bullies, children with strong emotional bonds to their parents were less likely to engage in online bullying, while those with poor emotional bonds were twice as likely to engage in online aggressive activities. The purpose of Study III was to extend these findings by examining whether it is parent solicitation, child disclosure, or parental control that more highly predicts online aggression among adolescents. 4.1.2. Peer and School Predictors As described in Study I, aggression and bullying are influenced by a number of peer and school predictors. For example, peers play a critical role in promoting positive behaviours, such as kindness, and learning (Harris, 1998; Hodges et al., 1997; Hymel et al., 1990; ZimmerGembeck et al., 2005), as well as negative behaviours, such as aggression and other problem behaviours (McFarland, 2001). Research has also found that feelings of acceptance and belonging among adolescents are critical during this developmental period (Brown, 2000; McPherson et al., 2001). That is, adolescents tend to congregate toward those who are similar to themselves in terms of demographics, personality, and key beliefs (Rodkin, 2004), and tend to divide themselves according to cliques and crowds (Brown et al., 1994). These cliques are 142  arranged hierarchically as adolescents vie for power and popularity within their school. This vie for power often results in aggressive behaviours among peers as they attempt to maintain control and status (Adler & Adler, 1998; Dodge et al., 1990; Estell et al., 2002; Rodkin et al., 2000; Vaillancourt et al, 2003; Xie et al., 2002). Moreover, in an attempt to remain in good standing with their high status peers, many adolescents who are not necessarily aggressive, may reinforce aggression that was instigated by a more popular peer by supporting them through approval, or laughter (O’Connell et al, 1999; Salmivalli et al., 1997). In conjunction with these peer predictors, research has also examined the importance for adolescents to feel they belong in their school. For example, schools which promote student engagement and which emphasize learning, caring and socio-emotional well-being are more likely to have students who engage in fewer aggressive behaviours, and who experience higher achievement (Ames, 1992; Griffith, 1995; Kasen et al., 2004; Luiselli et al., 2005). Please refer to Study I for a more thorough discussion of peer and school predictors. The Internet offers a unique environment for adolescents to identify with others and feel accepted, even if they may not possess these feelings offline. For example, adolescents who feel isolated offline because of personality, socio-economic status, ethnicity, or language barriers may feel accepted online, where they are comfortable communicating with others because their personal and physical characteristics are concealed. Unfortunately, this also means that adolescents who may not feel comfortable gossiping or engaging in aggressive behaviour offline, may feel inclined to do so online. These problem behaviours may even facilitate their feelings of acceptance, when they may not otherwise feel they belong. Given this, it was hypothesized that adolescents who reported lower levels of belonging would engage in higher levels of online aggression.  143  The purpose of this study was to examine whether it is disclosure to parents, parental control (limit setting on Internet use), or parental knowledge of their adolescent’s Internet use that is associated with an adolescent’s engagement in online aggression. Furthermore, this study investigates how belonging in school and a sense of similarity to one’s peers is related to online aggression.  4.2. METHODS The data for this study were drawn from the same data as Study II. As described in Study II, the participants were 733 (282 males and 451 females) elementary and high school students between the ages of 10 and 18, from the Lower Mainland of British Columbia; they completed a paper and pencil questionnaire titled Social Responsibility on the Internet (Please refer to Study II for more information on data collection). In addition to demographic information, students were asked to respond to questions pertaining to (a) Online Aggression, (b) Parenting, (c) Sense of School Belonging, and (d) School Similarity. Online Aggression. As described in the previous study, the students were asked to respond to questions pertaining to online aggression. Recall that an exploratory factor analysis (EFA) was run on these items in order to help develop the composite variables; three unique factors emerged: (1) Aggressive Messaging (2) Hostile Website Development and (3) Posting/Commenting on Embarrassing Photos or Videos (Refer to Study II for more information) Parenting. To determine whether parental solicitation, parental control, or adolescent self-disclosure of Internet use predicted online aggression or victimization, Stattin and Kerr’s Parenting Questionnaire (2000) was modified to address how parents might parent around ICT use. Students were first asked to respond to questions about their parents’ knowledge (i.e., “To what extent do your parents actually know about what you do on the Internet and where you are 144  going; who you have as friends on the Internet; what you are posting online; what you are texting online or on your cell phone?”). Next, participants responded to questions pertaining to parental control (i.e., “To what extent do you have to tell your parents when you are going on the Internet; what you will be doing on the Internet, who you will be chatting with on MSN; what messages, pictures, or videos you will be posting on public websites like Facebook, Nexopia, or YouTube”). Child Disclosure questions followed the parental control questions (i.e., “How often do you spontaneously tell your parents about what you and your friends are doing on the Internet; want to tell your parents about what you are chatting about or posting on the Internet; hide from your parents about what you are doing on the Internet?”). Finally, Parental solicitation questions were asked (e.g., “How often do your parents talk to your friends on MSN or on the Internet; you about what you are doing online; you about who you are chatting with online?”; See Appendix C question, 27). The responses to the items of the questionnaire were on a 5-point Likert scale ranging from Never to All of the time. A factor analysis was run to determine the factor structure of these items. Sense of School Belonging. The Psychological Sense of School Membership scale (PSSM), which is comprised of 20 items, was used to determine students’ sense of school belonging (e.g., “I feel a real part of my school”, “People here notice when I am good at something”, and “I feel very different from most other students here”). Responses were given on a 5-point Likert scale, ranging from 1 = Not at all True to 5 = Completely True (See Appendix C, question 48). Negatively-worded items were reverse coded. Internal consistency was found to be α = .89. A School Belonging composite was derived by calculating the mean score of the 20 items.  145  School similarity. This construct was assessed by asking students to respond to eight questions about how similar they are to their peers with respect to their ethnicity and to the language they speak at home (examples include: “Of the peers at school you feel comfortable going to for personal help, how many of them speak the same language as you?” and “…are the same ethnicity as you?”). Specifically, students were asked to rate, on a 5-point Likert scale ranging from 1 = None of them to 5 = All of them, how ethnically and linguistically similar the their peers are to them; peers that they go to for school-related help, for personal help, and for hanging out with (See Appendix C questions, 40, 42, 44). “I don’t know” responses were coded as missing.  4.3. RESULTS 4.3.1. Determining the Factor Structure for Parenting Items. Because the parenting measure was modified from Stattin and Kerr’s parenting questionnaire (2000), an EFA was conducted to evaluate the factor structure of the modified items. Unweighted least squares extraction method was used because it does not assume multivariate normality. With the exception of question 27j, “How often do you hide from your parents about what you are doing on the Internet”, which did not correlate with any of the parental control items (See Appendix C questions 27d to f), the correlation matrix indicated significant correlations among all items (See Table 4.1); as such, Oblimin rotation was employed on all 13 parenting items. The initial EFA revealed that four items (See Appendix C questions 27 h, i, j, and k) loaded poorly or cross-loaded on several factors; thus, they were removed from the next iteration, as recommended by Pett and colleagues (2003). For the second EFA, the Kaiser-Meyer-Oklin value was .8, exceeding the recommended minimum value of .6 (Kaiser,  146  1970, 1974), and the Bartlett’s Test of Sphericity (Bartlett, 1954) reached statistical significance (p < .001); thus supporting the factorability of the items. The screeplot illustrated a clear break after the second factor and the pattern matrix clarified the factor loadings for two factors: (a) Parent Solicitation and (b) Child Disclosure (See Table 4.2). Composite variables were created by taking the average across the items for each factor; higher scores represented higher parental solicitation or child disclosure. The reliability was found to be α = .87 and α = .80 for Parental Solicitation and Child Disclosure, respectively. Participants were also asked three additional questions related to parental control (i.e., “Do your parents install programs that keep track of where you are going online; Yes/No?”) and parental rule setting/control (i.e., “Do your parents limit the amount of time you spend on the computer” and “Do your parents limit where you can go and things you can do online?”). Students responded to the latter two questions on a 5-point Likert scale ranging from Never to All of the time, where 1 = Never and 5 = All of the time. “I don’t know” responses were coded as missing. (See Appendix C, questions 22 and 23).  147  Table 4.1. Correlation matrix for parenting items. 27b 27c 27d 27e 27f 27g 27h 27i 27j 27k 27l 27m  27a .58*** .55*** .20*** .23*** .21*** .34*** .35*** .35*** -.09* .08* .31*** .29***  27b  27c  27d  27e  27f  27g  27h  27i  27j  27k  27l  .54*** .22*** .23*** .30*** .33*** .38*** .32*** -.16*** .16*** .32*** .35***  .22*** .24*** .28*** .29*** .33*** .35*** -.11** .08* .28*** .27***  .72*** .53*** .44*** .37*** .28*** -.32 .20*** .40*** .33***  .67*** .56*** .46*** .37*** -.06 .24*** .47*** .41***  .62*** .43*** .38*** -.06 .26*** .49*** .56***  .42*** .42*** -.13** .15*** .42*** .42***  .72*** -.10* .15*** .46*** .45***  -.09* .17*** .42*** .40***  .01*** .03*** .05***  .26*** .29***  .75***  *p < .05. **p < .01. ***p < .001  148  Table 4.2: Factor loadings for the unweighted least squares factor analysis of parenting items. Factor Item Parent Solicitation To what extent do you have to tell your parents what you are doing online? To what extent do you have to tell your parents when you are online? To what extent do you have to tell your parents when you are online? To what extent do you have to tell your parents what you are posting online? How often do your parents talk to you about what you do online? How often do your parents talk to you about who you are chatting with online? Child Disclosure To what extent do you parents actually know about what you do and post online? To what extent do your parents actually know about who you have as friends online? To what extent do your parents actually know about what you are texting online or on your cell phone?  1  2  .910 .849 .762 .657 .590 .577 .766 .740 .719  *Note: Only factor loadings of .32 or greater are reported here as is consistent with reporting procedures for Factor Analyses (Tabachnich & Fidell).  In order to determine the impact of parenting behaviours and similarity/belonging in school, six separate hierarchical multiple regression analyses were conducted for the three online aggression constructs that were established in Study II (i.e., Aggressive Messaging, Hostile Website Development, and Commenting/Posting Embarrassing Photos/Videos). Similar to Study II, variables for having a computer in their bedroom, grade, sex, and coming from a single parent family were entered as covariates in Block 1 of the regression. The first set of regression analyses examined the extent to which belonging and similarity contributed to the model over and above parenting behaviours. That is, Parent Solicitation, Child Disclosure, and Parental Control variables were entered in Block 2, and Belonging and Similarity constructs were entered 149  in Block 3. As there is no clear theoretical reason why these items should be entered into the model in this particular order, a second set of regressions investigated the extent to which parenting behaviours contributed to the model over and above belonging and similarity; that is, for this model Belonging and Similarity were entered in Block 2, and Parent Solicitation, Child Disclosure, and Parental Control variables were entered in Block 3. By running the models in this way it was possible to confirm that there was no difference in predictive ability depending on how the items were entered. The assumptions for each of the three models were largely met. However, across all three models, minor violations were found in that Lone Parenting and Similarity to Peers did not significantly correlate with any of the dependent variables. Tolerance levels were in keeping with the assumption for multicollinearity, as all three models had tolerance levels between .67 and .98. Normal Probability Plots and the Residual Scatterplots were assessed for linearity, normality, and outliers. The assumptions were met for Aggressive Messaging and Commenting/Posting Embarrassing Photos/Videos as the points lay roughly in a straight diagonal line and the residuals were rectangular and centering along 0. For creating Hostile Websites, minor deviations were found, in that the distribution seemed to be slightly negatively skewed. However, given that regression analyses are robust to these minor assumption violations (Howell, 1997), and given the large sample size, it was deemed safe to proceed. Prior to reporting the outcomes, the relative contributions of each set of predictor variables were assessed for each of the models. As can be seen in Table 4.3, no matter what order the predictor variables were entered, the model came out relatively similar for each of the outcome variables with the exception of Aggressive Messaging. These analyses indicate that,  150  with the exception of Aggressive Messaging, the order in which the variables are entered is irrelevant. Table 4.3. Comparison of the R2-value for hierarchical multiple regression model with different orders of entry.  Dependent Variable Aggressive Messaging Hostile Website Development Commenting/Posting Embarrassing Photos/Videos  1 2 3 1 2 3 1 2 3  Block 2: Parenting Items Block 3: Belonging & Similarity Items Sig. F R2 ∆R2 Change .057 .057 .000 .077 .020 .035 .081 .004 .259 .041 .041 000 .045 .005 .714 .051 .005 .193 .110 .110 000 .113 .099 .887 .113 .000 .961  Block 2: Belonging & Similarity Items Block 3: Parenting Items Sig. F R2 ∆R2 Change .057 .057 .000 .065 .008 .086 .081 .016 .081 .041 .041 000 .047 .006 .157 .051 .004 .773 .110 .110 .000 .110 .000 .984 .113 .003 .881  For Aggressive Messaging, the variables for having a computer in the bedroom, grade, and sex significantly predicted adolescents’ engagement in this form of aggression, regardless of whether Parenting, Belonging, and Similarity items were placed in Block 2 or 3 (see Table 4.4 and 4.5). Specifically, older female adolescents who had a computer in their bedroom were more likely to send aggressive messages to others. These findings held when the parenting constructs were entered into the model (R2 = .08, ∆R2 = .02, p<.05). With regard to parenting behaviours, Parental Solicitation and Parental Control items did not significantly predict the sending of aggressive messages. However, Child Disclosure did significantly predict whether adolescents sent aggressive messages to others; this relationship was found to be negative (β=-.04, p<.05). The more often adolescents tell their parents about their online activities, the fewer aggressive messages they send online. Sense of Belonging and Similarity to Peers did not improve the 151  model when they were added in Block 3. However, even in the presence of these two constructs, having a computer in the bedroom, grade, and sex remained significant and Child Disclosure tended toward significance (β=-.09, p = .06). Slightly different outcomes emerged when Belonging and Similarity were entered into Block 2. Specifically, Belonging in school significantly predicted whether adolescents sent aggressive messages or not, and this relationship was found to be negative (β=-.09, p <.05). In other words, the more adolescents feel they belong in school, the less they participated in sending aggressive messages. This outcome did not hold when the parenting outcomes were entered into Block 3. Again, Child Disclosure tended toward significance (β=-.09, p = .06). Interaction effects for sex, grade, and each of the parenting constructs were also examined. These, however, did not emerge as significant and were not included in the final model. For Hostile Website Development, no matter whether the Parenting, Belonging and Similarity items were entered into Block 2 or 3, having a computer in the bedroom and grade were the only items to emerge as significant in the final model (See Table 4.6 and 4.7) and the relationship was found to be positive. Thus, older adolescents were more likely to develop a hostile website if they had a computer in their bedroom. None of the parenting, belonging, or similarity items significantly contributed to the model. For Posting/Commenting on Embarrassing Photos and Videos, similar outcomes emerged no matter whether Parenting, Belonging, and Similarity items were entered into Block 2 or 3 (See Tables 4.8 and 4.9). In this case, grade was the only significant predictor of this form of aggression, with sex tending toward significance in Blocks 2 and 3 for both models (p = .06). In other words older adolescents were more likely to engage in this form of aggression than younger adolescents.  152  Table 4.4. Summary of hierarchical multiple regression for Belonging & Similarity’s predictive value over & above Parenting Behaviours for Aggressive Messaging. Aggressive Messaging Block 1: Covariates Computer in bedroom Grade Sex Lone Parent Block 2: Parenting Computer in bedroom Grade Sex Lone Parent Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online Block 3: Belonging & Similarity Computer in bedroom Grade Sex Lone Parent Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online Sense of Belonging Similarity to Peers  B .08 .03 .09 .03  SE .03 .01 .03 .03  β  r .10** .17*** .11** .04 *  UR2  .057  .057***  .077  .020*  .081  .004  .11 .18 .11 .05  .07 .03 .10 .03 -.02 -.04  .03 .01 .03 .03 .02 .02  .09 .15** .13** .03 -.05 -.10*  .11 .18 .11 .05 -.16 -.12  .06  .06  .04  .08  .004  .01  .04  -.09  .01  -.06  -.09  -.01  R2  .07 .03 .10 .02 -.02 -.03  .03 .01 .03 .03 .02 .02  .09* .15** .13** .03 -.04 -.09  .11 .18 .11 .05 -.16 -.12  .06  .06  .04  .08  .004  .01  .04  -.09  -.01 -.04 -.002  .01 .03 .02  -.06 -.07 -.01  -.09 -.10 .000  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  153  Table 4.5. Summary of hierarchical multiple regression for Parenting Behaviour’s predictive value over & above Belonging and Similarity for Aggressive Messaging. Aggressive Messaging Block 1: Covariates Computer in bedroom Grade Sex Lone Parent Block 2: Parenting Computer in bedroom Grade Sex Lone Parent Sense of Belonging Similarity to Peers Block 3: Belonging & Similarity Computer in bedroom Grade Sex Lone Parent Sense of Belonging Similarity to Peers Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online  B  SE .08 .03 .09 .03  .08 .03 .09 .02 -.05 -.001  .03 .01 .03 .03 .03 .01 .03 .03 .03 .02  β  r .10* .17*** .11** .04 *  .10 .17*** .11** .03 -.09* -.002  *  UR2  .06  .06***  .07  .01  .08  .02  .11 .18 .11 .05 .11 .18 .11 .05 -.10 .000  .07 .03 .10 .02 -.04 -.002 -.02 -.03  .03 .01 .03 .03 .03 .02 .02 .02  .09 .15** .13** .03 -.07 -.01 -.04 -.09  .11 .18 .11 .05 -.10 .000 -.16 -.12  .06  .06  .04  .08  .004  .01  .04  -.09  .01  -.06  -.09  -.01  R2  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  154  Table 4.6. Summary of hierarchical multiple regression for Belonging & Similarity’s predictive value over & above Parenting Behaviours for Hostile Website Development. Hostile Website Development Block 1: Covariates Computer in bedroom Grade Sex Lone Parent Block 2: Parenting Computer in bedroom Grade Sex Lone Parent Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online Block 3: Belonging & Similarity Computer in bedroom Grade Sex Lone Parent Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online Sense of Belonging Similarity to Peers  B  SE  .06 .02 -.03 .02  .02 .01 .02 .03  β  R2  r  .10* .15*** -.05 .04 *  UR2 .041  .041***  .05  .005  .051  .005  .12 .16 -.05 .05  .05 .02 -.02 .02 -.01 -.01  .02 .01 .02 .03 .01 .01  .09 .14** -.04 .03 -.03 -.03  .12 .16 -.05 .05 -.11 -.08  .01  .05  .00  .03  -.01  .01  -.14  -.06  .01  .01  .13  -.05  .06 .02 -.03 .02 -.01 -.01  .02 .01 .02 .03 .01 .01  .09* .13** -.05 .03 -.03 -.03  .12 .16 -.05 .05 -.11 -.08  .01  .05  .01  .03  -.01  .01  -.14  -.06  .01 -.02 -.02  .01 .02 .01  .12 -.05 -.06  -.05 -.06 -.05  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  155  Table 4.7. Summary of hierarchical multiple regression for Parenting Behaviour’s predictive value over & above Belonging and Similarity for Hostile Website Development. Hostile Website Development Block 1: Covariates Computer in bedroom Grade Sex Lone Parent Block 2: Parenting Computer in bedroom Grade Sex Lone Parent Sense of Belonging Similarity to Peers Block 3: Belonging & Similarity Computer in bedroom Grade Sex Lone Parent Sense of Belonging Similarity to Peers Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online  B  SE  .06 .02 -.03 .02 .06 .02 -.03 .02 -.03 -.02  .02 .01 .02 .03 .02 .01 .02 .03 .02 .01  β  r  .10* .15*** -.05 .04 *  .10 .15*** -.05 .03 -.05 -.06  *  R2  UR2  .041  .041***  .047  .006  .05  .004  .12 .16 -.05 .05 .12 .16 -.05 .05 -.06 -.05  .06 .02 -.03 .02 -.02 -.02 -.01 -.01  .02 .01 .02 .03 .02 .01 .01 .01  .09 .13** -.05 .03 -.05 -.06 -.03 -.03  .12 .16 -.05 .05 -.06 -.05 -.11 -.08  .01  .05  .01  .03  -.01  .01  -.14  -.06  .01  .01  .12  -.05  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  156  Table 4.8. Summary of hierarchical multiple regression for Belonging & Similarity’s predictive value over & above Parenting Behaviours for Commenting/Posting Embarrassing Photos/Videos. Commenting/Posting Photos/Videos Block 1: Covariates Computer in bedroom Grade Sex Lone Parent Block 2: Parenting Computer in bedroom Grade Sex Lone Parent Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online Block 3: Belonging & Similarity Computer in bedroom Grade Sex Lone Parent Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online Sense of Belonging Similarity to Peers  B .04 .09 .09 .05  SE .05 .01 .05 .05  β  r .04 .32*** .07 .04  R2  UR2  .110  .110***  .113  .003  .113  .000  .06 .32 .08 .05  .04 .09 .09 .05 -.02 .00  .05 .01 .05 .05 .03 .02  .03 .30*** .08 .04 -.03 -.01  .06 .32 .08 .05 -.14 -.04  -.05  .09  -.02  .01  .01  .02  .05  -.09  -.01  .02  -.08  -.09  .04 .09 .09 .05 -.02 .00  .05 .01 .05 .05 .03 .02  .03 .30*** .08 .04 -.03 -.01  .06 .32 .08 .05 -.14 -.04  -.05  .09  -.02  .01  .01  .02  .06  -.09  -.01 .01 .00  .02 .04 .02  -.08 .01 .00  -.09 .00 -.02  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  157  Table 4.9. Summary of hierarchical multiple regression for Parenting Behaviour’s predictive value over & above Belonging and Similarity for Commenting/Posting Embarrassing Photos/Videos. Commenting/Posting Photos/Videos Block 1: Covariates Computer in bedroom Grade Sex Lone Parent Block 2: Parenting Computer in bedroom Grade Sex Lone Parent Sense of Belonging Similarity to Peers Block 3: Belonging & Similarity Computer in bedroom Grade Sex Lone Parent Sense of Belonging Similarity to Peers Parent Solicitation Child Disclosure Parents keep track of where you are going Parents limit time on computer Parents limit where you go online  B  SE .04 .09 .09 .05 .04 .09 .08 .05 .01 .00  .05 .01 .05 .05 .05 .01 .05 .05 .04 .02  β  r .04 .32*** .07 .04 .04 .32*** .07 .04 .01 .00  R2  UR2  .11  .11***  .11  .000  .113  .003  .06 .32 .08 .05 .06 .32 .08 .05 .00 -.02  .04 .09 .09 .05 .01 .00 -.02 .00  .05 .01 .05 .05 .04 .02 .03 .02  .03 .30*** .08 .04 .01 .00 -.03 -.01  .06 .32 .08 .05 .00 -.02 -.14 -.04  -.05  .09  -.02  .01  .01  .02  .06  -.09  -.01  .02  -.08  -.09  Note. *p < 0.05. **p < 0.01 ***p < 0.001.  158  4.4. DISCUSSION Based on the outcomes of the regression models, it is clear that having a computer in the bedroom strongly predicted whether adolescents reported engaging in aggressive messaging and developing hostile websites. This finding is not surprising given previous work which has found that adolescents are more likely to engage in aggressive behaviours when there is less supervision (Ridout et al, 2005). Surprisingly, parenting behaviours played a minimal role in allaying online aggressive behaviours. In fact, aggressive messaging was the only outcome variable to be predicted by a parenting item: Child disclosure. This result is not surprising as it is consistent with previous research on parenting (Stattin & Kerr, 2000), where child disclosure of what they are doing was the only construct to significantly predict the sending of aggressive messages. Specifically, the more adolescents disclosed to their parents about what they are doing online, the less likely they were to send aggressive messages to others. Parental Control and Parent Solicitation were not found to significantly predict any of the forms of online aggression. Although the finding that Child Disclosure was the only construct to predict whether adolescents sent aggressive messages makes sense and links well to “traditional” parenting research, it is interesting to note that this construct had no effect on whether adolescence created hostile websites or posted/commented on embarrassing pictures. Possible non-significance for creating websites might be due to the low sample size of individuals who participated in this form of aggression. Recall from Study II that only 12% of adolescents actively create websites to harm others; as such, it is possible that a Type II error is responsible for the lack of significance. That said, from Study II we learned that adolescents who created hostile websites did so for proactive reasons. Thus, another potential reason why parenting behaviours had no 159  significance on this type of online hostility is because these individuals seem to be a special group who are willing to intentionally spend time engaging in this form of aggression. Perhaps these aggressors come from families where parents exhibited behaviours that were linked to traditional aggression. For example, research has found that parents who withdrew love, threatened abandonment, had poor communication, or exercised aggression at home had children who were more likely to engage in aggressive behaviours (Finnegan et al, 1998; Rigby, 1994; Strassberg et al., 1994). Perhaps this is also true of adolescents who engage in hostile website development. Future research must examine the parent/child relationship of this special group more closely. Interestingly, none of the parental control items significantly predicted any form of online aggression. These findings are important when considered in an age where parents tend to focus on controlling and monitoring their adolescents’ online behaviours – techniques which this study has shown has no impact on decreasing online aggression. This finding makes sense given that previous research on parental control has found only a marginal association between parents controlling their children’s behaviour and a decrease in aggressive and deviant conduct (Stattin & Kerr, 2000). Sense of School Belonging only emerged as significant for sending aggressive messages, with greater feelings of belonging predicting fewer submissions of aggressive messages. These findings make sense given that feelings of school belonging have been linked to lower levels of aggressive behaviours (Kasen, Berenson, Cohen, & Johnson, 2004; Vaillancourt et al., 2003). A possible reason for this parallel could be that sending aggressive messages can be seen as another type of social/relational aggression, where adolescents can feel comfortable gossiping and spreading rumours.  160  Conversely, school belonging and similarity to peers did not emerge as significant for hostile website development or commenting/posting embarrassing pictures. This is contrary to work which has found a relationship between aggression and these constructs (Dodge et al., 1990; Vaillencourt et al.). A possible explanation for this is the idea that online aggression typically happens outside of school (Agatston et al., 2007). At school, adolescents may feel pressure to behave according to their peers in order to feel they belong; as such, they may engage in or reinforce offline aggression. When it comes to online aggression, however, adolescents may not feel this external pressure from peers to say or do mean things online – they may simply engage in this form of aggression because they, themselves, feel justified or want to. 4.4.1. Conclusions Overall, it appears that the influence of parenting behaviours, peer, and school factors in predicting online aggression varies according to the form of aggression adolescents are engaging in online. These outcomes support those which emerged from Study I and II which found that online aggression seems to be a construct unique from traditional forms of aggression and victimization. This study also emphasizes the importance of keeping computers out of adolescents’ bedrooms as this was found to be a predictor for both sending aggressive messages and developing hostile websites. Given the inverse relationship between school belonging and the sending aggressive messages, it is important for schools to foster an environment that promotes school belonging (Ames, 1992; Griffith, 1995; Kasen et al., 2004; Luiselli et al., 2005). Like traditional aggression, Child Disclosure was a predictor of Aggressive Messaging, however, it was not found to predict Hostile Website Development or Commenting/Posting Embarrassing Photos/Videos.  161  4.4.2. Limitations There are some limitations to this study. As mentioned in Study II, although the sample size was large (N=733), only a few adolescents engaged in specific forms of online aggression, namely Hostile Website Development, which elevated the risk for Type II errors. Future work must examine adolescents who create hostile websites more closely. In addition to this, the parenting measures for this work did not account for the large East Asian population and how parenting and parenting constructs might differ according to culture. Additional work in this area should be mindful in considering how parenting around the Internet might differ among cultures. Another limitation of this study was the questions that were asked regarding similarity to peers. The questions in this study focused primarily on ethnic and language similarity and did not examine similarity across other homophilic domains (i.e., key beliefs, personality characteristics, and shared interests; McPherson, et al., 2001; Rodkin, 2004). It is critical for future work to study other areas of similarity to properly determine whether and how similarity predicts online aggression. Despite these limitations, this work has established that it is good communication and good relationship between parents and adolescents that decreases aggressive messaging. These findings are critical to emphasize in an age where parents tend to focus on controlling and monitoring their adolescents’ online behaviours – techniques which have now been shown to have no significance in decreasing online aggression.  162  4.5. REFERENCES Adler, P.A., & Adler, P. (1998). Peer power: Preadolescent culture and identity. New Brunswick, NJ: Rutgers University Press. Agatston, P.W., Kowalski, R. & Limber, S. (2007). Students’ perspectives on Cyberbullying. Journal of Adolescent Health, 41, 59-60. Ames, C, 1992). Classrooms: Goals, structures, and student motivation. Journal of Educational Psychology, 84, 261-271. Bartlett, M.S. (1954). A note on the multiplying factors for various chi square approximations. Journal of the Royal Statistical Society, 16 (Series B), 296-298. Baumrind, D. (1968). Authoritarian versus authoritative parental control. Adolescence, 3, 255272. Benda, B.B., & Whiteside. L. (1995). Testing an integrated model of delinquency using LISREL. Journal of Social Science Research, 21, 1-32. Biglan, A., Duncan, T.E., Ary, D.B., & Smokowski, K. (1995). Peer and parental influences on adolescent tobacco use. Journal of Behavioural Medicine, 18, 315-330. Bowers, L., Smith, P.K., & Binney, V. (1994). Perceived family relationships of bullies, victims, and bully/victims in middle childhood. Journal of Social and Personal Relationships, 11, 215-232. Brown, B.B., Mory, M.S., & Kinney, D. (1994). The complex world of peer crowds: Casting crowds in a relational perspective: Caricature, channel, and context. In R. Montemayor, G.R. Adams, & T.P. Gullotta (Eds.), Advances in adolescent development: Vol. 6. Personal relationships during adolescence (pp.123-167).  163  Brown, R. (2000). Social identity theory: Past achievements, current problems and future challenges. European Journal of Social Psychology, 30, 745-778. Chao, R.K. (2003). Extending research on the consequences of parenting style for Chinese Americans and European Americans. Child Development, 72, 1832-1843. Crouter, A.c., MacDermid, S.M., McHale, S.M., & Perry-Jenkins, M. (1990). Parentla monitoring and perceptions of children’s school performance and conduct in dual and single-earner families. Developmental Psychology, 26, 649-657. Dishion, T.J., Capaldi, D., Spracklen, K.M., & Li, F. (1995). Peer ecology of male adolescent drug use: Developmental processes in peer relations and psychopathology [Special Issue]. Development and Psychopathology, 7, 803-824. Dishion, T.J., & McMahon, R.J. (1998). Parental monitoring and the prevention of child and adolescent problem behaviour: A conceptual and empirical formulation. Clinical Child and family Psychology Review, 1, 61-75 Dodge, K.A., Coie, J.D., Pettit, G., & Price, J. (1990). Peer status and aggression in boys’ groups: Developmental and contextual analyses. Child Development, 61, 1289-1309. Duncan, R.D. (2004).  The impact of family relationships on school bullies and victims.  Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 227 – 244). Mahwah, NJ: Lawrence Erlbaum Associates. Estell, D.B., Cairns, R.B., Farmer, T.W., & Cairns, B.D. (2002). Aggression in inner-city early elementary classrooms: Individual and peer-group configurations. Merrill-Palmer Quarterly, 48, 52-76.  164  Finnegan, R.A., Hodges, E.V.E., & Perry, D.G. (1998). Victimization by peers: Associations with children’s reports of mother-child interaction. Journal of Personality and Social Psychology, 75, 1076-1086. Flannery, D.J., Vazsonyi, a.T., Torquati, J., & Fredirich, A. (1994). Ethnic and gender differences in risk for early adolescent substance use. Journal of Youth and Adolescence, 23, 195-213. Fletcher, A.C., Darling, N., & Steinberg, L. (1995). Paretnal monitoring and peer influences on adolescent substance abuse. In J. McCord (Ed.), Coercion and Punishment in Long-Term Perspectives (pp. 259-271). Cambridge, MA: Cambridge University Press. Fosse, G.K., & Holen, A. (2002). Childhood environment of adult psychiatric outpatients in Norway having been bullied in school. Child Abuse & Neglect, 26, 129-137. Gorman, J.C. (1998). Parenting attitudes and practices of immigrant Chinese mothers of adolescents. Family Relations, 47, 73-80. Griffith, J. (1995). An empirical examination of a model of social climate in elementary schools. Basic and Applied Social Psychology, 17, 97-117. Harris, J.R. (1998). The Nurture Assumption: Why children turn out the way they do. New York, NY: Touchstone. Hodges, E.V.E., Malone, M.J. & Perry, D.G., 1997. Individual risk and social risk as interacting determinants of victimization in the peer group. Developmental Psychology, 33, 10321039. Hymel, S., Rubin, K.H., Rowden, L., & LeMare, L. (1990). Children’s peer relationships: Longitudinal prediction of internalizing and externalizing problems from middle to late childhood. Child Development, 61, 2004-2021.  165  Kaiser, H. (1970). A second generation Little Jiffy. Psychometrika, 35, 401-415. Kaiser, H (1974). An index of factorial simplicity. Psychometrika, 39, 31-36. Kasen, S., Berenson, K., Cohen, P., & Johnson, J.G. (2004). The effects of school climate on changes in aggressive and other behaviors related to bullying. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 107 – 117). Mahwah, NJ: Lawrence Erlbaum Associates. Luiselli, J.K., Putnam, R.F., Hndler, M.W., & Feinberg, A.B. (2005). Whole-school positive behaviour support: Effects on student discipline problems and academic performance. Educational Psychology, 25, 183-198. McFarland, D.A. (2001). Student resistance: How the formal and informal organization of classrooms facilitate everyday forms of student defiance. American Journal of Sociology, 107, 612-678. McPherson, M., Smith-Lovin, L., & Cook, J.M. (2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415-444. Metzler, C.W., Noell, J., Biglan, A., Ary, D., & Smolkowski, K. (1994). The social context for risky sexual behaviour among adolescents. Journal of Behavioural Medicine, 17, 419438. Nickel, R. (2006). Staying safe in a wired world: A parent’s guide to Internet safety. Ontario: Nickel Publishing. O’Connell, P., Pepler, D., & Craig, W. (1999). Peer involvement in bullying: Insights and challenges for intervention. Journal of Adolescence, 22, 437-452.  166  Olweus, D. (1992). Victimization by peers: Antecedents and long term outcomes. In L.H. Rubin & J.B. Asendorph (Eds.), Social withdrawal, inhibition, and shyness in childhood. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc Olweus, D. (1993a). Bullying at school. Cambridge, MA: Blackwell. Olweus, D. (1994). Annotation: Bullying at school: Basic facts and effects of a school-based intervention program. Journal of Child Psychology and Psychiatry, 35, 1171-1190. Pettit, G.S., Laird, R.D., Dodge, K.A., Bates, J.E., & Criss, M.M. (2001). Antecedents and behavior-problem outcomes of parental monitoring and psychological control in early adolescence. Child Development, 72, 583-598. Pulkkinen, L. (1982). Self-control and continuity from child hood to adolescence. In P.B. Baltes & O.G. Brim (Eds.), Life Span development and behaviour (Vol. 4, pp. 63-105). New York: Academic Press. Reid, J.B., & Paaterson, G.R. (1989). The development of antisocial behavior patterns in childhood and adolescence: Personality and aggression [Special Issue]. European Journal of Personality, 3, 107-119. Rigby, K. (1994). Psychosocial functioning in families of Australian adolescent schoolchildren involved in bully/victim problems. Journal of Family Therapy, 16, 173-187. Rodkin, P.C. (2004). Peer ecologies of aggression and bullying. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 87 – 106). Mahwah, NJ: Lawrence Erlbaum Associates. Rodkin, P.C., Farmer, T.W., Pearl, R., & Van Acker, R. (2002). Heterogeneity of popular boys: Antisocial and prosocial configurations. Developmental Psychology, 36, 14-24.  167  Salmivalli, C., Huttunen, A., & Lagerspetz, K.M.J. (1997). Peer networks and bullying in schools. Scandinavian Journal of Psychology, 38, 305-312. Snyder, J., & Patterson, G. (1987). Family interaction and delinquent behavior. In H.C. Quay (Ed.), Handbook of Juvenile Delinquency. New York: Wiley. Stattin, H. & Kerr, M. (2000). Parental monitoring: A reinterpretation. Child Development, 71, 1072-1085. Steinberg, L., Lamborn, S.D., Darling, N., Mounts, N.S., & Dornbush, S.M. (1994). Over-time changes in adjustment and competence among adolescents from authoritative, authoritarian, indulgent, and neglectful families. Child Development, 65, 754-770. Strassberg, Z., Dodge, K.A., Pettit, G.S., & Bates, J.E. (1994). Spanking in the home and children’s subsequent aggression toward kindergarten peers. Development and Psychopathology, 6, 445-461. Tabachnick, B. G., & Fidell, L.S. (2001). Using Multivariate Statistics (4th Ed). MA: Allyn & Bacon. Vaillancourt, T., Hymel, S., & McDougall, P. (2003). Bullying is power: Implications for school intervention programs. Journal of Applied School Psychology, 19, 157-176. Weintraub, K.J., & Gold, M. (1991). Monitoring and delinquency. Criminal Behaviour and Mental Health, 1, 268-281. Xie, H, Swift, D.J., Cairns, B.D., & Cairns, R.B. (2002). Aggressive behaviors in social interaction and developmental adaptation: A narrative analysis of interpersonal conflicts during early adolescence. Social Development, 11, 205-224.  168  Ybarra, M.L. & Mitchell, K.J. (2004). Online aggressor/targets, aggressors, and targets: A comparison of associated youth characteristics. Journal of Child Psychology and Psychiatry, 45, 1308-1316. Zimmer-Gembeck, M.J., Geiger, T.C., & Crick, N.R. (2005). Relational and physical aggression, prosocial behavior, and peer relations: Gender moderation and bidirectional associations. Journal of Early Adolescence 25, 421- 452.  169  CONCLUDING CHAPTER The overall purpose of this work has been to provide a more in-depth analysis of online aggression. In contrasting ‘Cyberbullying’ with more traditional forms of bullying, this research utilized a socio-ecological perspective which has shown to be a useful framework for guiding us in understanding the reciprocal effects of individual, parental, peer, school, and community factors on traditional bullying and victimization (see Espelage & Swearer, 2004 and Swearer & Doll, 2001 for more information).  5.1. STUDY I The intention of Study I was to make comparisons between online bullying and victimization, and traditional bullying and victimization in order to ascertain whether Cyberbullying is a unique construct, and one which needs to be examined in its own right. Drawing on data collected for a larger study on safe schools and social responsibility, the results of Study I showed that adolescents seem to view the construct of Cyberbullying differently from traditional bullying and traditional victimization. Specifically, it was found that although adolescents made clear distinctions between offline bullying and offline victimization, they did not make similar distinctions between online bullying and online victimization. It appears that the distinction between being a bully and being a victim is far less distinguishable when the aggression is occurring via ICT. Although the perceptions of online vs. offline bullying and victimization differed for participants, results from this study also found that similar factors predicted both forms of bullying and victimization. That is, for all forms of bullying and victimization, there were similar grade and sex differences – with grades 9’s and 10’s having more experience with these forms of aggression than grade 8’s, 11’s, or 12’s; and with boys reporting more incidences of 170  bullying experiences. Furthermore, all five of the other individual and school characteristics examined, significantly predicted all forms of bullying and victimization. Moreover, analyses showed that school-level factors were the biggest predictors of all forms of aggression and victimization (specifically, Belonging in School and School Engagement). A distinction of note was that academic performance was not related to offline victimization, but was inversely related to Cyberbullying/victimization and Traditional Bullying. Taken together, these results showed that while there do appear to be some similarities in the factors that predict bullying and victimization (irrespective of the venue through which it is occurring), there are also important distinctions in how adolescents conceptualize aggression when it is not happening in a face-toface situation but in a virtual situation. Clearly, more work is warranted to unravel these distinctions.  5.2. STUDY II In an attempt to do this, Study II employed a mixed-method approach for understanding the construct of online aggression. This study also explored several factors deemed to predict online aggression, and from a socio-ecological perspective, focused on individual-level characteristics. Several important outcomes emerged from this work. First, using more robust measures, findings were consistent with Study I in that adolescents did not differentiate between being bullied online or being victimized online. Specifically, it was demonstrated that, unlike traditional forms of bullying, where adolescents clearly differentiate between bullies, victims, and witnesses, for online aggression adolescents made distinctions between the methods for aggressing online. That is, the constructs which emerged from the factor analyses distinguished between the mode of the aggression, not the role of the people involved in the aggressive acts. Specifically, participants in this study identified three forms of online aggression: Aggressive 171  Messaging, Hostile Website Development, and Commenting/Posting Embarrassing Photos/Videos. Grade and sex differences were found to vary according to the form of online aggression the adolescents engaged in. For example, girls and older students were more likely to be involved in sending aggressive messages, but only grade differences were found for creating hostile websites and for commenting and posting embarrassing photos. As with sending aggressive messages, these forms of aggression were also found to increase with age. Regarding other individual characteristics, self-esteem was found to have a positive relationship with sending aggressive messages and comment/posting embarrassing photos/videos but this outcome did not hold true for creating hostile websites. Study II also looked at the reasons why adolescents aggressed online – whether it was proactive or reactive in nature. For sending aggressive messages and for posting/commenting on photos/videos, it appears that adolescents were primarily reactive in their aggression. In contrast, in terms of creating hostile websites, adolescents reported being primarily proactive in their aggression. These findings suggest that online aggressive behaviours seem to be much more reciprocal in nature than traditional bullying and, as such, do not possess the same power differential, intention to harm, or pattern of repetitiveness as characterized by traditional bullying. For example, the evidence from Study I and Study II showed that individuals are oftentimes both a bully and a victim in the same aggressive incident. Similarly, the interview findings from Study II revealed that individuals tended to engage in both proactive and reactive aggression and were inclined to feel that the aggressive behaviours of others, online, were proactive in nature, while they viewed their own behaviours as retaliatory and justified. With these differences in mind, labeling this phenomenon Cyberbullying may not be appropriate. A  172  more appropriate label is likely something more generic such as online aggression (which has been adopted here). It should be noted that one form of online aggression did stand out as perhaps being more closely related to traditional forms of bullying; namely, the creation of hostile or aggressive websites. Evidence from Study II indicated that this form of online aggression appears to be more proactive, intentional, and malicious in nature. Further work needs to specifically explore this form of aggression, as well as the characteristics of the individuals engaging in it (both bullies and victims). Another terminology problem that arises as a result of this work is the use of the term ‘bully-victim’ (individuals who bully and are bullied). Despite the fact that many participants indentified as engaging in bullying/aggressive behaviours and in being victimized, in traditional bullying literature this term refers to a very small, unique group of individuals who actively taunt and provoke more powerful individuals into bullying them, and then initiate aggressive behaviours toward others who are less competent than they are (Kaukiainen et al., 2002; Pelligrini, 2001a). These individuals are also at the highest risk for maladaptive outcomes associated with bullying (Pellegrini, 2001b; Swearer, et al., 2004). Data from this current work suggests that online bully-victims are very different from what is traditionally thought of as a bully-victim. Not only does it appear as though being a bully-victim in an online situation may be the norm (not the exception), but in most instances, there does not seem to be a power differential between the individuals who are involved in the aggressive acts (as is the case with traditional bully-victims). In addition, online, it appears that, although individuals are aggressing in reaction to perceived provocation, they are not turning around and being aggressive to  173  someone else. Rather, they tend to be aggressive back to the person who was aggressive to them. Future work in this area is clearly required to identify how these groups of individuals differ. Interview data from Study II also revealed that adolescents typically use a combination of non-confrontational aggression online and offline, and confrontational aggression online. These trends demonstrate that perhaps the Internet provides an alternative option for allowing individuals to be indirectly confrontational in their aggression. That is, via ICT, adolescents can make belligerent comments directly to someone else, without needing to see them face-to-face. Previous work on confrontational and non-confrontational aggression has found differences in individuals who engage in these two different types of aggression (Xie, Swift, Cairns, & Cairns, 2002). Future research is needed to explore whether this distinction also exist in an online situation.  5.3. STUDY III Study II provided an initial framework for deconstructing online aggression and for investigating some of the individual and motivational factors that constitute this phenomenon. Study III extended this by examining some of the home and school predictors of online aggression – from a socio-ecological perspective, the influence of the contexts in which bullying takes place. Using the same online aggression measures identified in Study II (Aggressive Messaging, Hostile Website Development, and Commenting/Posting Embarrassing Photos/Videos), this study revealed that adolescents who had a computer in their bedroom were more likely to engage in aggressive messaging, and hostile website development. In addition, regarding parental influences on online aggression, child disclosure was found to be a significant predictor. For example, the more adolescents in this study reported voluntarily disclosing to their parents about what they were doing online, the less likely they 174  were to engage in aggressive messaging (although this outcome was not found for hostile website creation or for commenting and posting on embarrassing photos and videos). Most important is the finding that increased parental monitoring (parent solicitation of information) and stricter parental control (about when and how often a child was allowed to use ICTs) did not predict lower reported levels of online aggression. These findings are consistent with those of recent work on parenting, which show that it is not parental control or parental solicitation which predicts better behavioural outcomes; rather it is the strength of the parent/child relationship (Stattin & Kerr, 2000). Regarding school and peer outcomes, sense of belonging in school did predict whether adolescents would engage in aggressive messaging. Conversely, similarity to one’s peers did not significantly predict any form of online aggression. This contradicts the findings of Study I, where school belonging was a significant predictor of Cyberbullying/victimization. Different measures of this construct were used for Study I and III which may explain this mixed finding. Regardless, peer and school influences on online aggression are factors that need to be explored further. Especially given that online aggression typically occurs outside of the school, but tends to be about the relationships which are formed at school (Agatston, Kowalski, & Limber, 2007).  5.4. LIMITATIONS OF THIS WORK There are a few limitations to this research. Most of these limitations have already been highlighted in detail at the end of each study; however, it is important to address some of the limitations that are relevant to the work as a whole. The first limitation involves sample size. Study I had a very large sample (N=19551). Although a large sample is desirable, it also puts the data at risk for Type I error, whereby statistical significance may have been achieved due to inflated power. In contrast, although the sample size for Studies II and III was appropriate (N = 175  733), participants were not equally distributed across grade, sex, and form of online aggression they engaged in. This means that the findings for these studies may have been confounded by Type II error, whereby the power for detecting statistical significance is hampered due to an unbalanced design. Another limitation of this work is that it was conducted crossectionally, which means we cannot be sure that the age effects represent actual within-person developmental changes that we would expect to see as an individual gets older. In future, it would be meaningful to collect data longitudinally. Collecting data longitudinally becomes even more paramount when we consider the context in which this work is occurring – namely the changing nature of the technological landscape. Due to the rapid changes in the ICTs being used for aggressive acts, it becomes incredibly important to collect multi-cohort longitudinal data so that cohort effects (related to the changes in technology) can be parsed from developmental effects. An additional limitation to this work is the large East Asian population for all three studies (54% for Study I and 45% for both Studies II and III), and that the measures for this study were largely based on or modified from measures used with primarily Euro-American populations. Because of this, there is possibility that the outcomes were not entirely reflective of what online aggression really is and how it manifests itself. As such, it is important for future research on parenting and the Internet to be mindful of cultural discrepancies when creating measures and constructs.  5.4. CONCLUSIONS AND IMPLICATIONS This work is among the first to examine online aggression in a theoretically grounded, thorough manner. More specifically, the strengths of this study are that it utilized a socioecological approach to examine the phenomenon of online aggression, as well as a mixed 176  method approach by incorporating data from both self report questionnaires and semi-structured interviews. Overall, this work has demonstrated that although there are some similarities to traditional bullying, online aggression is a unique construct and should be examined as such. One of the most important findings is that adolescents interpret online aggression differently from more traditional forms of bullying. That is, they differentiated themselves as individuals who participated in specific forms of online aggression, rather than as individuals who played a particular role in online aggression. This has important implications for prevention, intervention, and education about social responsibility on the Internet. Although a digital divide often exists between parents and children or teachers and students (Fox & Madden, 2005), it is imperative that we immerse ourselves in the environments that young people participate in so that we can better educate and promote responsible behaviour. This is especially so for dealing with aggression online. Focus group data has revealed that online bullying incidents decreased when victims of online aggression reported the incidents to adults (Hinduja & Patchin, 2008). Moreover, Study II demonstrated that adolescents who self-disclosed their online experiences to their parents were less likely to engage in aggressive messaging. Unfortunately, focus group data also revealed that many children do not come forward due to embarrassment, while others admit that they “wanted to tell my parents but I was afraid that they would never let me chat again and I know that's how a lot of other kids feel” (Share Your Story section). In addition, social responsibility programs need to be extended and developed specifically for aggression that is occurring via ICTs. These programs should not only teach children and adolescents to appropriately handle being the target of online aggression, but also to  177  help reduce the likelihood of retaliative behaviour, and, in so doing, teach them alternatives to aggressing online in response to a perceived threat. The findings from this study make it very clear that any education program needs to be developed specifically for this type of aggression, and we cannot assume that existing anti-bullying programs will also be effective for reducing this type of aggression. When online aggression is understood from a socio-ecological perspective, it becomes evident that it is comprised of a myriad of factors, including individual characteristics, peer relationships, and parent and teacher involvement. A full understanding of the phenomenon will only be obtained by including all these social systems in research, and looking at the interplay among them. Moreover, the development of intervention and prevention programs may also benefit from attending to individual characteristics, peer relations, and the broader family and school ecology. Clearly, much work remains to be done in order to address the problem of online aggression among children and adolescents, and the importance of online social responsibility.  178  5.5. REFERENCES Agatston, P.W., Kowalski, R. & Limber, S. (2007). Students’ perspectives on Cyberbullying. Journal of Adolescent Health, 41, 59-60. Espelage, D.L. & Swearer, S.M. (Eds.). (2004). Bullying in American Schools: A SocioEcological Perspective on Prevention and Intervention. Mahwah, NJ: Lawrence Erlbaum Associates, Inc., Publishers. Hinduja, S., & Patchin, J.W. (2008). Cyberbullying.us. Retrieved November 2008, from http://www.cyberbullying.us/shareyourstory.php. Kaukiainen, A., Slamivalli,C., Lagerspetz, K., Tamminen, M., Vauras, M., & Poskiparta, E. (2002). Learning difficulties, social intelligence, and self-concept: Connections to bullyvictim problems. Scandinavian Journal of Psychology, 43(3), 269-278. Pelligrini, A.D. (2001a). The roles of dominance and bullying in the development of early heterosexual relationships. In R.A. Geffner, M. Loring, & C. Young (Eds.), Bullying behaviour: Current issues, research, and interventions (pp. 63-73). New York: The Haworth Press, Inc. Pellegrini, A.D. (2001b). Sampling instances of victimization in middle school: A methodological comparison. In J. Juvonen & S. Graham. (es.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 125-146). New York: The Guilford Press. Stattin, H. & Kerr, M. (2000). Parental monitoring: A reinterpretation. Child Development, 71, 1072-1085. Swearer, S.M., & Doll, B. (2001). Bullying in schools: An ecological framework. Journal of Emotional Abuse, 2, 7-23. 179  Swearer, S.M., Grills, A.E., Haye, K.M. & Cary, P.T. (2004). Internalizing problems in students involved in bullying and victimization: Implications for intervention. Espelage & S. Swearer (Eds.), Bullying in American schools: A social-ecological perspective of prevention and intervention (pp 63-83). Mahwah, NJ: Lawrence Erlbaum Associates. Xie, H, Swift, D.J., Cairns, B.D., & Cairns, R.B. (2002). Aggressive behaviors in social interaction and developmental adaptation: A narrative analysis of interpersonal conflicts during early adolescence. Social Development, 11, 205-224.  180  APPENDICES  APPENDIX A  181  182  183  184  185  186  187  188  189  190  191  192  193  194  195  196  197  198  199  200  201  APPENDIX B  202  203  PARENT CONSENT FORM SOCIAL RESPONSIBILITY ON THE INTERNET Principal Investigator: My name is Danielle M. Law and I am a PhD student who is working under the supervision of Dr. Jennifer Shapka (an assistant professor at the University of British Columbia). Both Dr. Shapka and I are from the Department of Educational and Counselling Psychology, and Special Education. If you have any questions regarding this project, please call: Danielle Law: (604)-822-3000 Dr. Jennifer Shapka: (604)-822-5253 Next week, your child will be learning about a type of bullying that has started to emerge on the Internet. They will also be invited to participate in a research study that we are conducting at your child’s school, called “Social Responsibility on the Internet”. This research is part of a dissertation for my Doctoral degree, and the final document will be made public. At the same time, steps will be taken to make sure that everything your child says and writes down will be kept completely anonymous, and only the above mentioned researchers will have access to the raw information. Findings from this project will help us to better understand how children’s and adolescent’s computer and online experiences are influencing their development, and how we can help them be safer on the Internet. Purpose: The purpose of the study is to ask your child to help us, as researchers and people who work with children and teens, to better understand what they are doing on the Internet so that we can make it a safer place for them. Specifically, this study will ask your child about their experiences with Internet bullying and what they have done to deal with this growing issue. Children and teenagers are the experts on themselves and their experiences are very important. Their participation can really help us better understand the needs of other teenagers in British Columbia like them. We also hope that the results of this study will help teachers and parents improve their understanding of the issues that children and teenagers in Canada are dealing with.  204  We are asking your child and other students between the grades of 5 and 12 to participate in this study, because this is the age group that uses the Internet the most. Study Procedures: This study involves two parts. The first part asks your child to complete, during class time, a questionnaire on their computer use, how they feel about themselves (their selfconcept), and what they know about Internet Aggression. Demographic questions such as age, gender and ethnicity will also be asked, so that we can gain a better sense of who the participants in this study are. This will take about 60 minutes. The second part of the study asks your child or teen to participate in an interview. This interview will take about an hour and may occur either face-to-face or via Instant Messenger. The purpose of the interview is to help us clarify their computer and online experiences. Questionnaires will be completed during class time. If your child does not receive your consent he/she will not be penalized in any way, but will be asked to read or draw quietly, or complete their homework at their desk as their classmates complete the questionnaire. There are no known risks associated with this study; however, should your child feel uncomfortable, he/she has the right to withdraw from the study without any penalty, at any time. In addition, your child will have the option to request counseling should they so choose. Confidentiality: All comments and responses, made by your child, will be kept strictly confidential, unless legally required for disclosure, and no identifying information will be included with the documents. All the documents will only be identifiable by a code number and will be kept in a locked filing cabinet. Any information obtained via the internet (Instant Messenger) will be stored in a password protected database on the researcher’s computer. Your child will not be identified by name in any of the reports of the completed study.  205  Contact for information about the study: If you have any questions or would like further information with respect to this study, you may contact Danielle Law at 604-822-3000, Dr. Jennifer Shapka at (604)-822-5253. Contact for concerns about the rights of research subjects: If you have any concerns about your treatment or rights as a research subject, you may contact the Research Subject Information Line in the UBC Office of Research Services at 604-822-8598. Consent: Your child’s participation in this study is entirely voluntary and refusal to participate or withdraw from the study is permitted at any time without jeopardy to their class standing. Your signature below indicates that you have received a copy of this consent form for your own records. Your signature indicates that you consent to your child’s participation in this study. -------------------------------------------------------------------------------------------------------Please return this bottom portion and keep the consent form for your own records.  Please identify to what extent you give consent to your child’s participation in this study by checking the appropriate box. I consent to my child's participation in the questionnaire portion of the study. I consent to my child’s participation in the interview portion of the study I do not consent to my child’s participation in any part of this study  ____________________________________________________ Subject Signature Date (or Parent or Guardian Signature) ____________________________________________________ Printed Name of the Subject or Parent or Guardian signing above.  206  STUDENT ASSENT FORM SOCIAL RESPONSIBILITY ON THE INTERNET Principal Investigator: My name is Danielle M. Law and I am a PhD student who is working under the supervision of Dr. Jennifer Shapka (an assistant professor at the University of British Columbia). Both Dr. Shapka and I are from the Department of Educational and Counselling Psychology, and Special Education. If you have any questions regarding this project, please call: Danielle Law: (604)-822-3000 Dr. Jennifer Shapka: (604)-822-5253 Next week, you will be learning about a type of bullying that is starting to happen on the Internet. You will also be invited to participate in a research study that called “Social Responsibility on the Internet”. This research is part of a dissertation for my Doctoral degree, and the final document will be made public. At the same time, we will make sure that everything you say and write down will be kept completely anonymous, and only myself and Dr. Shapka will have access to the original information. Findings from this project will help us to better understand how children’s and teenager’s Internet use is influencing their development so we can make things safer for them on the Internet. Purpose: The purpose of the study is to ask you to help us, as researchers and people who work with children and teens, to better understand what you are doing on the Internet so that we can make it a safer place for you and future children and teens. Specifically, this study will ask you about your experiences with Internet bullying and what you have done to deal with this problem. Children and teenagers are the experts on themselves and their experiences are very important. Your participation can really help us better understand the needs of other teenagers in British Columbia. We also hope that the results of this study will help teachers and parents improve their understanding of the issues that children and teenagers in Canada are dealing with.  207  We are asking you and other students between the grades of 5 and 12 to participate in this study, because this is the age group that uses the Internet the most.  Study Procedures: This study involves two parts. The first part asks you to complete, during class time, a survey on your computer use, how you feel about yourself, and what you know about Internet bullying. Questions such as age, gender and ethnic background will also be asked, so that we can gain a better sense of who the participants in this study are. This will take about 60 minutes. The second part of the study asks you to participate in an interview. This interview will take about an hour and can happen either face-to-face or via Instant Messenger (MSN). The purpose of the interview is to help us clarify your computer and online experiences. Surveys will be completed during class time. If you do not want to participate in this study, or if your parents do not want to participate you will not get in trouble. We will just ask you to read or draw quietly, or complete homework at your desk as your classmates complete their survey. There are no known risks for participating in this study; but, if you feel uncomfortable, you have the right to stop participating in the study, at any time, and you will not get in trouble. In addition, you will have the option to request counseling if you want. Confidentiality: All comments and responses that you make will be kept strictly confidential, unless legally required for disclosure, and no identifying information will be included with the documents. This means your parents, teachers, and friends will not know what you have written down. All the documents will be given a code number and will be kept in a locked filing cabinet. Any information obtained via the internet (Instant Messenger) will be stored in a password protected database on the researcher’s computer. Your name will not be in any of the reports of the completed study.  208  Contact for information about the study: If you have any questions or would like further information about this study, you may contact Danielle Law at 604-822-3000, Dr. Jennifer Shapka at (604)-822-5253. Contact for concerns about the rights of research subjects: If you have any concerns about your treatment or rights as a research subject, you may contact the Research Subject Information Line in the UBC Office of Research Services at 604-822-8598. Consent: Your participation in this study is your choice. If you do not want to participate or want to stop participating in this study this decision will not affect your grades in any way. Your signature below indicates that you are willing to participate in this study. -------------------------------------------------------------------------------------------------------Please return this bottom portion and keep the assent form for your own records.  Please tell us what part of the study you would like to take part in by putting a check mark next to your choice. I would like to take part in the questionnaire part of the study. I would like to take part in the interview part of the study I do not want to take part in any part of this study  ____________________________________________________ Student Signature Date  ____________________________________________________ Printed Name of Student Grade  209  APPENDIX C  210  Social Responsibility on the Internet Questionnaire 2007  Survey for Elementary and High School Students The information you give us about your experiences is very important for the school so that we can help make school a better and safer place for you.  Some key things to remember: o DO NOT write your name on this survey. o This survey is voluntary and your answers are anonymous. No one will know what your answers are. o This is NOT a test and there are no right or wrong answers, but it is important that you answer honestly. o Whether or not you answer the questions will not affect your grade in this class. o Make sure to read every question. o Please do not look at other students’ answers. o If you are not comfortable answering a question just leave it blank. o By filling out this survey you are agreeing to participate in this study Thank you very much for your help. 211  1. What is the name of your school? __________________________________________ 2. What grade are you in? 3. What is your birthday? 4. Are you a boy or a girl?  5  6  7  8  9  10  11  12  Day____Month____Year_______ Boy  Girl  5. What is your racial/ethnic background? Choose one. My racial/ethnic background is: Aboriginal (e.g. First Nations, Non‐Status Indian, Inuit, Métis) African/Caribbean (e.g. Black) Asian (e.g. Cambodian, Chinese, Japanese, Korean, Taiwanese, Thai, Vietnamese, Filipino) South Asian (e.g. East‐Indian, Pakistani) Caucasian (e.g. White, European, Russian) Latin American (e.g. Mexican, Portuguese, South American) Middle Eastern (e.g. Arabic, Iranian, Kuwaiti, Persian, Turkish, Israeli, Palestinian) My racial/ethnic background is mixed. Please describe. ______________________________________________________________________ I don’t know. 6. What country were you born in? ______________________________________________ 7. What country were your parents born in? _______________________________________ *If you were born in Canada, skip to question 10. 8. If you were NOT born in Canada, how long have you lived here? I have lived in Canada for ____________________ years.  212  9. If you were born in another country, did you come to Canada as a (please check one): Immigrant (chose to come to Canada) Refugee (could not stay in your native country) International Student (studying in Canada) I don’t know 10. What language(s) are spoken at home?___________________________________ 11. What are the first 3 digits of your postal code at your home? __________________ 12. What is the highest level of education that you would like to complete? Choose one. Not finish high school High school graduation Training/apprenticeship program (like carpentry, computer training, legal assistant) Some college/university classes College diploma University/bachelor degree (undergraduate) Masters degree Professional degree (like lawyer, nurse, architect) Doctoral degree 13. Which of these adults do you live with MOST OF THE TIME? (Check all the adults you live with). Mother  Grandmother  ½ Mom, ½ Dad  Father  Grandfather  Foster parent(s)  Stepfather  Stepmother  Other adults (please tell us who:________________________________________________  213  14. What level of education do your parents have? Mom (or female caregiver)  Dad (or male caregiver)  I don’t have a mom or female caregiver  I don’t have a dad or male caregiver  Not finished high school  Not finished high school  High school graduation  High school graduation  Training/apprenticeship program (like  Training/apprenticeship program (like  carpentry, computer training)  carpentry, computer training)  Some college/university classes  Some college/university classes  College diploma  College diploma  University/bachelor degree  University/bachelor degree  (undergraduate)  (undergraduate)  Masters degree  Masters degree  Professional degree (like lawyer, nurse,  Professional degree (like lawyer, nurse,  architect)  architect)  Doctoral degree  Doctoral degree  I don’t know  I don’t know  15. What job(s) do your parents have? Mom (or female caregiver)  Dad (or male caregiver)  214  The following questions ask about your computer and cell phone access 16. How many computers does your family have at home? I do not have a computer at home 1 computer 2 computers 3 or more computers 17. Do you have your own computer?  Yes  No  18. Do you have High Speed Internet access at home (for example, Shaw or Telus ADSL)? Yes  No  I don’t know  19. Do you have your own cell phone?  Yes  No  20. Do you have a computer in your bedroom?  Yes  No  21. Do your parents install programs that keep track of where you are going online (e.g. Net Nanny)? Yes  No  I don’t know  22. Do your parents limit the amount of time you spend on the computer (eg. Only on weekends, only for a certain number of hours a day)? Please Circle one  Never  Rarely  Some of the time  Most of the time  All of the time  I don’t know  23. Do your parents limit where you can go and things you can do online (eg. Don’t let you play certain games or visit certain websites)? Please check one.  Never  Rarely  Some of the time  Most of the time  All of the time  I don’t know  215  These questions ask how much time you spend per day doing the following things on the Internet 24. Please pick the box that best describes how much time you spend doing the following things per day on WEEKDAYS Instant Messaging (e.g. MSN, ICQ, a Yahoo Chat, etc.) b Sending and reading email Posting messages, updating information, and/or creating c profiles on public websites (e.g. Facebook, Nexopia, LiveJournal, Xanga, etc.) Playing online games with other d people e Chatting in public chatrooms f Creating websites g Texting on your cell phone 25. Please pick the box that best describes how much time you spend doing the following things per day on WEEKENDS Instant Messaging (e.g. MSN, ICQ, a Yahoo Chat, etc.) b Sending and reading email Posting messages, updating information, and/or creating c profiles on public websites (e.g. Facebook, Nexopia, LiveJournal, Xanga, etc.) Playing online games with other d people e Chatting in public chatrooms f Creating websites g Texting on your cell phone  Never  1 hour or less/da y  Between 2 &5 hours/day  5 or more hours/day  Never  1 hour or less/da y  Between 2 &5 hours/day  5 or more hours/day  216  MARSH’S SELF­DESCRIPTION QUESTIONNAIRE (SELF­CONCEPT MEASURE) Please read the question to the left and choose the answer that best fits how you feel about yourself. Remember to only check ONE answer for each question. 26. False  Mostly False  More False than True  More True than False  Mostly True  True  1. Nobody thinks that I’m good looking. 2. Overall, I have a lot to be proud of. 4. I enjoy things like sports, gym, and dance. 5. I am usually relaxed. 6. It is difficult to make friends with members of my own sex. 7. People of the opposite sex whom I like don’t like me. 8. I have a nice looking face. 9. Overall, I am no good. 10. I am lazy when it comes to things like sports and hard physical exercise. 11. I worry more than I need to. 12. I make friends easily with boys. 13. I make friends easily with girls. 14. Most of my friends are better looking than I am. 15. Most things I do, I do well. 16. I’m good at things like sports, gym, and dance. 17. I don’t get upset very easily. 18. Not many people of my own sex like me. 19. I’m not very popular with members of the opposite sex. 20. I am good looking. 21. Nothing I do ever seems to turn out right. 22. I am awkward at things like sports, gym, and dance. 217  False  Mostly False  More False than True  More True than False  Mostly True  True  23. I am often depressed and down in the dumps. 24. I am popular with the boys. 25. I am popular with the girls. 26. I hate the way I look. 27. Overall, most things I do turn out well. 28. I’m better than most of my friends at things like sports, gym, and dance. 29. Other people get more upset about things than I do. 30. I do not get along very well with boys. 31. I do not get along very well with girls. 32. Other people think I am good looking. 33. I try to get out of the sports and physical education classes whenever I can. 34. I am a nervous person. 35. I have good friends who are members of my own sex. 36. I have lots of friends of the opposite sex. 37. I can do things as well as most people. 38. I can run a long way without stopping. 39. I often feel confused and mixed up. 40. Most boys try to avoid me. 41. Most girls try to avoid me. 42. I have a good looking body. 43. I feel that my life is not very useful. 44. I hate things like sports, gym, and dance. 45. I get upset easily. 218  False  Mostly False  More False than True  More True than False  Mostly True  True  46. I make friends easily with members of my own sex. 47. I get a lot of attention from members of the opposite sex. 48. If I really try I can do almost anything I want to do. 49. I am calm person. 50. I have a few friends of the same sex as myself. 51. Overall, I’m a failure. 52. People can really count on me to do the right thing. 53. I worry about a lot of things. 54. I enjoy spending time with my friends of the same sex. Please turn to the next page…  219  The following questions ask about your computer access at home 27. To what extent do your parents actually know about… what you do and what do you a post on the Internet? who you have as friends on b the Internet what you are texting online c or on your cell phone To what extent do you have to tell your parents … When you are going on the d Internet? what you will be doing on the e Internet? Who you will be chatting f with on MSN What messages, pictures, or videos you will be posting on g public websites like Facebook, Nexopia, or YouTube How often do you … Spontaneously tell your parents about what you and h your friends are doing on the Internet? Want to tell your parents about what you are chatting i about or posting on the Internet? Hide from your parents j about what you are doing on the Internet? How often do your parents talk to … Your friends on MSN or on k the Internet you about what you are l doing online? you about who you are m chatting with online?  Nothing at all  Very little  Some  Quite a bit  Everything  I don’t know  Never  Rarely  Some of the time  Most of the time  All of the time  I don’t know  Never  Rarely  Some of the time  Most of the time  All of the time  I don’t know  Never  Rarely  Some of the time  Most of the time  All of the time  I don’t know  220  Think about your experiences using the Internet, Email, and/or Cell Phone Text Messaging.  28. How often have you had experience with…  a  b  HAD THIS DONE TO ME Never  Once or a few times  b  c  d  e  Every week or more  Never  Once or a few times  About once a month  Every week or more  SAW OR HEARD ABOUT IT Never  Once or a few times  About once a month  Mean things, rumours, or gossip being said through the Internet, websites, email, or text messaging? Embarrassing pictures or video clips of yourself or people you know being sent or posted on the Internet?  29. How often have you had experience with…  a  About once a month  TOOK PART IN DOING IT TO OTHERS  WITH FRIENDS Never  Once or a few times  Every week or more  ALONE Never  Once or a few times  Every week or more  Using the Internet or text messaging to send mean messages, spread rumours, or gossip about others? Replying to mean messages about yourself using the Internet or text messaging? Receiving mean messages about somebody you know over the Internet or text messaging Replying to mean messages about somebody else over the Internet or text messaging Creating websites or posting messaging on Nexopia, Facebook etc. that embarrass or make fun of other people?  221  Every week or more  The following questions asks what you have done when you have been picked on, or bullied on the Internet or through cell phone text messaging. 30. Have you been picked on, or bullied, over the Internet or through cell phone text messaging? Yes  No (If no, please go to question 32)  31. When you have been picked on or had mean things said or done to you on the Internet or through cell phone text messaging, how often have you… Replied to the message and told the a person(s) to stop? b Talked to the person(s) about it? Ignored or blocked the bullying people c from your IM or email Ignored or avoided the person(s) who d bullied when you were at school Got your friends to help solve the e problem? f Talked to an adult at home? g Talked to a friend about it? h Told an adult at school? Talked to the bullying persons friends i about it j Talked to the person who was being hurt Got back at the bullying people by doing k something mean to them or posting mean messages about them online? Got a group of people together to fight the l people doing the bullying? Did something else (please explain):  Neve r  Hardl y ever  Some of the time  Most of the time  Always  m  222  The following questions will ask about how you handle certain situations 32. How often have you… a. b. c. d. e. f. g. h. i. j. k. l. m.  Never  Sometimes  Often  Posted or said mean things to others online when they have annoyed you? Posted or said mean things on the Internet to show who was on top? Reacted angrily online when provoked by others? Said or posted mean things about others on the Internet just for fun? Said or posted mean things online because you felt mad? Posted or said mean things on the Internet to be cool? Become angry or mad when you don’t get your way and then taken it out online by posting or saying mean things to others? Posted or said mean things online because others have threatened you? Threatened and bullied someone on the Internet? Posted or said mean things on the Internet to defend yourself or someone else. Gotten others to say or post mean things on the Internet about someone else Said or posted mean things online when you were teased Threatened, posted, or said mean things on the Internet so others would do things for you.  33. Please share some other comments you might have about Cyberbullying. If you don’t have any comments to share, please move on to the next page. _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ ___________________________________________________________________________  223  The following questions will ask about the types of support you feel you get from adults, your peers, and your school Adult 34. How many adults at school are you comfortable going to for help with school work? (For example, help with homework or projects):________________________________. 35. Of the ADULTS at school you feel comfortable going to for help with SCHOOL WORK, how many of them: Were born in another country? For example, they were immigrants or a refugees. Speak the same language at home b as you? Are the same ethnicity as you? For c example Asian or Caucasian. Were born in the same country as you? d  None of them  Mostly none  Half of Them  Most of them  All of them  I don’t know  36. How many adults at school are you comfortable going to for personal help? (For example, when you need someone to talk to when you are mad, sad, or stressed):_________________________________________________. 37. Of the ADULTS AT SCHOOL you feel comfortable going to for PERSONAL help, how many of them: Were born in another country? For example, they were immigrants or a refugees. Speak the same language at home b as you? Are the same ethnicity as you? For c example Asian or Caucasian. Were born in the same country as d you? 38. Are the adults you would go to for school work help and personal help the same people? (Circle One)  No, not at all  None of them  Mostly none  Mostly no  Half of Them  Half of Them  Most of them  Mostly yes  All of them  Yes, all of the them  I don’t know  Not applicable  224  Peer­At School 39. How many peers at school are you comfortable going to for school­related help? For example, help with homework or projects:____________________________________ 40. Of the PEERS AT SCHOOL you feel comfortable going to for SCHOOL RELATED help, how many of them: Were born in another country? For A example, they were immigrants or refugees. Speak the same language at home B as you? Are the same ethnicity as you? For C example Asian or Caucasian. Were born in the same country as you? D  None of them  Mostly none  Half of Them  Most of them  All of them  I don’t know  41. How many peers at school are you comfortable going to for personal help? For example, when you need to talk to someone when you are mad, sad, or stressed._______. 42. Of the PEERS AT SCHOOL you feel comfortable going to for PERSONAL HELP, how many of them: Were born in another country? For A example, they were immigrants or refugees. Speak the same language at home B as you? Are the same ethnicity as you? For C example Asian or Caucasian. Were born in the same country as D you?  None of them  Mostly none  Half of Them  Most of them  All of them  I don’t know  43. How many peers are you comfortable hanging out with at school? For example, at lunch breaks:______________________________________________.  225  44. Of the PEERS AT SCHOOL do you feel comfortable HANGING OUT WITH, how many of them: Were born in another country? For a example, they were immigrants or refugees. Speak the same language at home as b you? Are the same ethnicity as you? For c example Asian or Caucasian. Were born in the same country as d you? 45. Are the peers you would go to for school help, personal help, or to hang out with, at school, the same people? (Circle One)  None of them  No, not at all  Mostly none  Mostly no  Half of Them  Half of Them  Most of them  Mostly yes  All of them  Yes, all of the them  I don’t know  Not applicabl e  School composition 46. Thinking about the STUDENTS in your ENTIRE SCHOOL, how many people would you say: Were born in another country? a For example, they were immigrants or refugees. Speak the same language at home b as you? Are the same ethnicity as you? c For example Asian or Caucasian. Were born in the same country as you? d  47. Thinking about all of the ADULTS in your school, how many of the ADULTS would you say: Were born in another country? a For example, they were immigrants or refugees. Speak the same language at home B as you? Are the same ethnicity as you? C For example Asian or Caucasian. Were born in the same country as D you?  None of them  Mostly none  Half of Them  Most of them  All of them  I don’t know  None of them  Mostly none  Half of Them  Most of them  All of them  I don’t know  226  48.  Not at all true (1)  (2)  Some­ what true (3)  (4)  Completel y true (5)  a  I feel like a real part of my school. People here notice when I am good at b something. It is hard for people like me to be c accepted here. Other students in this school take my d opinions seriously. Most teachers at my school are e interested in me. Sometimes I feel as if I don’t belong f here. There’s at least one teacher or adult in g this school I can talk to if I have a problem. People in this school are friendly to h me. Teachers here are not interested in i people like me. I am included in lots of activities at my j school. I am treated with as much respect as k the other students. I feel very different from most other l students here. m I can really be myself at this school. I am treated with as much respect as o the other students. I feel very different from most other p students here. r The teachers here respect me. s People here know I can do good work. t I wish I were in a different school. u I feel proud of belonging to my school. Other students here like me the way I v am.  49. Please share any other comments you might have _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ _________________________________________________________________________________________________________ Thank you so much for your help!  ☺ 227  ***THIS PAGE IS OPTIONAL*** THIS MEANS THAT YOU ONLY NEED TO FILL IT OUT IF YOU WANT TO  If you tell us that you would like to see your school counselor and you write you name below, then we will pass your name and request for help on to your school counselor.  ***If you do not want to, then do NOT fill out this page*** □ YES, I would like to see my school counselor. Name: (PLEASE PRINT) ____________________ Grade: _______ Counsellor’s Name: ________________________ We will rip this page off and give your counselor your name. After that, this page will be destroyed, but we will keep the rest of the questionnaire (without your name attached).  228  APPENDIX D  229  Social Responsibility on the Internet Semi-Structured Interview Questions for Elementary Students As Victim 1. Have you ever been hurt by things people of done or said to you on the Internet? a. How does it usually happen? b. Why do you think it mostly happens over __________? 2. Do you know why these people said or did mean things to you online? a. Do you think they were mean on purpose? 3. Do you know the people who did it? 4. Are the people who have been mean online also mean to you at school? 5. Have you done anything to try and stop them from being mean? What have you tried? Did it work? 6. Do you think mean things that are said on the Internet are more hurtful, less hurtful or just as hurtful as mean things that are said in person? How come? As a Bully 7. Have you ever said or done mean things to people over the Internet? a. What does doing mean things on the Internet look like? b. Why have you done it? 8. Do you mostly do or say mean things online by yourself or with friends? Why? 9. Do you find it easier to say mean things online? Why? 10. Do you normally say mean things or talk behind people’s backs not on the Internet? 11. Think about the last couple times you have said mean stuff to people online, what was the reason? How did it start? 12. Do you think mean things that are said on the Internet are more hurtful, less hurtful or just as hurtful as mean things that are said in person? How come? 13. Is being mean online the same as being mean offline? 14. Would you say the same things online as you would offline? As a Witness 15. Have people ever sent you messages that were saying mean things about somebody else? What did you do with it? 16. Have you ever tried to stop the mean behaviour? What have you tried? Did it work? 17. Why do you think people say mean things online? 18. Any other comments?  230  Social Responsibility on the Internet Semi-Structured Interview Questions for Secondary Students As Victim 19. Have you ever been hurt by things people of done or said to you on the Internet? a. How does it usually happen? b. Why do you think it mostly happens over __________? 20. Do you know why these people said or did these hurtful or insulting things to you online? a. Do you think they were doing and saying these things on purpose? 21. Do you know the people who did it? 22. Are the people who have been intentionally hurtful to you online also hurtful to you at school? 23. Have you done anything to try and stop them from being mean or hurtful? What have you tried? Did it work? 24. Do you think hurtful or insulting things that are said on the Internet are more hurtful, less hurtful or just as hurtful as mean things that are said in person? How come? As a Bully 25. Have you ever said or done hurtful or insulting things to people over the Internet? a. What does doing mean things on the Internet look like? b. Why have you done it? 26. Do you mostly do or say hurtful, insulting, or threatening things online by yourself or with friends? Why? 27. Do you find it easier to say hurtful or threatening things online? Why? 28. Do you normally say insulting things or talk behind people’s backs when you are not on the Internet? 29. Think about the last couple times you have done insulting, hurtful or embarrassing things to people online, what was the reason? How did it start? 30. Do you think hurtful things that are said or done on the Internet are more hurtful, less hurtful or just as hurtful as mean things that are said in person? How come? 31. Is being hurtful or threatening online the same as being hurtful or threatening offline? 32. Would you say the same things online as you would offline? As a Witness 33. Have people ever sent you messages that were saying hurtful or embarrassing things about somebody else? What did you do with it? 34. Have you ever tried to stop this behaviour? What have you tried? Did it work? 35. Why do you think people say or do insulting or threatening things online? 36. Any other comments?  231  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0054370/manifest

Comment

Related Items