UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Bullying in schools : how children think, feel, and respond Funke, Kirstin A. 2003

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
[if-you-see-this-DO-NOT-CLICK]
ubc_2004-0068.pdf [ 5.69MB ]
Metadata
JSON: 1.0054505.json
JSON-LD: 1.0054505+ld.json
RDF/XML (Pretty): 1.0054505.xml
RDF/JSON: 1.0054505+rdf.json
Turtle: 1.0054505+rdf-turtle.txt
N-Triples: 1.0054505+rdf-ntriples.txt
Original Record: 1.0054505 +original-record.json
Full Text
1.0054505.txt
Citation
1.0054505.ris

Full Text

BULLYING IN S C H O O L S : HOW CHILDREN THINK, F E E L , AND R E S P O N D by KIRSTIN A. F U N K E B.Ed., The University of Manitoba, 1996 A THESIS SUBMITTED IN PARTIAL FULFILMENT O F THE R E Q U I R E M E N T S F O R THE D E G R E E OF M A S T E R O F A R T S in THE F A C U L T Y OF EDUCATION D E P A R T M E N T O F EDUCATIONAL AND C O U N S E L L I N G P S Y C H O L O G Y AND S P E C I A L EDUCATION We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December, 2003 © Kirstin A. Funke, 2003 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t t h e U n i v e r s i t y o f B r i t i s h C o lumbia, I agree t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e head o f my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n ot be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f /^d^ccchion^ Ia^o( C>^ni-< /li'n^ fsyeko103 y a^d £ f e c i * I £d"Cation The U n i v e r s i t y o f B r i t i s h C o lumbia Vancouver, Canada Date Oece.r~ /. ZL0O3> Abstract Research has frequently documented significant associations between empathy and behaviour. Typically, empathy has been found to be higher in individuals who behave prosocially, and lower in individuals who behave aggressively. Extending this research, the present study examined links between empathy and bullying/victimization behavior, with particular interest in whether children who behave differently in the face of bullying (i.e., assume different participant roles) differ in the nature of empathy they experience. To this end, 409 students (210 females, 199 males) from grades 5 to 7 completed both peer-nominations and self-reports of various bullying participant role behaviours, as well as self-reports of tendencies to experience various facets of empathy. Results revealed that self-reports of empathic concern (i.e., empathy in the form of concern for others), and perspective-taking (i.e., empathy in the form of understanding others' perspectives) were significantly lower in children nominated by peers for roles that supported bullying, as compared to children who assumed roles that defended against or stayed away from bullying or who were victims of bullying. No significant differences were observed among participant roles for empathic distress (i.e., empathy in the form of personal distress). This pattern was evident regardless of whether more generic or dispositional forms of empathy were considered, or whether more specific forms of empathy in response to bullying situations were considered. Overall, females reported higher levels of all facets of empathy than males. Given these findings, empathy appears to be an important distinguishing factor among children who behave differently within bullying situations. In order to encourage children's prosocial responses to bullying, and discourage behaviours that support bullying, empathy appears to be an important variable to target in designing anti-bullying programs for schools. Table of Contents Abstract ii Table of Contents iii List of Tables vi Acknowledgments ix Introduction P-1 Defining Empathy p. 3 Measuring Empathy p. 5 Linking Empathy to Behaviour p .8 Defining and Measuring Bullying p. 14 Participant Roles in Bullying p. 17 Empirical Evidence of an Empathy-Bullying Link p. 19 Statement of the Problem p. 24 Primary Research Questions p. 26 Question #1 p. 26 Question #2 p. 26 Question #3 p. 27 Secondary Research Question p. 28 Method p. 29 Participants p. 29 Procedures p. 31 Measures p. 32 Demographic Information p. 32 Bullying Behaviour p. 32 Dispositional Empathy p. 33 Results P- 35 Preliminary Analyses P- 35 Identifying Participant Roles P- 35 Classification into Participant Roles P- 43 Overlap Between Peer-Nominated and Self-Reported Participant Roles P- 45 Characteristics of Participant Roles P- 47 Characteristics of Empathy Scales P- 50 Primary Analyses P- 58 Overview P- 58 Question #1 P- 58 Question #2 P- 61 Question #3 P- 63 Analyses with Single-Item Understanding Factor P- 66 Secondary Analyses P- 68 Empathic Concern in Response to Bullying Scenarios P- 68 Empathic Distress in Response to Bullying Scenarios P- 71 Perspective-taking in Response to Bullying Scenarios P- 73 Discussion P- 77 Limitations of the Current Study P- 88 Future Applications for Interventions in Schools P- 91 Conclusion P- 94 References P- 95 Appendix A: Letter of Initial Contact & Informed Consent Form P- 107 Appendix B: Student Consent Form P- 109 Appendix C: Demographic Information Questionnaire p. 110 Appendix D: Participant Role Questionnaire (Peer) p. 111 Appendix E: Participant Role Questionnaire (Self) p. 119 Appendix F: Thoughts and Feelings Questionnaire p. 121 Appendix G : Factor Loadings by Gender for Peer-Nominations of Participant Roles p. 126 Appendix H: Factor Loadings by Gender for Self-Reports of Participant Roles p. 128 Appendix I: Factor Loadings by Gender for Primary Empathy Scales p. 130 Appendix J : Factor Loadings by Gender for Secondary Empathy Scales p. 132 VI List of Tables Table 1. Distribution of Participants by Grade, Gender, and Mean Age. p. 30 Table 2. Factor Loadings for Pro-bully, Defender, Victim, and Outsider Factors (Peer-Nominations) p. 36 Table 3. Factor Loadings for Pro-bully, Defender, Victim, (Single Item) Reinforcer, and Outsider Factors (Self-Ratings) p. 39 Table 4. Correlations Between Averages of Peer-Nominations and Self-Reports in Participant Roles p. 42 Table 5. Distribution of Participants by Participant Role (Peer- and Self- Reported) p. 44 Table 6. Proportion of Self-Reported Participant Roles (%) that Matched Peer-Nominated Participant Roles p. 46 Table 7. Distribution of Participants Among the Four Peer-Nominated Participant Roles by Gender, Grade p. 48 Table 8. Factor Loadings for Empathic Distress, Perspective-taking, Empathic Concern, and Understanding (Whole Sample) p. 53 Table 9. Correlations Between Primary Empathy Subscales p. 54 Table 10. Factor Loadings for Bullying-Specific Perspective-taking, Empathic Distress, and Empathic Concern (Whole Sample) p. 56 Table 11. Correlations Between Bullying-Specific Empathy Subscales p. 57 Table 12. Correlations Between Primary and Bullying-Specific Empathy Subscales p. 57 Table 13. Mean Levels of Empathic Concern (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample p. 59 Table 14. Mean Levels of Empathic Concern (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample p. 60 Table 15. Mean Levels of Empathic Concern (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample p. 61 Table 16. Mean Levels of Empathic Distress (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample p. 62 Table 17. Mean Levels of Empathic Distress (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample p. 62 vii Table 18. Mean Levels of Empathic Distress (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample p. 63- Table 19. Mean Levels of Perspective-taking (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample p. 64 Table 20. Mean Levels of Perspective-taking (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample p. 65 Table 21. Mean Levels of Perspective-taking (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample p. 66 Table 22. Mean Levels of Understanding (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample p. 67 Table 23. Mean Levels of Understanding (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample p. 67 Table 24. Mean Levels of Understanding (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample p. 68 Table 25. Mean Levels of Empathic Concern in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample p. 69 Table 26. Mean Levels of Empathic Concern in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample p. 70 Table 27. Mean Levels of Empathic Concern in Response to Bullying Scenarios (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample p. 71 Table 28. Mean Levels of Empathic Distress in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample p. 71 Table 29. Mean Levels of Empathic Distress in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample p. 72 Table 30. Mean Levels of Empathic Distress in Response to Bullying Scenarios (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample p. 73 Table 31. Mean Levels of Perspective-taking in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample p. 74 viii Table 32. Mean Levels of Perspective-taking in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample p. 75 Table 33. Mean Levels of Perspective-taking in Response to Bullying Scenarios (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample p. 76 ix Acknowledgments I would like to thank my supervisor, Dr. Shelley Hymel for her amazing attention to detail, reassurance through all stages of the thesis process, and above all, for teaching me how to think like a researcher. I would also like to express my thanks to my committee, Dr. Kimberly Schonert-Reichl, and Dr. Marion Porath for their involvement and support. This study would not have been possible without the participation and interest of students and staff from my schools. I would especially like to thank the school principals, Heather Bronson, Bob Padgham, Ron Sandland, Marie Harel, and Roger Douglas. Thank-you not only for taking part, but also for taking a keen interest in the content of this study. I have relied on the guidance and favours of close friends throughout the writing of this thesis. Tanya, your generosity covered such a wide range, I could never thank you enough. You returned library books, helped me with statistics, provided me with accommodation when I came to campus, and helped with proofreading. Thank-you from the bottom of my heart. Karen, you have heard enough about this thesis that you could probably have written it yourself. Thank-you for being the greatest friend and support throughout the process. I could never have produced this thesis without the support of my family. To my sisters and parents, thank-you for all the late night phone calls, and your constant words of encouragement. Jim, thanks for being my biggest fan. I know the completion of this thesis is like a personal victory for you too. 1 Introduction The phenomenon of bullying among school-aged children presents a prototypic opportunity for witnesses to experience empathy. Studies of empathy, defined as one's emotional reaction when witnessing others in distress, has repeatedly demonstrated significant associations with behaviour (Hoffman, 2001). Specifically, empathy has been found to be higher in individuals who behave prosocially (for a review, see Eisenberg & Miller, 1987), and lower in individuals who behave aggressively (for a review, see Miller & Eisenberg, 1988). Both of these associations are relevant to the phenomenon of bullying in that various children are present when bullying occurs (O'Connell, Craig, & Pepler, 1999), including those who opt to behave aggressively (i.e., actively bully others, support a bully), those who behave prosocially (i.e., defend victims), and those who remain uninvolved (Salmivalli, 1999). The study of empathy has not previously been applied to all children who are present when bullying occurs. When empathy or related processes have been considered with respect to the phenomenon of bullying, the focus has been on the extent to which individuals perpetuate the victimization of others, rather than on other behaviours enacted by children who are present when bullying occurs (e.g., Endresen & Olweus, 2001). This is a critical gap in the research, given that peers are present in 85% of bullying episodes, yet they only intervene 11 to 13% of the time (Craig & Pepler, 1997). Accordingly, interventions involving peers have been identified as one of the most important avenues for bullying interventions (Salmivalli, 1999). The need for effective bullying interventions is particularly critical, given that both bullies and victims are at risk for a number of school-related and psychosocial difficulties (Pellegrini, 1998). Significant relationships have been demonstrated between bullying behaviour at school age, and increased risk for later criminal behaviour (Olweus, 1992), involvement with substance abuse, theft, property damage, violation of parents' rules (Haynie et al., 2001), and an inability to develop and maintain positive interpersonal relationships (Oliver, Hoover, & Hazier, 1994). 2 The devastating effects of bullying on its victims include increased depression and lower self- esteem in adulthood (Olweus, 1993), loneliness, unhappiness at school, having few good friends, and avoidance of school (Batsche & Knoff, 1994; Boulton & Underwood, 1992; Olweus, 1991). The current study takes a closer look at empathy among elementary school-aged children, and compares empathy in children who assume various behavioral roles within scenarios of bullying. Specifically, the study answers the question: Are there differences in the empathy of children who behave in different, and in some cases distinctly prosocial or aggressive ways in response to bullying? As revealed in the literature review that follows, there is some evidence to suggest that specific facets of empathy are differentially linked to behaviour. Accordingly, children who behave differently in the face of bullying may in fact exhibit (or fail to exhibit) empathy in ways that contribute to their distinct behavioural roles. The remainder of this thesis is organized according to the following sequence of topics. First, a thorough review of relevant literature is presented regarding (a) the study of empathy, including problems of definition and measurement; (b) the relevance of empathy to various forms of behaviour; (c) a definition of bullying and consideration of measurement issues; (d) the participant role approach to studying bullying; and (e) empirical evidence examining the link between empathy and bullying. The review concludes with a summary of how the extant literature points to the need for the current study, as well as an overview of the present study and the hypotheses investigated. Following this review is a discussion of the methodology used in the present study, and a presentation of the analyses used to examine the research questions, along with the results obtained. Finally, the implications, limitations, and avenues for future exploration that have arisen in light of the current data are presented. 3 Defining. Empathy Reviews of the evolution of the construct of empathy in research (e.g., Davis, 1994; Duan & Hill, 1996; Eisenberg & Strayer, 1987; Goldstein & Michaels, 1985; Omdahl, 1995), have identified a broad range of definitions across studies. In fact, as Omdahl (1995) notes, many researchers have conducted studies under the premise of empathy, but have actually measured constructs that they define in mutually exclusive ways. For example, some researchers have defined empathy as a primarily affective process, concerned with a person's emotions in response to observing another person in an aroused state (e.g., Stotland, 1969). Other researchers have studied empathy as a cognitive process, concerned with a person's ability to understand the emotions and general condition of another person (e.g., Borke, 1971; Dymond, 1949; Hogan, 1969). More recently, researchers have investigated empathy as a multidimensional process involving both affect and cognition, essentially a combination of feeling what another feels, and understanding how another feels (e.g., Davis, 1983, 1994; Davis, Hull, Young, & Warren, 1987; Deutsch & Madle, 1975; Feshbach, 1975, 1982; Hoffman, 1975, 1984, 1987, 2001). Acknowledging the interplay of both affect and cognition in empathy (e.g., Eisenberg, 2000; Endresen & Olweus, 2001; Olweus & Endresen, 1998), the current study considers an empathic response to occur when one experiences an emotional response to another (target) person's emotional state. The affective response may be congruent with (i.e., if target is sad, person feels bad), or identical to a target's state (Eisenberg, 2000; Mussen & Eisenberg, 2001). Affective empathy responses may also be either other- or self-oriented (Eisenberg, 2000). For example, in the case of witnessing another in distress, one may experience other-oriented feelings of sorrow or concern (i.e., empathic concern), or one may experience self-oriented feelings of anxiety or distress (i.e., empathic distress). In comparison to self-oriented feelings of distress (herein referred to as empathic distress), other-oriented feelings of concern (herein 4 referred to as empathic concern) are thought to represent a more mature form of empathy (Hoffman, 2001). In his influential model of the development of empathy, Hoffman (1984, 1987, 2001) delineated a series of levels through which humans progress as one's cognitive sense of self and others develops. The levels are based upon humans' experiences as bystanders, when they witness others in distress. Empathy is first evident as a primitive, global form that occurs when a newborn baby hears another's distressful cry, and responds by crying himself/herself. The newborn's cry has been shown to be more than mere imitation or response to a noxious stimulus. Rather, it is a distressful, vigorous cry that is identical to that of the stimulus infant (Sagi & Hoffman, 1976). More advanced forms of empathy are believed to develop as a child learns to recognize others' feelings, and learns that these inner states are independent of his/her own. The most advanced form of empathy typically occurs around age 9 when a child realizes that people have stable identities that are based upon personal factors such as gender and ethnicity and that these identities may affect one's empathy for others (e.g., one might feel more empathy for someone who is similar to oneself). As Hoffman's model illustrates, developmentally mature empathic responses (i.e., empathic concern) are enhanced by one's cognitive understanding of the target's condition or state. Yet, rudimentary forms of empathy are believed to occur without cognitive mediation (Hoffman, 2001). Perspective-taking is a critical cognitive process involved in the experience of empathic concern. When one imagines oneself in someone else's place and subsequently feels concern for the other, perspective-taking has aroused empathy (Hoffman, 1984). Much research has focused on this cognitive process (e.g. Borke, 1971; Dymond, 1949; Hogan, 1969). Although many researchers have used the term perspective-taking interchangeably with empathy, or used terms like cognitive empathy, Hoffman (2001) proposes that it is merely one potential means by which empathy may be aroused. Empathy may occur, albeit in a more primitive form, 5 without an individual having yet achieved the ability to assume another's perspective. That is, one may experience empathic distress, but not empathic concern without first being able to adopt another person's perspective. Despite the apparent clarity of Hoffman's model, what remains to be proven empirically is whether perspective-taking ability necessarily leads to empathic concern. That is, could it be that some individuals can experience high levels of perspective-taking, but not feel high levels of empathic concern? Clearly, any assessment of empathy must address not only affective components such as feelings of concern and distress, but also the influential cognitive component of perspective-taking. People are not universal in their tendencies to experience empathy (i.e., an emotional reaction to another's emotional state). Instead, people display noteworthy individual differences in empathy (Davis, 1994). These predispositions are thought to be based upon "temperamental emotionality", or a "general tendency to experience strong emotional responses" (Olweus & Endresen, 1998, p. 371). Indeed, research on the relation between emotionality and empathy has demonstrated that individuals who tend to experience intense negative emotions are those who experience more empathy for others in distress (Eisenberg et al., 1994). Specifically, those who maintain their emotional arousal at a level that is not aversive in empathy-inducing contexts are likely to experience empathic concern. In contrast, those who become emotionally over- aroused and experience their emotion as aversive subsequently experience empathic distress (Eisenberg et al., 1998). Given the apparent links of empathy to dispositional factors such as emotionality and emotion regulation, empathy may be conceptualized as a trait-like tendency to experience emotional responses to other people's emotional states. Measuring Empathy Measures that assess empathy as a trait address dispositional tendencies to be empathic across settings (Davis, 1994). This is typically achieved by one of two methods. The first method involves ratings by others, such as parents and teachers (e.g., Barnett, Howard, 6 Melton & Dino, 1982; Eisenberg, et al., 1998), or peers (e.g., Bjorkvist, Osterman & Kaukiainen, 2000; Kaukiainen et al., 1999). Other-reports require individuals with extensive knowledge of target individuals to estimate the extent to which the targets are empathic across settings. This method has not been used extensively, perhaps because empathy is a highly internal and personal experience that may not always be apparent to others. In fact, close inspection of a recent peer-report measure used to assess empathy (Kaukiainen et al., 1999) revealed that ratings are often based upon observable behaviours that are assumed to be motivated by empathy, rather than actual experiences of empathy (for example, "Helps classmates in trouble", "Comforts others when they are sad", Kaukiainen et al., 1999, p. 85). Given their reliance on ratings based on observable behaviours, or ratings based on inferences about how another feels, other-reports of empathy may not permit access to the personal, internal nature of the experience of empathy. The second of the dispositional measurement methods, the self-report questionnaire (e.g., Bryant, 1982; Davis, 1983; Mehrabian & Epstein, 1972), requires that individuals report their own tendencies to experience empathy for others. Although this method was initially designed for adults (Davis, 1983; Mehrabian & Epstein, 1972), it was later adapted for children (Bryant, 1982; P. Miller, personal communication, March 20, 2002). Participants are typically presented with empathy-evoking scenarios, and are asked to rate the extent to which they would feel a certain way in that scenario (e.g., "I get upset when I see a girl being hurt", Bryant, 1982). Although self-report, dispositional measures provide indices of empathy across various circumstances, they may also be used to ascertain empathy under specific circumstances. For example, the sex of the stimulus person in the scenarios may be manipulated (e.g., Bryant, 1982), or a specific type of scenario (e.g., bullying) may be described (e.g., Endresen & Olweus, 2001). In addition, questions may be designed to specifically address various components of empathy (e.g., other-oriented empathic concern, self-oriented empathic distress, perspective- 7 taking). For example, the item, "Seeing a boy who is sad makes me want to comfort him" is clearly other-oriented, and contrasts with the self-oriented item, "It often makes me distressed when I see something sad on TV" (Endresen & Olweus, 2001, p. 151). Self-report questionnaire measures are most applicable to the current study, for many reasons. First, in order to assess and differentiate between the specific facets of empathy of interest to the current study (i.e., empathic concern, empathic distress, perspective-taking), a self-report questionnaire measure directly addresses each of these components. Second, unlike younger children (i.e., preschoolers), late elementary school children are quite capable of completing a self-report questionnaire that involves some reading. As Bryant (1982) found, measures of reading ability were not found to significantly correlate with fourth and seventh- grade children's scores on a questionnaire measure of empathy. Third, questionnaires are convenient to administer to large groups. Indeed, self-report questionnaires are the preferred method of data collection with school-age children and adolescents (Olweus & Endresen, 1998). Despite the aforementioned advantages, self-report measures of empathy have been criticized in that a respondent can misrepresent himself or herself, responding in a socially desirable manner (Eisenberg, 1986). Links between self-reports of empathy and social desirability, however, have not been demonstrated empirically (e.g., Bryant, 1982; Mehrabian & Epstein, 1972). In contrast to dispositional measures, situational-specific measures of empathy have been developed to assess people's responses to very realistic, controlled stimuli. Situational measures of empathy are often obtained via experimental manipulations in a laboratory or other contrived setting. Examples of techniques include asking children to report their feelings in response to an emotionally evocative story or video (e.g., Cohen & Strayer, 1996; Feshbach & Roe, 1968; Hughes, Tingle & Sawin, 1981; Strayer, 1993), rating individuals' facial expressions, tone of voice, or gestures while they are exposed to an emotionally arousing situation (e.g., 8 Eisenberg et al., 1994; Eisenberg et al., 1988; Marcus, Roke, & Bruner, 1985), or monitoring individuals' physiological changes, such as heart rate (e.g., Eisenberg et al., 1988; Eisenberg et al., 1998) in response to arousing (usually distressful) stimuli. Another related technique, experimental induction, involves construing the experimental setting in a manner that is assumed to induce empathy (e.g., telling the child to imagine what it would feel like in a needy other's position), so that its impact on behaviour may be measured (Eisenberg & Miller, 1987). Although very specific, controlled empathy-evoking stimuli were not used in the current study, they have been used in many other empirical investigations. The results of these studies form an important part of the existing research linking empathy to various forms of behaviour, as reviewed in the next section. With regard to the present study, of primary interest were the empathic tendencies of children across settings, reflecting dispositional empathy. Dispositional empathy is best assessed by posing many different, empathy-evoking scenarios to respondents (Davis, 1994). However, of secondary interest was the nature of empathy in scenarios where bullying occurs, reflecting more context-specific empathy. To this end, children's empathy was also assessed using hypothetical bullying scenarios, allowing for a comparison of children's empathy when bullying occurs, versus more general situations. Linking Empathy to Behaviour The distinction between self-oriented empathic distress and other-oriented empathic concern is of particular importance for the motivational and behavioural consequences of empathy. While both have historically been thought to inhibit aggressive behaviour, and encourage prosocial behaviour (Davis, 1994), recent empirical study has sought to differentiate between the behaviour-motivating capacities of each. Prosocial behaviour is defined as "voluntary behaviour intended to benefit another" (Mussen & Eisenberg, 2001). The relationship between empathy and prosocial behaviours 9 such as altruism and helping has been widely investigated (see Eisenberg, 2000; Eisenberg & Miller, 1987 for reviews). Essentially, the proposed link between empathy and prosocial behaviour is that when witnessing someone in distress, "vicariously induced arousal generated from apprehension of the other's emotional state (or general situation) produces sympathetic concern for the other, aversive arousal within the self, or both" (Eisenberg & Miller, 1987, p. 92). It is further assumed that out of concern for the other (i.e., empathic concern), or in an effort to reduce personal feelings of discomfort (i.e., empathic distress), an individual will be motivated to help the person in distress (Davis, 1994). Unfortunately, interpretation of much of the existing research on empathy and prosocial behaviour is clouded by the fact that (a) empathy is defined as either empathic concern or distress, or both, (b) empathy has been measured using a variety of methods, and (c) empathy research has involved samples of different ages (Eisenberg, 1986). Nevertheless, in Eisenberg and Miller's meta-analysis (1987) an overall positive, significant association was found between dispositional empathy (defined as either concern, distress, or both) and prosocial behaviour. This association was evident even when the meta-analysis was re-calculated using only samples of children and adolescents. Certainly, this result supports the position that prosocial behaviour increases along with increased affective empathy. When situational measures were employed, only studies that assessed empathy by rating children's facial expressions in response to distressful stimuli still demonstrated significant, positive relationships with prosocial behavior (Eisenberg & Miller, 1987). Batson (1998) proposed that affective empathic arousal in the form of empathic concern that is focused on others (i.e., other-oriented) is associated with an altruistic motive to help others. He contrasted this with self-focused, empathic distress reactions that may only be linked to helping others in order to relieve the personally aversive state that one is experiencing. In the case of distress, if escape is possible, it may be more likely than helping (Hoffman, 2001). 10 Research with child samples has supported Batson's theory, demonstrating that empathic concern typically leads to helping, whereas empathic distress typically does not (Eisenberg, 2000). For example, Knight, Johnson, Carlo, and Eisenberg (1994) found that children aged 6 to 8 years who reported high tendencies to experience empathic concern for others more frequently helped a peer in need than children who reported low tendencies to experience empathic concern. Similarly, Carlo, Roesch, and Melby (1998) found that adolescents who reported high dispositional empathic concern also reported that they frequently helped others. Fabes, Eisenberg, and Eisenbud (1993) found that boys in grades 3 and 6 who reported high tendencies to experience empathic concern were described as helpful by their parents. In contrast, negative or non-significant relations have been demonstrated between empathic (personal) distress and prosocial behaviours in children. For example, Fabes et al. (1993) found that girls in grades 3 and 6 who experienced high levels of self-focused, empathic distress were not described by their parents as helpful. Similarly, Miller, Eisenberg, Fabes, and Shell (1996) found that preschool children who experienced high levels of self-focused distress when exposed to others in distress, tended to offer help to another child less often than those preschoolers who experienced high levels of other-oriented concern. In addition to the relation of affective empathic processes to prosocial behaviours, the empathy-related cognitive process of perspective-taking has also been investigated with respect to prosocial behaviours. Reviews of this research (Davis, 1994; Eisenberg, 2000; Eisenberg & Miller, 1987; Underwood & More, 1982) provide summative evidence of a positive association between cognitive empathy (i.e., perspective-taking) and prosocial (e.g., helping) behaviour. That is, as perspective-taking increases, so does the likelihood of behaving prosocially. For example, 7-year-old children who were able to anticipate others' perspectives in conversation in a non-egocentric way tended to share and help others (Rubin & Schneider, 1973). College students who were instructed to imagine how a target person was feeling, were observed to 11 offer help to that person more frequently than comparison participants who were not given instructions (Batson et ai., 1989). Other studies have found similar results (e.g., Batson et al, 1991; Dovidio et al., 1990). Much empirical attention has been dedicated to exploring the link between empathy and aggression (for a review, see Miller & Eisenberg, 1988). The experience of empathy is thought to inhibit aggression by allowing an individual to share in the victim's emotional state and context, thereby inhibiting individuals from continuing to aggress (Feshbach, N., 1987). Given this, it should be expected that those individuals who are highly aggressive are low in their tendency to experience affective empathy (Cohen & Strayer, 1996). Similar to the research on empathy and prosocial behaviour, studies linking empathy to aggression are plagued by inconsistent operationalizations and measurement difficulties (Miller & Eisenberg, 1988). Nevertheless, results of a meta-analysis (Miller & Eisenberg, 1988) of studies that examined the relationship between dispositional measures of affective empathy (including both empathic concern and empathic distress measured by a variety of methods) and aggressive behaviour revealed a significant, negative relationship. The more affective empathy reported, the less aggressive behavior. It should be noted that none of the overall relationships between empathy as assessed by situational-specific measures and aggression were significant. Multiple studies have found significant negative associations between dispositional empathy (measured by questionnaires) and aggression in children and adolescents (Eisenberg, 2000). For example, peer-ratings of dispositional empathy (not clearly distinguishing empathic concern and empathic distress) correlated negatively and significantly with verbal, physical, and indirect forms of aggression as rated by peers in 10 to 14-year olds 1 (Kaukiainen et al., 1999). Similarly, self-reported dispositional empathic distress was found to correlate negatively with 1 Indirect aggression in 12-year olds was an exception, as the negative correlation did not reach statistical significance. 12 teacher-rated physical and verbal aggression in first and fourth-grade boys (Bryant, 1982). In their work with adolescents, Carlo et al. (1998) found a significant, negative relationship between self-reported, dispositional empathy (considered as a composite of empathic concern and perspective-taking) and self-rated aggressive tendencies. Since the publication of the meta-analysis described above, research evidence with child and adolescent samples (e.g., Cohen & Strayer, 1996; Endresen & Olweus, 2001) increasingly points to empathic concern as the aspect of affective empathy that is most strongly associated with reductions in aggressive behaviour. For example, a recent study that examined the relation between adolescents' self-reported, dispositional empathic tendencies and bullying behavior (a unique form of aggression), demonstrated that it was empathic concern, and not empathic distress, that was significantly, negatively related to bullying others (Endresen & Olweus, 2001). Another study (Cohen & Strayer, 1996) found that, although aggressive, delinquent adolescents scored lower than comparison, non-aggressive youth on a measure of dispositional empathic concern, they actually scored higher on a measure of dispositional empathic distress. In this study, not only was empathic concern associated with less aggression, but empathic distress was associated with more aggression. Clearly, the nature of the link between dispositional empathic concern, empathic distress, and behaviour (both aggressive and prosocial), requires clarification. For this reason, it was important to assess the tendencies to experience both empathic concern and distress, within the current study. In addition to affective empathy-aggression links, the cognitive process of perspective- taking is considered by some (e.g., Richardson, Hammock, Smith, Gardner, & Signo, 1994) to be linked to the inhibition of aggression. The theory behind this link is that an enhanced ability to adopt others' viewpoints makes it more difficult to aggress against others. There is some research evidence on perspective-taking and aggression that has examined this link using dispositional measures. Research conducted with college students (Richardson et al., 1994) 13 demonstrated that self-reported, dispositional tendencies to take the perspectives of others were negatively and significantly related to self-reported tendencies to behave aggressively. Instructing participants to take the perspectives of others in experimentally manipulated aggressive scenarios also reduced aggressive responses in experimental tasks (Richardson et al., 1994). Letoumeau (1981) found that mothers who used physical aggression as a form of abuse of their children scored significantly lower than comparison (i.e., non-abusive) mothers on a self-report measure of the tendency to adopt others' perspectives. Cohen and Strayer (1996) found that among aggressive, conduct-disordered adolescents, the self-reported tendency to adopt the perspectives of others was much lower than that of comparison youth. Although the results described above apply to the particular populations examined (e.g., randomly sampled college students, conduct-disordered youth), they reveal important evidence of a link between perspective-taking and aggression. Certainly, perspective-taking is an influential process in the motivation of behaviour, and was important to directly assess in the current investigation of empathy and behaviour. In sum, there is ample evidence of a significant relationship between empathy and behaviour (Eisenberg, 2000). In particular, the experience of empathic concern appears to be linked to greater prosocial, and less aggressive behaviour. By comparison, experience of empathic distress appears to be associated with increased prosocial behaviour only to a limited extent (i.e., when escape of the aversive situation is not possible). The relationship between empathic distress and aggression is less clear, but some evidence that increased distress leads to increased aggression does exist (e.g., Cohen & Strayer, 1996). Additionally, skill in perspective-taking appears to be critically linked to behaviour, although the nature of the relationship is less clear than that observed with affective components of empathy. Given the interest of the current study in comparing the nature of empathy in children who behave in various ways, it was critical to assess three distinct facets of empathy: empathic concern, 14 empathic distress, and perspective-taking. The differential nature of affective and cognitive components of empathy in children who assume various behavioural roles within bullying scenarios has not yet been empirically investigated. For this reason, it is critical to turn now to a discussion of the phenomenon of bullying, in order to demonstrate the importance of the empathy-bullying link. Defining and Measuring Bullying Childhood bullying, also known as peer harassment or peer victimization, is a unique form of aggression in which repeated, negative actions of either a physical, psychological, and/or sexual nature, are carried out by one or more students who are perceived to hold power over the victim (Olweus, 1993, 1999; Smith & Brain, 2000). A number of critical components within this definition distinguish bullying from other forms of aggression and antisocial behaviours, including: (a) the purposefulness (intentionality) of the harm, (b) the establishment of bully power over the victim, and (c) the repetition of aggression over time. Typically, bullying is manifested in three major forms: verbal (e.g., name-calling, taunting), physical (e.g., hitting, pushing), and indirect or relational (e.g., spreading rumours, excluding socially) (Bjorkvist, Lagerspetz & Kaukiainen, 1992). In reviewing the research methods most often used to sample school bullying/victimization, three groups of predominant measures emerge: observational, ipsative, and normative methods (Pellegrini, 2001; Pellegrini & Bartini, 2000). Observations are often conducted in hallways and classrooms, or on school playgrounds (e.g., Craig & Pepler, 1997; O'Connell, Pepler, & Craig, 1999; Pellegrini, 2001; Pepler, Craig & Roberts, 1998). The greatest advantage of an observational method is its relatively unbiased account of bullying as it occurs in naturalistic situations (Pellegrini, 2001). Despite the face-validity of observational methods, they require a great deal of time to obtain an adequate sample of behaviour across time and settings (Pellegrini & Bartini, 2000). In addition, observations require training for those 15 who observe, sample, and code behaviours (Pellegrini, 2001). A further limitation of observational methods is that they are often limited to public settings where outside observers may be present. This is a considerable disadvantage, because research shows that the majority of bullying occurs in locations away from adult witnesses (Boulton & Underwood, 1992; Craig & Pepler, 1997; Whitney & Smith, 1993), and takes on covert and indirect forms that are difficult for adults to detect (Crick & Grotpeter, 1995; Osterman et al., 1994). For these reasons, direct observation of bullying behaviour is often difficult. Given the limitations of observational methods, including issues such as complex administration and time requirements, as well as difficulty in accessing incidents of bullying, the sampling of peer victimization/bullying most frequently involves informants' ratings (Pellegrini & Bartini, 2000). Ipsative methods of assessing bullying involvement are based on individuals' perceptions of bullying involvement (Pellegrini, 2001). Most ipsative measures are questionnaires that begin with a clear operational definition of what bullying is, followed by questions related to both bullying others and being bullied by others (e.g., Whitney & Smith, 1993). Parent-reports are ipsative measures that require parents to report their perception of their child's involvement in bullying/victimization. Such measures have not been widely used in bullying research, likely because they have been found to provide little unique information when compared with self and peer-reports (Ladd & Kochenderfer-Ladd, 2002, Study 2). Much research on bullying has made use of self-report data (e.g., Charach, Pepler & Ziegler, 1995; Nansel et al., 2001; Oliver, Hoover & Hazier, 1994; Olweus, 1991, 1993; Rivers & Smith, 1994; Roland, 2002; Whitney & Smith, 1993), often collected in anonymous questionnaires following the research of Dan Olweus of Norway (1978, 1991, 1993). Self-report measures are advantageous because they provide personal perspectives based on a broad range of direct experiences, including those which foster strong emotional reactions (e.g., when victimized) and therefore, vivid memories (Ladd & Kochenderfer-Ladd, 2002). Self-reports are 16 also an efficient means of gathering a large and detailed quantity of information from many individuals. However, self-report measures may reflect a "systematic bias" in how the self is presented (Pellegrini, 2001). For example, a student may respond in a socially desirable manner to project a positive self-image (e.g., to avoid identifying oneself as a bully (Pellegrini & Bartini, 2000). Similarly, a student may be reluctant to admit to being victimized, due to embarrassment. For these reasons, self-report ratings may underestimate the extent of bullying (Pellegrini & Bartini, 2000). In contrast, self-reports may actually inflate the extent of victimization that occurs, due to individuals' perceptions that they are being victimized, rather than actual incidence (Juvonen et al., 2001). Despite pros and cons, self-report measures are likely to continue to be used in bullying research, given their efficiency and ease of administration. In order to address the limitations discussed above, self-reports are often collected along with reports from others, so that similarities and differences between sources can be compared (e.g., Juvonen et al., 2001). Normative methods tap group perceptions of those who bully and those who are victimized, through ratings or nominations relative to peers or students in a classroom (Pellegrini & Bartini, 2000). Both peers (e.g. Perry, Kusel, & Perry, 1988) and teachers (e.g., Sutton, Smith, & Swettenham, 1999a) are popular sources of information for normative methods. Although teacher-reports are often used to assess students' behaviour (e.g., Coie & Dodge, 1988; Sutton, Smith, & Swettenham, 1999a), there are a number of threats to their reliability and validity. Examples include a limited range of opportunities to witness bullying interactions between children (as compared to children, themselves), reporting biases (e.g., underreporting to avoid admitting their inadequate supervision), and relational biases (e.g., halo effects based on prior knowledge and expectations of individual students) (Ladd & Kochenderfer-Ladd, 2002). In the case of peer-nomination methods, students are typically provided with a class list (e.g., Salmivalli, Karhunen, & Lagerspetz, 1996), or a picture of their entire class (e.g., 17 Osterman et al., 1994), and are asked to select from their class individuals who fit particular behavioural criteria. Students are then classified into specific bully roles (e.g., bully or victim) on the basis of specified cut-off scores for the number of nominations in each role. Peer- nomination measures are popular in childhood bullying research (e.g., Camodeca, Goossens, MeerumTerwogt, & Schuengel, 2002; Salmivalli & Nieminen, 2002) as they have the particular advantage of providing the group view of behaviour, based upon the aggregated nominations of many individuals. Another advantage of peer-nomination measures is that they are based upon accumulated observations over extended periods, providing a wide sampling of bullying behaviours across various circumstances (Pellegrini, 2001). Peer-nominations are also a useful measure of bullying behaviours that occur away from adults (Ladd & Kochenderfer-Ladd, 2002). A potential disadvantage of peer-nominations is that they may reflect a child's established social reputation (e.g., as a bully or victim) rather than whether or not he or she actually behaves in particular ways related to bullying/victimization (Juvonen et al., 2001). Given this, despite many benefits of peer-nominations, the recommended method for sampling bullying-related behaviours is to obtain separate indices from both self- and peer-reports, so that the two can be compared (Juvonen et al., 2001; Ladd & Kochenderfer-Ladd, 2002; Pellegrini, 1998, 2001; Pellegrini & Bartini, 2000). Considering evidence that correlations between peer and self-report measures are typically modest, ranging from .18 to .50 (Crick & Bigbee, 1998; Ladd & Kochenderfer-Ladd, 2002, Study 2; Osterman et al., 1994; Pellegrini, 2001; Pellegrini & Bartini, 2000), it appears as though each type of measure provides some unique, non-overlapping information. For this reason, both self-reports and peer-nominations were collected within the current study, so that results from both sampling methods could be compared. Participant Roles in Bullying The importance of identifying and studying the roles that all students play, particularly in supporting or opposing bullying, is highlighted by research evidence that shows that peers are 18 present in 85 to 88% of bullying episodes, but only intervene 11 to 13% of the time (Craig & Pepler, 1997). Additional research shows that although most students possess anti-bullying attitudes (Whitney & Smith, 1993), in the face of bullying the majority of them acts in ways that are prone to maintain bullying, rather than counter it (Salmivalli, 1999). Given the presence of peers during much bullying, it is imperative to identify and understand what roles peers play in the bullying process, so that their influences on the proliferation, or discontinuation of bullying may be understood. The roles typically assumed by children when bullying occurs have been systematically identified and studied (Salmivalli, 1999). Apart from the more common practice within peer nominations of differentiating students simply on the basis of bully or victim status, researchers have further delineated the roles involved in bullying (Salmivalli, Lagerspetz, Bjorkvist, Osterman, & Kaukiainen, 1996). Salmivalli, Lagerspetz, et al. (1996) developed a peer-report instrument known as the Participant Role Questionnaire (PRQ). The scale consists of 22 examples of behaviours that children are required to consider, and determine the extent to which the descriptions match the behaviour of their peers (Salmivalli, Lappalainen, & Lagerspetz, 1998). The P R Q consists of six subscales, each assessing a specific role within bullying scenarios. The subscales include roles that are supportive of bullying, including the Bully role, which is defined by behaviour that is "active, initiative-taking, (and) leader-like" (Sutton & Smith, 1999, p. 98), the Assistant role that is "active, but more follower than leader- like" (Sutton & Smith, 1999, p. 98), and the Reinforcer role that involves "providing an audience, and inciting the bully" (Sutton & Smith, 1999, p. 98). The Defender role is a distinctly prosocial role, characterized by "sticking up for or consoling the victim" (Sutton et al., 1999a, p. 438). In contrast, the Outsider role represents a benign, uninvolved role, described as "doing nothing in bullying situations, staying away" (Sutton et al., 1999a, p. 438). Identification of Victims is also often included with the P R Q , by simply asking students to name who gets bullied. The peer- 19 nomination format of the P R Q has often been accompanied by self-nominations for the various participant roles (e.g., Sutton & Smith, 1999). Research conducted using peer-nominations from the P R Q (Salmivalli et al., 1998) has found that approximately 8 to 10% of students in grades 6 and 8, respectively, were identified as Bullies, 6 to 13% as Assistants, 17 to 16% as Reinforcers, 17 to 20% as Defenders, 26 to 30% as Outsiders, and 11 to 5% as Victims. Similar role distributions were expected in the present sample, considering that children of a similar age group were involved. Given the interest of the current study in examining empathy in all children who are present when bullying takes place, including distinctly aggressive (i.e., Bully role), distinctly prosocial (i.e., Defender role), and all other children who witness bullying episodes (including Victims), the Participant Role approach was especially appropriate. In consideration of the aforementioned importance of collecting separate peer- and self- reports of bullying-related behaviours, both peer-nominations of participant role behaviour, as well as parallel self-ratings of participant role behaviour were included in the current study. It was hoped that gaining an understanding of how children who are present when bullying occurs tend to feel and think when they see others in distress (i.e., how they experience empathy) would provide valuable insight into the phenomenon of bullying. Indeed, the body of research that has previously examined empathy and bullying, or closely related constructs, suggests that the relationship between empathy and the phenomenon of bullying is a significant one. Empirical Evidence of an Empathy-Bullying Link Research linking empathy to the phenomenon of bullying (e.g., Endresen & Olweus, 2001) has typically focused on the perpetration of bullying (i.e., the Bully role), to the exclusion of other participant roles assumed by children within bullying episodes. In addition, most research (e.g., Sutton et al., 1999a) has dealt with only affective or cognitive facets of empathy, rather than focusing on multiple dimensions, as current empathy theory recommends (Davis, 20 1994; Eisenberg, 2000; Hoffman, 2001). Nevertheless, such research establishes the relevance of considering empathy with respect to bullying, and illustrates a number of important considerations to be made in designing such a study. In their study of empathy, measured separately as both empathic concern and empathic distress via self-report questionnaires with students between the ages of 13 and 16, Endresen and Olweus (2001) demonstrated that children who experienced higher levels of other-oriented empathic concern tended to bully less than their peers. In contrast, no significant relationship was found between the tendency to experience self-focused empathic distress and bullying. Although the age group examined in Endresen and Dlweus's study is slightly older than that r included in the current study, the findings indicateithe importance of assessing both empathic concern and distress separately within the current study in order to reveal their differential relationships to bullying behaviour. Despite the importance of Endresen and Olweus's (2001) study for providing evidence of a link between affective empathy and bullying, the study did not address the role of cognition in producing empathy. Given recent evidence and debate (e.g., Arsenio & Lemerise, 2001; Sutton et al., 1999a, b, c) regarding cognitions such as perspective-taking in children who bully, it is critical to address cognitive factors as well as affective ones. In contrast to evidence that aggressive individuals tend to lack skills in assuming the cognitive perspectives of others (i.e., cognitive empathy) (e.g., Richardson et al., 1994), recent theory suggests that competency in perspective-taking skill is advantageous to bullies (Sutton et al., 1999a, b, c). Bullying is social in nature in that it is carried out in social settings, where peers serve as an audience that can accept or reject a bully's behaviours (Arsenio & Lemerise, 2001). In order to be a successful bully, Sutton et al. (1999a, b, c) propose that psychological manipulation of both victims and the larger peer audience is advantageous. Manipulation of the larger peer group might take the form of convincing others to join in the bullying, or at least, not to inform an adult. By 21 comparison, manipulation of the victim might include inflicting psychological torment through excluding the victim socially, and convincing him or her that he/she has no friends. Support for the theory that bullies seek to manipulate other people within social situations is provided by a study that demonstrated that self-identified bullies (aged 9 to 12 years) reported higher Machiavellian attitudes than comparison peers (Sutton & Keogh, 2000). Machivellianism is defined as the "attitudinal personality predisposition to see people as manipulable in interpersonal situations" (p. 445). Sutton and colleagues (1999a) directly investigated their theory that bullies are proficient in their capacity to assume the perspectives of others by applying the "Theory of Mind" paradigm. Theory of Mind refers to "the ability of individuals to attribute mental states to themselves and others in order to explain and predict behaviour" (Sutton et al., 1999a, p. 436). In their study, schoolchildren aged 7 to 10 years listened to stories and answered questions designed to assess their understanding of the mental states and emotions of the story characters. Within the study, a combination of both self and peer nominations were used to determine each child's typical role in bullying (Bully, Assistant, Reinforcer, Defender, Outsider, Victim) using the Participant Role Questionnaire (Salmivalli, Lagerspetz et al., 1996). Results revealed that students identified by a combination of peer and self-reports as habitual bullies had consistently higher scores on Theory of Mind stories than all other children sampled, with the exception of those who stayed away from bullying. In comparison, children within the Reinforcer role tended to perform the most poorly on the Theory of Mind task. Overall, performance on the Theory of Mind task was positively and significantly correlated with the extent of behaviours associated with leader-like bullying. In contrast, Theory of Mind was negatively and significantly correlated with behaviours associated with remaining uninvolved in bullying, or being victimized, and was unrelated to behaviours associated with defending the victim. Given that perspective-taking is thought to become fully developed around age 9, which 22 is in the middle of the ages sampled in the study described above, results might differ in an older sample. The current study sought to determine the nature of perspective-taking (similar to the Theory of Mind construct) in each of the participant roles in a slightly older age group than that described here (11 to 13 years versus 7 to 10 years). Two studies (Bjorkvist, Osterman, & Kaukiainen, 2000; Kaukiainen et al., 1999) that have attempted to assess both affective empathy and related cognitive processes in relation to various forms of in-school aggression have employed measures of questionable validity. Specifically, aggression was assessed via peer-estimations, and included various aggressive behaviours that are often included in definitions of bullying, although they were not clearly representative of bullying. Empathy was assessed by peer-estimations of the extent to which same-sex classmates engaged in prosocial behaviours (presumably thought to be motivated by empathy), or felt a variety of emotions. Social intelligence was the construct that most closely resembled the cognitive construct of interest to the current study, perspective-taking. Peer estimates of social intelligence were obtained with ratings of the extent to which classmates thought or behaved in ways that demonstrated insight into others' feelings. Clearly, the major constructs within these studies were not operationalized consistently with other studies of empathy described in this review. Nevertheless, the results of the studies outlined above are relevant to a review of existing empathy-bullying research. In terms of affective empathy, significant negative associations between physical, verbal, and indirect forms of aggression as estimated by classmates, and affective empathy as estimated by classmates, were found in 10- to 14-year-olds (Kaukiainen et al., 1999) and in 10- to 12-year-olds (Bjorkvist et al., 2000). In other words, students who were seen by peers as being involved in physical, verbal, and indirect forms of aggression were also seen by peers as lacking in affective empathy. With respect to social intelligence, those 10- to 14-year-old children identified as indirectly aggressive were also rated by their peers as more socially 23 intelligent in comparison to children who used more overt forms of aggression (i.e., physical and verbal) (Kaukiainen et al., 1999). Social intelligence was found to be positively and significantly correlated with all forms of aggression (verbal, physical, indirect) in 10- to 12-year old children (Bjorkvist et al., 2000). In both studies, the strongest correlations were observed between indirect forms of aggression and social intelligence (Bjorkvist et al., 2000; Kaukiainen et al., 1999). These findings certainly seem to support Sutton and colleagues' (1999a) theory that children who bully are neither socially unintelligent, nor unskilled. In summary, there are still a number of gaps in the existing literature on bullying and empathy. First, the research has largely been focused on children who perpetrate bullying, rather than on all participant roles. Second, many studies have limited their examination to one facet of empathy (i.e., affective or cognitive). Third, those studies that have addressed both affect and cognition have employed measures that are not necessarily valid to the study of empathy. The current study aimed to close the gap in the existing literature by addressing all participant roles, examining both affective and cognitive facets of empathy, and using measures that are well-validated for the current research. 24 Statement of the Problem The importance of mobilizing the entire peer group to counter bullying in schools is supported by research that demonstrates that although many peers are typically present when bullying occurs, very few intervene on behalf of victims (Craig & Pepler, 1997). In fact, many children act in ways that support the bullying (Salmivalli, Lagerspetz, et al., 1996). Although we now understand how children behave in the face of bullying, we do not yet understand why children behave as they do. There is well-established empirical evidence linking increased empathy to both decreased aggression (Miller & Eisenberg, 1988) and increased prosocial behaviour (Eisenberg & Miller, 1987). Given this, it is logical to propose that those children who are aggressive in the face of bullying tend to experience little empathy, and those who prosocially defend victims in the face of bullying tend to experience high empathy. Indeed, there is some research evidence to support the point with respect to perpetration of bullying (Endresen & Olweus, 2001). Yet, the link between empathy and bullying becomes more complex when critical cognitive and. affective processes are considered separately. In fact, there is some evidence that an increase in perspective-taking ability (a cognitive facet of empathy) is associated with increased perpetration of bullying (Sutton et al., 1999a). Unfortunately, the research that exists is plagued by inconsistencies in how empathy has been defined and measured. In addition, multiple dimensions of empathy (i.e., empathic concern, empathic distress, perspective-taking ability) have not been assessed and compared in the same sample for their differential links to behaviour. Taking a closer look at how all children who are present when bullying occurs feel and think about others (i.e., dispositional empathy) is a critical step towards explaining why children do what they do in the face of bullying. It is imperative to understand the specific nature of both cognition and affect so that behavioural interventions may be designed to capitalize on children's strengths, and build upon their weaknesses. In this way, not only can children's 25 perpetration of bullying be decreased, but also all children within bullying scenarios can be mobilized to take active, prosocial action against bullying. Accordingly, the major purpose of the current study was to explore empathy in children who behave in various ways within bullying scenarios. Students in grades 5 to 7 were classified into their typical participant roles in bullying separately, using peer-nominations on an adapted form of the P R Q (Sutton & Smith, 1999), and through a parallel self-report version of the P R Q adapted for the current study. For peer-nominations, children nominated classmates who fit each of 25 behavioural descriptions. For self-reports, children rated themselves in terms of the extent to which they fit each behavioural description. Behavioural descriptions depicted characteristic Bully, Assistant, Reinforcer, Outsider, Defender, and Victim roles. Empathy was assessed using self-reports of how children usually think and feel about others. Three different aspects of empathy were addressed, including other-focused empathic concern, self-focused empathic distress, and perspective-taking as a reflection of more cognitive aspects of empathy. It was expected that each aspect of empathy would be differentially related to the roles children adopt in response to bullying. The current study extended the literature on both empathy and bullying by examining the empathic tendencies of children who assume various roles in the bullying process. Of primary interest were variations in general, self-reported tendencies to experience empathic concern and distress, and to take the perspectives of others. However, variations in situation-specific empathy in response to bullying scenarios were also examined across students who adopted different roles in response to bullying. The following questions serve as an outline of the design, rationale, and hypothesized findings of the study. 26 Primary Research Questions Question #1. Do children who assume various participant roles (i.e., Bully, Assistant, Reinforcer, Defender, Outsider, Victim) differ according to their tendency to experience other- oriented empathic concern (i.e., a facet of affective empathy) for others in distress? Although no previous research had directly assessed empathic concern according to participant roles, general predictions about the nature of empathy in some of the participant roles were based upon the extensive research on empathy and aggression, and empathy and prosocial behaviour. Bullies were expected to report less empathic concern than other participant roles, based on evidence of a negative relation between empathic concern and bullying (Endresen & Olweus, 2001), and other evidence of a negative relation between affective empathy and various types of aggression (e.g., Bjorkvist et al., 2000; Cohen & Strayer, 1996; Kaukiainen et al., 1999; Miller & Eisenberg, 1988). Defenders were expected to report higher empathic concern than all other participant roles, based upon the fact that their behaviour is distinctly prosocial, and prosocial behaviour has demonstrated empirical links to high affective empathy (defined in various ways) (see Eisenberg & Miller, 1987 for a review). The research evidence supports the theoretical position that individuals who experience other-oriented concern for others in distress are motivated to help those individuals out of a genuine concern for the other person (Batson, 1998). Predictions concerning the other participant roles were more difficult, given that no research existed upon which to base such predictions. However, it was expected that those roles that are pro-bullying (i.e., Assistant, Reinforcer) would follow a similar trend to Bullies, reporting lower empathic concern in comparison to other roles. Question #2. Do children who assume various participant roles differ in terms of their tendency to experience self-focused empathic distress (i.e., a facet of affective empathy) in response to others in distress? 27 Although no previous research had directly assessed empathic distress among participant roles, general predictions about the nature of empathy in some of the participant roles were based upon the research linking empathy to bullying, aggression and prosocial behaviour. It was expected that Bullies' reports of self-focused empathic distress would not differ significantly from the other participant roles, based on previous failure to find a significant relation between empathic distress and bullying (Endresen & Olweus, 2001). Outsiders' reports of empathic distress were expected to be high, relative to the other participant roles. This expectation was based upon the theory (Batson, 1998) that individuals who experience high levels of self-focused empathic distress will choose to escape distressing situations (e.g., when someone is being bullied), in order to relieve their own aversive state. In contrast, it was expected that Defenders would report low levels of empathic distress, because their attempts to stay and intervene to stop bullying suggest that they do not experience high levels of self- focused distress, but rather they experience other-focused, genuine caring for the victim's welfare, according to Batson's (1998) theory. No further predictions were made concerning the remaining participant roles. Question #3. Do children who assume various participant roles differ in terms of their tendency to spontaneously adopt the perspectives of others (i.e., experience a cognitive facet of empathy)? In general, perspective-taking was expected to be relatively higher in the Bully role, given a similar finding in Sutton et al.'s study (1999a). By comparison, Sutton et al.'s findings of negative or non-significant correlations between perspective-taking and behaviours associated with the Outsider, Victim, and Defender roles, made it difficult to offer predictions about perspective-taking in those roles. However, given that there was much research evidence to support a positive relation between increased perspective-taking and various forms of prosocial behaviour (see Underwood & Moore, 1982 for a review), it was expected in the current study 28 that Defenders, as highly prosocial children, would also report high tendencies to assume others' perspectives. Although that result would differ from that of Sutton et al.'s study (1999a), that is possible, given that perspective-taking is thought to become fully developed around age 9, which is in the middle of the ages sampled in the study described above. It was predicted that higher perspective-taking skills would be reported in the Defenders within the older sample considered in the current study. Secondary Research Question How do children in various participant roles respond (in terms of empathic concern, empathic distress, and perspective-taking) when another child is being bullied? Although to the knowledge of the investigator, this question had not previously been addressed empirically, it was expected that the same pattern of results predicted for dispositional forms of empathy would also result for the situational, bullying-specific forms of empathy. Given the interest of the current study in the phenomenon of bullying, children's responses towards victims of bullying were of specific interest. 29 Method Participants Participants in the current study were 409 students from urban Vancouver Island elementary schools in grades 5 to 7 (10 to 13 years of age). The students came from 20 different classes, in five elementary schools. The distribution of participants according to grade, gender, and mean age is presented in Table 1. The ethnicity of the sample was as follows: (a) 80% White/Caucasian, (b) 8% Mixed Origin (i.e., Asian-Caucasian, First Nations-Caucasian), (c) 5% First Nations, (c) 3% Asian-Canadian, (d) 1% Indo-Canadian, and (e) 3% Other (i.e., Black, Middle-Eastern, Latin, non-specified). Of the students who participated, 90% spoke only English at home, 5% spoke English and another language at home, and 5% spoke only another language at home 2 . The family composition of participants was as follows: (a) 58% lived with both their mother and father, (b) 17% lived with only their mother or father, (c) 15% lived with one parent and a step-parent, (d) 5% lived in shared custody between parents, (e) 5% lived in other arrangements (i.e., foster care, grandparents, aunts or uncles). Only those students who received parental permission (Appendix A), and who themselves agreed to participate (Appendix B) took part. Overall, 82% of the students targeted for participation took part in the data collection process. 2 All participants were deemed to be proficient in English after consulting with classroom teachers. 30 Table 1 Distribution of Participants by Grade, Gender, and Mean Age Grade Number of Girls Number of Boys Mean Age Grade 5 65 60 10.23 years Grade 6 67 70 11.21 years Grade 7 78 69 12.19 years Total 210 199 11.26 years The 10- to 13-year-old age group was selected as optimal for consideration in the present study on the basis of a few key principles. First, bullying is a prevalent phenomenon during late childhood-early adolescence (Charach, Pepler & Ziegler, 1995; Haynie et al., 2001; Nansel et al., 2001; Whitney & Smith, 1993), steadily declining in later, secondary school years (Whitney & Smith, 1993). For this reason it was expected that bullying would be a relatively common and personally salient phenomenon among the target participants. Second, in British Columbia schools, students change from elementary to junior high or secondary schools after grade 7. This means that students enter a new school, where many factors relevant to social interactions may change. In order to avoid any differences in students' bullying behaviour upon entry into secondary school, students were assessed within elementary school only. Third, the current study's goal of encompassing relational forms of bullying necessitated that target students had reached sufficient levels of cognitive and social sophistication required to carry out such bullying. Indirect (relational) forms of aggression commonly used in bullying others (e.g., gossip, becoming friends with others as revenge) have been demonstrated to increase substantially at ages 11 and 15, in comparison to age 8, particularly in girls (Bjorkvist et al., 1992). Fourth, although development of children's reading ability was not significantly 31 related to dispositional empathy on questionnaires in previous research (e.g., Bryant, 1982), minimal reading was required to complete the empathy questionnaire within the current study. Additionally, the investigator checked with the classroom teacher to determine if there were any students for whom the minimal reading requirements of the questionnaire measures were prohibitive. Finally, no previous study of multidimensional empathy as it is defined here and bullying had been conducted with the current age group, despite the fact that children's abilities to distinguish between others' emotions and perspectives, a skill that is critical to the experience of empathy (Eisenberg, 2000) is typically well-established. Procedures Students in grades 5 to 7 were initially visited in their home classrooms by the principal investigator to recruit students to participate in the proposed study. The purpose of the study and the nature of the activities involved were explained. At that time, information letters/parent consent forms (Appendix A) were distributed. All students who had received written parental consent, and who themselves had agreed to participate (see Student Consent form Appendix B) took part in a single 45-minute testing session. For those students who had received consent, but who were absent on the day of testing, alternate arrangements were made with the teachers to conduct testing at another time within a week of the rest of the class. The primary testing session took place in the spring of the school year, during regular school hours at a time that was deemed convenient for the classroom teacher. During the session, participating students first completed a demographic information sheet (Appendix C), followed by a self-report questionnaire about the prevalence of bullying in school which was not relevant to the study described here, the Participant Role Questionnaire (Appendices D and E), and the Thoughts and Feelings Questionnaire (Appendix F). Anonymity was assured to all participants. Participants' desks were moved apart and dividers were placed on the desks to prevent answers from being seen by other participants. The pace of completion was set by the 32 examiner's reading aloud of each item on the questionnaires. Participants' questions were addressed directly. Once participants finished the questionnaires, they were put into sealed envelopes. All data collection was performed by the principal investigator. Measures Demographic information. Participating students' demographic information was collected by having them complete a brief, personal information sheet (Appendix C). This included items regarding gender, age, grade, birthdate, family composition, primary language spoken, and racial/ethnic identity. Bullying behaviour. Active participation in bullying, as well as other behavioural roles assumed by children within bullying scenarios was measured by parallel peer-nomination and self-rating measures. The Participant Role Questionnaire (PRQ) (Sutton & Smith, 1999)3 (Appendix D) is a peer-nomination measure. On the PRQ, participants were initially presented with an explicit definition of bullying behaviour in order to guide them to consider various forms of bullying (physical, verbal, relational), and only bullying, as opposed to other aggressive acts (e.g., fighting between opponents of equal power). The bullying definition was also printed at the top of the page of the questionnaire, for easy reference to remind students as they completed the survey. Children were instructed to think about what they, and their classmates typically do in situations in which someone is being bullied. The names of all pupils in the class who had received permission to participate in the current study, divided into columns by gender, were printed beneath each of 25 questions about particular behaviours related to bullying (which comprised the six different participant role subscales). For each question, the students nominated as many peers as they could think of who fit the behavioural description, by circling their names. Peer-nominations of participant roles were standardized within class so that the mean for each class was 0, and the standard deviation was 1. This ensured that variations in 3 Adaptations included splitting single items into two or three items, and retaining some items from Salmivall i 's version (1996). 33 sco res were not attributable to varying character ist ics ac ross c l a s s e s (e.g., c lass s ize , number of nominators) . A factor ana lys is and internal cons is tenc ies were computed to ascer ta in the psychometr ic a d e q u a c y of the adapted vers ion of the P R Q used in the current study (see Resu l ts sect ion). Fol lowing complet ion of the peer-nominat ion sect ion of the P R Q , students responded to 25 parallel i tems that asked them to indicate whether they, themse lves , e n g a g e d in the behaviours descr ibed (Appendix E) . A s opposed to the peer-nominat ion format which a s k e d " W h o . . . ? " , the self-rat ing format a s k e d "Do you . . . ? " . The response cho ices ranged from 1=NO (indicating "Never") , 2=no ("Not often"), 3=somet imes, 4=yes ("Often"), 5 = Y E S ("Always"). S c o r e s were s u m m e d for each subsca le (Bully, Ass is tant , Reinforcer , Defender , Outs ider , Vict im), for e a c h respondent , and the total sco res were divided by the number of i tems within the subsca le , to obtain average sco res . Th is self-report format of the P R Q w a s adap ted speci f ical ly for the current study, in order to obtain parallel peer- and self-reports. T o be certain of the psychometr ic propert ies of this adapted measure , both a factor ana lys is and internal cons is tenc ies were computed (see Resu l ts sect ion). Val idi ty has previously been demonstrated in signif icant, posit ive correlat ions be tween var ious self-report quest ionnai res and measu res of bullying, such as peer, teacher , and parent- reports (Ladd & Kochender fe r -Ladd , 2002). Addit ionally, sel f-reports of bullying have demonst ra ted theoretical ly appropriate relat ionships to lonel iness, dep ressed affect, anxiety, and peer rejection (Ladd & Kochender fe r -Ladd , 2002). Dispositional empathy. The tendency to be empath ic ac ross situations (i.e., d isposi t ional empathy) w a s a s s e s s e d using a self-report quest ionnaire. A n adapted form of the Interpersonal Reactivity Index (IRI) (Davis , 1983) entitled Thoughts and Feelings Questionnaire (TFQ) (P . Miller, persona l communica t ion , March 20, 2002) was used to obtain indices of both disposi t ional empath ic concern , and disposi t ional perspect ive-taking (Appendix F) . T h e T F Q is a 34 14-item, self-report questionnaire that consists of two, 7-item subscales, assessing empathic concern and perspective-taking. The empathic concern scale assesses "the tendency to experience other-oriented feelings of sympathy and concern for unfortunate others" (Davis, 1983, p. 114). The perspective-taking scale assesses "the tendency to spontaneously adopt the psychological point of view of others" (Davis, 1983, p. 114). In addition, a 7-item personal (empathic) distress subscale that assesses "self-oriented feelings of personal anxiety and unease in tense interpersonal settings" (Davis, 1983. P. 114) was adapted from the original I RI. It was necessary to adapt this measure in order to create an index of empathic distress that was suitable for use with pre-adolescents, in terms of language complexity, in the current study. An additional empathy measure was created for exploratory purposes, assessing empathy (i.e., empathic concern, perspective-taking, empathic distress) in response to specific bullying scenarios (i.e., verbal, physical, relational). This situation-specific empathy measure was created by adapting items from both the IRI (Davis, 1983), and the Empathic Responsiveness Questionnaire (ERQ) (Olweus, & Endresen, 1998). All empathy items (from all subscales) were presented in random order within the same questionnaire (Appendix F). Children responded to each item on the empathy scales by indicating the extent to which a statement described them, using a scale of 1=Not at all like me, 2=A little bit like me, 3=Kind of like me, 4=A lot like me, 5=Always like me. Scores were summed for each subscale (i.e., Empathic Concern, Perspective-Taking, Empathic Distress, Empathic Concern (Bullying), Perspective-Taking (Bullying), and Empathic Distress (Bullying) subscales), for each respondent. The total scores were then divided by the number of items to obtain average scores for each of the different forms of empathy. Thus, six different self-report indices of empathy were computed for each participant, tapping both dispositional empathy as well as situational empathy (i.e., empathy in response to bullying scenarios). Higher scores were indicative of greater empathy in each case. 35 Results Preliminary Analyses Prior to considering the primary research questions, preliminary analyses were conducted to examine the psychometric adequacy of both the independent variables (i.e., peer- and self-reported participant roles) and the dependent variables (i.e., dispositional and situational empathic concern, empathic distress, and perspective-taking). Identifying participant roles. Of initial interest was whether the bullying participant roles identified through peer-assessments and self-reports in the present study were consistent with each other, and with those identified in previous studies. To this end, separate factor analyses were conducted 4 on both peer- and self-assessments of participant roles in order to determine whether items would form the composites for the Bully, Assistant, Reinforcer, Defender, Outsider, and Victim subscales that had been analyzed in previous research (Salmivalli, Lagerspetz, et al., 1996; Sutton et al., 1999a). First, a principal components factor analysis with varimax rotation was performed on the standardized peer-nominations received for each of the 25 participant role behaviours. From this analysis, four distinct factors emerged: Pro-bully, Defender, Victim, and Outsider (see Table 2 for loadings). Factor analyses were necessary, given that parallel peer- and self-reports had not been used in previous research, and because some of the items taken from Sutton and Smith (1999) were adapted, while others were added for the current study. 36 Table 2 Factor Loadings for Pro-bully, Defender, Victim, and Outsider Factors (Peer-Nominations) Factor Item Pro- bully Defender Victim Outsider Who starts bullying? Who joins in the bullying if someone else has started it? Who gets others to join in the bullying? Who always thinks of new ways of picking on the victim? Who takes the bully's side? Who encourages the bully by shouting? Who leads a gang? Who gets others to come and watch? Who laughs at the person being bullied? Who helps the bully by holding the victim? Who says things to the bully like "Show him!"? Who helps to catch the victim for the bully? Who is usually there, even if not doing anything? Who sticks up for the victim? Who tries to make others stop the bullying? Who says to the others that bullying is stupid? Who tries to cheer the victim up? Who tells an adult about the bullying? Who is the one to get called mean names? Who is the one to get hit or kicked? Who is the one to get left out of the group on purpose? Who doesn't even know about the bullying? Who isn't usually there, and stays away? Who doesn't do anything or take sides? .900 .892 .880 .877 .872 .872 .858 .852 .852 .852 .840 .731 .512 .855 .848 .827 .817 .722 .914 .858 .846 .777 .695 .649 Note. Factor loadings that were not significant (i.e., less than .400) are not included. All cross-loadings were less than .400. The Pro-bully factor was composed of 13 items that have been used in previous research (e.g., Sutton & Smith, 1999; Sutton et al., 1999a) to represent bullying, assisting, and reinforcing bullying behaviours. The items "Who starts the bullying?", "Who gets others to join in the bullying?", "Who always thinks of new ways of picking on the victim?", and "Who leads a gang?" represented active bully behaviours. The items "Who joins in the bullying if someone 37 else has started it?", "Who takes the bully's side?", "Who helps the bully by holding the victim?", and "Who helps to catch the victim for the bully?" represented assisting behaviours. The questions "Who encourages the bully by shouting?", "Who gets others to come and watch?", "Who laughs at the person being bullied?", "Who says things to the bully like 'Show him!", and "Who is usually there, even if not doing anything?" represented reinforcing behaviours. Collectively, these items grouped together to form the Pro-bully factor, which accounted for 39.1% of the variance (eigenvalue=9.78). Previous research using the P R Q (Salmivalli, Lagerspetz, et al., 1996; Sutton & Smith, 1999) has also found a single, Pro-bully factor that encompasses all of the Bully, Assistant, and Reinforcer items. Nevertheless, all previous studies using the P R Q have examined the Bully, Assistant, and Reinforcer roles separately, because of their conceptual interest. A second factor was composed of five items representative of defending behaviours. The items "Who sticks up for the victim?", "Who tries to make others stop the bullying?", "Who says to the others that bullying is stupid?", "Who tries to cheer the victim up?", and "Who tells an adult about the bullying?" grouped together to form the Defender factor, which accounted for 16.6% of the variance (eigenvalue=4.15). A third factor was made up of three items indicating verbal, physical, and relational victimization. The items "Who is the one to get called mean names?", "Who is the one to get hit or kicked?", and "Who is the one to get left out of the group on purpose?" grouped together to form the Victim factor, accounting for 10.3% of the variance (eigenvalue=2.59). The fourth factor was composed of items representing uninvolved, or outsider behaviours. The items "Who doesn't even know about the bullying?", "Who isn't usually there, and stays away?", and "Who doesn't do anything or take sides?" clustered together to compose the Outsider factor, which accounted for 9.1% of the variance (eigenvalue=2.28). An additional item "Who pretends not to notice what is happening?" did not load significantly onto any of the 38 factors, and was therefore dropped from the analyses. Further analyses verified that, with minor exceptions, the same participant roles were found for both girls and boys, when considered separately5. As with the peer-nominated participant roles, a principal components factor analysis with varimax rotation was conducted on the self-reports of participant roles (see Table 3 for factor loadings). Although five factors emerged rather than four, with few exceptions, the results of the factor analysis for the self-report data essentially replicated the results obtained for peer- nomination data. 5 In the case of girls, the item "Who is usually there, even if not doing anything?" did not load significantly (i.e., factor loading of .321) onto the Pro-bully factor. In addition, "Who pretends not to notice what is happening?" loaded significantly (.431) onto the Victim factor in the case of boys, but not significantly (.370) onto the Outsider factor in the case of girls. See Appendix G for complete analyses examining gender differences in peer-nominated participant roles. 39 Table 3 Factor Loadings for Pro-bully, Defender, Victim, (Single Item) Reinforcer, and Outsider Factors (Self-Ratings) Item Do you encourage the bully by shouting? Do you get others to join in the bullying? Do you say things to the bully like 'Show him!'"? Do you always thinks of new ways of picking on the victim? Do you join in the bullying if someone else has started it? Do you start the bullying? Do you take the bully's side? Do you laugh at the person being bullied? Do you lead a gang? Do you help the bully by holding the victim? Do you help to catch the victim for the bully? Do you get others to come and watch? Do you try to make the others stop the bullying? Do you stick up for the victim? Do you try to cheer the victim up? Do you say to the others that bullying is stupid? Do you tell an adult about the bullying? Are you the one to get called mean names? Are you the one to get left out of the group on purpose? Are you the one to get hit or kicked? Are you usually there, even if not doing anything? Do you do nothing, or refuse to take sides? Do you pretend not to notice what is happening? Do you stay away? Pro- bully Factor (Single Defender Victim Item) Outsider Reinforcer .753 .753 .735 .709 .697 .694 .686 .670 .653 .622 .540 .506 .805 .800 .788 .774 .704 .886 .804 .782 .739 .744 .709 .632 Note. Factor loadings that were not significant (i.e., less than .400) are not included. All cross-loadings were less than .400. 40 The primary difference between the factor analyses with peer- and self-reported data concerned a single item, "Are you usually there, even if not doing anything?", which is conceptually included as part of the Reinforcer role. In the present sample, this item did not load on the Pro-bully factor for self-report data. Instead, this was the only item included in a fourth factor, accounting for 7.6% of the variance (eigenvalues .90). When considered as a single-item "role", the meaning of this item is rather ambiguous and alone does not clearly reflect the role of someone who "reinforces" bullying. Hence, this item (and the factor identified) was dropped from subsequent analyses. As presented in Table 3, the remaining four factors reflected the same roles as were identified using peer-nomination data. Specifically, self-report data yielded a Pro-bully factor, (22.9% of the variance, eigenvalue=5.72), comprised of 12 of the 13 items from the Pro-bully factor on the peer-nominated format, as well as the same Defender factor (14.3% of the variance, eigenvalue=3.57), and the same Victim factor (accounting for 8.5% of the variance, eigenvalue=2.12). The Outsider factor (accounting for 6.2% of the variance, eigenvalue=1.56) varied somewhat from that observed for peer-nomination data. The item, "Do you pretend not to notice what is happening?" was included for self-reported roles, but not for peer-nominations, whereas the item, "Do you know about the bullying?" was not included for self-reported roles, but was for peer-nominations, even though parallel versions of both items were included in the Outsider role in previous research (e.g., Salmivalli, Lagerspetz, et al., 1996; Sutton & Smith, 1999). When further analyses were conducted for both genders separately, the same participant roles were found for both girls and boys, with the exception of a few differences among items within the Pro-bully roles 6. 6 Inconsistencies among items by gender were likely due to extremely low numbers of nominations in these roles. See Appendix H for complete analyses examining gender differences in self-reported participant roles. 41 Comparing results of factor analyses conducted with both peer-nominations and self- reports, it is noteworthy, but hardly surprising, that few students rated themselves as adopting Pro-bully roles. In fact, the number of individuals in the self-reported Pro-bully role was insufficient for analyses involving either the overall Pro-bully role or the three separate Pro-bully roles (Bully, Assistant, Reinforcer). Accordingly, for self-reported participant roles, all further analyses involved the Defender, Outsider, and Victim roles only. The internal consistencies for each of the four peer-nominated participant role scales were high. Cronbach's alpha scores obtained for the present sample were a=.97 for the Pro- bully scale, a=.91 for the Defender scale, a=.81 for the Outsider scale (with one item dropped), and a=.86 for the Victim scale. Despite the fact that Pro-bully items did not load separately onto Bully, Assistant, and Reinforcer subscales, internal consistencies were nevertheless computed, as these subscales were of theoretical interest. These Cronbach's alpha scores were as follows: a=.95 for the Bully scale, a=.91 for the Assistant scale, and a=.91 for the Reinforcer scale. The internal consistencies demonstrated for the current sample are consistent with those obtained by Salmivalli and colleagues (1998) in a sample of children in the late childhood to early adolescent range, in which Cronbach's alphas ranged from .84 to .94. The internal consistencies for each of the three self-reported participant role scales were moderate to good, with Cronbach's alpha scores as follows: Defender scale, a=.87; Outsider scale (with one item dropped), a=.53; and Victim scale, a=.78. Although the internal consistencies observed for self-reported participant roles were considerably lower than those obtained with the peer-nomination format of the P R Q , they were considered adequate for research purposes. Pearson product moment correlations were conducted to examine the relations among peer-nominations for the various participant roles. As seen in Table 4 (below the diagonal), peer perceptions of victimization were not significantly related with other roles. However, 42 significant correlations were observed among the remaining peer-nominated roles in expected directions. Not surprisingly, given the results of the factor analysis, the strongest correlations were observed among the Pro-bully participant roles (Bully, Assistant, Reinforcer). Defender and Outsider roles were modestly correlated, but, as might be expected, students who were perceived as Defenders or Outsiders were significantly less likely to also be viewed in any of the Pro-bully roles. Table 4 Correlations Between Averages of Peer-Nominations and Self-Reports in Participant Roles 1 2 3 4 5 6 1. Bully Score ™ 2. Assistant Score .92** — 3. Reinforcer Score .90** .91** — 4. Defender Score -.43** -.38** -.41** .20** -.05 .13 5. Outsider Score -.57** -.57** -.61** .50** .08 .05 6. Victim Score -.05 -.06 -.02 -.12* .07 .42** Note. *p<.01 (one-tailed tests), **p<.05 (one-tailed tests). Intercorrelations among self-reports are presented above the diagonal, while intercorrelations among peer-nominations are presented below the diagonal. Correlations between self-reports and peer-nominations are along the diagonal. Associations among the three self-reported participant roles (presented in Table 4, above the diagonal) were all low and non-significant. Thus, self-reported participant roles appear to be quite distinct. Some degree of consistency was evident in the positive, significant correlations between peer- and self-assessments (see Table 4, along the diagonal) in both Defender and Victim roles, although these correlations were modest. In contrast, the peer- and self-reports within the Outsider role were not significantly related to one another, suggesting that different individuals may be identified as Outsiders across peer- and self-reports. 43 Given the results of preliminary analyses, it was apparent that the four participant role solution was more psychometrically sound than the six role solution. For this reason, all primary analyses were performed using the four roles. Following previous researchers, however, additional analyses were also performed with all six participant roles, given their theoretical interest. In terms of self-reported participant roles, the three roles (Defender, Outsider, Victim) were separately examined. Classification into participant roles. For peer-nomination data, all students were classified into a specific participant role following procedures used by Salmivalli and colleagues (1998) to distinguish between participant roles on the P R Q . Specifically, after summing and standardizing the number of peer-report nominations for each participant role by class (so that the mean score = 0, SD = 1), these sums were divided by the number of items in each participant role scale (to control for varying numbers of items in each composite). Using these standardized scores, students were assigned to particular participant roles if (1) their score for that role was above the overall mean for that participant role, and (2) they scored higher on that participant role than on any of the other participant roles. Students who scored below the mean, who received raw scores of zero on all of the scales, or who had a difference of less than 0.1 between his or her highest and second highest score, were determined to have no clearly discernible participant role. Victims 7 were identified as those students who were nominated by at least 30% of their classmates as victims. Table 5 presents a complete distribution of participants according to the classification procedure described. 7 Although victims were not classified by the primary nature of their victimization within subsequent analyses, it is interesting to note that of the 66 participants identified as victims, 21 (32%) were identified as victims of primarily physical bullying, 16 (24%) as victims of verbal bullying, and 29 (44%) as victims of primarily relational bullying. 44 Table 5 Distribution of Participants by Participant Role (Peer- and Self-Reported) Pro- R .. Assist- Reinf- Defen- Outsi- Victim No bully y ant orcer der der Role Peer- 120 35 22 63 99 87 66 37 Nominated (29.4%) (8.6%) (5.4%) (15.4%) (24.2%) (21.3%) (16.1%) (9.0%) Self- 6 1 1 4 183 59 85 76 Reported (1.4%) (0.2%) (0.2%) (1%) (44.7%) (14.4%) (20.8%) (18.6%) The classification of participants into roles using self-reports, as presented in Table 5, varied slightly from the procedure used with peer-nominations. After converting responses to numerical values (see Measures section for details), all responses were summed for each respondent within each participant role subscale8, and divided by the number of items within the subscale, yielding an average score for each student, for each participant role subscale. If a respondent's average score on a participant role subscale was equal to or greater than three (indicating that he or she behaved in the manner described, at least "sometimes"), and that score was the highest subscale score for that child, he or she was assigned to that participant role. If a child scored at least three or higher on any one of the Victim items (indicating various types of victimization), he or she was assigned to a Victim role, unless another role score was higher9. If a child's average participant role score was less than three (i.e., he or she did not behave in described ways at least "sometimes"), or if he or she had equal scores across more than one participant role, it was determined that no clear participant role could be assigned to that child. 8 Although all participant roles were used for classification purposes, only the three with adequate n's were retained for exploratory analyses. 9 Victims were not classified by the primary nature of their victimization within subsequent analyses. However, it is interesting to note that of the 85 participants identified as victims, 24 (20%) were identified as victims of primarily physical bullying, 55 (46%) as victims of verbal bullying, and 40 (34%) as victims of primarily relational bullying. 45 Overlap between peer-nominated and self-reported participant roles. Next considered was the degree to which students were similarly classified in terms of participant roles across peer- and self-reports. Examination of the percent of students who were similarly classified across self- and peer-reports revealed considerable discrepancies. Overall, 72.4% of all participants emerged within different participant roles, when peer and self-reports were compared. This left only 27.6% of the sample who reported that they behaved in a manner that was consistent with the participant role for which they were nominated by peers. Table 6 indicates the proportion (%) of individuals within each peer-nominated participant role who self- reported the various participant roles, with consistent role classifications indicated in bold on the diagonal of the table. 46 Table 6 Proportion of Self-Reported Participant Roles (%) that Matched Peer-Nominated Participant Roles Self-Reported Participant Role Bully Assistant Reinforcer Defender Outsider Victim No Role Peer-Nominated Participant Role Bully 40 14 17 23 Assistant 36 23 18 23 Reinforcer 33 16 17 29 Defender 0 1 60 13 13 13 Outsider 0 0 46 22 15 17 Victim 0 0 26 8 45 21 No Role 0 0 65 22 8 47 It is apparent from the table that the Defender role contained the highest proportion of agreement between peer and self-reports. That is not surprising, given that almost half (44.7%) of all participants were categorized into the Defender role by their self-reports. It is also not surprising to note that the second highest proportion of agreement occurred within the Victim role, which contained the second highest percentage (20.8%) of self-reported participant roles. Of the participant roles with the highest agreement between peer- and self-reports, it is interesting that none of the roles is a distinctly antisocial role. The lack of self-reports falling into the Pro-Bully roles may indicate that students are reluctant to identify themselves as behaving in antisocial ways. In this respect, self-reports may be biased by participants' efforts to present themselves in a socially desirable manner. The current findings are consistent with past research (e.g., Salmivalli, Lagerspetz, et al., 1996; Sutton & Smith, 1998) in which self-evaluations and peer-nominations of participant roles obtained by the PRO. have been found to differ. Specifically, four out of five children identified by peers as being involved in bullying others claimed that they were not, with most claiming to be Defenders (Sutton & Smith, 1998). Given the extremely low numbers within the self-reported Pro-bully role(s), these roles could not be incorporated into further analyses using the self-reported participant roles. Instead, only the Defender, Outsider, and Victim roles were examined. The participants who were classified within the "No Role" category were dropped from all further analyses. Although those individuals who were identified in consistent participant roles across peer- and self-reports were of interest, there were insufficient numbers within cells to examine them statistically. Due to lack of consistency between peer-nominations and self- ratings, both were used separately to create and compare the participant roles required for the independent variables in the major statistical analyses. Characteristics of participant roles. Preliminary, univariate analyses were conducted to determine variations across peer-nominated participant roles in terms of demographic variables 48 that might be potential confounds to the primary analyses. Results of chi-square analyses indicated that the four participant roles did not differ significantly in terms of ethnicity 0c2(18, N - 361) = 27.28, ns), or school (%2(12, N = 372) = 14.19, ns). In terms of gender (see Table 7), it was found that boys were significantly over-represented in the Pro-Bully category, as well as the Victim role. In contrast, girls were significantly under-represented in those roles, and over- represented in the prosocial Defender and uninvolved Outsider role (%2(3, N = 372) = 52.29, p < .001). Given the significant differences in gender across the peer-nominated participant roles, gender was added as an independent variable for examination within the primary analyses of the study. Table 7 Distribution of Participants among the Four Peer-Nominated Participant Roles by Gender, Grade Girls Boys Grade Pro- Bully Defender Outsider Victim Pro- Bully Defender Outsider Victim 5 10 27 15 2 24 10 14 6 6 11 24 20 7 29 7 10 19 7 14 23 17 16 32 8 11 16 Total 35 74 52 25 85 25 35 41 A one-way A N O V A was performed to determine whether or not the four peer-nominated participant roles differed significantly in terms of grade (see Table 7). A significant difference was found between participant roles, according to grade F(3, 368) = 4.20, p < .01 (four roles). Pairwise comparisons revealed that individuals within the Pro-bully role (when Bully, Assistant, and Reinforcer roles are combined) tended to come from lower grades than Victims (f(184) = - 49 2.22, p =.003, two-tailed). Defenders also tended to come from lower grades than Victims 63) = -3.43, p = .001, two-tailed). In contrast, Victims tended to come from lower grades than Outsiders (£(151) = -3.01, p =.003, two-tailed). Despite a significant overall association between grade and participant role, when participants were grouped by their gender as well as their grade, the relationship between grade and participant role did not reach statistical significance for either gender (peer-nominated four roles). Despite variations in grade across participant roles, grade was not added to further analyses as an additional independent variable, given limited sample size. In future research, however, it may be important to consider variations of both grade and gender. In the current study, analysis focused only upon groupings by gender and participant role. As with the four primary, peer-nominated participant roles, preliminary, univariate analyses were also conducted to check for differences in demographic variables across the six peer-nominated participant roles, and the three self-nominated participant roles 1 0 . Chi-square tests indicated that variations in terms of ethnicity in the participant roles were not significant (X2(30, N = 361) = 34.91, ns for the six peer-nominated roles, x2(14, N = 327) = 10.19, ns for the three self-reported roles). Differences as a function of school among participant roles were also non-significant (%2(20, N = 372) = 21.17, ns for the six peer-nominated roles, x2(8, N = 327) = 8.20, ns for the three self-reported roles). In terms of gender, significant differences were again observed among the participant roles when peer-nominations of all six roles were compared (x2(5, N = 372) = 56.22, p < .001, peer-nominated six roles). Boys were significantly over-represented in the Bully, Assistant, Reinforcer, and Victim roles. Girls were significantly under-represented in those roles, and over-represented in the prosocial Defender and uninvolved Outsider role. When the three self- Only Defender, Outsider, Victim roles were examined, due to low n's in Pro-Bully role, and inconsistency in factor analyses. 50 reported roles were compared, a significant difference was also found by gender (x2(2, N = 327) = 6.71, p =.04), attributable to an over-representation of girls in the Defender role. Given the significant differences in gender across the six peer-nominated participant roles, as well as the three self-reported roles, gender was added as an additional independent variable to the exploratory analyses. The one-way A N O V A conducted to check for differences across the six peer-nominated participant roles as a function of grade level revealed significant differences between roles, F(5, 366) = 4.28, p = .001. Pairwise comparisons revealed that Bullies tended to come from lower grades than Assistants (t(55) = -2.27p = .03, two-tailed), Reinforcers (f(96) = -2.73, p < .01, two- tailed), and Victims (f(99) = -3.86, p < .001, two-tailed). Overall, Defenders tended to come from lower grades than Reinforcers (f(160) = 2.17, p = .03, two-tailed), and Victims (£(163) = -3.43, p = .001, two-tailed). Victims tended to come from lower grades than Outsiders (£(151) = -3.01, p =.003, two-tailed). Despite the significant association between grade and participant role, when participants were grouped by their gender as well as their grade, the relationship between grade and participant role only remained significant for girls (%2(10, N = 186) = 24.29, p < .01). Although variations in grade across peer-nominated participant roles did exist for girls, grade could not be added to further analyses as an additional independent variable, given limited sample size. As previously mentioned, it may be relevant for future research to consider variations as a function of both grade and sex. Interestingly, the one-way A N O V A done to check for differences by grade across self-reported participant roles revealed no significant differences between roles F(2, 324) = .461, ns (self-reported three roles). Characteristics of empathy scales. A principal components factor analysis was conducted to evaluate the independence of the three empathy scales included in the present 51 study 1 1. Results of this analysis, as presented in Table 8, indicate that all seven items on the Empathic Distress subscale loaded consistently onto a single factor, which accounted for 20.4% of the variance (eigenvaiue=4.29). Six of the seven items included on the Perspective-taking subscale loaded consistently onto a second factor, which accounted for 19% of the variance (eigenvalue=3.99). The one item from the Perspective-taking subscale that did not load consistently with the rest of the subscale was "It's easy for me to understand why other people do the things they do". This item loaded independently onto a fourth factor (Understanding) that accounted for 5.39% of the variance (e igenvalues. 13). Six of the seven items included on the Empathic Concern subscale loaded significantly onto a third factor, which accounted for 17.8% of the variance (eigenvalue=3.74). However, four of these items were cross-loaded onto other factors (although the primary loading for three of the items was on the Empathic Concern subscale). As shown in Table 8, two of the seven items from the Empathic Concern subscale were cross-loaded onto the Perspective-taking factor ("/ often feel sorry for other children who are sad or in trouble", and "When I see someone being treated mean, it bothers me"). Another item from the Empathic Concern subscale was significantly cross-loaded onto the first, Empathic Distress factor ("/ often have strong feelings about things that happen around me"). Despite cross-loadings, the six items with their highest loadings on the Empathic Concern factor were retained for the Empathic Concern subscale in subsequent analyses. An additional item from the Empathic Concern subscale ("/ am a person who cares about the feelings of others") actually had its primary loading on the Perspective- taking factor (but cross-loaded onto the Empathic Concern factor). Given the inconsistency of this item, and the difficulty in interpretation it would have caused, it was removed from further 1 1 It was necessary to conduct factor analyses because such data were unavailable for the adapted version (P. Miller, personal communication, March 20, 2002) of the Interpersonal Reactivity Index (Empathic Concern and Perspective-taking), and because items from the Empathic Distress subscale were adapted for use in the current study. 52 analyses. Further factor analyses demonstrated that results were quite consistent across gender, with a couple of exceptions 1 2 (see Appendix I for complete analyses by gender). 1 2For boys, the item "I often have strong feelings about things that happen around me" did not load significantly (i.e., factor loading was less than .400) onto the Empathic Concern factor, as its loading was .391. In addition, "I am a person who cares about the feelings of others" loaded significantly (.599) onto the Empathic Concern factor for boys, but loaded significantly (.597) onto the Perspective-taking factor in the case of girls. 53 Table 8 Factor Loadings for Empathic Distress, Perspective-taking, Empathic Concern, and Understanding (Whole Sample) Item I have a hard time dealing with emergencies. I tend to lose control during emergencies. In emergency situations, I feel very uneasy and nervous. Being in very tense and emotional situations scares me. When I see someone get hurt, I can't stay calm and I get upset. I feel helpless when I am in the middle of a tense and emotional situation. When I see someone who badly needs help in an emergency, I just can't take it and I get upset. I try to understand how other kids feel before I decide what to say to them. Before I say anything bad about anyone, I try to imagine how I would feel if I were that person. Even when I'm mad at someone, I try to understand how they feel. Even when I know I'm right I listen to what other people think. There are different ways to think about a problem and I try to look at all of them. I am a person who cares about the feelings of others. Sometimes I try to understand my friends better by imagining how they think about things. When I see someone being picked on, I feel kind of sorry for them. Sometimes I feel sorry for other people when they are having problems. I often feel sorry for people who don't have the things I have. I often feel sorry for other children who are sad or in trouble. When I see someone being treated mean, it bothers me. I often have strong feelings about things that happen around me. It's easy for me to understand why other people do the things they do. Factor Empathic Perspective Empathic Understand Distress -taking Concern -ing .840 .795 .782 .717 .672 .667 .654 .783 .715 .684 .674 .610 .582 (.532) .562 .752 .745 .726 (.418) .711 (.428) .672 (.438) .460 .922 Note. All other factor loadings were less than .400. Cross-loadings greater than .400 are shown in parentheses. 54 The internal consistencies (Cronbach's alpha scores) for the three primary subscales were as follows: Empathic Concern scale, a=.88; Perspective-taking scale, a=.83; and Empathic Distress scale, a=.88. These results were comparable, and slightly higher than those observed in a previous sample (P. Miller, personal communication, March 20, 2002). Thus, each of the scales demonstrated adequate internal consistency for the purposes of the present study. Table 9 presents the inter-correlations among the three primary empathy subscales, as well as the single item factor (Understanding). Inter-correlations were moderate between the primary empathy scales, with the exception of the relationship between Empathic Concern and Perspective-taking, which was moderately high, consistent with past research (e.g., Carlo et al., 1998). The single item factor "It's easy for me to understand why other people do the things they do" was barely inter-correlated with the primary scales, suggesting that it may represent a unique facet of empathy. Following previous research (e.g., Endresen & Olweus, 2002; Eisenberg, Miller, Shell, McNalley, & Shea, 1991; Olweus & Endresen, 1998), each of the three primary empathy measures were considered separately in subsequent analyses despite their significant inter-correlations. Table 9 Correlations Between Primary Empathy Subscales 1 2 3 4 1. Empathic Concern 2. Empathic Distress .51** 3. Perspective-taking 72** .40** 4. Understanding .15** .02 .15** Note. **p=.01 (one-tailed tests). 55 The psychometric adequacy of the bullying-specific subscales of empathy was examined next. First, a principal components factor analysis was conducted on the items included in the bullying-specific empathy scales. Results (see Table 10) indicated that all nine items clustered as expected into separate Empathic Distress (Bullying), Empathic Concern (Bullying), and Perspective-taking (Bullying) factors. All three items on the Perspective-taking (Bullying) subscale loaded consistently onto that factor, which accounted for 27.3% of the variance (eigenvalue=2.46). Similarly, all three items on the Empathic Distress (Bullying) subscale loaded consistently, accounting for 26.8% of the variance (eigenvalue=2.41), as did all three items from the Empathic Concern (Bullying) subscale, accounting for 25.7% of the variance (eigenvalue=2.31), with the exception of one cross-loading. Results by gender were consistent with those obtained with the whole sample (see Appendix J for factor analytic results by gender). 56 Table 10 Factor Loadings for Bullying-Specific Perspective-taking, Empathic Distress, and Empathic Concern (Whole Sample) Item When I see someone getting verbally bullied (called names, threatened), I try to understand how they feel. When I see someone getting socially bullied (gossiped about, left out of the group), I try to understand how they feel. When I see someone getting physically bullied (hit, kicked), I try to understand how they feel. When I see someone getting socially bullied (gossiped about, left out of the group), I feel very uneasy and nervous. When I see someone getting physically bullied (hit, kicked), I feel very uneasy and nervous. When I see someone getting verbally bullied (called names, threatened), I feel very uneasy and nervous. I feel sorry for a student who is getting verbally bullied (called names, threatened). I feel sorry for a student who is getting physically bullied (hit, kicked). I feel sorry for a student who is getting socially bullied (gossiped about, left out of the group). Factor Perspective- Empathic Empathic taking Distress Concern (Bullying) (Bullying) (Bullying) .827 .772 .772 .817 .814 .812 (.438) .841 .783 .708 Note. All other factor loadings were less than .400. Cross-loadings greater than .400 are shown in parentheses. Cronbach's alpha scores for the three bullying-specific, secondary subscales were as follows: Empathic Concern (Bullying) scale, a=.86; Perspective-taking (Bullying) scale, a=.88; and Empathic Distress (Bullying) scale, a=.86. These results indicate good internal reliability across subscales. Inter-correlations (Pearson Product Moment Correlations) among the three bullying- specific empathy subscales were also computed. Results, presented in Table 11, revealed moderately high inter-correlations among scales. Despite these correlations, each of the three 57 bullying-specific subscales were retained for subsequent analyses, given their conceptual distinctiveness. Table 11 Correlations Between Bullying-Specific Empathy Subscales 1 2 3 1. Empathic Concern (Bullying) 2. Empathic Distress (Bullying) .61** 3. Perspective-taking (Bullying) .72** .63** Note. **p=.01 (one-tailed tests). Correlations between the bullying-specific empathy subscales and the primary empathy subscales are presented in Table 12. Ai r correlations were significant, and ranged from moderate to high. As expected, the highest correlations were typically found between the dispositional and situational measures of the same facets of empathy. An exception to this was the equal correlation between Perspective-taking (Bullying) and both Perspective-taking and Empathic Concern. This strong inter-relationship is consistent with the high inter-correlations found between Empathic Concern and Perspective-taking for both the general scales, and the bullying-specific scales. Table 12 Correlations Between Primary and Bullying-Specific Empathy Subscales Empathic Empathic Perspective- Concern Distress taking (Bullying) (Bullying) (Bullying) Empathic Concern .85** .67** .79** Empathic Distress .39** .74** .45** Perspective-taking .68** .55** .79** Note. **p=.01 (one-tailed tests). 58 Primary Analyses Overview. A series of two-way, univariate A N O V A ' s 1 3 were performed to address the primary research questions of the current study. The aim of the study was to determine whether or not students identified in each of the bullying-related participant roles varied in terms of three forms of empathy (Empathic Concern, Empathic Distress, Perspective-taking). This was primarily evaluated using a Participant Role x Gender A N O V A design, first examining participant roles as identified by peers (4 X 2), and second examining participant roles as identified through self-evaluations (3 X 2). For the peer-identified roles, the four participant roles (Pro-bully, Defender, Outsider, Victim) were of primary interest, but the six broader participant roles (including Bully, Assistant, Reinforcer) were also of interest. For self-reported participant roles, a similar Participant Role (Defender, Outsider, Victim) by Gender A N O V A was conducted. Dispositional measures tapping three different aspects of empathy (Empathic Concern, Empathic Distress, Perspective-Taking), as well as the single item "It's easy for me to understand why other people do the things they do" (Understanding factor) 1 4 and bullying- specific empathy measures tapping three aspects of empathy (Empathic Concern, Empathic Distress, Perspective-Taking) were included as dependent variables. Significant main effects in all analyses were followed up with the Tukey-Kramer 1 5 post-hoc test, unless otherwise stated. Question #1. Do children who assume various participant roles in response to bullying differ according to their tendency to experience other-oriented empathic concern (i.e., a facet of affective empathy) for others in distress? 1 3 Huberty and Morris (1989) support the use of multiple A N O V A ' s with exploratory research examining equivalence of groups, such as the current study, and when previous studies have used univariate analyses. Given this, as well as the empirical evidence that describes the theoretical distinctions between facets of empathy, a M A N O V A approach was considered inappropriate. 1 4 This single item was included in analyses because it loaded independently in factor analyses. 1 5 The Tukey-Kramer test is recommended for use by Jaccard, Becker, and Wood (1984) when cell sizes are unequal, but assumptions of the statistics are not otherwise violated, as is the case here. 59 Differences in Empathic Concern by participant role and gender were examined via a 4 (Peer-reported Participant Role: Pro-bully, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 13). Significant main effects were observed for both Peer-nominated Participant Role F(3, 349) = 3.14, p =.025, r\2 = .03 (four roles), as well as Gender F(1, 349) = 19.80, p <.001, r| 2 = .05. The interaction between Participant Role (four roles) and Gender was not statistically significant F(3, 349) = 2.24, ns. The main effect of Gender on Empathic Concern was significant, with girls reporting higher Empathic Concern than boys {Mgirls = 3.50, SD = 0.89; /W b o y s = 2.91, SD = 1.00). Follow-up post-hoc analyses demonstrated that participants in the Pro-Bully role reported significantly lower Empathic Concern than all other roles, as shown in the table below. Table 13 Mean Levels of Empathic Concern (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample Pro-Bully Defender Outsider Victim M 2.86 a 3.53 b 3.20 b 3.34 b (SD) (1.02) (0.89) (0.97) (0.92) n per cell n = 117 n = 97 n = 82 n = 61 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Results of a 6 (Peer-reported Participant Role: Bully, Assistant, Reinforcer, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A revealed significant differences in Empathic Concern as a function of both Gender, F(1, 345) = 25.40, p < .001, r\2 = .07, and Participant Role, F(5, 345) = 2.28, p = .047, i f = .03 (six roles) (see Table 14). The interaction between Participant Role (six roles) and Gender was not statistically significant, F(5, 345) = 1.44, ns. In 60 terms of gender, girls again reported higher Empathic Concern than boys (Mgiris = 3.50, SD = 0.89; /Wboys = 2.91, SD = 1.00). With respect to participant roles, post-hoc analyses revealed that participants in the Assistant role reported significantly lower Empathic Concern than Victims and Defenders. Those in the Reinforcer and Bully roles reported significantly lower Empathic Concern than participants in the Defender role. Table 14 Mean Levels of Empathic Concern (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample Bully Assistant Reinforcer Defender Outsider Victim M 3.00 a, b 2.69 a 2.84 a, b 3.53 c 3.20 a i b , c 3 .34 b c (SD) (0.96) (0.98) (1.07) (0.89) (0.97) (0.92) n per cell n = 35 n = 21 n = 61 n = 97 n = 82 n = 61 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Variations in Empathic Concern by self-reported participant role and gender were examined via a 3 (Self-reported Participant Role: Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) ANOVA. Significant main effects were found for both Self-reported Participant Role, F(2, 321) = 25.87, p < .001, r i 2 = .14 (three roles), as well as Gender, F(1, 321) = 16.41, p < .001, r\2 = .05 (see Table 15). The interaction was not statistically significant, F(2, 321) = 1.36, ns. Girls reported higher overall Empathic Concern than boys (Mgiris = 3.56, SD = 0.85; Mbovs = 3.09, SD = 0.93). Results of post-hoc analyses demonstrated that Outsiders reported significantly lower Empathic Concern than both Victims and Defenders. Victims also reported significantly lower Empathic Concern than Defenders. 61 Table 15 Mean Levels of Empathic Concern (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample Defender Outsider Victim M 3.71c (0.86) n = 183 2.74a (0.85) n = 59 3.20b (0.88) n = 85 (SD) n per cell Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Question #2. Do children who assume various participant roles in response to bullying differ in terms of their tendency to experience self-focused empathic distress (i.e., a facet of affective empathy) in response to others in distress? Variations in Empathic Distress by participant role and gender were examined via a 4 (Peer-reported Participant Role: Pro-bully, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 16). It was found that a significant main effect existed for Gender, F(1, 345) = 10.14, p .002, i f = .03, with girls reporting higher Empathic Distress than boys (Mgir,s = 2.62, SD = 0.98; Mboys = 2.20, SD = 0.97). However, the main effect of Peer-nominated Participant Role (four roles) was not significant, F(3, 345) = .966, ns, nor was the interaction between Participant Role (four roles) and Gender (F(3, 345) = 1.04, ns). 62 Table 16 Mean Levels of Empathic Distress (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample Pro-Bully Defender Outsider Victim M 2.20 2.56 2.42 2.58 (SD) (0.97) (0.96) (0.89) (1.18) n per cell n = 116 n = 96 n = 81 n = 60 Next considered were variations in Empathic Distress across all six peer-identified participant roles, analyzed using a 6 (Peer-reported Participant Role: Bully, Assistant, Reinforcer, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 17). Results revealed a significant main effect for Gender F(1, 341) = 7.82, p = .005, r\2 = .02, but not Participant Role F(5, 341) = 1.40, ns (six roles). The interaction between Participant Role (six roles) and Gender was also not statistically significant F(5, 341) = 1.06, ns. As before, girls reported higher overall Empathic Distress than boys (Mgiris = 2.62, SD = 0.98; M b o y s = 2.20, SD = 0.97). Table 17 Mean Levels of Empathic Distress (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample Bully Assistant Reinforcer Defender Outsider Victim M 2.48 2.09 2.09 2.56 2.42 2.58 (SD) (1.12) (0.8) (0.90) (0.96) (0.89) (1.18) n per cell n = 35 n = 21 n = 60 n = 96 n = 81 n = 60 63 The variations in Empathic Distress across self-reported participant role and gender were examined via a 3 (Self-Re ported Participant Role: Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 18). Significant main effects were observed for both Self- Reported Participant Role F(2, 318) = 3.99, p = .02, i f = .02 (three roles), as well as Gender, F(1, 318) = 9.10, p = .003, i f = .03. The interaction between Self-reported Participant Role (three roles) and Gender was not statistically significant, F(2, 318) = 2.42, ns. As was found for peer-nominated participant roles, girls reported significantly higher overall Empathic Distress than boys (Mgiris = 2.71, SD = 1.01; Mboys = 2.34, SD = 0.97). Post-hoc analyses 1 6 were performed to determine which participant role groups differed in terms of Empathic Distress. These analyses revealed that Outsiders reported significantly lower Empathic Distress than did Victims or Defenders, with no significant differences between the latter two groups. Table 18 Mean Levels of Empathic Distress (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample Defender Outsider Victim M 2.57b 2.22a 2.71b (SD) (0.99) (0.90) (1.14) n per cell n = 182 n = 59 n = 83 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Question #3. Do children who assume various participant roles in response to bullying differ in terms of their tendency to spontaneously adopt the perspectives of others (i.e., experience a cognitive facet of empathy)? 1 6 Given unequal variances within the current analysis, the Games Howell post-hoc test was employed. Jaccard, Becker, and Wood (1984) recommend that Games Howell is most appropriate when variances and cell sizes are unequal, but assumptions of A N O V A are not otherwise violated. 64 Perspective-taking differences as a function of participant role and gender were examined via a 4 (Peer-reported Participant Role: Pro-bully, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 19). Significant main effects were found for both Peer-Nominated Participant Role, F(3, 349) = 4.66, p =.003, r | 2 = .04 (four roles), as well as Gender, F(1, 349) = 15.12, p <.001, r\2 = .04. The interaction between Participant Role (four roles) and Gender was not statistically significant, F(3, 349) = 1.40, ns. Girls reported higher overall Perspective-taking than boys (Mgir,s= 3.08, SD = 0.84; M b o y s = 2.58, SD = 0.87). Follow- up post-hoc analyses indicated that participants in the Pro-Bully role reported significantly lower Perspective-taking than all other roles. Table 19 Mean Levels of Perspective-taking (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample Pro-Bully Defender Outsider Victim M 2.51a 3.14b 2.86b 2.92b (SD) (0.87) (0.83) (0.80) (0.93) n per cell n = 117 n = 97 n = 82 n = 61 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Results of a 6 (Peer-reported Participant Role: Bully, Assistant, Reinforcer, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A revealed significant differences in Perspective-taking as a function of Gender, F(1, 345) = 18.14, p < .001, r\2 = .05, as well as Participant Role, F(5, 345) = 4.11, p = .001, TI2 = .06 (six roles) (see Table 20). The interaction between Participant Role (six roles) and Gender was not statistically significant F(5, 345) = 1.18, ns. As before, girls reported higher Perspective-taking than boys (Mgiris = 3.08, SD = 0.84; 65 /W b o y s= 2 . 58 , SD = 0 .87) . With regard to Participant Role, results of post-hoc analyses revealed that participants in the Assistant role reported significantly lower Perspective-taking than those in the Bully, Outsider, Victim, and Defender roles. In addition, those in the Reinforcer role reported significantly lower Perspective-taking than those in the Defender role. Table 2 0 Mean Levels of Perspective-taking (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample Bully Assistant Reinforcer Defender Outsider Victim M 2 . 6 9 b c 2 . 1 8 a 2 .51a b 3 . 1 4 C 2 . 8 6 b c 2 . 9 2 b > c (SD) (0 .95) (0 .71) (0 .86) (0 .83) (0 .80 ) (0 .93) n per cell n = 3 5 n = 21 n = 61 n = 9 7 n = 8 2 n = 61 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Differences in Perspective-taking by self-reported participant role and gender were examined via a 3 (Self-Reported Participant Role: Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 21 ) . Significant main effects were found for both Self-Reported Participant Role, F ( 2 , 3 2 1 ) = 1 0 . 6 3 , p < . 0 0 1 , r\2 = .06 (three roles), as well as Gender, F ( 1 , 3 2 1 ) = 2 0 . 0 4 , p < . 0 0 1 , if = .06 . The interaction between Self-Reported Participant Role (three roles) and Gender was not statistically significant F ( 2 , 3 2 1 ) = 0 .82 , ns. Girls reported higher Perspective-taking than boys {Mgirls = 3 . 1 7 , SD = 0 .83 ; Mboys = 2 . 7 1 , SD = 0 .84 ) . Post-hoc analyses for Participant Role showed that Outsiders and Victims reported significantly lower Perspective-taking than Defenders. 6 6 Table 21 Mean Levels of Perspective-taking (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample Defender Outsider Victim M 3.15b 2.57a 2.80a (SD) (0.84) (0.83) (0.82) n per cell n = 183 n = 59 n = 85 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Analyses with single-item understanding factor. Given findings of preliminary factor analyses that the item "It's easy for me to understand why other people do the things they do" did not load significantly onto any of the other empathy scales, it was added to primary analyses as a dependent variable, referred to as the Understanding factor. Accordingly, differences in responses to this item across children who assumed various participant roles were examined. Variations in Understanding by participant role and gender were examined via a 4 (Peer- reported Participant Role: Pro-bully, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 22). No significant main effects were found (Gender F(1, 346) = .06, ns, Peer-nominated Participant Role (four roles) F(3, 346) = 2.54, ns). The interaction was not significant either (Participant Role (four roles) x Gender F(3, 346) = 1.00, ns). 67 Table 22 Mean Levels of Understanding (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample Pro-Bully Defender Outsider Victim M 2.84 2.92 2.59 2.58 (SD) (1.17) (1.26) (1.05) (1.15) n per cell n = 116 n = 97 n = 82 n = 59 Variations in Understanding across all six peer-identified participant roles were analyzed using a 6 (Peer-reported Participant Role: Bully, Assistant, Reinforcer, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 23). No significant main effects were revealed (Gender F(1, 342) = .36, ns, Peer-nominated Participant Role (six roles) F(5, 342) = 1.76, ns). The interaction was also non-significant (Participant Role (six roles) X Gender F(5, 342) = .65, ns). Table 23 Mean Levels of Understanding (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample Bully Assistant Reinforcer Defender Outsider Victim M 2.89 2.55 2.90 2.92 2.59 2.58 (SD) (1.08) (1.23) ' (1.21) (1.26) (1.05) (1.15) n per cell n = 35 n = 20 n = 61 n = 97 n = 82 n = 59 The variations in Understanding across self-reported participant role and gender were examined via a 3 (Self-Reported Participant Role: Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 24). No significant main effects were found (Gender F(1, 318) = 68 .37, ns, Self-reported Participant Role (three roles) F(2, 318) = 2.50, ns). The interaction was not significant either (Participant Role (three roles) x Gender F(2, 318) = .29, ns). Table 24 Mean Levels of Understanding (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample Defender Outsider Victim M 2.79 2.38 2.70 (SD) (1.17) (1.07) (1.18) n per cell n = 182 n = 58 n = 84 Secondary Analyses How do children in various participant roles respond (in terms of Empathic Concern, Empathic Distress, and Perspective-taking) when another child is being bullied? As a topic of secondary interest, differences in empathy in response to bullying were examined. As with the primary analyses, the same two-way univariate A N O V A format was employed. Although the four participant roles (Pro-bully, Defender, Outsider, Victim) were the primary roles under evaluation, the six participant roles (including Bully, Assistant, Reinforcer), as well as the three self-reported participant roles (Defender, Outsider, Victim) were also evaluated as independent variables, in conjunction with gender. Significant main effects in all analyses were followed up with the Tukey-Kramer post-hoc test, unless otherwise stated. Empathic concern in response to bullying scenarios. Variations in Empathic Concern in response to bullying were evaluated in a 4 (Peer-reported Participant Role: Pro-bully, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) ANOVA (see Table 25). Peer-nominated Participant Role, F(3, 345) = 7.25, p <001, r\2 = .06 (four roles), as well as Gender, F(1, 345) = 9.71, p =.002, r| 2 = .03 had significant main effects on Empathic Concern responses to bullying 69 scenarios. The interaction between Participant Role (four roles) and Gender was not statistically significant, F(3, 345) = 2.28, ns. The main effect of Gender was evident in that girls reported higher overall Empathic Concern in response to bullying (M g i r|S= 3.74, SD = 0.99; M b o y s = 3.15, SD = 1.16). Post-hoc analyses demonstrated that participants in the Pro-Bully role reported significantly lower Empathic Concern in response to bullying than all other roles within the four-role solution. In addition, those within the Outsider role reported significantly lower Empathic Concern in response to bullying than those within the Defender role. Table 25 Mean Levels of Empathic Concern in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample Pro-Bully Defender Outsider Victim M 2.97 a 3.94 c 3.43 b 3.58 b i c (SD) (1.14) (0.92) (1.09) (1.02) n per cell n = 115 n = 96 n = 82 n = 60 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. In terms of the 6 (Peer-reported Participant Role: Bully, Assistant, Reinforcer, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) univariate A N O V A performed on Empathic Concern-Bullying, significant main effects were found for both Participant Role, F(5, 341) = 4.67, p <001, i f = .06 (six roles), as well as Gender, F(1, 341) = 15.01, p <001, i f = .04 (see Table 26). As with the four participant roles, the interaction between Participant Role (six roles) and Gender was not significant, F(5, 341) = 1.65, ns. With respect to gender, girls again reported overall higher levels of Empathic Concern in response to bullying than boys, Mgiris = 3.74, SD = 0.99; Mboys = 3.15, SD = 1.16). Follow-up post-hoc analyses on the main effect of participant 70 role revealed that Assistants reported significantly lower Empathic Concern in response to bullying than Outsiders, Victims, and Defenders. Reinforcers' reports were also significantly lower than Victims and Defenders. Bullies' reports of Empathic Concern were significantly lower than Defenders'. Table 26 Mean Levels of Empathic Concern in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample Bully Assistant Reinforcer Defender Outsider Victim M 3.07 a, b, c 2.83a 2.96a, b 3.94d 3.43b c d 3.58c d (SD) (1.13) (1.23) (1.14) (0.92) (1.09) (1.02) n per n = 35 n = 21 n = 59 n = 96 n = 82 n = 60 cell Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. A 3 (Self-Reported Participant Role: Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A was performed on Empathic Concern in response to bullying (see Table 27). Significant main effects were found for both Self-Reported Participant Role, F(2, 319) = 28.67, p < .001, i f = .15 (three roles), as well as Gender, F(1, 319) = 12.02, p = .001, r| 2 = .04. The interaction was not statistically significant, F(2, 319) = 0.84, ns. Girls reported higher overall levels of Empathic Concern than boys {Mgirls= 3.80, SD = 0.95; Mboys = 3.35, SD = 1.09). Results of post-hoc analyses revealed significant differences in Empathic Concern in response to bullying reported by participants across all three roles. Outsiders reported the lowest Empathic Concern-Bullying, followed by Victims, and Defenders. 7 1 Table 27 Mean Levels of Empathic Concern in Response to Bullying Scenarios (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample Defender Outsider Victim M 3.93 c 2.85 a 3.38 b (SD) (0.92) (1.01) (0.98) n per cell n = 182 n = 59 n = 84 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Empathic distress in response to bullying scenarios. A 4 (Peer-reported Participant Role: Pro-bully, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A was conducted to examine differences in Empathic Distress in response to bullying scenarios (see Table 28). Gender had a significant main effect on Empathic Distress in response to bullying, F(1, 348) = 11.78, p =.001, n,2 = .03, but Peer-nominated Participant Role, F(1, 348) = 1.80, ns (four roles), did not. The interaction between Participant Role (four roles) and Gender was not statistically significant either, F(3, 348) = 1.12, ns. The significant main effect of Gender on bullying-specific Empathic Distress revealed that girls reported higher Empathic Distress-Bullying than boys (Mgirls= 2.72, SD= 1.08; M b o y s =2 .17 , S D = 1.05). Table 28 Mean Levels of Empathic Distress in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample Pro-Bully Defender Outsider Victim M 2.13 2.75 2.48 2.51 (SD) (1.01) (1.06) (1.09) (1.20) n per cell n = 117 n = 97 n = 81 n = 61 72 A 6 (Peer-reported Participant Role: Bully, Assistant, Reinforcer, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A was performed on Empathic Distress in response to bullying (see Table 29). A significant main effect was found for Gender. F(1, 344) = 10.48, p = 001, n,2 = .03 (six roles), but not Participant Role, F(5, 344) = 1.41, ns. The interaction between Participant Role (six roles) and Gender was not significant, F(5, 344) = 1.00, ns. With respect to gender, girls reported higher mean levels of Empathic Distress in response to bullying than boys (Mgiris = 2.72, SD = 1.08; M b o y s = 2.17, SD = 1.05). Table 29 Mean Levels of Empathic Distress in Response to Bullying Scenarios (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample Bully Assistant Reinforcer Defender Outsider Victim M 2.15 1.97 2.16 2.75 2.48 2.51 (SD) (0.93) (0.84) (1.11) (1.06) (1.09) (1.20) nper n = 35 n = 21 n = 61 n = 97 n = 81 n = 61 cell The 3 (Self-Reported Participant Role: Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A performed on Empathic Distress in response to bullying revealed significant main effects for both Self-reported Participant Role, F(2, 320) = 9.35, p < .001, rr2 = .06 (three roles), as well as Gender, F(1, 320) = 13.08, p < .001, r\2 = .04 (see Table 30). The interaction between Self-reported Participant Role (three roles) and Gender was not statistically significant, F(2, 320) = 1.07, ns. Girls reported higher overall Empathic Distress in response to bullying than boys (M g / r i s = 2.81, SD = 1.11; /W b o y s= 2.30, SD = 1.05). Post-hoc analyses demonstrated that Outsiders reported significantly lower Empathic Distress in response to bullying than Victims and Defenders. 73 Table 30 Mean Levels of Empathic Distress in Response to Bullying Scenarios (and Standard Deviations) by Self-Reported Participant Role (3 Roles) for Whole Sample Defender Outsider Victim M (SD) n per cell 2.71 b (1.10) n = 182 1.99a (0.91) n = 59 n = 85 2.68 b (1.14) Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. Perspective-taking in response to bullying scenarios. Variations in Perspective-taking in response to bullying by participant role and gender were examined via a 4 (Peer-reported Participant Role: Pro-bully, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A (see Table 31). Significant main effects were found for both Peer-nominated Participant Role, F(3, 347) = 4.78, p =.003, n,2 = .04 (four roles), as well as Gender, F(1, 347) = 10.86, p = .001, i f = .03. The interaction between Participant Role (four roles) and Gender was not statistically significant, F(3, 347) = 1.71, ns. Girls demonstrated significantly higher levels of Perspective- taking, as compared to boys {Mgirls = 3.18, SD = 1.03; M b o y s = 2.62, SD = 1.14). Post-hoc analyses demonstrated that participants within the Pro-bully role reported the lowest level of bullying-specific Perspective-taking, which was significantly lower than that reported by Outsiders, Victims, and Defenders. 74 Table 31 Mean Levels of Perspective-taking in Response to Bullying (and Standard Deviations) by Peer-Nominated Participant Role (4 Roles) for Whole Sample Pro-Bully Defender Outsider Victim M 2.50 a 3.29 b 2.95 b 2.98 b (SD) (1.07) (1.06) (1.04) (1.18) n per cell n = 115 n = 97 n = 82 n = 61 Note. *p = .05 Significant differences between groups from post-hoc analyses are indicated by differing subscripts. As in the analysis with four roles, the 6 (Peer-reported Participant Role: Bully, Assistant, Reinforcer, Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) A N O V A found significant main effects for Gender, F(1, 343) = 11.38, p =.001, xf = .03 (six roles), and Participant Role, F(5, 343) = 3.18, p = .008, n 2 = .04 (see Table 32). The interaction between Participant Role (six roles) and Gender was not significant, F(5, 343) = 1.26, ns. As before, girls' reports of Perspective-taking in response to bullying were higher than boys' (M g / r f s = 3.18, SD = 1.03; Mbovs = 2.62, SD = 1.14). Post-hoc analyses revealed that participants in the Assistant role reported significantly lower Perspective-taking in response to bullying than Outsiders, Victims, and Defenders. In addition, those in the Reinforcer role reported significantly lower Perspective- taking than those in the Defender role. 75 Table 32 Mean Levels of Perspective-taking in Response to Bullying (and Standard Deviations) by Peer-Nominated Participant Role (6 Roles) for Whole Sample Bully Assistant Reinforcer Defender Outsider Victim M 2 .73 a , b , c 2.33 a 2.44 a , b 3.29 c 2 .95 b , c 2.98 b , c (SD) (1.11) (1.05) (1.05) (1.06) (1.04) (1.18) n per cell n = 34 n = 21 n = 60 n = 97 n = 82 n = 61 Note. *p = .05 Significant differences between groups from post-hoc analyses are indicated by differing subscripts. Using a 3 (Self-Reported Participant Role: Defender, Outsider, Victim) X 2 (Gender: Girls, Boys) ANOVA, differences in Perspective-taking in response to bullying were examined (see Table 33). Significant main effects were found for both Self-reported Participant Role, F(2, 319) = 18.65, p < .001, n,2 =.11 (three roles), as well as Gender, F(1, 319) = 10.16, p = .002, TI2 = .03. The interaction between Self-reported Participant Role (three roles) and Gender was not statistically significant, F(2, 319) = 0.08, ns. Girls reported significantly higher overall Perspective-taking in response to bullying scenarios than boys {Mgirls = 3.27, SD = 1.00; Mboys = 2.81, SD = 1.12). Post-hoc analyses showed that Outsiders reported significantly lower Perspective-taking in bullying scenarios than Victims and Defenders. 76 Table 33 Mean Levels of Perspective-taking in Response to Bullying (and Standard Deviations) by Self- Reported Participant Role (3 Roles) for Whole Sample Defender Outsider Victim M 3.32b 2.34a 2.97b (SD) (1.05) (0.92) (1.03) n per cell n = 183 n = 59 n = 83 Note. *p = .05 Significant differences between groups from post-hoc analyses indicated by differing subscripts. 77 Discussion The current study sought to investigate whether or not differences in various facets of empathy were evident across bullying participant roles. To this end, students in grades 5 to 7 nominated individuals from their classrooms who fit particular behavioural descriptions depicting the participant roles, and also completed self-report indices of participant role behaviour and three different aspects of empathy (empathic distress, empathic concern and perspective taking). Results of the current study demonstrated significant differences in empathy across participant roles, but those differences were not always in the directions predicted, based upon past research and current theory. Accordingly, the current findings provide confirmation of much of the existing literature on both empathy and bullying, and yet extend this literature by illuminating relevant new avenues for both research and practice. The participant role approach to studying bullying developed by Salmivalli and colleagues (1996) has broadened the focus of bullying research, beyond merely examining the bully and victim roles, to addressing the critical role played by all individuals who are present when bullying occurs. Participant roles within past research have included those who behaved in ways that were supportive of bullying (i.e., Bullies, Assistants, Reinforcers), oppositional to bullying (i.e., Defenders), uninvolved in bullying (i.e., Outsiders), and victimized by bullying (i.e., Victims) (Salmivalli et al., 1996, 1998; Sutton & Smith, 1999). The proportion of individuals identified within the various roles from the current sample were quite similar to past research with a similarly-aged sample of students (grades 6 and 8), involving peer-nominations of participant role behaviours (Salmivalli et al., 1998). For ease of comparison, the proportions of individuals in each of the participant roles in the current study are presented here, with those found by Salmivalli and colleagues (1998), in parentheses (grades 6 and 8, respectively). Within the current sample, Victims represented 16.1% (11%, 5%), Outsiders 21.3% (26%, 30%), Defenders 24.2% (17%, 20%), Bullies 8.6% (8%, 10%), Assistants 5.4% (6%, 13%), and 78 Reinforcers 15.4% (17%, 16%). Thus, the distribution of students across these participant roles is strikingly similar across samples from different Western countries. As in previous studies (Salmivalli, Lagerspetz, et al., 1996; Salmivalli et al., 1998; Sutton & Smith, 1999), the six conceptual participant roles did not emerge separately in factor analyses within the current study. Instead, four distinct roles emerged: Pro-bully, Defender, Outsider, and Victim. Although four composites emerged from factor analysis, the current study, like previous studies with the P R Q which have also found four composites (Salmivalli, Lagerspetz, et al., 1996; Salmivalli et al., 1998; Sutton & Smith, 1999), sought to investigate the theoretical distinctions between each of the six participant roles. The rationale behind these distinctions is essentially that the three roles which compose the Pro-bully factor (Bully, Assistant, Reinforcer) represent levels of bullying behaviour that vary in terms of directness (Sutton & Smith, 1999). Certainly, the face validity of the items supports the distinction between the specific roles. For example, "Who starts the bullying?" and "Who leads a gang?" refer to specifically instigator or leader-like behaviours, as opposed to "Who helps to catch the victim for the bully?" or "Who joins in the bullying if someone else has started it?", which are distinctly less ringleader-like, and more supporter or helper-like. Additionally, "Who laughs at the person being bullied?" and "Who gets others to come and watch?" refer to relatively passive or indirectly supportive behaviours. Despite this conceptual argument, the results of factor analyses in this and other studies (e.g., Salmivalli, Lagerspetz, et al., 1996; Sutton & Smith, 1999) do not support the uniqueness of three separate pro-bullying roles (i.e., Bully, Assistant, Reinforcer). Perhaps children do not differentiate between pro-bullying behaviours according to whether they are instigator-like, or not, and seem to see all individuals who support bullying in some way as bullies. Given that peer evaluations are used to identify participant roles, it may also be the case that the same children participate across these three "pro-bully" roles, making distinctions difficult to observe in factor analytic results. Nevertheless, given expected differences between pro-bully participant 79 roles within the primary analyses of this study, the Pro-bully role was treated as unitary in primary analyses, but also as three conceptually distinct pro-bully roles in secondary analyses. The emergence of the same four participant roles, rather than six, was consistent when participant role behaviours were peer-nominated versus self-reported. Although the nature of the roles was the same with self-reports, the proportion of individuals who identified themselves within each role varied considerably from peer-nominations. This finding confirmed results with self-reports of participant roles found in previous research (Salmivalli, Lagerspetz, et al., 1996; Sutton & Smith, 1999), in that there was an apparent over-representation of individuals within the Defender, Outsider, and Victim roles, and an extreme under-representation of individuals within the Pro-bully roles when self reports were considered. This disparity between peer- and self-reports suggests that self-reported roles may not truly reflect participants' behaviours. That is, children may want to represent their behaviour as prosocial, or at least not aggressive or distinctly anti-social, and therefore respond that they behave in Defender, Outsider, or Victim ways. As outlined in the literature review of this thesis, there is unresolved debate as to whether or not self-reports may under-estimate the true extent of bullying (Juvonen, Nishina, & Graham, 2001; Ladd & Kochenderfer-Ladd, 2002; Pellegrini, 1998, 2001; Pellegrini & Bartini, 2000). Given the inconsistency between peer and self-reports, future research using the Participant Role Questionnaire may find peer-reports to be the most informative. Having said that, collecting both for the purposes of comparison is useful in revealing interesting discrepancies. An avenue for future investigation might be to analyze the variables linked to consistency or inconsistency of self- and peer-reports on the P R Q . Conceptually, the three aspects of empathy examined within the current study are distinct in terms of their links to behaviour. Specifically, empathic concern (i.e., other-oriented concern) is thought to motivate helping behaviour out of an altruistic concern for the welfare of others (Batson, 1998), and reduce aggression (Miller & Eisenberg, 1988). Empathic distress 80 (i.e., self-focused distress) is thought to motivate helping only in order to relieve personal distress, and consequently, only when escaping another person in distress is not an option (Hoffman, 2001). Some research has found increased empathic distress to be associated with increased antisocial and aggressive behaviours (Cohen & Strayer, 1996). Enhanced perspective-taking (i.e., a cognitive facet of empathy) is thought to motivate prosocial behaviours (Eisenberg, 2000) and decrease aggression (Richardson et al., 1994) through understanding of others' viewpoints. In contrast, recent research has found that bullies performed better than other children on a task assessing skills similar to perspective-taking (Sutton et al., 1999a), suggesting that heightened perspective-taking skills may be linked to increased aggression in the case of bullies. Despite conceptual distinctions between the three aspects of empathy investigated here, the current study found that all three were significantly, positively inter-correlated. The inter- correlations were typically moderate, suggesting some degree of non-shared variance, although the relationship between Empathic Concern and Perspective-taking was moderately high and there were several items that cross-loaded across these two factors in the present sample. This is perhaps not surprising, given the developmental theory that affective concern for others increases as a consequence of increased perspective-taking that develops with age (Hoffman, 2001). However, given the aforementioned finding suggesting a positive relationship between bullying and perspective-taking (Sutton et al., 1999a), as well as a negative relationship between bullying and empathic concern (Endresen & Olweus, 2002), it was expected that Empathic Concern and Perspective-taking would be independent of each other. Certainly, they were expected to display unique links to bullying-related behaviours. Given some conceptual similarities and the statistical evidence found here, it may be most relevant for future empathy research to treat Empathic Concern and Perspective-taking as a unitary construct. 81 Alternatively, differences in these facets of empathy might be evident if finer, more varied indices of empathy were employed (e.g., physiological measures, experimental manipulations). The present study examined multiple dimensions of empathy, comparing dispositional (empathy across various scenarios) as well as situational (empathy in response to bullying situations per se) empathy measures. Overall, the situational empathy measures demonstrated results that were highly consistent with the more general empathy responses. Such consistency was evident in high correlations between the situational and dispositional measures of the same facets of empathy, as well as the same pattern of results in terms of the three primary aspects of empathy occurring across participant roles. The similarity of results across the primary and bullying-specific empathy scales has implications for the broader study of empathy, in that it suggests that empathy is universal across situations. That is, empathy is not situation-specific. However, there is some research evidence to refute this, and to demonstrate that empathy changes as a result of specific factors within the stimulus scenario (Bryant, 1982; Endresen & Olweus, 2001). Further research should examine whether or not empathy changes as a result of the personal characteristics of the bully and victim (i.e., gender, age, ethnicity) and the nature of the type of bullying (i.e., physical, verbal, relational). It may be that these factors influence how empathy is experienced. A limitation of the current bullying-specific empathy items was that they did not make such distinctions. Across empathy measures, female participants reported higher empathy than male participants. These findings are consistent with past research with children using dispositional, self-report empathy indices (e.g., Bryant, 1982; Fabes et al., 1993; Olweus & Endresen, 1998). Although some researchers have suggested that gender differences in favour of females' self- reported empathy are due, in part, to social desirability biases (i.e., females have a greater desire than males to appear empathic to others), research has generally failed to find a 82 significant relationship between measures of empathy and social desirability (e.g., Bryant, 1982; Mehrabian & Epstein, 1972). In addition, recent evidence points to further association between gender and empathy, in that interactive effects between sex of participant and sex of stimulus (i.e., person in distress) have been found (Olweus & Endresen, 1998). That is, while females were found to be more empathic towards both boys and girls, boys tended to be more empathic towards girls, but less empathic towards other boys. This complex pattern of results seems unlikely to be attributable to socially desirable responding. Certainly, a limitation of the current study is its neglect to assess differences in empathy according to the gender of the stimulus person in distress. Given results indicating empathy differences by gender of stimulus, as well as the results from the current study indicating empathy differences by gender of respondent with both cognitive and affective facets of empathy, it may be important to measure empathy by differentiating the stimulus by gender in future empathy research. That is, constructing the empathy-eliciting scenarios so that some involve boys in distress, and others, girls in distress. Primary analyses within the current study revealed that students across the four major, peer-nominated participant roles reported significantly different levels of Empathic Concern. Specifically, students nominated for the Pro-bully role reported significantly lower Empathic Concern than students within the Defender, Victim, and Outsider roles. The significant finding of lower Empathic Concern in children who behave in antisocial ways with respect to bullying, confirms the results from a previous meta-analysis of empathy-aggression research that found decreased affective empathy to be associated with increased aggression (Miller & Eisenberg, 1988). It also supports previous research that found empathic concern to be negatively associated with bullying behaviour (Endresen & Olweus, 2002). The current findings extend the existing research on empathy and aggression/bullying by considering not only bullying behaviour, but also a range of both pro- and anti-social bullying-related behaviours. 83 Across all participant roles, students' reports of self-focused Empathic Distress did not differ significantly accordingly to the participant roles they assumed. It is interesting to note that the mean of Empathic Distress responses for each of the four primary participant roles was below 3.00, indicating average responses to empathy-eliciting questions were low, in that they were not even "Kind of like me" (the response scored as a 3.00). By comparison, average responses on the Empathic Concern and Perspective-taking subscales were above 3.00 in at least one of the roles. The lack of significant differences in Empathic Distress across bullying- related participant roles supports previous research that failed to find a significant association between empathic distress and bullying behaviour (Endresen & Olweus, 2002). Primary analyses revealed that students reported significantly different Perspective- taking depending on their participant role. In particular, students nominated by their peers as supportive of bullying (i.e., Pro-bully role) reported lower tendencies to consider others' perspectives, as compared to students who took a stance against bullying (i.e., Defenders), those who were victimized (i.e., Victims), and those who remained uninvolved (i.e., Outsiders). This finding, and its consistency with the finding for Empathic Concern certainly supports Hoffman's (2001) theory that posits that increased perspective-taking leads to increased empathic concern. That is, those individuals who reported high levels of one of these facets of empathy, tended to report high levels of the other facet as well. Overall, the same general pattern of results was detected across both general, dispositional scales of empathy, and bullying-specific, situational empathy scales for all peer- nominated participant roles. In keeping with the general pattern of results found with the primary analyses, participants within the Pro-bully roles reported significantly lower Empathic Concern and Perspective-taking than those within the Outsider, Defender, and Victim roles, but not Empathic Distress. Given that situational-specific empathy responses to bullying scenarios had not been previously examined, the evidence provided within this secondary component of 84 the current study extends the existing literature by suggesting that empathy responses to bullying scenarios do not differ from more generic empathy responses. As an addition to the major findings of the current study involving peer-nominated participant roles, it is interesting to note that when self-reported participant roles were examined, significant differences were revealed across roles for all three facets of empathy. Defenders reported the highest Empathic Concern and Perspective-taking, followed by Victims and Outsiders, respectively. In terms of Empathic Distress, reported levels by Victims and Defenders did not differ significantly, but did differ from that of Outsiders, which was significantly lower. Approximately the same pattern of results was revealed with bullying-specific scales of empathy, with the exception that for Perspective-taking (Bullying) the reports of Defenders and Victims did not differ significantly. The finding that participants nominated by peers as supportive of bullying (i.e., Pro- bullies) reported significantly lower Empathic Concern for others than participants nominated by peers as defending against bullying, staying away from bullying, and being the targets of bullying, confirmed expectations. Given that children who carry out or support bullying are typically causing distress in others (i.e., their victims), it is not surprising to note that they report caring less about others' welfare. This result confirms research that has found decreased empathic concern associated with increased bullying (Endresen & Olweus, 2001), as well as numerous studies that have found decreased empathic concern to be associated with increased aggression (e.g., Carlo et al., 1998; Cohen & Strayer, 1996; Kaukiainen et al., 1999). The finding that Pro-bullies' self-reported Empathic Distress did not differ significantly from that reported by children within other roles, confirmed expectations. The prediction was based upon recent research findings that empathic distress did not differ according to the extent of self-reported bullying behaviour that adolescents engaged in (Endresen & Olweus, 2001). Accordingly, the current study extends the literature by demonstrating that individuals identified 85 by their peers as behaving in various ways that are supportive of bullying (as opposed to self- reports of only very direct, instigator-like bullying behaviours) did not report significantly different empathic distress than other children, within a slightly younger sample. Pro-bullies' reports of relatively lower Perspective-taking than all other roles contradict what was predicted within the current study. Based upon the recent theoretical argument that in order to be a successful bully, an individual must possess superior skills of psychological manipulation, including the ability to understand others' perspectives (Arsenio & Lemerise, 2001; Sutton et al., 1999a, b, c), students who were seen by their peers as Pro-bullies were expected to report relatively high Perspective-taking tendencies. Although previous results demonstrated that children who were identified as bullies had stronger skills in imagining others' perspectives on a limited set of tasks (Sutton et al., 1999a), they did not provide evidence of children's tendencies to imagine others' thoughts and feelings (i.e., cognitive empathy) across different situations. The current study went beyond past research by employing a dispositional, self-report measure of children's tendencies to imagine others' perspectives across social situations (i.e., cognitive empathy). Within this study, the lack of an overall finding in the direction predicted may be due to a failure within the Participant Role Questionnaire to distinguish between different types of bullying behaviour. That is, when bullying is differentiated by the nature of the behaviour (i.e., physical, verbal, relational), and therefore children's participant roles become increasingly specialized, perhaps differences in perspective-taking skills according to the bullying behavioural roles exist. There is some research evidence to suggest that this may be the case. Two separate studies that examined similar age groups to the current study have found positive, significant, relationships between social intelligence (a construct that they defined in a manner that was distinct from, and yet overlapped considerably with perspective-taking) and indirect aggression (a construct that they defined in a manner that was very similar to relational bullying, 86 discussed in the current thesis) (Bjorkvist et al., 2000; Kaukiainen et al., 1999). In contrast, these same studies did not find significant relationships between social intelligence and verbal and physical forms of aggression. This evidence suggests that perhaps a significant difference would be found within the Pro-bully role if that role were further differentiated by the nature of bullying behaviour. That is, perhaps individuals involved as Pro-bullies in relational bullying would report higher Perspective-taking, given that type of bullying is especially social in nature, and often requires "mastery" of the social arena, for which strong perspective-taking skills would presumably be beneficial. In future research, this will be a critical factor to address, and one that will require a much larger sample size than that used here. The finding that Defenders reported the highest level of Empathic Concern (although not significantly higher than that of Outsiders and Victims) essentially supported predictions based upon previous findings of a positive association between empathic concern and helping others (e.g., Eisenberg, 2000; Eisenberg & Miller, 1987). The current finding suggests that Batson's (1998) theory that it is empathic concern, and not empathic distress that motivates prosocial behaviour, is accurate with respect to behaviour within bullying scenarios. This is further supported by a lack of significant differences in Empathic Distress among participant roles. Defenders, as individuals who behave prosocially, reported higher tendencies to feel concern for others (i.e., Empathic Concern), which, according to Batson, is the source of an altruistic motive to help others. In providing evidence of empathy with respect to prosocial behaviours in response to bullying, the current study has made a novel contribution to the study of both empathy and bullying. As expected, Defenders reported relatively high Perspective-taking. That is, Defenders reported Perspective-taking that was significantly higher than Pro-Bullies (Assistants and Reinforcers when all six roles were examined). Given that Defenders' behaviour is highly prosocial, and in the case of bullying, represents the opposite of aggression, this result confirms 87 findings from a number of existing studies that have found inverse relationships between aggression and perspective-taking (e.g., Cohen & Strayer, 1996; Letourneau, 1981; Richardson et al., 1994). Although the high Perspective-taking reported by Defenders within the current study was not found in Sutton et al.'s study (1999a), this might be explained by the older sample employed here. That is, evidence that perspective-taking typically reaches full development around age nine (Hoffman, 1984, 1987, 2001) suggests that participants within the current study have reached their developmental potential in terms of perspective-taking skills, and might therefore present a different pattern of results as compared to younger samples. Although Outsiders did report higher Empathic Concern and Perspective-taking than Pro-bullies, their reports did not differ significantly from those of Defenders or Victims. With respect to the Outsider role, no clear predictions were initially made about the nature of Empathic Concern and Perspective-taking to be found. This was largely due to a lack of existing research involving students who remain uninvolved in bullying. In contrast, the majority of research has focused on linking facets of empathy to behaviour. For Outsiders, their lack of direct involvement in bullying may have previously been perceived by researchers as "non- behaviour". Considering the assertion by many empathy researchers (e.g., Batson, 1998; Eisenberg, 2000; Hoffman, 2001) that high empathic concern and perspective-taking lead to helping, and low levels of such facets of empathy lead to aggression, it is not surprising to note that Outsiders, who do not engage in particularly helpful or aggressive behaviour, according to their peers, reported Empathic Concern and Perspective-taking that were not extremely high or low, as compared to other children. Empathic Distress was expected to be lower in Outsiders within the current study; however, it was found not to differ significantly from reports by those within other roles. The prediction regarding Outsiders was based upon the contention that someone high in empathic distress would escape an aversive, distressing situation if given the chance (Hoffman, 2001), a 88 behaviour that appears to be similar to an Outsider's response (i.e., escape) to a bullying scenario. Given that the predicted finding did not emerge for Outsiders, it may be that Outsiders' efforts to stay uninvolved in bullying reflect a motive other than escape. Victims' reports of Empathic Concern and Perspective-taking were significantly higher than those of Pro-bullies, but did not differ significantly from those of Outsiders and Defenders. As with all of the other participant roles, no significant differences were found with respect to Empathic Distress. No specific predictions about the nature of empathy in children who are victimized by bullying were made before conducting the current study. Nevertheless, it is perhaps not surprising that Victims reported empathy that was relatively higher than that reported by Pro-bullies, given that Victims are the targets of Pro-bullies' aggression, and experience distressful situations very directly. Perhaps going through such experiences oneself leads Victims to greater concern for the welfare of others, and an increased tendency to imagine just how others feel when they are in distress. Limitations of the Current Study Despite the important contributions the current study has made to an understanding of the empathy-bullying link, there are a number of cautions to be exercised when interpreting the data from this research. Given that the empathy data within the current study was derived from self-reports, it may have been subject to socially desirable responding. This problem is rather difficult to overcome in that empathy is by nature, a very personal, and internalized experience. Although, as previously mentioned, empirical evidence has not found a significant relationship between empathy and socially desirable responding (e.g., Bryant, 1982), with self-report data, the risk does exist. A potential solution to the risk of participants responding according to how they would like to appear would be to employ a wide variety of measures, both situational (i.e., direct responses to real or simulated stimuli) and dispositional (i.e., responses about empathic 89 tendencies across stimuli), and compare, and possibly combine, results from the two types of measures. Another possible limitation of the current study is that of just over 500 students targeted for participation, only 82% actually took part. Although the percentage of students contacted who took part is sufficient to identify participant roles through peer-nominations, it is disappointing to note that of the 82% of students, an additional 9% could not be ultimately assigned to a participant role, given no clear pattern of nominations. In this respect, some participants' data could not be used in all analyses, and were lost. Limited sample size impacted the analyses performed in that some analyses could simply not be performed, while others must be interpreted with caution given low cell numbers, and may have failed to find significant results where they do exist, due to insufficient numbers. Unfortunately, analyses involving grade as an additional independent variable should have been performed, given significant differences among participant roles by grade. They could not be performed however, given that numbers in cells were too small to fulfill power requirements for statistical analyses (see Table 7 for numbers per cell). In this respect, error variance in the primary analyses was increased, and subsequently, variance that could have been otherwise explained, was likely lost. Overall, sample size would have to be increased in order to accommodate differences by grade in the future. Analyses that were performed to test for significant interaction effects of gender x participant role were all non-significant within the current study. Despite this, it may be that true differences did exist, but were obscured by insufficient numbers within some of the cells in the analyses. Certainly, future tests of such interactive effects should employ larger sample sizes in order to accommodate statistical power. Another limitation of the current study is that all self-reported participant roles could not be fully analyzed. This was due to extremely low numbers in some of the self-reported participant roles (i.e., Bully n = 1, Assistant n = 1, Reinforcer n = 4). Three of the self-reported 90 roles (Defender, Outsider, Victim) had sufficient frequencies to analyze them, but the information reported by children who did not fit into these roles was lost. Although some studies using the Participant Role Questionnaire have collected self-report data in the form of self- nominations (e.g., Sutton & Smith, 1999), no previous study had used the P R Q as a parallel self-report, examining the extent to which each student engaged in each of the participant role behaviours. Future use of the P R Q as a parallel self-report may provide informative comparisons to peer-nominated data; however its exclusive use may under-represent participants' true role behaviours. Caution should be taken when interpreting results from analyses using all six peer- nominated participant roles. This is because factor analysis results clearly showed that four factors emerged from both peer-nominations and self-reports. Although the decision to analyze both four and six role solutions was based on the theory behind the participant role approach to studying bullying (Salmivalli et al., 1996), and for comparison purposes to existing research, results from analyses using the six roles must still be interpreted with caution. This is critical because measurement error is substantially increased by maintaining six roles within analyses. An example of this is the high inter-correlations found between average number of nominations in each role, which are particularly high for those within Pro-bully roles. Nevertheless, there was remaining scope for non-shared variance, and this was of interest within this study. The final limitation of the current study is that the magnitude of effect sizes was small to moderate for all significant differences between group means within the current study. This is a limitation of the results from the current study, in that only minimal to moderate variance within the dependent variables was explained by the independent variables. It may have been that such effect sizes were detected due to inadequacies in instrumentation and sample size. For example, the significant differences in empathy between participant roles might be increased if participant roles were more finely differentiated, according to suggestions made earlier in the 91 Discussion section of this thesis (e.g., distinguishing between different types of bullies— physical, verbal, relational). Obviously, making such changes would require significantly increasing sample size. In so doing, interaction effects of participant role x gender, and perhaps even participant role x grade may be found, accounting for even more variance. Future Applications for Interventions in Schools The current study points to a number of factors critical to preventing and intervening effectively to curb bullying in school settings. Given the results of the current study, increasing students' empathy may be an effective way of motivating uninvolved children (Outsiders), and aggressive children (Bullies, Assistants, Reinforcers) to act prosocially when faced with bullying. Similarly, making use of highly prosocial children's (Defenders) empathy may be an effective way of mobilizing all students to take prosocial action against bullying. The questions that remain are "How can we build empathy in school children? Can it be done?" and "How can we involve prosocial children to mobilize others against bullying?". In response to the first question, there is certainly some empirical evidence that empathy can be fostered in children through explicit intervention strategies. For example, in their review of empathy-inducing socialization techniques, Mussen and Eisenberg (2001) found that inductions (using reasoning to try to influence children's behaviour), victim-oriented discipline (emphasizing concern for the victim), preachings (verbally highlighting the needs of others), and modeling (actually demonstrating all of the aforementioned skills, and thereby demonstrating both affective and cognitive aspects of empathy) have been found to increase children's empathy in empirical studies. Although this research evidence was based primarily on empathy-inducing techniques used by primary caregivers with children, some additional evidence exists to support the role that teachers and peers can play within schools to foster empathy. Feshbach and Feshbach (1982, 1986) did seminal work on the development, implementation, and evaluation of a school program designed to raise children's empathy 92 through small group activities such as role-plays and discussions of dilemmas and conflicts. They found that kindergarten to grade four children who were exposed to the empathy-training program demonstrated more cooperation, helping, and generosity (behaviours thought to indicate high levels of empathy) than children from control classrooms (Feshbach & Feshbach, 1986). More recently, Schonert-Reichl and Zaidman-Zait's research (as cited in Roots of Empathy, 2003) with schoolchildren in grades one to three found that children who took part in a comprehensive empathy-building program focused on the needs of a dependent infant, as well as emotional and social understanding showed improved emotion knowledge, social understanding, prosocial behaviour, decreased aggression and proactive aggression (e.g., bullying) from pre- to post-test following the empathy program. It is curious to note that although minimal research evidence exists to support the relationship between empathy and bullying, as well as the effectiveness of programs that aim to build empathy in order to counter bullying in schools, many programs of this nature have nevertheless been designed and implemented. Examples of such programs may be found in many countries, including the United Kingdom (e.g., the No Blame approach) (Robinson & Maines, 1997), the United States (e.g., Bully Busters) (Newman, Home, & Bartolomucci, 2000), and Canada (e.g., Roots of Empathy) (Roots of Empathy, 2001). Clearly, the effectiveness of efforts designed to foster empathy, as well as the subsequent impact of empathy-building programs on school bullying are important areas for future research. Prosocial children can be enlisted in helping to curb bullying school-wide. As the students identified as Defenders by their peers clearly have social reputations of being "prosocial", and "anti-bullying", they are likely to have a great deal of face validity with their peers when they work to promote anti-bullying efforts. In other words, these are students who "walk the walk and talk the talk" when it comes to countering bullying. Specific strategies that could be used to enlist these students' help include: training them to be peer counsellors or 9 3 mediators (Cowie & Sharp, 1996; Naylor & Cowie, 1999), having them "befriend" victimized children (Boulton, Trueman, Chau, Whitehand, & Amatya, 1999), having them lead an anti- bullying committee (Peterson & Rigby, 1999), or having them talk to less active children within the Pro-bully role, such as Assistants and Reinforcers, in order to illustrate to them the impact of bullying on others (Sutton et al., 1999a). The concept behind this strategy is that the social support network of the Bullies is targeted, and these children's behaviours might be more susceptible to change than the more entrenched behaviours of the leader Bully, or the repeated Victim. Efforts to increase the success of Defenders as anti-bullying peer helpers might involve combining use of the Participant Role Questionnaire with a sociometric assessment (Sutton et al., 1999a). In this way, Defenders seen by their peers as well-liked and prosocial could be enlisted, thereby increasing the likelihood that other students will listen and respond to them. 94 Conc lus ion Tradit ionally, research on the phenomenon of bullying focused on the perpetrators (i.e., bull ies) and vict ims (Sutton & Smi th , 1999). A l though past research provided much insight into the preva lence and c o n s e q u e n c e s of bul lying/vict imization, recent research increasingly identified the peer group a s a critical factor in the main tenance of bullying (Salmival l i , 1999). Al though such research uncovered the nature of chi ldren's behaviour in bullying scenar ios (i.e., participant roles), the factors that motivated and inf luenced such behaviour to occur were less c lear. T h e separate research tradition of empathy revealed inf luences on prosocia l (E isenberg & Miller, 1987), and aggress i ve (Miller & E i senbe rg , 1988) behaviour, yet until the current study, the link between empathy and bul lying-related behaviours had been largely unexplored. T h e current s tudy 's f indings of lower empath ic concern and perspect ive-taking in chi ldren who a s s u m e pro-bully roles, a s opposed to those who a s s u m e other behaviora l roles, and particularly prosocia l roles, have meaningful impl icat ions for the planning of anti-bullying interventions. Clear ly , those chi ldren with wel l -deve loped empathy and related prosocia l behaviours can be util ized proactively to counter bul lying. T h o s e chi ldren with poorly deve loped empathy and related ant isocia l or uninvolved behaviours can be targeted for empathy-bui ld ing. For the purposes of better informing school -wide bullying interventions, it is imperat ive that the empathy of all chi ldren who are present when bullying occurs be fully unders tood. Th is is precisely the empir ical contribution of the current study. 95 References Arsenio, W.F., & Lemerise, E.A. (2001). Varieties of childhood bullying: Values, emotion processes, and social competence. Social Development, 70(1), 59-73. Barnett, M.A., Howard, J.A., Melton, E.M., & Dino, G.A. (1982). Effect of inducing sadness about self or other on helping behaviour in high and low empathic children. Child Development, 53, 920-923. Batsche, G.M. , & Knoff, H.M. (1994). Bullies and their victims: Understanding a pervasive problem in the schools. School Psychology Review, 23(2), 165-175. Batson, C D . (1998). Altruism and prosocial behavior. In D.T. Gilbert, S T . Fiske, & G. Lindzey (Eds.), The Handbook of Social Psychology, 2, 282-316. Boston: McGraw-Hill. Batson, C D . , Batson, J .G . , Griffitt, C.A., Barrientos, S., Brandy, J.R., Sprengelmeyer, P. et al. (1989). Negative-state relief and the empathy-altruism hypothesis, dournal of Personality and Social Psychology, 56(6), 922-933. Batson, C D . , Batson, J .G . , Slingsby, J.K., Harrell, K., Peekna, H.M., & Matthew, T.R. (1991). Empathic joy and the empathy-altruism hypothesis. Journal of Personality and Social Psychology, 67(3), 413-426. Bjorkvist, K., Lagerspetz, K., & Kaukiainen, A. (1992). Do girls manipulate and boys fight?: Developmental trends in regard to direct and indirect aggression. Aggressive Behavior, 18, 117-127. Bjorkvist, K., Osterman, K., & Kaukiainen, A. (2000). Social intelligence - empathy = aggression?. Aggression and Violent Behavior, 5(2), 191-200. Borke, H. (1971). Interpersonal perception of young children: Egocentrism or empathy?. Developmental Psychology, 5(2), 263-269. 96 Boulton, M., Trueman, M., Chau, C , Whitehand, C , & Amatya, K. (1999). Concurrent and longitudinal links between friendship and peer victimization: Implications for befriending interventions. Journal of Adolescence, 22, 461-466. Boulton, M.J., & Underwood, K. (1992). Bully/victim problems among middle school children. British Journal of Educational Psychology, 62, 73-87. Bryant, B.K. (1982). An index of empathy for children and adolescents. Child Development, 53, 413-425, Carlo, G. , Roesch, S.C. , & Melby, J . (1998). The multiplicative relations of parenting and temperament to prosocial and antisocial behaviors in adolescence. Journal of Early Adolescence, 18{3), 266-290. Retrieved July 10, 2002, from the Academic Search Premier database. Caspi, A. (1998). Personality development across the life course. In N. Eisenberg (Ed.), Handbook of child psychology, (Vol. 3, pp.311-388). New York: Wiley. Charach, A., Pepler, D., Ziegler, S. (1995). Bullying at school: A Canadian perspective. Education Canada, (Spring 1995), 12-18. Cohen, D., & Strayer, J . (1996). Empathy in conduct-disordered and comparison youth. Developmental Psychology, 32(6), 988-998. Coie, J.D., & Dodge, K.A. (1988). Multiple sources of data on social behavior and social status in the school: A cross-age comparison. Child Development, 59, 815-829. Craig, W., & Pepler, D. (1997). Observations of bullying and victimization in the school yard. Canadian Journal of School Psychology, 13(2), 41-60. Cowie, H. & Sharp, S. (1996). Peer counselling in school: A time to listen. London: David Fulton. 97 Crick, N.R., & Bigbee, M.A. (1998). Relational and overt forms of peer victimization: A multi-informant approach. Journal of Consulting and Clinical Psychology, 66, 337- 347. Crick, N., & Grotpeter, J.K. (1995). Relational aggression, gender, and social- psychological adjustment. Child Development, 66, 710-722. Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44 (1), 113- 126. Davis, M. H. (1994). Empathy: A social psychological approach. Madison, WI: Brown & Benchmark. Davis, M.H., Hull, J .G . , Young, R.D., & Warren, G .G . (1987). Emotional reactions to dramatic film stimuli: The influence of cognitive and emotional empathy. Journal of Personality and Social Psychology, 52(1), 126-133. Deutsch, F.M., & Madle, R.A. (1975). Empathy: Historic and current conceptualizations, measurement, and a cognitive theoretical perspective. Human Development, 18, 267-287. Dovidio, J .F. , Allen, J.L., & Schroeder, D.A. (1990). Specificity of empathy-induced helping: Evidence for altruistic motivation. Journal of Personality and Social Psychology, 59(2), 249-260. Duan, C. & Hill, C. (1996). The current state of empathy research. Journal of Counseling Psychology, 43 (3), 261-274. Dymond, R. (1949). A scale for the measurement of empathic ability. Journal of Consulting Psychology, 13, 127-133. Eisenberg, N. (1986). Altruistic emotion, cognition, and behavior. Hillsdale, NJ : Lawrence Erlbaum Associates, Publishers. 9 8 Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review of Psychology, 51, 665-697. Eisenberg, N., Fabes, R., Bustamante, D., Mathy, R., Miller, P., & Lindholm, E. (1988). Differentiation of vicariously induced emotional reactions in children. Developmental Psychology, 24(2), 237-246. Eisenberg, N., Fabes, R., Murphy, B., Karbon, Maszk Patl, Smith, Melanie, et al. (1994). The relations of emotionality and regulation to dispositional and situational empathy-related responding. Journal of Personality and Social Psychology, 66(4), 776-797. Eisenberg, N., Fabes, R., Shepard, S., & Murphy, B. (1998). Contemporaneous and longitudinal prediction of children's sympathy from dispositional regulation and emotionality. Developmental Psychology, 34(5), 910-924. Eisenberg, N. & Miller, P. (1987). The relation of empathy to prosocial and related behaviors. Psychological Bulletin, 101 (1), 91-119. Eisenberg, N., Miller, P., Shell, R., McNalley, S., & Shea, C. (1991). Prosocial development in adolescence: A longitudinal study. Developmental Psychology, 27(5), 849-857. Eisenberg, N., & Strayer, J . (Eds.). (1987). Empathy and its development. New York: Cambridge University Press. Endresen, I.M., & Olweus, D. (2001). Self-reported empathy in Norwegian adolescents: Sex differences, age trends, ands relationship to bullying. In A .C . Bohart, & D.J. Stipek (Eds.), Constructive & destructive behavior: Implications for family, school, & society (pp. 147-165). Washington, DC: American Psychological Association. 99 Espelage, D.L., Mebane, S .E . , Adams, R.S. (in press). Bullying in American schools: A social- ecological perspective on prevention and intervention. Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers. Fabes, R., Eisenberg, N., & Eisenbud, L. (1993). Behavioral and physiological correlates of children's reactions to others in distress. Developmental Psychology, 29(4), 655-663. Feshbach, N.D. (1975). Empathy in children: Some theoretical and empirical considerations. Counseling Psychologist, 5(2), 25-30. Feshbach, N. D. (1982). Sex differences in empathy and social behavior in children. In N. Eisenberg (Ed.), The development of prosocial behavior. New York: Academic Press. Feshbach, N. D. (1987). Parental empathy and child adjustment/maladjustment. In N. Eisenberg, & J . Strayer (Eds.) Empathy and its development. New York: Cambridge University Press. Feshbach, N.D., & Feshbach, S. (1982). Empathy training and the regulation of aggression: Potentialities and limitations. Academic Psychology Bulletin, 4, 399-413. Feshbach, S., & Feshbach, N.D. (1986). Aggression and altruism: A personality perspective. In C. Zahn-Waxler, E.M. Cummings, & R. lannotti (Eds.) Altruism and aggression: Biological and social origins (pp. 189-217). Cambridge, England: Cambridge University Press. Feshbach, N. D., & Roe, K. (1968). Empathy in six- and seven-year-olds. Child Development, 39, 133-145. Goldstein, A .P . , & Michaels, G.Y. (Eds.). (1985). Empathy: Development, training, and consequences. Hillsdale, NJ : Lawrence Erlbaum Associates, Publishers. 100 Haynie, D.L., Nansel, T., Eitel, P., Crump, A.D., Saylor, K., Yu, K., & Simons-Morton, B. (2001). Bullies, victims, and bully/victims: A distinct group of at-risk youth. Journal of Early Adolescence, 27(1), 29-49. Retrieved February 17, 2002, from the Academic Search Elite database. Hoffman, M. (1975). Developmental synthesis of affect and cognition and its implications for altruistic motivation. Developmental Psychology, 11 (5), 607-622. Hoffman, M. (1984). Interaction of affect and cognition in empathy. In C E . Izard, J . Kagan, & Zajonc, R.B. (Eds.), Emotions, cognition, and behavior (pp. 103-131). New York: Cambridge University Press. Hoffman, M. (1987). The contribution of empathy to justice and moral judgment. In N. Eisenberg, & J . Strayer (Eds.), Empathy and its development (pp. 47-80). New York: Cambridge University Press. Hoffman, M. (2001). Toward a comprehensive empathy-based theory of prosocial moral development. In A .C . Bohart, & D.J. Stipek (Eds.), Constructive and destructive behavior: Implications for family, school, & society (pp. 61-86). Washington, DC: American Psychological Association. Hogan, R. (1969). Development of an empathy scale. Journal of Consulting and Clinical Psychology, 33(3), 307-316. Huberty, C .J . , & Morris, J.D. (1989). Multivariate analysis versus multiple univariate analyses. Psychological Bulletin, 705,302-308. Hughes, R., Tingle, B.A., & Sawin, D.B. (1981). Development of empathic understanding in children. Child Development, 52, 122-128. Jaccard, J . , Becker, M.A., & Wood, G. (1984). Pairwise multiple comparison procedures: A review. Psychological Bulletin, 96, 589-596. 101 Juvonen, J . , Nishina, A., & Graham, S. (2001). Self-views versus peer perceptions of victim status among early adolescents. In J . Juvonen, & S. Graham (Eds.) Peer harassment in school: The plight of the vulnerable and victimized (pp. 105-124). New York: The Guilford Press. Kaukiainen, A., Bjorkvist, K., Lagerspetz, K., Osterman, K., Salmivalli, C , Rothberg, S. et al. (1999). The relationships between social intelligence, empathy, and three types of aggression. Aggressive Behaviour, 25, 81-89. Knight, G., Johnson, L., Carlo, G., & Eisenberg, N. (1994). A multiplicative model of the dispositional antecedents of a prosocial behavior. Social Psychology, 66(1). Ladd, G.W., & Kochenderfer-Ladd, B. (2002). Identifying victims of peer aggression from early to middle childhood: Analysis of cross-informant data for concordance, estimation of relational adjustment, prevalence of victimization, and characteristics of identified victims. Psychological Assessment, 14(A), 74-96. Letourneau, C. (1981). Empathy and stress: How they affect aggression. Social Work, 26, 383-389. Marcus, R.F., Roke, E.J. , & Bruner, C. (1985). Verbal and nonverbal empathy and prediction of social behavior of young children. Perceptual and Motor Skills, 60, 299-309. Mehrabian, A. & Epstein, N. (1972). A measure of emotional empathy. Journal of Personality, 40, 525-543. Miller, P. A., & Eisenberg, N. (1988). The relation of empathy to aggressive and externalizing/antisocial behaviour. Psychological Bulletin, 103(3), 324-344. Miller. P.A., Eisenberg, N., Fabes, R., & Shell, R. (1996). Relations of moral reasoning and vicarious emotion to young children's prosocial behavior toward peers and adults. Developmental Psychology, 32(2), 210-219. 102 Mussen, P., & Eisenberg, N. (2001). Prosocial development in context. In A . C . Bohart, & D.J. Stipek (Eds.), Constructive and destructive behavior: Implications for family, school, & society (pp. 61-86). Washington, DC: American Psychological Association. Nansel, T.R., Overpeck, M., Pilla, R.S., Ruan, W.J . , Simons-Morton, B., Scheidt, P. (2001). Bullying behaviors among US youth: Prevalence and association with psychosocial adjustment. Journal of the American Medical Association (JAMA), 285, (16), 2094-2100. Naylor, P. & Cowie, H. (1999). The effectiveness of peer support systems in challenging school bullying: The perspectives and experiences of teachers and pupils. Journal of Adolescence, 22, 467-479. Newman, D.A., Home, A.M. , & Bartolomucci, C.L. (2000). Bully Busters: A teacher's manual for helping bullies, victims, and bystanders. Champaign, Illinois: Research Press. O'Connell, P., Pepler, D., & Craig, W. (1999). Peer involvement in bullying: Insights and challenges for intervention. Journal of Adolescence, 22, 437-452. Oliver, R., Hoover, J . , & Hazier, R. (1994). The perceived roles of bullying in small-town midwestern schools. Journal of Counseling and Development, 72(4), 416-421. Olweus, D. (1978). Aggression in the schools: Bullies and whipping boys. Washington, DC: Hemisphere Publishing Corporation. Olweus, D. (1991). Bully/victim problems among schoolchildren: Basic facts and effects of a school-based intervention program. In D.J. Pepler, & K.H. Rubin (Eds.), The development and treatment of childhood aggression (pp. 411-448). Hillsdale, NJ Lawrence Erlbaum Associates, Publishers. 103 Olweus, D. (1992). Bullying among schoolchildren: Intervention and prevention. In R.DeV. Peters, R.J. McMahon, & V.L. Quinsey (Eds.), Aggression and violence throughout the lifespan (pp. 100-125). Newbury Park, CA: Sage Publications. Olweus, D. (1993). Bullying at school: What we know, and what we can do. Oxford: Blackwell Publishers. Olweus, D. (1999). Norway. In P.K. Smith, Y. Morita, J . Junger-Tas, D. Olweus, R. Catalano, & P. Slee (Eds.), The nature of school bullying: A cross-national perspective (pp.28-48). New York: Routledge Olweus, D., & Endresen, I.M. (1998). The importance of sex-of-stimulus object: Age •trends and sex differences in empathic responsiveness. Social Development, 7(3), 370-388. Omdahl, B.L. (1995). Cognitive appraisal, emotion, and empathy. Mahwah, NJ: Lawrence Erlbaum Associates, Publishers. Osterman, K., Bjorkvist, K., Lagerspetz, K.M., Kaukiainen, A., Huesmann, L.R., & Fraczek, A. (1994). Peer and self-estimated aggression and victimization in 8-year old children from five ethnic groups. Aggressive Behavior, 20, 411-428. Pellegrini, A.D. (1998). Bullies and victims in school: A review and call for research. Journal of Applied Developmental Psychology, 19(2), 165-176. Pellegrini, A.D. (2001). Sampling instances of victimization in middle school: A methodological comparison. In J . Juvonen, & S. Graham (Eds.), Peer harassment in school: The plight of the vulnerable and victimized (pp. 125-143). New York: Guilford Press. Pellegrini, A.D., & Bartini, M. (2000). An empirical comparison of methods of sampling aggression and victimization in school settings. Journal of Educational Psychology, 92(2), 360-366. 104 Pepler, D.J., Craig, W., & Roberts, W.L. (1998). Observations of aggressive and nonaggressive children on the school playground. Merrill-Palmer Quarterly, 44(\), 55-76. Perry, D.G., Kusel, S. J . , & Perry, L.C. (1988). Victims of peer aggression. Developmental Psychology, 24(6), 807-814. Peterson, L. & Rigby, K. (1999). Countering bullying at an Australian secondary school with students as helpers. Journal of Adolescence, 22, 481-492. Richardson, D., Hammock, G., Smith, S., Gardner, W., & Signo, M. (1994). Empathy as a cognitive inhibitor of interpersonal aggression. Aggressive Behaviour, 20, 275- 289. Rivers, I. & Smith, P.K. (1994). Types of bullying behaviour and their correlates. Aggressive Behavior, 20, 359-368. Robinson, G. & Maines, B. (1997). Crying for Help: The no blame approach to bullying. Bristol: Lucky Duck Publishing. Roland, E. (2002). Aggression, depression, and bullying others. Aggressive Behavior, 28, 198-206. Roots of empathy: Training manual. (2001). Toronto, Ontario, Canada: Roots of Empathy. Roots of Empathy. (2003). Effectiveness of the roots of empathy program [Brochure]. Toronto, Ontario, Canada: Author. Rubin, K.H., & Schneider, F.W. (1973). The relationship between moral judgment, egocentrism, and altruistic behavior. Child Development, 44, 661-665. Sagi, A., & Hoffman, M. (1976). Empathic distress in newborns. Developmental Psychology, 12, 175-176. Salmivalli, C. (1999). Participant role approach to school bullying: Implications for interventions. Journal of Adolescence, 22, 453-459. 105 Salmivalli, C , Karhunen, J . , & Lagerspetz, K. (1996). How do the victims respond to bullying?. Aggressive Behavior, 22, 99-109. Salmivalli, C , Lagerspetz, K., Bjorkvist, K., Osterman, K., & Kaukiainen, A. (1996). Bullying as a group process: Participant roles and their relations to social status within the group. Aggressive Behaviour, 22, 1-15. Salmivalli, C , Lappalainen, M., & Lagerspetz, K. (1998). Stability and change of behavior in connection with bullying in schools: A two-year follow-up. Aggressive Behavior, 24,205-218. Smith, P.K., & Brain, P. (2000). Bullying in schools: Lessons from two decades of research. Aggressive Behavior, 26, 1-9. Stotland, E. (1969). Exploratory studies in empathy. In L. Berkowitz (Ed.), Advances in experimental social psychology, 4 (pp. 271-314). New York: Academic Press. Strayer, J . (1993). Children's concordant emotions and cognitions in response to observed emotions. Child Development, 64, 188-201. Sutton, J . , & Keogh, E. (2000). Social competition in school: Relationships with bullying, Machiavellianism and personality. British Journal of Educational Psychology, 70, 443-456. Sutton, J . , & Smith, P.K. (1999). Bullying as a group process: An adaptation of the participant role approach. Aggressive Behavior, 25, 97-111. Sutton, J . , Smith, P. K., & Swettenham, J . (1999a). Social cognition and bullying: Social inadequacy or skilled manipulation?. British Journal of Developmental Psychology, 17, 435-450. Sutton, J . , Smith, P. K., & Swettenham, J . (1999b). Bullying and 'theory of mind': A critique of the 'social skills deficit' view of anti-social behaviour. Social Development, 8 (1), 117-134. 106 Sutton, J . , Smith, P.K., & Swettenham, J . (1999c). Socially undesirable need not be incompetent: A response to Crick and Dodge. Social Development, 8(1), 132- 134. Underwood, B., & Moore, B. (1982). Perspective-taking and altruism. Psychological Bulletin, 97(1), 143-173. Whitney, I., & Smith, P.K. (1993). A survey of the nature and extent of bullying in junior/middle and secondary schools. Educational Research, 35(1), 3-25. v. 110 Appendix C TELL US ABOUT YOURSELF Instructions: We are interested in learning a little about your background. Read each of the following questions to yourself, and write your answers in the spaces provided. Don't skip any questions. Please follow the directions carefully. 1. Are you a boy or a girl? (Check one) Girl Boy 2. How old are you? years old 3. When is your birthday? month day year 4. What GRADE are you in? (Check one) 5 t h 6th 7* 5. How would you describe your ethnic or cultural background? (Check one) White (Anglo, Caucasian, etc.) First Nations (Native Canadian) Indo-Canadian (East Indian) Asian )Chinese, Japanese, Korean, etc.) Latin (Spanish, Mexican, South American, etc.) Black (African, Haitian, Jamaican, etc.) Other (please describe) 6. Which of these adults do you live with MOST of the time? (Check all the adults that you live with). Mother Father Stepmother Stepfather Grandmother Grandfather Other adults (for example, aunt, uncle, mom's boyfriend, friend's parents, etc.) 111 Appendix D BULLYING AT YOUR SCHOOL There are lots of different ways to bully someone, but a bully wants to hurt the other person (it's not an accident), and does so unfairly (the bully has some advantage over the victim). Sometimes a group of students will bully another student. Bullying takes many forms: PHYSICAL BULLYING VERBAL BULLYING SOCIAL BULLYING COMPUTER BULLYING -when someone hits, shoves, kicks, spits, or beats up on -when someone damages or steals another student's property -name-calling, mocking, hurtful teasing -humiliating or threatening someone -making people do things they don't want to do -excluding others from the group -gossiping or spreading rumours about others -setting others up to look foolish -making sure others don't associate with the person -using computer or email messages or pictures to hurt someone's feelings, make someone look bad, or threaten someone Instructions: When you answer these questions, think about this school year and remember that bullying can take many different forms. We would like you to think about what you and your classmates usually do in situations in which someone is being bullied. For each of the questions on the next few pages, please circle the names of girls and boys from the list of people in your class who behave in that way. You can choose as many people as you can think of for each question, and people can be chosen for more than one question. REMEMBER, DO NOT SHOW YOUR ANSWERS TO ANYONE EXAMPLE: When someone is being bullied... Who runs away? BOYS: GIRLS: ' 112 When someone is being bullied. 1. Who pretends not to notice what is happening? 2. Who helps to catch the victim for the bully? BOYS: GIRLS: • 3. Who doesn't even know about the bullying 1 BOYS: •'. GIRLS: BOYS: GIRLS: 4. Who tries to make the others stop the bullying? BOYS: ' • • GIRLS: 113 When someone is being bullied. 5. Who says things to the bully like "Show him!"? 6. Who is usually there, even if not doing anything? .BOYS: :,: GIRLS: 7. Who is the one to get hit or kicked? BOYS: GIRLS BOYS: GIRLS: 8. Who takes the bully's side? BOYS: GIRLS 114 When someone is being bullied... 9. Who doesn't do anything or take sides? 10. Who always thinks of new ways of picking on the victim? BOYS: GIRLS: 11. Who joins in the else has starter. bullying if someone lit? I f l lpYS: ' GIRLS: BOYS: GIRLS- 12. Who isn't usually there, and stays away? IlllllYSivl'',̂ , GIRLS: 115 When someone is being bullied... 13. Who tells an adult about the bullying? 14. Who gets others to join in the bullying? GIRLS: -/ 15. Who is the one to get left out of the group on purpose? BOYS: GIRLS: ; : BOYS: GIRLS: 16. Who gets others to come and watch? BOYS: GIRLS: 116 When someone is being bullied. 17. Who leads a gang? 18. Who starts the bullying? BOYS: ';• GIRLS: 19. Who encourages the bully by shouting? BOYS: • ., GIRLS:' BOYS: GIRLS: 20. Who tries to cheer the victim up? BOYS: GIRLS: 1 1 7 When someone is being bullied... 21. Who helps the bully by holding the victim? 22. Who sticks up for the victim? BOYS: G I R L S : - , 23. Who laughs at the person being bullied? GIRLS: BOYS: GIRLS: 24. Who says to the others that bullying is stupid? BOYS: 118 When someone is being bullied... 25. Who is the one to get called mean names? BOYS: GIRLS: STOP AND WAIT 119 Appendix E Instructions: Now, we would like you to think about what you usually do in situations in which someone is being bullied. For each of the questions listed on this page, please circle the answer that shows how you behave. 1. When someone is being bullied, do you pretend not to notice what is happening? NO no sometimes yes YES 2. When someone is being bullied, do you help to catch the victim for the bully? : ; NO no sometimes ! yes Y E S 3. When someone is being bullied, do you know about the bullying? NO no sometimes yes YES A.. When someone is being bullied, do you try to make the others stop the bullying? : NO no sometimes yesi-YES 5. When someone is being bullied, do you say things to the bully like "Show him!"? NO no sometimes yes YES 6. When someone is being bullied, are you usually there, even if not doing anything? NO j i o sometimes yes. Y£§ 7. When someone is being bullied, are you the one to get hit or kicked? NO no sometimes yes YES 8, When someone is being bullied, do you take the bully's side? y; :V I- !>'. \ NO' w f sometimes;! yesfWES 9. When someone is being bullied, do you do nothing or refuse to take sides? NO no sometimes yes YES 10. When someone is being bullied, do you always think of new ways of picking on the victim? . : NO no sometimes • yes ' :.YES 11. When someone is being bullied, do you join in the bullying if someone else has started it? NO no sometimes yes YES if2l ;yyhen§^ stay away? , NO_no sometimes, yes . YES, 13. When someone is being bullied, do you tell an adult about the bullying? NO no sometimes yes YES i14. When someone is being bullied, do you get others to join in the bullying? v : NO no sometimes ^yes YES 15. When someone is being bullied, are you the one to get left out of the group on purpose? NO no sometimes yes YES J16. When someone is beingbullied, do you get others to watch? ... NO no sometimes yes YES' 17. When someone is being bullied, do you lead a gang? NO no sometimes yes YES i18/When somepheJsbejng>bullied,;do you, start 19. When someone is being bullied, do you encourage the bully by shouting? NO no sometimes yes YES 20. WhehtSoMbhe. js bejhg bullied, do you try to cheerthew ^ : ; j : - . »NO!i ;nd '^s j j j i f lg i i^^yes^ES, 120 21. When someone is being bullied, do you help the bully by holding the victim NO no sometimes yes YES 22. When someone is being bullied, do you stick up for the victim? ~ _ NO. no sometimes y e s ' Y E f 23. When someone is being bullied, do you laugh at the person being bullied? NO no sometimes yes YES 2 £ ? W h e n ^ 25. When someone is being bullied, are you the one to get called mean names? NO no sometimes yes YES STOP AND WAIT 121 Appendix F Thoughts and Feelings Questionnaire Instructions: Read the following list of ways that children might feel about others. After you've read each one, please tell us how much this is true for you, by checking one of the choices on your paper. 1. I often feel sorry for people who don't have the things I have. Check one: 2. When I see someone who badly needs help in an emergency, I just can't take it and I get upset. Check one: 3. It's easy for me to understand why other people do the things they do. Check one: 4. I feel sorry for a student who is getting verbally bullied (called names, threatened). Check one: 5. Sometimes I try to understand my friends better by imagining how they think about things. Check one: 6. Sometimes I feel sorry for other people when they are having problems. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me 1 2 2 7. In emergency situations, I feel very uneasy and nervous. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 8. Even when I'm mad at someone, I try to understand how they feel. Check one: 9. When I see someone being physically bullied (hit, kicked), I try to understand how they feel. Check one: 10. When I see someone being picked on, I feel kind of sorry for them. Check one: 11. I feel sorry for a student who is getting socially bullied (gossiped about, left out of the group). Check one: 12. I try to understand how other kids feel before I decide what to say to them. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me 13. Being in very tense or emotional situations scares me. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 123 14. I often feel sorry for other children who are sad or in trouble. Check one: 15. When I see someone getting verbally bullied (called names, threatened),! feel very uneasy and nervous. Check one: 16. When I see someone being socially bullied (gossiped about, left out of the group), I try to understand how they feel. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me 17. When I see someone being treated mean, it bothers me. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 18. When I see someone get hurt, I can't stay calm and I get upset. Check one: 19. I feel sorry for a student who is getting physically bullied (hit, kicked). Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me 20. Even when I know I'm right I listen to what other people think. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 124 21. I have a hard time dealing with emergencies. Check one: 22. I often have strong feelings about things that happen around me. Check one: 23. When I see someone getting verbally bullied (called names, threatened), I try to understand how they feel. Check one: 23. When I see someone getting socially bullied (gossiped about, left out of the group), I feel very uneasy and nervous. Check one: 25. Before I say anything bad about anyone, I try to imagine how I would feel if I were that person. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me Not at all like me A little bit like me Kind of like me A lot like me Always like me 26. I am a person who cares about the feelings of others. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 27. I feel helpless when I am in the middle of a tense and emotional situation. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 125 28. When I see a student who is getting physically bullied (hit, kicked) I feel very uneasy and nervous. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 29. There are different ways to think about a problem and try to look at all them. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 30. I tend to lose control during emergencies. Check one: Not at all like me A little bit like me Kind of like me A lot like me Always like me 126 Appendix G Factor Loadings for Pro-bully, Defender, Outsider, and Victim Factors (Peer-Nominations) (Girls Only) Factor Item Pro-bully Defender Outsider Victim Who starts bullying? .874 Who gets others to join in the bullying? .847 Who gets others to come and watch? .835 Who encourages the bully by shouting? .826 Who leads a gang? .820 Who joins in the bullying if someone else has started it? .814 Who takes the bully's side? .810 Who always thinks of new ways of picking on the .794 victim? Who laughs at the person being bullied? .787 Who helps the bully by holding the victim? .786 Who says things to the bully like "Show him!"? .778 Who helps to catch the victim for the bully? .513 Who tries to make others stop the bullying? .844 Who sticks up for the victim? .832 Who says to the others that bullying is stupid? .832 Who tries to cheer the victim up? .824 Who tells an adult about the bullying? .735 Who pretends not to notice what is happening? Who doesn't even know about the bullying? .808 Who isn't usually there, and stays away? -.432 .697 Who doesn't do anything or take sides? .680 Who is usually there, even if not doing anything? -.510 Who is the one to get called mean names? Who is the one to get left out of the group on purpose? Who is the one to get hit or kicked? .866 .850 .836 Note. Factor loadings that were not significant (i.e., less than .400) are not included. All cross-loadings were less than .400. 127 Factor Loadings for Pro-bully, Defender, Victim, and Outsider Factors (Peer-Nominations) (Boys Only) Factor Item Pro-bully Defender Victim Outsider Who joins in the bullying if someone else has started it? .912 Who starts bullying? .908 Who always thinks of new ways of picking on the .901 victim? Who gets others to join in the bullying? .886 Who takes the bully's side? .876 Who encourages the bully by shouting? .869 Who leads a gang? .865 Who laughs at the person being bullied? .863 Who gets others to come and watch? .860 Who helps the bully by holding the victim? .849 Who says things to the bully like "Show him!"? .846 Who helps to catch the victim for the bully? .792 Who sticks up for the victim? .853 Who tries to make the others stop the bullying? .850 Who tries to cheer the victim up? .843 Who says to the others that bullying is stupid? .812 Who tells an adult about the bullying? .687 Who is the one to get called mean names? .897 Who is the one to get hit or kicked? .878 Who is the one to get left out of the group on .835 purpose? Who pretends not to notice what is happening? .431 Who isn't usually there, and stays away? .749 Who doesn't even know about the bullying? .697 Who doesn't do anything or take sides? .634 Who is usually there, even if not doing anything? .504 -.504 Note. Factor loadings that were not significant (i.e., less than .400) are not included. All cross-loadings were less than .400. 128 Appendix H Factor Loadings for Pro-bully, Defender, Pro-bully 2, Victim, and Outsider Factors (Self-Ratings) (Girls Only) Factor Item Do you always thinks of new ways of picking on the victim? Do you help the bully by holding the victim? Do you encourage the bully by shouting? Do you say things to the bully like 'Show him!'"? Do you get others to come and watch? Do you get others to join in the bullying? Do you take the bully's side? Do you start the bullying? Do you help to catch the victim for the bully? Do you lead a gang? Do you laugh at the person being bullied? Do you stick up for the victim? Do you try to cheer the victim up? Do you try to make the others stop the bullying? Do you say to the others that bullying is stupid? Do you tell an adult about the bullying? Are you usually there, even if not doing anything? Do you know about the bullying? Do you join in the bullying if someone else has started it? Are you the one to get called mean names? Are you the one to get left out of the group on purpose? Are you the one to get hit or kicked? Do you do nothing, or refuse to take sides? Do you stay away? Do you pretend not to notice what is happening? Pro-Pro-bully Defender ^ Victim Outsider .805 .749 .699 .685 .629 .621 .619 .606 .569 .557 .539 .414 .412 .798 .785 .766 .711 .666 .442 .723 -.645 .505 .849 .807 .655 .430 .806 .700 .467 Note. Factor loadings that were not significant (i.e., less than .400) are not included. All cross-loadings were less than .400. 129 Factor Loadings for Pro-bully, Defender, Victim, Outsider, Reinforcer, and Assistant Factors (Self-Ratings) (Boys Only) Factor Item Do you say things to the bully like 'Show him!"'? Do you get others to join in the bullying? Do you start the bullying? Do you join in the bullying if someone else has started it? Do you lead a gang? Do you take the bully's side? Do you encourage the bully by shouting? Do you always thinks of new ways of picking on the victim? Do you laugh at the person being bullied? Do you try to make the others stop the bullying? Do you say to the others that bullying is stupid? Do you stick up for the victim? Do you try to cheer the victim up? Do you tell an adult about the bullying? Pro- bully Victim Reinfor- Assist- cer ant .790 .773 .760 .751 .744 .696 .695 .674 .670 Are you the one to get called mean names? .900 Are you the one to get hit or kicked? .863 Are you the one to get left out of the group on .817 purpose? Do you pretend not to notice what is .753 happening? Do you do nothing, or refuse to take sides? .723 Do you stay away? .662 Do you know about the bullying? -.760 Are you usually there, even if not doing .663 anything? Do you get others to come and watch? .401 .506 Do you help to catch the victim for the bully? .773 Do you help the bully by holding the victim? .728 .836 .810 .793 .765 .721 Note. Factor loadings that were not significant (i.e., less than .400) are not included. All cross-loadings were less than .400. 130 Appendix I Factor Loadings for Empathic Distress, Empathic Concern, Perspective-taking, and Understanding (Girls Only) Item I have a hard time dealing with emergencies. I tend to lose control during emergencies. In emergency situations, I feel very uneasy and nervous. When I see someone get hurt, I can't stay calm and I get upset. Being in very tense and emotional situations scares me. When I see someone who badly needs help in an emergency, I just can't take it and I get upset. I feel helpless when I am in the middle of a tense and emotional situation. When I see someone being picked on, I kind of sorry for them. I often feel sorry for other children who are sad or in trouble. Sometimes I feel sorry for other people when they are having problems. When I see someone being treated mean, it bothers me. I often feel sorry for other people who don't have the things I have. I often have strong feelings about things that happen around me. Before I say anything bad about anyone, I try to imagine how I would feel if I were that person. I try to understand how other kids feel before I decide what to say to them. Even when I know I'm right I listen to what other people think. Even when I'm mad at someone, I try to understand how they feel. I am a person who cares about the feelings of others. There are different ways to think about a problem and I try to look at all of them. Sometimes I try to understand my friends better by imagining how they think about things. It's easy for me to understand why other people do the things they do. Factor Empathic Empathic Perspective Understand Distress Concern -taking -ing .837 .801 .780 .742 .735 .640 .610 .762 .752 .727 .724 .681 .455 .761 .755 .732 .643 (.515) .597 .564 .882 Note. All other factor loadings were less than .400. Cross-loadings greater than .400 are shown in parentheses. 131 Factor Loadings for Empathic Distress, Empathic Concern, Perspective-taking, and Understanding (Boys Only) Item I have a hard time dealing with emergencies. I tend to lose control during emergencies. In emergency situations, I feel very uneasy and nervous. I feel helpless when I am in the middle of a tense and emotional situation. Being in very tense and emotional situations scares me. When I see someone who badly needs help in an emergency, I just can't take it and I get upset. When I see someone get hurt, I can't stay calm and I get upset. I often have strong feelings about things that happen around me. When I see someone being picked on, I feel kind of sorry for them. Sometimes I feel sorry for other people when they are having problems. I often feel sorry for people who don't have the things I have. I often feel sorry for other children who are sad or in trouble. When I see someone being treated mean, it bothers me. I am a person who cares about the feelings of others. There are different ways to think about a problem and I try to look at all of them. I try to understand how other kids feel before I decide what to say to them. Even when I know I'm right I listen to what other people think. Sometimes I try to understand my friends better by imagining how they think about things. Even when I'm mad at someone, I try to understand how they feel. Before I say anything bad about anyone, I try to imagine how I would feel if I were that person. It's easy for me to understand why other people do the things they do. Factor Empathic Empathic Perspective- Understand Distress Concern taking -ing .832 .788 .755 .727 .708 .660 .570 .525 (.446) .771 .762 .726 .670 .636 .599 (.422) (.465) (.483) (.422) (.496) .713 .670 .664 .633 .622 .587 .919 Note. All other factor loadings were less than .400. Cross-loadings greater than .400 are shown in parentheses. 132 Appendix J Factor Loadings for Bullying-Specific Perspective-taking, Empathic Distress, and Empathic Concern (Girls Only) Item When I see someone getting verbally bullied (called names, threatened), I feel very uneasy and nervous. When I see someone getting socially bullied (gossiped about, left out of the group), I feel very uneasy and nervous. When I see someone getting physically bullied (hit, kicked), I feel very uneasy and nervous. I feel sorry for a student who is getting verbally bullied (called names, threatened). I feel sorry for a student who is getting physically bullied (hit, kicked). I feel sorry for a student who is getting socially bullied (gossiped about, left out of the group). When I see someone getting physically bullied (hit, kicked), I try to understand how they feel. When I see someone getting verbally bullied (called names, threatened), I try to understand how they feel. When I see someone getting socially bullied (gossiped about, left out of the group), I try to understand how they feel. Empathic Distress (Bullying) Factor Empathic Concern (Bullying) Perspective- taking (Bullying) .839 .834 .826 .831 .787 .738 .826 .785 .705 Note. All other factor loadings were less than .400. Cross-loadings greater than .400 are shown in parentheses. 133 Factor Loadings for Bullying-Specific Perspective-taking, Empathic Distress, and Empathic Concern (Boys Only) Item When I see someone getting socially bullied (gossiped about, left out of the group), I try to understand how they feel. When I see someone getting verbally bullied (called names, threatened), I try to understand how they feel. When I see someone getting physically bullied (hit, kicked), I try to understand how they feel. I feel sorry for a student who is getting verbally bullied (called names, threatened). I feel sorry for a student who is getting physically bullied (hit, kicked). I feel sorry for a student who is getting socially bullied (gossiped about, left out of the group). When I see someone getting physically bullied (hit, kicked), I feel very uneasy and nervous. When I see someone getting socially bullied (gossiped about, left out of the group), I feel very uneasy and nervous. When I see someone getting verbally bullied (called names, threatened), I feel very uneasy and nervous. Perspective- taking (Bullying) Factor Empathic Concern (Bullying) Empathic Distress (Bullying) . 8 4 7 . 8 4 5 . 7 1 3 ( .463) ( .431) . 8 5 3 . 7 7 4 . 7 0 5 . 8 1 3 . 8 0 4 . 7 6 8 Note. All other factor loadings were less than .400. Cross-loadings greater than .400 are shown in parentheses.

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 40 0
Philippines 34 0
Germany 7 2
France 4 0
China 3 0
Japan 2 0
Greece 1 1
Canada 1 0
United Kingdom 1 0
Unknown 1 0
City Views Downloads
Unknown 34 2
Washington 16 0
Ashburn 11 0
San Juan 4 0
Beijing 3 0
Makati City 3 0
Seattle 2 0
Menlo Park 2 0
Tokyo 2 0
Minneapolis 2 0
Mountain View 2 0
Nashville 1 0
Buting 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

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}]}"
                            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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0054505/manifest

Comment

Related Items