INVESTIGATING THE RELATIONSHIP BETWEEN ...

6 downloads 0 Views 695KB Size Report
Michelle Faye Wright, daughter of Sharon L. Wright, was born October 3, .... an increasing concern for educators and parents (Verlinden, Herson, & Thomas,.
INVESTIGATING THE RELATIONSHIP BETWEEN INTENT ATTRIBUTIONS AND VICTIMIZATION STATUS

Michelle Faye Wright

Certificate of Approval:

______________________________ Bridgette D. Harper Thesis Chair/Assistant Professor

_____________________________ Glen E. Ray Professor Interim Dean

______________________________ Sheila Mehta Associate Professor

_____________________________ Director of Graduate Studies

i

INVESTIGATING THE RELATIONSHIP BETWEEN INTENT ATTRIBUTIONS AND VICTIMIZATION STATUS

Michelle Faye Wright

A Thesis Submitted to the Graduate Faculty of Auburn University at Montgomery in Partial Fulfillment of the Requirements for the Degree of Master of Science

Montgomery, Alabama August 22, 2008

ii

INVESTIGATING THE RELATIONSHIP BETWEEN INTENT ATTRIBUTIONS AND VICTIMIZATION STATUS

Michelle Faye Wright

Permission is granted to Auburn University Montgomery to make copies of this thesis at its discretion, upon the request of individuals or institutions and at their expense. The author reserves all publication rights.

Signature of Author

_______________________

_________________________ Date Copies sent to: Name

Date

iii

VITA Michelle Faye Wright, daughter of Sharon L. Wright, was born October 3, 1982, in Orlando, FL. She graduated from Apopka High School in 2001. She attended the University of Central Florida, graduating with a Bachelor of Science in Psychology in Spring 2005. In August of 2006, she entered Graduate School at Auburn University Montgomery. She will attend DePaul University (Chicago, Illinois) as a doctoral student in Experimental Psychology beginning September of 2008.

iv

THESIS ABSTRACT

INVESTIGATING THE RELATIONSHIP BETWEEN INTENT ATTRIBUTIONS AND VICTIMIZATION STATUS

Michelle Faye Wright Master of Science, August 22, 2008 (B.S., University of Central Florida) 67 Typed Pages Directed by Bridgette D. Harper Social evaluations are important to all aspects of children’s development. Social evaluations help children interact in social situations and determine in part, children’s peer status. Once peer status is determined, it is relatively stable. For example, children that are victimized continue to be so once their status is determined. It is important to understand victimized children because peer victimization has been linked to psychosocial maladjustment. However, not all children that experience peer victimization experience psychosocial maladjustment. Because not all victimized children experience psychosocial maladjustment, it has been theorized that children’s attributions may be a better predictor of adjustment to peer victimization than the frequency of victimization. Thus, attributions may help determine children’s adjustment to peer victimization. While the study of attributions in relation to peer victimization is important, work is needed to investigate how attributions may vary as a function of the evaluator-target peer relationship. The present study investigates evaluator-target relationships. In addition, the present study also investigates whether victimized children are more vulnerable to relationship effects than nonvictimized children. That is, the present study

v

investigates whether victimized children’s attributions are influenced more by relationship effects compared to nonvictimized children.

vi

ACKNOWLEDGEMENTS I would first like to express my thanks to Dr. Bridgette Harper for her encouragement throughout this entire thesis. Additionally, I would also like to thank her for giving up her time in order to complete this project. I would also like to extend my thanks to Dr. Glen Ray and Dr. Sheila Mehta for their contributions throughout this entire project. Additionally, they showed me a lot of support and encouragement for this project which also helped to make it easier. Next, I would also like to extend my love to my mother for her help encouragement in my dreams of furthering my education. Also, like always, she expressed a lot of support for me throughout this entire process even when I was not at my best. If not for her, I am certain this thesis would not even be in existence. Finally, I would like to extend my love and thanks to my family. To my wonderful partner, I would like to thank him for allowing me plenty of time to work on this project. Additionally, I would like to thank him for maintaining his sanity when all I could do was obsess over the project. Lastly, I would like to thank my daughter if not for her smiles and kisses, this project would have been even harder to complete.

vii

TABLE OF CONTENTS VITA……………………………………………………………………………………...……....iv ABSTRACT…………………………………………………………………………………….....v ACKNOWLEDGEMENTS…………...……………………………………...………………….vii LIST OF TABLES..………………………………………………………………………………ix INTRODUCTION……………………………………………………………………………….10 Evaluating Social Situations……………………………………………………………..11 The Importance of Investigating Peer Victimization…………………………………….16 Attributions………………………………………………………………………………17 Research on Peer Victimization, Children’s Attributions and Children’s Psychological Adjustment……………………………………………………………………….20 The Relationship Between Peer Victimization and Attributions………………...………21 The Relationship Between Attributions and Psychosocial Adjustment……...………….22 Attributions and Relationship Types…………………………………………………….24 The Present Study………………………………………………………………………..26 METHOD…………………………………………………………………..……………………28 Participants………………………………………………………………………………28 Design……………………………………………………………………………………28 Materials…………………………...…………………………………………………….28 Procedure…...……………………………………………………………………………31 RESULTS……………………………………………………………………….…………...….33 DISCUSSION…………...………………………………………………………………………37 REFERENCES…………………………………………………………………….…………….42 APPENDICES……………………...……………………………………………………………46

viii

LIST OF TABLES TALBE 1: Attributions x Relationship Interaction………………….……………………………….34 TABLE 2: Relationship x Attribution Interaction………………………………………………...…35 TABLE 3: Victimization Status x Attributions x Relationship Type………………………………..36

ix

Investigating the Relationship between Intent Attributions and Victimization Status Because of many high profile cases of school shootings and stabbings, peer victimization has become an increasing concern for educators and parents (Verlinden, Herson, & Thomas, 2000). Peer victimization is quite common among children with an estimated 160,000 children not attending school daily because they fear being the target of peer victimization (Fried & Fried, 1996). In fact, approximately eighty percent of sixth through tenth graders have experienced some form of peer victimization (Nansel, Overpeck, Pilla, Ruan, and SimonsMorton, 2001). Even though high profile cases of school violence are associated with a history of peer victimization, responding to peer victimization with violence is rare; however, negative psychosocial adjustment in relation to peer victimization is not rare and is associated with many negative outcomes for the victimized child. Research has consistently linked peer victimization with a host of psychosocial maladjustment problems (Hawker & Boulton, 2000). For example, victimization is moderately associated with an increased risk for loneliness, depression, anxiety, and low self-esteem (Bond, Carlin, Thomas, Rubin, & Patton, 2001; Olweus, 1993). Compared to less victimized children, frequently victimized children are more likely to avoid school, to have poorer academic performance and to be disliked by their peers (Kochenderfer, & Ladd, 1996a). Additionally, frequently victimized children report more psychosomatic health problems such as stomachaches than do nonvictimized children (Williams, Chambers, Logan, & Robinson, 1996). Interestingly, not all children who experience frequent peer victimization experience these negative consequences. Instead, attributions about peer victimization have been theorized to influence children’s adjustment after an episode of peer victimization more than the frequency of peer victimization (e.g., Graham, & Juvonen, 1998). Studies conducted that investigate children’s attributions in relation to peer victimization have found consistent, moderate

10

correlations between frequency of peer victimization and self-blame attributions (characterological and behavioral) and between self-blame attributions and children’s negative psychosocial adjustment (e.g., Bellmore, Wikow, Graham, & Juvonen, 1998; Graham, & Juvonen, 1998; Graham, & Juvonen, 2002; Hawker, & Boulton, 2000). In addition, some researchers (e.g, Ray, & Cohen, 1997) argue that while attributions play a role in subsequent adjustment to peer victimization, it is also important to consider the relationship between the evaluator and the target peer being evaluated. That is, the relationship between the evaluator and the target influence the attribution made. The major purpose of the present study is to extend previous research (e.g., Peets, Kikas, Hodges, & Salmivalli, 2007; Ray, & Cohen, 1997) by investigating the influence of the evaluator-target relationship on children’s attributions of intent. Children will evaluate an actual friend, an acquaintance, and a disliked peer in various hypothetical situations. Another purpose of this present study is to investigate the possibility that differences in intent attributions may exist between victimized and nonvictimized children. This review begins with theory and research on the important role of children’s evaluations in understanding their social situations. The second section focuses on the importance of studying peer victimization, such as the associated negative consequences of peer victimization such a depression, loneliness, and anxiety. The final section reviews theory and research on the relationship between children’s self-perceptions of peer victimization and causal attributions. Within this final section, theory and research on the attributions of intent in social situations will be examined. Evaluating Social Situations Whether in the classroom, on the playground, in the lunch room, or on the school bus, children are constantly evaluating their peers’ social behavior. Evaluations of peer social behavior is developmentally significant for the evaluator and for the peer being evaluated (Ray,

11

Graham, & Cohen, 2001). Because children examine and evaluate the social behavior of peers daily, their evaluations determine their relationship choices not only with the target peer being evaluated but also with other peers. For example, children seen engaging in peer conflict may be avoided while children viewed as being prosocial and not engaging in peer conflict will be sought out as friends. It may also be important to understand how children evaluate unfamiliar social situations because the majority of the time children must evaluate the behavior of others when the situation is novel and the intentions are unclear (Ray et. al, 2001). That is, children do not always know the intentions of others that they interact with or even how to act in social situations that are unfamiliar to them. When children are uncertain about how to act in social situations, they may rely on the actions of others to gauge how they should interact with the target peer and interpret the peers intentions. As a result, children may begin to model the prosocial actions of others and thus learn how to manage conflicts in socially acceptable ways. It is theorized (e.g., Arsenio, & Lemerise, 2000; Crick, & Dodge, 1994) that children store social information about other peers in a knowledge base that influences their current and future interactions. Whether directly interacting with peers or observing others interacting with peers, social information is encoded and stored to be used for future interactions involving target peers. Social information processing (SIP) theory asserts that children actively encode and store mental representations of their interactions involving target peers for the current and future social situations (Crick & Dodge, 1994). This model argues that the way in which children conceptualize and evaluate social information determines how they will react or behave towards peers. Children use these mental processes to evaluate (make attributions about) problematic interactions with other children. According to this model, there is an array of cues that children

12

receive as input and their subsequent behavioral responses are a function of processing each of those cues. Specifically, before children respond to a particular situation, they progress through six distinct steps: (1) encoding of external and internal cues, (2) interpretation and mental representation of those cues, (3) clarification and selection of a goal, (4) response access or construction, (5) response decision, and (6) behavioral enactment. Below, each step will be examined and demonstrated with an example. During the encoding and interpretation of social cues (Step 1 and Step 2), children selectively attend to certain situational and internal cues and then interpret them (Crick & Dodges, 1994). During interpretation, there are many independent cognitive processes including mental representations of situational cues that are stored in long-term memory. Next, children engage in a causal analysis of the events that have occurred. Children also make inferences about the perspectives of others involved in the situation. Then, an evaluation of outcome expectations and self-efficacy predictions are made (e.g., “I am capable of doing this”). Inferences are made regarding self-evaluations and the evaluations by peers. Each of these processes may be influenced by the knowledge base stored in memory (e.g., social schemata, scripts, and social knowledge). Changes and revisions are made to the knowledge base based on these interpretations. Children then select a goal or a desired outcome for the present situation and will also revise or create new goals (Step 3). Next, children consider all possible responses to the situation (step 4). If children do not have a set of responses stored in their knowledgebase, they will construct a novel behavior response. After considering all possible responses, children select the best possible response which is influenced by outcome expectations, self-efficacy, and response evaluations (Step 5). Finally, children behaviorally enact the chosen response (Step 6). It is important to note that the current study will focus on steps (1) and (2). To illustrate an example of this model, imagine children sitting in a lunchroom eating

13

their lunches and drinking their milk. John walks by drinking milk and the next thing that happens is that John spills his milk all over Andrew’s back. Now, Andrew must encode this situation (Step 1) before interpreting it (Step 2). During the interpretation process, Andrew will decide why John spilled milk on his back. Andrew will think about many possible reasons to explain John’s behavior such as, John was making fun of him, John tripped on accident, or maybe Andrew may think he did something to cause John to spill milk on his back. For this example, let’s consider that Andrew believes John tripped which resulted in the milk being spilt on his back. Andrew may consider what the other children are thinking. Assuming Andrew believes the other children believe it was an accident, too, Andrew now must decide what he wants the outcome of this situation to be (Step 3). Ultimately, Andrew wants to have the milk cleaned from his back or to change his shirt (Step 4) which is equivalent to the outcome goal of this situation. Andrew will then decide which behaviors must be enacted in order to get his shirt cleaned (Step 5). Next, Andrew decides that he should ask John to grab some paper towels to achieve the goal of having his shirt cleaned. Finally, Andrew must enact this behavior and to do this, Andrew asks John to get some paper towels from the lunchroom workers (Step 6). Lemerise and Arsenio (2000) integrated emotional processes into the SIP model. They proposed that the emotional valence of the relationship between the two individuals involved affects how information is processed, and thus how a behavioral response is enacted. While emotions influence the interpretation of social situations, other factors also play a role in the influence of children’s evaluations of social situations as well. Consider the previous example of the milk being spilt on the Andrew’s back. Throughout the whole process when Andrew is deciding the cause of the situation, which goals to enact, and how to enact these goals, emotional processes are also being relied upon much like the knowledge base in the Crick and Dodge (1994) SIP model. The emotional temperament, ability to regulate one’s emotions, and the moods or emotions of the other children play a role in these processes. That is, Andrews

14

emotions and feelings about the situation may influence his reactions and his thoughts about why this situation occurred. Other factors that influence children’s attributions of other’s behavior include previous social experiences, gender, grade, and social status to name a few (Crick & Dodges, 1994). For example, researchers (e.g., Burgess, Wojslawowicz, Rubin, Rose-Krasnor, & BoothLaForce, 2006) have found that aggressive children attribute hostility to others, particularly in ambiguous situations. Aggressive children expect more positive outcomes for behaving aggressively in social situations and have higher self-efficacy beliefs because they believe aggression is an acceptable way of behaving (Erdley & Asher, 1998; Perry, Perry, & Rasmussen, 1986). Aggressive children expect reciprocated hostility and for hostility to continue following a social event that has a negative outcome (Burgess et al., 2006). Aggressive children are also less likely to trust the target peer in the future. Likewise, sociometrically rejected children also misattribute hostility to others in nonhostile situations (accidental and benign) and in ambiguous situations (Crick, & Ladd, 1990). Sociometrically rejected children believe that aggressive strategies used in social encounters with peers produce desired outcomes. Likewise, Bell and Dolen (1995) examined the attributions of anxious children, especially girls, and found that anxious girls were more likely to propose maladaptive strategies in response to perceived hostility. Furthermore, shy/withdrawn children process social information differently than aggressive children. Shy/withdrawn children are more likely than nonwithdrawn and aggressive children to blame themselves under ambiguous circumstances (Burgess et al., 2006). Specifically, shy/withdrawn children are more likely to attribute the cause to something negative about themselves. Some of the strategies employed by aggressive children may seem to potentially set them up for continuing the cycle. That is, the way they interpret the situation and the strategies they use seem to only set them up for future negative interactions with their peers. These children

15

may be the target for future peer victimization. Thus, it is important to examine peer victimization. Additionally, the continuation of the cycle of peer victimization, may set children up for future problems because children that are frequently victimized are at an increased risk for future psychosocial maladjustment problems. The Importance of Investigating Peer Victimization Investigating peer victimization is important because previous researchers (e.g., Boulton, & Underwood, 1992; Kochenderfer, & Ladd, 1996b) demonstrate that children who experience frequent peer victimization are at risk for higher levels of depression, anxiety, and loneliness than are less frequently victimized children. Hawker and Boulton (2000) further examined the increase risk of psychosocial maladjustment problems in frequently victimized children. Their meta-analysis encompassed cross-sectional studies published between 1978 and 1997 evaluating thirteen studies that assessed the association between the frequency of peer victimization and children’s depression. Results indicated that depression was the most highly correlated adjustment problem experienced by victims of chronic peer victimization. When depression was assessed by self-report, the mean effect size was r = .45 when assessed by peerreports, the mean effect size was r = .29. Five studies were used to evaluate loneliness. Loneliness was associated with frequent peer victimization but not as strongly as it was for depression. The mean effect size of loneliness assessed by self-reported victimization was r = .32, whereas the mean effect size for peer-reported victimization was r = .25. Twelve studies were examined that assessed the association between anxiety and frequency of peer victimization. Self-reported victimization was correlated with higher levels of anxiety with an overall mean effect of r = .25. While there have been consistent links found between the frequency of victimization and negative psychosocial adjustment problems, not all frequently victimized children suffer these negative consequences. Attributions have been theorized to account for individual differences in

16

adjustment to peer victimization (Graham, & Juvonen, 1998). That is, it is not exactly the frequentcy of victimization that may determine how children adjust to peer victimization. Specifically, adjustment to peer victimization may be related to children’s interpretation and labeling of causes in their interactions with their peers. Attributions When children are victimized, they may ask: “Why was I picked on today?” To answer this question, they essentially must decide whom or what to blame for this situation (e.g., themselves or others). Attribution theory attempts to explain how people make decisions about the perceived causes of events and how these decisions affect them, their thoughts, and their behaviors (Weiner, 1986). The way an individual processes social information (e.g., negative events) is an important determinant of their feelings and behaviors about why events occur (Graham & Juvonen, 2001). There are many factors that influence the attributions children make when faced with negative situations (Weiner, 1986). Each of these concepts will be addressed in the next section. Also, an example will be included to further elaborate and explain each of these concepts. Weiner (1995) argues that there are three different conceptual properties of causal attributions: a) locus; b) controllability, c) stability. Locus involves determining whether the cause of an event is internal or external (Weiner, 1995). For example, imagine driving to the store and just as you prepare to turn into the parking lot, another driver cuts in front of you and turns. You were seconds from rear-ending him. When you get to the store, you began to wonder why the driver cut you off. Your goal is to determine the cause of why the driver had cut you off. Perhaps, he or she was in a rush or simply was not paying enough attention which resulted in him or her not noticing you. Or does the situation have something to do with you? You may began to believe that maybe if you were more of a defensive driver and turned faster instead of hesitating, the driver would have never cut you off. You must determine which explains the behavior of the other driver. You must

17

determine if the cause of the event is related to you (internal) or something outside of your control (external). Controllability refers to the beliefs about the changeability of the event (Weiner, 1995). Again returning to the driving example, you may generate beliefs about whether the situation is able to be changed or if it is outside of your control. If you believe the driver was just not paying attention and in a rush, there probably is not much you can do to control his or her behavior; but if you believe that perhaps you need to learn to drive more defensively and not hesitate as much, you are able to control your behavior by changing the way you drive. Stability refers to whether the cause is constant or varying (Weiner, 1995). Referring back to the driving example, you may began to think that this event happens way too often to you or you may determine that this event rarely happens to you. The locus dimension is related to self-esteem and to depression. The individual experiences depression because they attribute failure to a shortcoming in their character; whereas, when failure is attributed to internal causes, self-esteem is lowered because they attribute the cause to something not in their control. Guilt and shame are associated with the controllability dimension. When failure is attributed to uncontrollable causes such as low aptitude, the individual may feel more shame and withdraw but if the failure is attributed to controllable factors such as lack of effort, the individual may feel guilty and attempt to address their failures. The stability dimension is linked to the individual’s expectations about their failures and successes in the future. Previous research (e.g., Weiner, 1995) on attributions has focused on the achievement of children and how their attributions can enhance or hinder their performance. For example, when children fail a test, they may ask themselves “Why did I just fail this test?” They may initially conclude that they did not put a lot of effort into studying for the test or enough effort taking it; they could also conclude that they have low aptitude and are just not “smart enough” to have received a good grade. Each of these attributions has an internal locus where children determine

18

that the cause of the event is “inside” (or internal) of him or her. When children attribute failure to low aptitude, the cause of their failure is perceived as stable and unchanging. If children attribute failure to their lack of effort, the cause of their failure is perceived as unstable and changeable. Other external attributions about poor performance may include children explaining their failure in accordance to their teacher’s poor abilities in explaining the material. For this explanation, the cause of their failure is perceived as unstable and unchangeable. All of the causal dimensions are linked to very specific psychological consequences. Based on the work of Weiner (1986, 1995), Graham and Juvonen (1998) developed an attributional model related to peer victimization. Graham and Juvonen argue that after children have experienced an episode of peer victimization, they could make either of two basic types of attributions: internal or external. Furthermore, there are two different internal attributions that can be made: characterological self-blame and behavioral self-blame. Graham and Juvonen (1998) hypothesized that frequently victimized children would make more characterological self-blaming (e.g., “If I were a cooler kid, I wouldn’t get picked on”) attributions than behavioral (e.g., “I should have been more careful”) or peer-blaming attributions (e.g., “These kinds of kids pick on everybody” ). When children engaged in characterological self-blame, they attributed the cause of victimization to internal, uncontrollable and stable characteristics about themselves. These victimized children may conclude that victimization occurred because they were not cool enough or because of some other negative characteristic about themselves. These children may decide that they were victimized because of some characteristic about themselves that is not in their control and is unchangeable. Graham and Juvonen (1998) also hypothesized that those children who engaged in behavioral self-blame would report better psychological adjustment after experiencing peer victimization. Behavioral self-blame involves children attributing the cause of victimization to internal, controllable, unstable and changeable aspects of their behavior. Children engaged in

19

behavioral self-blame attribute victimization to something they may have said or some way they have acted. Children who make the decision that they were picked on because of “the way they acted,” may have hope for change that leads to time-limited distress and better long-term psychological adjustment. However, it was found that behavioral self-blame was also related to maladjustment in that children experienced increases in loneliness and social anxiety, and decreases in self-worth. Ultimately, through the use of regression analysis, Graham and Juvonen could not conclude that blaming one’s behavior was adaptive but they also could not conclude that it was maladaptive. For external attributions, Graham and Juvonen proposed that peerblaming attributions (e.g., “These kinds of kids pick on everybody”) may protect victimized children from maladaptive psychosocial adjustment because they set the locus of blame on external rather than internal characteristics. By doing this, victimized children’s self-worth remains intact because they realize no matter what they do or how they act, they are not able to change the characteristics of the peer. Research on peer victimization, children’s attributions and children’s psychological adjustment. Few studies have examined Graham and Juvonen’s (1998) model. In their first study, Graham and Juvonen (1998) were interested in three major areas: a) whether the frequency of peer victimization influenced children’s self-blame attributions; b) whether self-blame attributions would subsequently influence children’s adjustment (loneliness, social anxiety, and self worth); and c) whether self-blame mediates the influence of peer victimization on children’s adjustment (loneliness, social anxiety, and self-worth). To investigate children’s attributions and adjustment, Graham & Juvonen (1998) assessed attributions about peer victimization and adjustment by presenting 6th and 7th graders with hypothetical vignettes involving episodes of peer victimization. Children were asked to imagine that the situation was actively happening to them and to report about what they would think if this situation happened. The response sets were designed to assess characterological self-blame

20

attributions, behavioral self-blame attributions, and external blame attributions. The response sets were organized with a five-point likert scale with one representing “definitely would think,” three representing “not sure,” and five representing “definitely would not think.” To assess children’s psychological adjustment, children were given self-reported measures of anxiety, loneliness, and self-worth. Graham and Juvenon (1998) found results consistent with their hypotheses. Frequently self-perceived victimization was linked to higher levels of characterological self-blaming attributions, r = .26; and characterological self-blame attributions were related to poorer psychological adjustment including self-reported loneliness (r = .32), social anxiety (r = .49), and self-worth (r = -.20). Frequently self-perceived victimization was also linked to higher levels of behavior self-blaming attributions, r = .20; and behavioral self-blame attributions were related to poorer psychological adjustment including self-reported loneliness (r = .18), and social anxiety (r = .25), and self-worth (r = -.15). The relationship between peer victimization and attributions. Since Graham and Juvonen’s (1998) study, a handful of studies have been conducted on the relationship between frequent peer victimization and characterological self-blame attributions. Bellmore, Wikow, Graham and Juvonen (2004) and Graham and Juvonen (2002) found that the frequency of peer victimization does make a difference in the type of attributions made by children. For example, the more frequently victimized children are, the more likely they are to endorse characterological self-blaming attributions. That is, frequently victimized children are more likely to blame themselves for peer victimization. They are more likely to blame a characteristic about themselves such as “not being cool enough.” Using hierarchical linear modeling (HLM) analysis, Graham (2005) found additional support for the original Graham and Juvonen (1998) hypotheses. She found that frequent peer victimization leads to characterological self-blame but not behavioral self-blame. Ultimately, they are more likely to believe they were picked on

21

because “they were not cool enough.” Conversely, they are less likely to believe they were picked on because “they were in the wrong place at the wrong time.” These studies demonstrate a moderate relationship between peer victimization and attributions. Longitudinal studies would help in further understanding this relationship by identifying any developmental trends in peer victimization, attributions, and adjustment across multiple school years. While there are some studies on characterological self-blame and behavioral self-blame, research is also lacking on peer-blaming attributions which would also elaborate further on this relationship. Interestingly, by examining peer-blame, it not only will offer more information on bullies, but it would also provide more information about the attributions and psychosocial adjustment of frequently victimized children. The relationship between attributions and psychosocial adjustment. Prinstein, Cheah, and Guyer (2005) examined causal self-referent attributions (a very similar construct to characterological self-blame) and child outcomes. They utilized a longitudinal research design to test for this association. The first portion of the study included 116 children from eight kindergarten classes. The measures used accessed attributions such as hostile intent attributions, and critical self-referent attributions. Additional measures included peer nominations, and selfreported measures of social-psychological functioning. Results demonstrated that when children interpreted ambiguous social cues negatively, they also were deriving critical self-referent casual attributions which could be related to internalizing symptoms such as depression, social anxiety, loneliness, and self-esteem. Additionally, when children made critical self-referent attributions from ambiguous experiences it was associated with peer rejection and victimization. Teachers also reported on the adjustment of children. Teacher-reports indicated that children who made critical self-referent attributions in ambiguous situations were more likely to withdraw from peers. This withdraw was related to depressive symptoms. Children that made characterological self-blame attributions had higher levels of depressive symptoms across the seventeen month

22

interval. Similar to the construct of critical self-referent attributions made by victims found by Prinstein et al. (2005), Graham, Bellmore, and Mize (2006) found that victims were more likely to endorse negative views of the self compared to non-victims. Additionally, Graham et al. (2006) found that characterological self-blame attributions mediated the relationship between peer victimization and school adjustment problems such as lower grades and poorer academic engagement. As suggested by these studies, characterological self-blame is an important factor in how children adjust to peer victimization. Therefore, more studies need to be conducted examining this concept. While few studies have been conducted on this topic, each indicates that more frequently victimized children are more likely to make characterological self-blaming attributions compared to children that make peer-blame attributions or behavioral self-blaming attributions. Similarly, other studies (e.g., Graham, 2005) have found that children that make characterological self-blaming attributions are also more vulnerable to higher levels of anxiety, higher levels of loneliness, higher level of depressive symptoms, and lower feelings of self-worth than children that make peer-blame or behavioral self-blaming attributions. Interestingly, little work has been done investigating how children’s attributions may change as a function of their relationship with the peer involved. Toward this end, it is important to examine children’s attributions in the context of their relationship (friend, enemy, unknown peer) with the target peer being evaluated. To date, relationship effects in children’s evaluations have been found by manipulating the relationship between peer being evaluated (Ray, & Cohen, 1997), and placing the observer in a relationship with one of the peers being evaluated (Ray, & Cohen, 2000). Results from these studies demonstrate that children use relationship knowledge while evaluating the behaviors of others and subsequently influence their attributions. That is, best friends are evaluated more positively than are enemies – particularly in ambiguous situations.

23

Attributions and Relationship Types. Understanding children’s differing attributions when they are engaged in several relationships with the target peer is important because attributions of intent can vary as a function of the behavioral reputation of the target peer (Ray et al., 1999). For example, when children evaluate current social situations, they take into account the social status or behavioral reputation of the target peer; thus, their attributions vary based on the target’s social status. Therefore, when children assign attributions to others behaviors, they are likely to assign prosocial intent to prosocial children and hostile intent to aggressive children independent of the actual social situation (e.g., Burgess, Wojslawowicz, Rubin, Rose-Krasnor, & Booth-LaForce 2006; Peets, Hodges, Kikas, & Salmivalli, 2007). While it is important to understand that the attributions children make depend on the social status of the target peer, it is also important to understand how the evaluator’s relationship to the target peer may also influence attributions. Ray et al. (1999) investigated the influence of social relationships on children’s evaluations of peers in potentially conflictual social situations and the influence intentions had on children’s evaluations of social behavior. Children were in either a hypothetical best friendship, an acquaintanceship or enemy relationship with the target peer being evaluated. The target peer’s behaviors were evaluated in the context of a hostile scenario, an accidental scenario or an ambiguous scenario. Results demonstrated that when children were unfamiliar with the victim or disliked the victim in the conflict situation, they reported liking the victim more after the conflict had occurred. There was no relationship found when the victim in the conflict situation was a best friend. There were age differences found for children’s attributions of the aggressor’s intentions. All children perceived the aggressor’s intentions and the victim’s behavioral response negatively in the hostile scenario and positively in the ambiguous and accidental scenarios. However, older children (fifth and sixth graders) evaluated the aggressor’s intentions and the victim’s behavioral responses to be more negative in the hostile scenario compared to younger children’s (second and third graders) evaluations.

24

While Ray et al. (1999) laid a good foundation for assessing attributions of intent in relation to the evaluators relationship with the target peers in the scenario, it left children in a purely evaluator role. That is, Ray et al. did not set the peer making the evaluations in the conflict situation but instead left the evaluator outside of the conflict situation. Recent studies have (e.g., Burgess, Wojslawowicz, Rubin, Rose-Krasnor, & Booth-LaForce 2006; Peets, Hodges, Kikas, & Salmivalli, 2007) placed the evaluator in the conflict situation with another known peer. For example, Burgess et al. (2006) found that shy/withdrawn children reported more prosocial attributions in potentially stressful social situations when their friends were involved than when unfamiliar peers were involved. The study by Burgess et al. (2006) only examined relationships to the provocateur as either an unfamiliar classmate or a friend. It did not examine the attributions of intent when the evaluator did not like the target. Peets, Kikas, Hodges, and Salmivalli (2007) addressed this limitation including a hypothetical situation involving a disliked peer. They were interested in whether children would change their hostile attributions and behavior strategies as a function of the relationship type between the evaluator and the peer being evaluated. Additionally, they were interested in whether the relationship effect would also hold despite the target peer’s social and behavioral reputation. Results demonstrated that children attributed more hostility to their enemies than they did to other peers (friend, unfamiliar classmate) which is consistent with previous research (e.g., Ray, & Cohen, 1997). Like Burgess et al. (2006), Peets and colleagues (2007) found that children gave their friends the benefit of the doubt in a hostile situation while enemies actions were seen as intentional and hostile. These results also held true when previous reputations of the target peers was controlled. This provides insight to debunk any belief that the results were not a product of the target peers reputation. It was also found that when children were identified as prosocial, children were less likely to attribute hostility. Additionally, friends may be seen as having good

25

intentions even when the situation is hostile. Enemies may be seen as hostile no matter the situation (hostile or not). However, these studies focused on children in general. Neither addressed the differences that may exist when victimized children are examined. The Present Study The present study extended previous work (e.g., Burgess et al., 2006) by utilizing ambiguous, hypothetical situations involving three different types of actual peer relationships (unfamiliar peer, best friend and disliked peer). That is, children were asked to write down the name of their best friend, and a disliked peer. Children were then asked to imagine these actual peers for each hypothetical situation. The present study examined whether children’s attributions vary as a function of their relationship with the target peer. Furthering these previous studies (e.g., Burgess et al., 2006; Peets et al., 2007; Ray et al., 2001), the present study examined whether there would be differences between victimized and nonvictimized children. It is first hypothesized that children’s attributions would vary depending on their relationship with the target peer being evaluated. Specifically, it was predicted that situations involving a disliked peer would trigger more hostile attributions while situations involving a best friend or anunknown peer would trigger more prosocial attributions. It was further predicted that situations evaluating best friends would trigger more prosocial attributions than situations involving unknown peers. The second hypothesis was related to victimized and nonvictimized children. That is, compared to nonvictimized children, victimized children’s attributions were predicted to be more influenced by their relationship with the target peer. Specifically, it is predicted that victimized children will make more hostile attributions when the target peer is a disliked peer or an unfamiliar peer compared to nonvictimized children. It is further predicted that victimized children will make more hostile attributions in ambiguous situations when the target peer is an acquaintance. Furthermore, it is predicted that victimized children will make more prosocial

26

attributions in ambiguous situations when the target peer is a best friend. Ultimately, it is hypothesized that victimized children are more vulnerable to the relationship bias than are nonvictimized children. That is, victimized children are more likely to consider their relationship to the target peer than nonvictimized children.

27

Method Participants Participants were 123 sixth grade children from a rural community in the southeastern United States. The return rate of parental consent forms (Appendix A) was 131 (65.5%) with a “yes,” 44 with a “no” (22%), and 25 were never returned with a yes or no (14.3%). Child assent (Appendix B) was also collected. Of the 131 permission slips returned with a “yes,” 8 students declined to participate before the initiation of the questionnaires. Sample Demographics Participants included 72 girls (58.5%) and 51 boys (41.5%). The age of the sample ranged from 11 to 13 years old. The make-up of ages in this sample is as follows: 18 were 11years-old (17.5%), 69 were 12-years-old (67%), and 16 were 13-years-old (15.5%). The mean age of the sample was 11.9 years-old. It is important to note that ages for 20 participants were not known since demographic information was obtained from the teacher only. The racial makeup of this sample is as follows: 51% were Caucasian (63 total), 47% were African American (57 total), and 2% were Hispanic. Design The current study includes one between participant variable: victimization status (victimized, nonvictimized). This study also includes two within-participant variables: relationship type (acquaintance, disliked peer, best friend), and attribution type (hostile, internal, prosocial). Materials Each child completed three questionnaires and a fun activity. This fun activity was not used for data analysis. The first questionnaire (Peer Nominations for Peer Status and Physical Aggression) identified children’s best friends. The second questionnaire (Social Experience Questionnaire) identified children’s perception of peer victimization which ultimately identified

28

the victimized and nonvictimized children. The third questionnaire (Attributions and Coping Questionnaire) included three different parts (best friend, acquaintance, disliked peer) presenting hypothetical scenarios. Scenarios were created with the intent to find out how children interpret the causes of peer victimization. The questionnaires ended with “My favorite thing is…” This measure was included to end the interview on a positive note. Peer Nominations for Peer Status and Social and Physical Aggression (see Appendix C). This measure assessed peer status, physical aggression, social aggression, prosocial behavior, best friends, and shy children using peer nominations (Crick & Grotpeter, 1995). Children were asked to answer the questions by providing the first and last name of the student they thought fit the description. Some examples of questions included: “who is nice to everyone,” “who starts fights,” “who gets along well with others,” “who gets picked on,” “who do you like the best,” and “who is shy?” Children nominated as many children for each description as they wanted. This present study only included questions pertaining to the identification of friendships (“who do you like the best”) and disliked peer (“who do you disliked”). To identify best friendships, participants were also asked to circle or write a “1” next to their best friend for the question “who do you like the best.” Social Experience Questionnaire (see Appendix D). This self-report measure assessed how often children believe they experience overt aggression, relational aggression, and prosocial behavior by their peers (Crick & Grotpeter, 1996). It measures the frequency with which peers attempt or threaten to harm their peer relationships (e.g., “How often do other children leave you out on purpose when it is time to play or do an activity”), other children’s attempt or threaten to harm their physical well-being (e.g., “How often do you get hit by another child at school”), and are targets of peers’ caring acts (e.g., “How often does another child say something nice to you”). The participants rated on a five point Likert-type scale how often they

29

believed the described behavior happens to them. “1” is never, “3” for sometimes, and “5” for all the time. For the present study, children that were identified as being one standard deviation above the mean were identified as victimized and one standard deviation below the mean were identified as nonvictimized. The present study utilized questions that specifically asked children’s experience of overt aggression. Attributions and Coping Questionnaire (ACQ) (see Appendix E-G). This assessment measures children’s attributions, emotional reactions, and coping strategies in hypothetical situations when the target is a general peer versus a friend (Burgess et al., 2006). The present study modified this measure in that vignettes involving friends were changed to actual best friends. Additionally, the present study added a third vignette to assess children’s attributions, emotional reactions, and coping strategies to an actual disliked peer. This measure involved children reading a series of short vignettes and then answering questions about why they think this happened to them, their emotional reactions, and how they would cope with the situation. In all scenarios and relationship types, children were asked to imagine that the situation involved an actual peer whether an acquaintance (Appendix E), best friend (Appendix F), and disliked peer (Appendix G). To ensure that children would think about an actual peer, peer nominations asked children to list their best friend, and a disliked peer. To determine each participants’ attribution score, the same procedure utilized by Rubin, Wojslawowicz, Rose-Krasnor, Booth-LaForce, and Burgress (2006) was followed. There are a total of five different hypothetical scenarios for each relationship type. The target peer involved is either an unfamiliar classmate, a best friend, or someone they do not like very much. This yielded a total of fifteen different scenarios requiring an attribution. Attributions were classified into one of three types. The three attribution types are: hostile, internal blame, and prosocial. After evaluating the hypothetical scenarios, participants received a total score for the

30

type of attributions they made. For example, if for two of the scenarios for unfamiliar peer the participant recorded hostile attributions and for the remaining scenarios recorded prosocial attributions, the participant would receive a score of two for hostile attributions and three for prosocial attributions. This procedure was repeated for each relationship type (unfamiliar classmate, best friend and disliked peer). Thus for attributions, each participant read three sets of scenarios: one for an acquaintance (Appendix E), one set for best friend (Appendix F), and one set for disliked peers (Appendix G). There were three portions to each scenario: attributions, coping responses, and emotions. For the current study, the attributions section was the only one used. Additionally, the original measure included four attributions while this study changed this total to three leaving out the neutral attribution. My Favorite Thing (see Appendix H). This measure was included to end the interview process on a happy note. Children were to think about something that makes them happy and to write or draw it. Procedures All participating children returned a signed parental consent form (Appendix A) and gave their own assent (Appendix B) to participate at the time of data collection. Children were informed that their answers would not be shared with anyone else and that they should not share their answers with any of their friends. Data collection extended over the course of four days with two classes per day. Group sessions consisted of about 20 children. Five research assistants were present throughout the group interviews to answer any questions that the children may have had. Children worked independently and at their own pace. Children were seated with enough space between them and were instructed to place folders in front of their surveys to ensure privacy. All children began the study giving their assent. Assent was followed by the peer

31

nominations questionnaire. After the peer nominations, children moved to the social experience questionnaire. The attributions (ACQ) questionnaire was administered next in the following order: first was the scenario about the acquaintance, followed by best friend, and ending with the disliked peer. After all surveys were administered, children then filled out the “My Favorite Thing.” All children remained in their classroom sitting quietly or reading a book until the final participant completed the measures.

32

Results Results are divided into two sections. The first section includes analysis that investigated the association between relationship type and attribution type. The second section investigated the link between relationship type and attribution type as a function of victimization status. Victimization was divided into two categories: victimized and nonvictimized. Follow-up tests of statistical significance interactions were conducted using Newman-Keuls post-hoc tests to determine sources of differences. The Association between Attributions and Relationship Type To investigate the association between attributions and different type of peer relations, a 3 (relationship type) x 3 (attribution type) repeated measure MANOVA was conducted. As predicted, results revealed a significant interaction, F (1, 120) = 7616.68, p < .001. Post-hoc follow-up test revealed that when evaluating best friends, children were more likely to make prosocial attributions than either internal or hostile attributions which did not differ. On the other hand when evaluating a disliked peer children were more likely to attribute hostile intent than prosocial or internal attributions. When evaluating an acquaintance, children were more likely to endorse hostile attributions than either prosocial or internal attributions. In summary, when the situation involved a best friend, more prosocial attributions were triggered while situations involving a disliked peer triggered more hostile attributions. Best friends were more likely to trigger prosocial attributions than hostile and internal attributions. Additionally, children were equally more likely to endorse hostile attributions than internal attributions in scenarios involving best friend. Disliked peers triggered more hostile attributions than prosocial and internal attributions. Specifically, children were equally likely to endorse either hostile or internal attributions in scenarios involving a disliked peer. When the situation involved an acquaintance, children were more likely to endorse hostile attributions than internal or prosocial attributions. Additionally, when the situation involved an acquaintance, children

33

were also more likely to endorse prosocial attributions than internal attributions. Acquaintances were less likely to trigger internal attributions. Table 1 lists all means and standard deviations for the interaction between relationship and attribution type.

Table 1: Attribution x Relationship Interaction Best Friend (BF)

Acquaintance (AQ)

Disliked Peer (DP)

Attributions

Mean (SD)

Mean (SD)

Mean (SD)

Internal (I)

.11 (.14)

.09 (.13)

.05 (.11)

Hostile (H)

.17 (.23)

.45 (.32)*

.76 (.34)*

Prosocial (P)

.36 (.13)*

.23 (.16)

.09 (.14)

Effect

P > H*; P > I*

H>P>I*

H > P*; H > I*

*p AQ > BF *

BF > AQ > DP *

*p