SEARCH

SEARCH BY CITATION

Keywords:

  • bullying;
  • victimization;
  • social goals;
  • social information processing

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Conclusion
  8. References

Understanding what social goals are associated with bullying and victimization behaviours, even after allowing for biases in interpretation of and affective responses to social events, is critical for understanding the socio-behavioural profile of bullies and victims. In the present study, 181 nine- to ten-year-olds' affective responses, attribution of intent, and social goals were assessed in the context of a series of ambiguous and overtly hostile provocation vignettes. Results showed that even after allowing for other social information processing biases, social goals were meaningfully associated with bullying and victimization scores. Bullying was inversely associated with relationship-building goals, and positively associated with goals to be assertive over the provocateur when provocation was overtly hostile. Being victimized was associated with having submissive goals even when provocation was ambiguous and after accounting for attribution of hostile intent. Findings are discussed in light of theoretical and practical implications.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Conclusion
  8. References

The role of social cognition in bullying has come under increased scrutiny of late, not least as a consequence of the discrepant explanations for bullies' aggressive behaviour put forward by proponents of the social information processing (SIP) approach and researchers who have emphasized the role of theory of mind in bullying. The former scholars argue that various information-processing biases precede the inappropriately aggressive behaviour inherent in bullying. For example, Camodeca and Goossens (2005) reported that bullies attributed malicious intent behind a story character's provocation even when the root cause of the provocation was ambiguous. By contrast, researchers who have identified superior theory of mind skills in ring leader bullies (especially in comparison to their followers and victims; see Sutton, Smith, & Swettenham, 1999b) point to bullies' competence in evaluating and manipulating social situations as facilitating successfully implemented, aggressive behaviour. It remains unclear whether bullies' aggressive behaviour is representative of skilful or deficient social cognitive processing, but a consensus seems to be forming between the two positions that one of the keys to unlocking the debate potentially lies in understanding the social goals that drive this aggression. For bullies to engage in ‘effective’ aggression that brings at least short-term gains, the skill required to do so must be accompanied by the will to perform antisocial aggressive acts, with the latter inferred in part from their social goals.

A Social Goals Perspective on Bullying and Victimization

Clear evidence exists that a child's social goals in response to a social situation are closely related to their subsequent behaviour (see Crick & Dodge, 1994), and thus they are likely to be integral in predicting a developmental trajectory of bullying behaviours. Indeed, evidence already indicates that social goals are likely to explain a degree of variance in bullying scores, thus meriting further investigation into the area (e.g., Sijtsema, Veenstra, Lindenberg, & Salmivalli, 2009). As Erdley and Asher (1999) point out, understanding social goals' influence on bullying behaviours is likely to be critical in shaping intervention work, as changing the behaviours (e.g., by introducing new policy, developing social skills, and so on) may be ineffective in the long term if the child's social goals remain the same.

In spite of this, the social goals of aggressive children in response to provocation have received relatively little attention compared with other aspects of their SIP (Lemerise, Fredstrom, Kelley, Bowersox, & Waford, 2006), and research on this topic has rarely focused on connections with bullying in particular (see Camodeca & Goossens, 2005, for an exception). Moreover, research from a social goals perspective that has considered the construct of bullying has tended to use global trait-like goal scales (e.g., the Interpersonal Goals Inventory for Children (IGI-C); Sijtsema et al., 2009). Although trait-like individual differences in goal orientation may indeed play a key role in influencing behavioural responses (e.g., reacting aggressively to provocation), such broad links are likely to be indirect; much remains unknown regarding the role played by situation-specific social goals in scenarios such as provocation. The present study is designed to help us gain a more precise understanding of what types of goals might be motivating the behaviour of bullies and victims in particular social contexts, including provocation scenarios where there is obvious hostile intent and other scenarios where there is more ambiguity about the provocation.

Social Goals in Response to Provocation

Although very little research has specifically considered the social goals of bullies in response to provocation, various inferences can be made from studies of SIP in aggressive and socially maladjusted children. Traditionally, a distinction has been drawn between relationship-building goals and instrumental goals, with aggressive children reported to demonstrate bias towards the latter in response to provocation (Crick & Dodge, 1996, 1999). However, as Sutton, Smith, and Swettenham (2001) point out, the distinction of relational vs. instrumental goals is problematic. The two goal dimensions are not necessarily mutually exclusive of each other, as children are able to entertain multiple goals simultaneously (Lemerise et al., 2006; Rabiner & Gordon, 1992). Consequently, it is difficult to develop a comprehensive understanding of what motivations lie behind bullies and victims' behaviour unless a wider range of goals are considered, with specific relevance to the specific provocation contexts at hand.

A good indication as to what goals may be worth investigating in the present research effort comes from existing literature that has focused on the social goals endorsed by socially maladjusted groups of children (for a review, see Erdley & Asher, 1999). For example, Erdley and Asher (1996) categorized eight- to ten-year-old children who favoured aggressive, withdrawn, and problem-solving behavioural responses to ambiguous provocation, and reported different patterns of social goal importance ratings across these categories. Aggressive children rated goals for revenge and retaliation significantly higher than the other groups and gave lower ratings for goals of getting along with the provocateur and resolving the problem peacefully, even when they did not infer hostile intent on the part of the provocateur. Withdrawn children attached more importance to avoiding the provocateur than the other groups, but only significantly so when they interpreted hostile intent behind provocation.

Other research has examined the influence that emotional displays of the provocateur may have on children's social goals in response to provocation. Lemerise et al. (2006) reported that regardless of whether the provocateur was depicted as happy, angry, or sad, aggressive-rejected children placed more importance on goals of assertiveness and dominance than non-aggressive children. These children also favoured revenge goals more than their peers when the provocateur was depicted as being angry or sad, and rated social relational goals less positively than popular non-aggressive children when the provocateur was depicted as being angry.

This socio-motivational profile of aggressive-rejected children may overlap with that of bullies. However, because bullies frequently display cool, callous, but adaptive behaviour (Boulton, 1999), we argue that they may not be immediately focused on a hot-headed desire for revenge, but instead may be more concerned with maintaining their image of dominance. Indeed, research that has considered trait-like goals has reliably found bullies to be motivated by dominance (Kiefer & Ryan, 2008; Vaillancourt, Hymel, & McDougall, 2003; Veenstra et al., 2007). When provoked, the dominance an individual has over his or her peers is threatened, and bullies are therefore likely to be motivated to maintain it by seeking to appear assertive towards their provocateur.

Erdley and Asher (1996) also cited evidence that aggressive children tend not to hold goals for getting along with the provocateur, which is in line with previous evidence from the aggression literature (Crick & Dodge, 1989; Renshaw & Asher, 1983; Slaby & Guerra, 1988). Low regard for relationship building is also part of the common perception of a bully, and we thus can expect bullying to be negatively associated with this goal. We would not expect any association with victimization, however. In fact, evidence is emerging that victims may actually possess an acute empathic responsiveness, and may be hyper-vigilant to the feelings of others (Malti, Perren, & Buchmann, 2010; Zahn-Waxler, Cole, & Barrett, 1991). It is possible that non-aggressive victims' withdrawn behaviour is partly the result of goals that revolve around the feelings of others, such as a concern not to upset or anger the provocateur. We believe these goals are conceptually distinct from proactive relationship-building goals and so are investigated separately in the present study.

It is worth noting here that several studies have indicated a subgroup of victims who are reactively aggressive and have retaliatory goals (e.g., Salmivalli & Nieminen, 2002). These children, also referred to as ‘bully-victims’, have a distinctive social-cognitive profile (Hanish & Guerra, 2004; Toblin, Schwartz, Gorman, & Abou-ezzeddine, 2005), and are likely to endorse different goals from both passive victims and bullies. For example, Salmivalli, Ojanen, Haanpaa, and Peets (2005) found that agentic goals are positively correlated with the proactive aggression observed in bullies, but not the reactive aggression that is associated with aggressive victims. Investigation into the social goals of these aggressive victims is an important avenue for future research, but it is considered beyond the scope of the present study.

Social-cognitive, Affective, and Situational Influences on Social Goals

Having established a range of goals that could conceivably be associated with bullying and victimization, it is now important to review the role played by other SIP biases. In Crick and Dodge's (1994) SIP model, children's social goals are placed within a cycle of factors that may contribute to behaviour. In particular, children who make hostile attributions about their peers are more likely to pursue retaliation goals (Erdley & Asher, 1996; Slaby & Guerra, 1988), but may also adopt goals that involve withdrawing from the confrontation (Erdley & Asher, 1999). Moreover, SIP biases that occur in each step of Crick and Dodge's (1994) model, including biases in goal formulation, can be energized by emotions (Lemerise & Arsenio, 2000). For example, being in an angry mood makes it more likely that a child will interpret threat in any form of provocation and that they will focus on instrumental over relationship-building goals (Harper, Lemerise, & Caverly, 2010).Thus, the present study investigates which social goals are able to explain variance in bullying/victimization scores independently from attribution of intent and affective responsiveness.

Finally, it is important to consider that although children's affective response and attribution of intent may hold some sway over the goals they endorse, the biases are in themselves influenced by the context from which they are assessed. In general, when provocation is ambiguous, children are less likely to attribute intent or to report feeling angry than when provocation is hostile. However, in such scenarios, aggressive and withdrawn children have been found to attribute more hostile intent than their peers and formulate different goals accordingly (Erdley & Asher, 1996, 1999). In order to allow for any variance in SIP biases across modes of provocation, the present study also investigated whether any associations between social goals and bullying and victimization are consistent across ambiguous and hostile provocation scenario types. Moreover, because it has been contended that bullies actually demonstrate adaptive behaviour to gain benefits from situations (Sutton, Smith, & Swettenham, 1999a), it is worth considering whether they are able to adapt their goals between the two scenario types. If they are adaptive, bullies should identify the risks that hostile provocation brings to their image of dominance and consequently place more importance on being assertive over the provocateur than when provocation is ambiguous. In line with the existing literature that has reported on victims' inability to switch their goals effectively, it is expected that the goals associated with victimization will be consistent across provocation types.

In summary, the current study has three main aims. Firstly, it is designed to investigate which of the social goals drawn from the existing SIP literature are associated with bullying and/or victimization in response to provocation scenarios. Secondly, the study assesses whether these goals remain associated with bullying/victimization after controlling for variance explained by affective response and attribution of intent. Thirdly, the role played by social goals is evaluated in two different types of provocation scenarios, distinguishing between provocation that is ambiguous or overtly hostile.

Method

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Conclusion
  8. References

Participants

The sample consisted of 90 boys and 91 girls, ages 7 years and 7 months to 10 years and 7 months (M = 9,1; SD = 10.28 months), recruited from six classes of a school within a middle socioeconomic status community. Children were predominantly White and all were native English speakers. Consent for participation was obtained from the school authorities, and parents received a letter informing them of the experimental procedure and giving them the option to withdraw their child from the study. None of the parents took up this option; hence, the participation rate was 100 percent subject to absences on the day of testing.

Measures

Two computerized measures were carried out in the present study: a bullying questionnaire, and an SIP measure depicting a series of scenarios that considered the affective response of participants to provocation, their attribution of intent, and the social goals endorsed.

Bullying Questionnaire

Children were asked to nominate classmates they had seen engaging in particular behaviours. Nominations were limited to three per question. Questions were designed to measure both physical and relational behaviours associated with bullying and being bullied. The questionnaire consisted of four physical bullying items (e.g., pushing or tripping another child on purpose), four relational bullying items (e.g., stopping another child from joining in games), four physical victimization items (e.g., being hit by other children), and four relational victimization items (e.g., having rumours made up about them behind their backs), along with four filler questions. The order of the questions was initially randomized and then kept in a fixed order for each child. As well as being presented on the computer screen, each question was read by a male experimenter, with clarification of any terms used that the child failed to understand given when necessary. Children entered nominations on the computer using code numbers that had been assigned to names on the class list.

Social Information Processing Task

Children heard eight provocation scenarios. In four of these vignettes, the provocateur's intent was ambiguous (adapted from Camodeca & Goossens, 2005):

Imagine you're on a school trip to a big adventure country park. You have built a really great den using branches from the woods. It's bigger than any you've ever built before and you're really happy with it. Someone from your class sees what a good job you've done and comes to have a better look. You are building a new roof so that the den will be a dry and safe place, even when it rains. But then the child from your class puts a really heavy branch on top and the whole den collapses under the weight!

In the other four, the provocation was clearly hostile (adapted from Lösel, Bliesener, & Bender, 2007):

Imagine you've been away on summer holiday and had a really nice time. You're really looking forward to seeing all your friends at school again. You arrive at school and go to class but none of your classmates are talking to you. Your best friend comes up to you and tells you that Claudine, one of the girls in your class, has been making up stories about you. She's been telling them to the whole class. You wished she hadn't because the stories aren't very nice and they're not even true.

All stories were of similar length and verbal complexity.

Participants were then asked a series of questions that assessed their SIP in response to the depicted provocation. The first two questions, taken from Camodeca and Goossens (2005), considered their affective response: ‘How angry/sad would you feel if this happened to you’ (answers on a 4-point scale from not at all angry/sad to really angry/sad). The following four questions, also from Camodeca and Goossens (2005), assessed attribution of intent: ‘Do you think the other child is mean? Do you think (s)he did it on purpose? Do you think (s)he was happy with what (s)he had done? Do you think (s)he should be blamed for doing it?’ (No, I do not know, Yes). In order to assess children's social goals, they were asked: ‘Do you think it is important… (1) to tell the other child that (s)he can't help with your den again? (assertiveness); (2) to get your own back for what the other child did? (revenge); (3) to avoid the other child? (avoid provocateur); (4) to avoid getting into trouble by not causing a fuss? (submissiveness); (5) that you get along with the other child? (relationship building); (6) that the other child does not feel upset about what happened? (prosocial concern for others’ feelings); (7) that the other child does not feel angry with you? (self-protective concern about others' feelings)'. Answers were on a 4-point scale ranging from not at all important for me to very important for me. The social goals items detailed above were adapted from a number of studies, most notably Lemerise et al. (2006).

Scoring

Bullying Questionnaire

With each nomination counted as one point, scores were obtained for physical bullying, relational bullying, physical victimization, and relational victimization. These scores were then standardized by class. Factor analysis supported a four-factor model explaining 75.63 percent of variance (αs = .97, .90, .75, .76, respectively), with loadings of items onto the expected factors exceeding .58 in all cases.

Social Information Processing Task

Scores for feeling angry and sad ranged from 1 (not at all angry/sad) to 4 (really angry/sad) in each scenario. The mean scores were calculated within scenario type. Hence, each child had scores that ranged between 1 and 4 for feeling angry in ambiguous scenarios, feeling sad in ambiguous scenarios, feeling angry in hostile scenarios, and feeling sad in hostile scenarios. Attribution of hostile intent scores were obtained for each scenario by counting the number of ‘Yes’ responses over the four questions. Scores were tallied over the four scenarios within each scenario type and then averaged over the number of scenarios completed. Consequently, each child then had scores ranging from 0 to 4 for perceived threat in ambiguous scenarios and for perceived threat in hostile scenarios. Scores for social goals were obtained by taking a mean of the scores within a scenario type for each socio-motivation. Fourteen scores (ranging from 1 to 4) were thus obtained, one for each social goal in ambiguous scenarios and one for each social goal in hostile scenarios. Participants were only considered to have completed a scenario if they answered all the questions that followed it. In all cases, participants had completed every question for at least two of the four stories within each scenario subtype, and thus none were excluded from analysis.

Design and Procedure

Children were seen by class in their school's computer room. They accessed each questionnaire online under the supervision of a male experimenter. For each of the measures, the experimenter read out the questions in order, and children were encouraged to wait for him to have done so before responding. On demand, any terms used in the questionnaires that caused confusion were explained to participants. Additionally, two teaching assistants were on hand to assist children with low reading ability. The bullying survey was administered first and was completed within a 30-minute session. The SIP task took place over two 30-minute sessions, with four scenarios (two ambiguous and two hostile provocation) completed in each session. The three sessions were completed within 2 weeks of each other in all cases.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Conclusion
  8. References

Bullying and Victimization Scores

Mean raw scores for bullying and victimization are displayed in Table 1. Mixed 2(gender) × 3(school year group) × 2(type: physical vs. relational) analyses of variance (ANOVAs) (with type of bullying/victimization as a within-subjects factor) were carried out to determine whether there were any effects of gender, year group, or type of bullying/victimization on bullying or victimization scores. The ANOVAs revealed a main effect of gender on bullying, F(1, 175) = 7.73, p < .01, as well as an interactive effect of gender by type of bullying, F(1, 175) = 18.27, p < .01. Specifically, physical bullying but not relational bullying scores were higher in boys than in girls. Girls scored higher in relational bullying than in physical bullying whereas both forms were evident to a similar level in boys. There were no gender differences in victimization scores, but an interaction effect existed between gender and type, F(1, 175) = 13.73, p < .01. Whereas girls tended to be victimized by relational rather than physical methods, boys were victimized evenly in both forms. There were no effects of school year group on bullying or victimization scores.

Table 1. Mean (SD) Bullying and Victimization Nominations by Gender (Bullying Nominations Ranged From 0 to 56 in Boys and From 0 to 37 in Girls, and Victimization Nominations Ranged From 0 to 30 in Boys and From 0 to 26 in Girls) and Year Group
 NBullyingVictimization
PhysicalRelationalPhysicalRelational
Boys907.09 (12.13)6.11 (8.78)5.03 (5.23)5.08 (5.63)
Girls911.56 (4.45)4.54 (7.11)2.89 (2.81)5.15 (4.47)
Overall1814.31 (9.51)5.32 (8.00)3.96 (4.32)5.12 (5.07)

Story Responses

The potential range and the mean scores for children's responses to each component of the SIP measure are displayed in Table 2. In order to test for differences in dependent variables between gender and story type, a series of mixed 2(gender) × 3(year group) × 2(story type: ambiguous or hostile) ANOVAs were carried out, with story type as a within-subjects factor. The ANOVAs indicated no gender differences in any of the dependent variables. The only significant effect of age was on attribution of intent, with year 5 children attributing significantly less hostile intent to the provocateur than year 3 and year 6 children, F(1, 175) = 6.69, p < .01. There were also several differences in SIP scores across scenario type. As expected, children attributed significantly more hostile intent in the hostile scenarios, F(1, 175) = 254.77, p < .01. Children were also significantly more likely to be motivated by being assertive, F(1, 175) = 83.32, p < .01, avoiding the provocateur, F(1, 175) = 82.40, p < .01, and less likely to be concerned with getting on with the provocateur, F(1, 175) = 12.38, p < .01, and not upsetting them F(1, 175) = 13.75, p < .01, in response to the hostile rather than the ambiguous scenarios. Additionally, the ANOVAs indicated an interaction between gender and story type for both the ‘stay out of trouble’ social goal, F(1, 155) = 5.40; p = .02, and the ‘not anger the provocateur’ social goal, F(1, 154) = 10.37; p < .01. In scenarios of ambiguous provocation only, boys showed less concern than girls not to anger the provocateur but more concern for staying out of trouble.

Table 2. Mean Scores for Each Social Information Processing Item by Gender and Story Type
 NAffective responseIntentSocial goals
Feel angryFeel sadAssertiveRevengeAvoid provocateurSubmissiveGet on with provocateurNot upset the provocateurNot anger the provocateur
Possible range 1–41–40–41–41–41–41–41–41–41–4
Boys913.483.392.213.352.262.672.862.592.802.98
Girls903.413.511.993.382.192.782.842.722.953.15
Ambiguous stories3.433.441.413.072.212.442.832.763.013.06
Hostile stories3.463.462.783.652.273.002.872.542.743.05
Overall1813.453.452.103.372.232.722.852.652.873.06

Relations between Responses to Social Scenarios and Bullying/Victimization

The main analysis focused on the associations between social goals and the bullying/victimization scores. Firstly, we noted that a subgroup of ‘bully-victims’ could potentially mask distinctive patterns of associations between social goals and bullying/victimization scores in the correlational analysis. In order to identify children belonging to this subgroup, overall bullying nominations were tallied (physical + relational nominations), and then standardized within class (in case of different patterns within individual classes) and gender (as initial analysis indicated gender differences in scores of physical vs. relational bullying and victimization). The process was repeated for victimization scores. Children who scored more than one standard deviation above the mean score in bullying and victimization were classified as bully-victims and were excluded from subsequent analysis. A total of 11 children were excluded in this way, leaving a remaining sample size of 170.

Table 3 shows the intercorrelations among all measures in both scenario types. There is some evidence for the expected associations between bullying and victimization on the one hand, and affective responses, attributions of hostile intent, and social goals on the other. As predicted, the goals associated with bullying and victimization differed, and there was some variance in associations across scenario type. In hostile scenarios, both forms of bullying were associated with feeling angry, and physical bullying was also inversely related to concerns for getting on with, and not upsetting, the provocateur. In ambiguous scenarios, physical bullying was again inversely correlated with concerns for getting on with, and not upsetting, the provocateur. Relational bullying was associated with feeling angry and a lack of concern for getting on with the provocateur. Physical victimization was associated with a goal for submissiveness in hostile scenarios, and a goal for avoiding the provocateur in response to ambiguous provocation. Relational victimization was associated with submissive goals in both scenario types.

Table 3. Correlation Matrix for Bullying/Victimization Scores and All Measures of the Hostile Provocation Stories Above and Ambiguous Provocation Stories Below the Diagonal
 1234567891011121314
  1. * p < .05, ** p < .01.

 1Physical bullying.71**.11.02.21**−.19*−.01.07−.04−.03−.03−.20*−.16*−.06
 2Relational bullying.71**.21**.16*.23**.00−.01.07−.01−.01.07−.11−.15−.02
 3Physical victimization.11.21**.60**.03.05.13.07−.01.08.18*−.05.03.04
 4Relational victimization.02.16*.60**−.17*.11.12.13−.08.08.20*−.04−.02.09
 5Feel angry.15.21**−.01−.07.28**.24**.15.28**.17*.19*−.13.02.10
 6Feel sad.09.15−.06.06.62**.09.27**−.03.29**.21**.12.25**.30**
 7Interpreted hostile intent.02.05.11.12.50**.40**.30**.28**.47**.33**−.33**−.19*.00
 Social goal to…
 8Be assertive.11.14.12.04.54**.52**.58**.02.12.21**.06.18*.24**
 9Get revenge on provocateur.05.07.02−.03.45**.32**.58**.49**.36**.24**−.22**−.09−.06
10Avoid the provocateur−.02.00.16*.11.46**.44**.69**.63**.57**.40**−.23**−.10−.04
11Be submissive.08.08.14.20*.21**.24**.34**.33**.30**.35**−.06.13.21**
12Get on with the provocateur−.24**−.20*−.04−.07−.32**−.23**−.39**−.28**−.31**−.36**−.06.55**.43**
13Not upset the provocateur−.20*−.09.04.01−.01.08−.13−.04−.07−.06.27**.42**.60**
14Not anger the provocateur−.16*.01.05.15.08.13.09.05.08.12.43**.31**.65**

The correlation matrix also shows that some social goals are associated with affective responses and intent attributions. Thus, the extent to which social goals could predict bullying and victimization characteristics over and above any significant effects of affective responses and hostile intent attributions was evaluated. Each bullying and victimization score (physical bullying, relational bullying, physical victimization, and relational victimization) was regressed on the other measures, which were entered in three blocks. The first block evaluated the predictive value of the affective response score, the second block included the intent attribution score, and the seven social goal scores were added into the third block using stepwise criteria for entry into the model.

Table 4 shows the results of these analyses for the ambiguous provocation scenarios, and Table 5 shows the results of these analyses for the hostile provocation scenarios. As predicted, the social goals associated with bullying differed from those associated with victimization. In the ambiguous provocation scenarios, both physical and relational bullying remained inversely related to a concern for getting along with the provocateur after allowing for affective response scores and attributions of hostile intent, and relational bullying was also inversely related to goals to avoid the provocateur. In the same scenarios, physical victimization was predicted by goals of avoiding the provocateur and relational victimization by a goal for submissiveness.

Table 4. Hierarchical Regression Analysis for Ambiguous Provocation Measures as Predictors of Bullying/Victimization Scores
 Physical bullyingRelational bullyingPhysical victimizationRelational victimization
  1. Note: Only significant predictors are shown in Block 3.

  2. * p < .05, ** p < .01.

Block 1ΔR2.02.04.00.02
Feel angryβ.15.21*.04−.17
Feel sadβ.01.01−.07.16
Block 2ΔR2.00.00.01.02
Feel angryβ.15.22*−.01−.24**
Feel sadβ−.01.01−.09.14
Interpreted hostile intentβ.00−.02.12.17
Block 3 (social goals)ΔR2.07.06.03.03
Feel angryβ   −.23**
Avoid the provocateurβ −.25**.23** 
Submissivenessβ   .19*
Get on with the provocateurβ−.22**−.19*  
Final model R2 .09.10.04.08
Table 5. Hierarchical Regression Analysis for Hostile Provocation Measures as Predictors of Bullying/Victimization Scores
 Physical bullyingRelational bullyingPhysical victimizationRelational victimization
  1. Note: Only significant predictors are shown in Block 3.

  2. * p < .05, ** p < .01.

Block 1ΔR2.11.06.00.06
Feel angryβ.28**.25**.02−.22**
Feel sadβ−.27**−.06.03.17
Block 2ΔR2.00.00.00.01
Feel angryβ.30**.26**.01−.25**
Feel sadβ−.27**−.06.03.16
Interpreted hostile intentβ−.08−.07.05.12
Block 3 (social goals)ΔR2.07.03.03.03
Feel angryβ.32**.26** −.26**
Feel sadβ−.29**   
Be assertiveβ.16*   
Submissivenessβ  .19*.20*
Get on with the provocateurβ−.22**   
Not upset the provocateurβ −.17*  
Final model R2 .18.09.03.10

Interestingly, there was a slightly different pattern in the social goals associated with bullying and victimization in scenarios of hostile provocation. In hostile provocation scenarios, physical bullying was again negatively predicted by a concern for getting along with the provocateur, but was also associated with goals for being assertive as hypothesized. Relational bullying, on the other hand, was predicted by a lack of concern for upsetting the provocateur only. Victimization was again predicted by goals of harm avoidance as expected, with both physical and relational victimization related to submissive goals.

Finally, we tested for potential moderation of social goal effects by gender. The four bullying and victimization scores were regressed on social goal scores, gender, and terms for the interaction between gender and each social goal score. No effects of moderation by gender were apparent (for all interaction terms, ps > .10).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Conclusion
  8. References

The present study offers clear evidence that social goals in provocation scenarios are systematically associated with levels of bullying and victimization behaviours, even after controlling for affective responses to the provocation and attribution of hostile intent. Moreover, the results suggest that bullies, but not victims, are able to adapt their social goals according to the nature of the provocation, endorsing goals for assertiveness only when provocation was overtly hostile.

Previous research that has considered the social goals of bullies reliably reports that such individuals place great importance on social dominance (Arsenio & Lemerise, 2001; Hawley, 2003; Rodkin, Farmer, Pearl, & Van Acker, 2000; Veenstra et al., 2007) and hold relationship damaging goals in response to provocation (Camodeca & Goossens, 2005; Crick & Dodge, 1999; Lochman, Wayland, & White, 1993; Rabiner & Gordon, 1992). Providing further support for these links, the present study found physical bullying to be associated with low levels of concern for getting on with the provocateur in both scenario types, and also with goals for assertiveness in response to the hostile provocation vignettes, even after allowing for SIP biases. Associations between social goals and relational bullying showed a similar pattern with regard to low relationship-building goals (relational bullying was associated with a low concern for getting on with the provocateur in response to ambiguous provocation, and low concern for not upsetting the provocateur in response to hostile provocation), but there was no association with goals for assertiveness. Because the above associations remained even after allowing for SIP biases, the present study provides good evidence that social goals may be particularly important in understanding how bullies respond to social situations.

The association between physical bullying and the goal for assertiveness was only evident in response to the overtly hostile scenarios. We argue that this is indicative of adaptive goal selection in bullies as their dominance over their peers is only realistically threatened in these scenarios. Interestingly, this association was not significant in the correlational analysis, implying that bullies hold goals for maintaining dominance in spite of the other SIP factors assessed, not because of them. Although relational bullying was not associated with assertiveness goals, this is arguably compatible with the conceptualization of the relational bully as a subtle manipulator of social groups who is concerned with their social image, and thus may not opt for the type of direct confrontation implied by assertive goals. Indeed, the relational bully may be more likely to prevent a repeat of hostility by delivering negative long-term consequences to the provocateur by means of gossip and peer-group manipulation. Physical bullying was also negatively associated with goals for avoiding the provocateur in ambiguous provocation scenarios. Again, we argue that this demonstrates an adaptive goal selection in bullies because there is little need to endorse a goal of avoidance in response to ambiguous provocation.

The social goals associated with victimization differed considerably from those associated with bullying. In line with the existing literature (Erdley & Asher, 1996; Veenstra et al., 2007), victimization was found to be positively associated with goals for harm avoidance (in response to ambiguous provocation) and submissiveness (in both scenario types). Erdley and Asher (1999; see also Juvonen & Graham, 2001) posited that the attribution of intent may lead withdrawn children to hold goals for avoiding the provocateur, and this is supported by the strong associations reported in the present study between attribution of intent and goals for avoiding the provocateur in both hostile and ambiguous scenarios. However, even after accounting for attributed intent and affective response, the goal for avoiding the provocateur remained associated with physical victimization, but only in ambiguous provocation scenarios where it is arguably misplaced. Ojanen, Aunola, and Salmivalli (2007) have suggested that victims have specific difficulties adapting their goals to the situation, and the present study has also indicated that victims were unable to adapt their goals in accordance with the nature of the provocation. It is not clear why the same association was not apparent with relational victimization, but it follows that because relational victims are not necessarily confronted by their harasser in their victimization experiences, they may not hold as strong a desire to avoid provocateurs especially when provocation is ambiguous in intent.

In hostile provocation scenarios, the submissive goal for ‘staying out of trouble by not causing a fuss’ predicted both forms of victimization even after allowing for affective responses and attribution of intent. That victims may hold submissive goals is well supported in the existing literature. For example, Ojanen et al. (2007) and Sijtsema et al. (2009) have reported that victims lacked status goals, perhaps because experiences of harassment leave victims with feelings of hopelessness and anxiety when dealing with provocation (Burgess, Wojslawowicz, Rubin, Rose-Krasnor, & Booth-Laforce, 2006).

Thus, the present study has provided a more detailed picture of the social goals endorsed by victims. The reported associations imply that physical victims' desire to avoid even ambiguously intended provocation exists over and above the influence of their affective response and attribution of intent. Victims also hold submissive goals (even after allowing for SIP biases) in response to overtly hostile provocation. With these findings taken together, the goals associated with victimization (perhaps the consequence of an avoidance schema developed through previous experiences of harassment) are arguably non-adaptive in resolving the situations successfully, and may increase the likelihood that the victim will become the target of subsequent aggressive acts.

Whereas the social goals associated with bullying/victimization reported here are fairly congruent with the existing literature, it is worth considering some of the associations among affective response, interpreted hostile intent and bullying/victimization that were less expected. Firstly, bullying was not associated with interpreted intent in either scenario type. Although this would be expected in ambiguous provocation scenarios, it is not clear why bullying was not associated with interpreted intent in response to overtly hostile provocation. Although this calls into question the validity of the attribution of intent measure, the associations that were reported among attribution of intent, affective response, and social goals are both logical and echo existing literature (see Erdley & Asher, 1999), lending criterion validity to the measure.

Secondly, victimization was not associated with feeling angry or sad in either scenario type. This is in contrast with previous research (e.g., Camodeca & Goossens, 2005). However, in the present study, we excluded the children who were identified as bully-victims—a subgroup considered to have the most difficulty in emotion regulation (Toblin, Schwartz, Gorman, & Abou-ezzeddine, 2005). It is likely that associations between victimization and affective response were not found here as a result, and demonstrates the importance of distinguishing this subgroup in any further research. Interestingly, relational victimization was actually negatively associated with ‘feeling angry’ in response to hostile provocation. This may be further indication that victims do not adapt their affective response across ambiguous and hostile provocation, or alternatively could imply that they show low levels of anger in response to provocation (perhaps because they have come to accept these occurrences). Either way, this would be an interesting avenue for further research.

Thirdly, victimization was not associated with interpreted hostile intent in either scenario type. Again, this may be in part because we excluded bully-victims from our correlational analysis, but it should also be considered that the associations between victimization and interpreted intent neared significance. Because the weight of the associations was very similar across scenario types, it could be considered further indication that victims do not distinguish ambiguous from hostile provocation in terms of their social cognitive processing.

Strengths and Limitations of the Current Study

This study has provided important evidence that social goals are likely to have a critical role to play in predicting bullying and victimization. To our knowledge, it is the first study to consider the social goals associated with bullying and victimization after accounting for affective response and interpreted hostile intent. It is also one of few studies in the field that has examined the social goals associated with victimization, and the only one that has attempted to identify and remove bully-victims from the sample so that a clearer picture of the social goals held by ‘pure’ bullies and victims can be painted. Additionally, by investigating situation-specific goals, the present study provides increasingly detailed information on the social motivation of bullies and victims, and adds a new perspective to associations reported in previous studies that have assessed children's trait-like social goals (e.g., Sijtsema et al., 2009). The next step is to determine the specific effects that these goals have on children's behaviour in response to provocation, potentially as mediators of the broader trait-like variations in social motivation that have been studied by Ojanen, Gronroos, and Salmivalli (2005) and Sijtsema et al. (2009). Specifically, do relationship damaging goals or goals for social dominance predict responses for confrontation, avoidance, or problem solving? And do submissive goals in victims have a direct influence on their response to provocation? Further research is needed that considers the social-cognitive factors investigated here, alongside the responses that children select in reaction to provocation.

There are, however, various limitations that need to be considered. Firstly, the association between bullying and the goal for assertiveness in response to hostile provocation has been held to indicate a general concern for maintaining an image of social dominance in bullies. It is not yet clear whether this inference is well founded as the same goal could equally demonstrate a short-term concern with protecting the self. Moreover, dominance is likely to be multifaceted, and there are several different reasons why an individual might desire it (e.g., individual self-enhancement vs. protecting one's social group). Further research that investigates the response strategies that are developed to fulfil goals for assertiveness or dominance in social situations may help us better understand the nature of their ambitions. Similarly, the goal for ‘staying out of trouble by not making a fuss’ was held to represent submissiveness, but could also indicate a self-protective concern for not getting into trouble with the teacher. Future research should take care to distinguish these response options.

Secondly, the present study assessed social goals in response to provocation only. Although provocation scenarios have been argued to provide a particularly strong situation from which to identify the social-cognitive biases that may be prevalent in certain groups of children (see Erdley & Asher, 1999, for a review), bullies are often the instigators rather than the receptors of provocation within their peer group (Coie, Dodge, Terry, & Wright, 1991). Thus, there is a clear need for future research to consider which social goals are associated with bullying and victimization in a wider range of social situations, including scenarios where the child casts him/herself as the aggressor or provocateur.

Thirdly, there was some indication that the social goals we assessed in the present study may not have been particularly salient to the children who participated in this study. Although it is possible to hold multiple social goals simultaneously, the correlations between goals for avoidance/submissiveness and goals for gaining revenge/being assertive across both scenario types do seem incongruent. Nevertheless, consultation with some similar-age children with hindsight of these associations suggested that they did understand the distinction, and that we had presented children with valid goals in the context of the scenarios. An important task for future research will, therefore, be to untangle the way in which multiple social information-processing factors (attributions of intent, emotional reactions, and various simultaneously held social goals) may interact with each other in influencing socio-behavioural outcomes.

Finally, it should be noted that the variance in bullying/victimization explained by the SIP predictors was low, especially for physical victimization. Consequently, there are likely to be other variables able to explain more variance in bullying and victimization scores. Indeed, an inspection of the existing literature on bullying and victimization suggests that there are significant associations with a range of behavioural, cognitive, emotional, and distal factors, with school ethos and climate also playing a significant role (Smith, 2004). Nevertheless, the present study contributes to this literature, demonstrating that the social goals children hold in response to provocation have an important role to play in understanding bullying and victimization.

Implications for Intervention Strategies

The present study offers various potential directions for intervention strategies. If bullies' motivation for social dominance is calculated rather than emotional, it may be worth demonstrating prosocial techniques to achieve positive social status, subsequently achieving popularity without relying on aggressive behaviours (which often results in peer rejection in the long term, see Rigby & Slee, 1993). In fact, various peer-mentoring intervention strategies have carefully selected individuals to serve as positive mentors and have met with success in reducing bullying (see Farrington & Ttofi, 2009).

With regard to victims, this study has found victims of bullying to entertain goals of submission. This is unsurprising given that they have most likely experienced harassment over a period of time, putting them at risk for emotional problems, such as depressive symptoms, lower self-worth, and social anxiety (Boivin, Dodge, & Coie, 1995; Craig, 1998; Graham, Bellmore, & Mize, 2006). However, effectively dealing with provocation is important for finding a route out of repeated provocation in the future. Programmes that develop the social skills of victims and empower them, not just to deal with provocation but also to engage with others positively in ambiguous situations, are therefore likely to meet with success (again, see Farrington & Ttofi, 2009, for a systematic review of social skills training schemes).

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Conclusion
  8. References

This study has provided important evidence that children's social goals are likely to be linked in distinctive ways to the behaviour of children who bully or become bullied, over and above the potentially influential emotional and social-cognitive contributors of affective response and attribution of intent. However, it is vital that further investigations proceed to assess the causal influence of these social goals on the emergence of bullying and other anti-social behaviour.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. Conclusion
  8. References