Siân Jones, Claire Fox, Simon Hunter, & Jon Kennedy [email protected] @ SianOxBrookes

BULLYING AND BELONGING: A LONGITUDINAL STUDY OF THE PROTECTIVE ROLE OF DEFENDERS IN FRIENDSHIP GROUPS

Siân Jones, Claire Fox, Simon Hunter, & Jon [email protected]

@SianOxBrookeshttp://throughtheacademiclookingglass.wordpress.com

GROUPS AND BULLYING

Within 51 cases reported by teachers, there were 23 cases where a single target was bullied by a group of children, but the target was then supported by other children.

Children (friends of the bullied) approached me and told me about what had happened, giving me names of the bullies, also of other children who could corroborate their story.[ ....] they had not approached any other teachers or informed their parents (P 19, 11-13 years)

I discovered that a group of girls in my class were bullying one particular child ... there were about 7 or 8 involved altogether (P30, 10-11 years).

GROUPS AND BULLYING

Jones et al., in prep.

DEFENDERS

Defenders are those children who “take sides with the victims, comforting and supporting them” (Salmivalli, 2010, p. 114).

Defenders tend to:

be emotionally stable (Tani, Greenman, Schneider, & Fregoso, 2003)

be cognitively skilled (Caravita, DiBlasio, & Salmivalli, 2009).

be empathic (e.g., Caravita et al., 2009) have high self-efficacy in their defending ability, (Thornberg & Jungert, 2013).

Pozzoli and Gini (2010) found high levels of defending behaviour under conditions of high perceived peer pressure, even when personal responsibility for intervening was low.

EFFECTS OF DEFENDERS

Salmivalli, Voeten, and Poskiparta (2011) showed that defending the victim was negatively associated with the frequency of bullying in a classroom.

Sainio et al. (2011) found that being defended was positively related to victims’ adjustment and social status.

What about the effect of defenders over time on bullying at the peer (friendship) group level?

CHILDREN’S FRIENDSHIPS Friendships are generally considered to be protective against peer victimization.

However, the sheer number of friends, and simply having a very best-friend may not be sufficient in protecting children (Hodges & Perry, 1999).

Fox and Boulton (2006) found that the number of friends, and the peer acceptance of a very best-friend attenuate experiences of victimization over time.

The identity of children’s friends (i.e. those who are not weak or victimised themselves) and the quality of the friendship (i.e. how much a child ‘sticks up for their friend’) are more important (Hodges et al., 1999).

INGROUP IDENTIFICATIONIngroup identification has a moderating influence on children’s reactions to intergroup events.

Nesdale, Durkin, Maass, and Griffiths (2005) found that children’s ethnic prejudice was positively related to strength of identification with their ethnic ingroup.

Jones, Manstead, and Livingstone (2009, 2011) showed that group-based reactions to bullying intensified as a function of group membership and in-group identification.

Levels of identification with the friendship group therefore influence group members’ willingness to stick with and support the group.

BULLYING AND SOCIAL IDENTITY

Social network research on bullying has shown that defenders are well-liked (Salmivalli et al., 1996) and popular among their peers (Caravita et al., 2009).

Sample of 1 234 UK children, aged 11-13 years (M = 11.68 years, SD = 0.64 years, 612 male, 93% white).

Data were collected at two time points – Autumn and Summer Terms.

Children completed peer nominations of peer victimization (e.g., Björkqvist, Lagerspetz, & Kaukiainen, 1992) .

ESRC SCHOOL BULLYING PROJECT

Children were asked to nominate a best friend, and their friends, in the class.

They were asked to give each classmate a rating from 1 “dislike very much”, to 5 “like very much”.

ESRC SCHOOL BULLYING PROJECT

IDENTIFYING FRIENDSHIP GROUPS

Based on Baines & Blatchford (2009).

A group is defined as a set of children (N ≥2) each of whom reciprocally nominates at least two others (or one other, where N=2) in the group as a friend or best friend, and reciprocally gives at least one of those a friendship rating of 5 (“like very much”).

11

HYPOTHESES

We hypothesized that having defenders in the friendship group, and having children with multiple friendship group associations in the friendship group, would be negatively related over time to the levels of peer victimization.

We controlled for the effects of gender, peer acceptance, number of membership groups, and class size.

Peer Victimization at Time 1

Peer Victimization at Time 2

Defenders in Friendship Groups

Multiple Group Associations of

Friends

LONGITUDINAL DATA ANALYSES

Adjusted R2 = .33, ß=-.155, t =-3.86, p<.001.

Hierarchical Linear RegressionEffects of defenders and networked-ness on later victimization

IMPLICATIONS

• The reduction in peer victimization as a function of the number of defenders and ‘networked’ children in a child’s friendship group emphasizes the need to investigate bullying as a group phenomenon at the level of the

friendship group in real friendship groups.

• How the effect of group-identification might vary depending on group norms could also be examined, given Polozzi and Gini’s (2010) finding that perceived peer group pressure predicted defending behaviour even when personal responsibility for intervening was low.  

PRACTICAL IMPLICATIONS

• An explicit focus on friendship groups would be a valuable addition to existing anti-bullying strategies.

• Encouraging and enabling children to defend and support victimized peers can help reduce levels of victimization.

• Our results support the importance of friendship groups within schools and emphasise the need to network individuals across those friendship groups to ensure they can have maximum impact.  

SUMMARY

We aimed to explore the effect of having defenders in one’s friendship group, and having networked friends in one’s friendship group, on peer victimization.

The longitudinal design allowed us to look at the cause-effect relationship between defending and later levels of victimization.

Previous research had not looked at the effect of defenders in friendship groups.

SUMMARY

Having more defenders in one’s friendship group, and having children in one’s friendship group who are networked across multiple friendship groups, reduces peer victimization across a nine month period.

We now need to better understand the social identity concerns  of (a) defenders, and (b) those who belong to multiple friendship groups, as a basis for developing anti-bullying interventions encouraging intragroup defending of victims.

Siân [email protected]

@SianOxBrookeshttp://throughtheacademiclookingglass.wordpress.com

ACKNOWLEDGEMENTS

Claire FoxSimon Hunter

Keele Research TeamJon Kennedy

Susan Pratley

All the children who took part, and the schools and parents who allowed them to do so

A CLASS NETWORK

N = 17

Group 1 Group 2

Group 3

768 770

765

756

759

764

777

775

771

766

772

767

769

774

778776

760

Group 4

19

CORRELATIONS  1 2 3 4 5 6 7 8 9 10 11

Mean 3.76 4.37 7.41 7.96 0.79 0.80 3.27 3.30 -- 0.29 0.30

SD 7.27 8.02 5.84 6.42 0.60 0.63 0.56 1.98 -- 0.32 0.30

1. Time 1 Peer Victimization                      

2. Time 2 Peer Victimization .748***                    

3. Time 1 % Defenders in Group .115** .038                  

4. Time 2 % Defenders in Group .138** .150** .248***                

5. Time 1 N of Membership Groups -.212*** -.212*** .055 .030              

6. Time 2 N of Membership Groups -.161*** -.216*** -.042 .027 .286***            

7. Time 1 Peer Acceptance -.385*** -.369*** .030 -.033 .365*** .259***          

8. Time 2 Peer Acceptance -.096* -.123** .043 .014 -.009 .143*** .122**        

9. Gender .133*** .112** -.045 -.059 -.115** -.065 -.103**        

10. Time 1 % of Networked Group Members in Membership Groups

.059 .014 .077 .085 .103* .045 -.078 -.013 -.041    

11. Time 2 % of Networked Group Members in Membership Groups

.146*** .102** .051 .122** -.028 .032 -.014 .068 .162*** .097  

*p < .05; **p < .01, ***p < .001.

GROUP STATS

  Time 1 Time 2

N of Groups Male 73 95

  Female 118 100

  Mixed 40 39

Total 231 234

Mean (SD) Group Size  

Males 2.67 (1.42) 2.67 (1.27)

Females 2.95 (1.60) 2.86 (1.76)

Mixed 5.35 (4.63) 4.56 (5.14)

Total 3.27 (2.55) 3.07 (2.59)

Size and Number of Friendship Groups Across Time Points