Thinking patterns, victimisation and bullying among adolescents in a South Australian metropolitan secondary school

Abstract

Responding to staff concerns about anti-social behaviour among students (n = 311, 50.5% boys, age range 13–16 years) at a low socio-economic Adelaide metropolitan school, we investigated victimisation and bullying and associated patterns of thinking. Two instruments were administered: the How I Think Questionnaire, which measures self-serving cognitive distortions; and the Bullying Experiences Questionnaire, which requires students to rate victimisation and bullying. The study revealed that: levels of distorted thinking were high; the most frequent forms of victimisation and bullying were verbal, indirect and physical; there were low levels of more extreme forms of victimisation and bullying; and there were higher levels of cognitive distortions among bullies and bully-victims. The research confirms the role of distorted thinking in the enactment of anti-social and bullying behaviours and provides a contemporary update of the types of victimisation/bullying in an Australian secondary school in 2011. Implications for interventions using social-cognitive approaches are addressed.

Keywords::

• adolescence
• bullying
• victimisation
• anti-social behaviour
• cognitive distortions
• social-cognitive interventions

Introduction

This study arose from a request by senior staff at an Adelaide metropolitan secondary school to researchers in the School of Education at Flinders University. The secondary school is in a low-income area with 70% of parents receiving government assistance for their children's education. There are high levels of unemployment and underemployment and pockets of high levels of crime within the community. Given this context, the school strives to provide alternative programmes and modified curricula to meet the students' social and educational needs. Despite these efforts by the school, the staff expressed their concerns about considerable disengagement by students and frequent outbreaks of vandalism, aggression, bullying and anti-social behaviour. Concern was also expressed that some girls were exhibiting atypical levels of aggressive and anti-social behaviours. The Principal and other school leaders approached us seeking help to better understand the students' behaviour and subsequently develop programmes to address this behaviour. Our view was that the student behaviour appeared to be associated with low levels of empathy and care, and lack of self-responsibility by individuals for their actions. This was consistent with a social-cognitive psychology approach that links behaviour to the way one thinks about situations. We also believed that behaviour needed to be explained with reference to students' social circumstances, meaning that our research approach ought to be informed by sociological theories. This paper reports work related to the cognitive psychology component while the sociological work is ongoing.

Research aims

Our purpose was to assist the school staff to understand the anti-social and aggressive behaviours of their students. Specifically for this paper we investigated the thinking patterns of the students and how these relate to their anti-social and bullying behaviours.

Literature review

Cognitive distortions describe the biased means used by some individuals to confer meaning to personal experiences, and as such they may contribute to responses that are emotionally and behaviourally problematic (Barriga & Morrison, 2001). If used repeatedly, these distortions can cause the development of psychopathological tendencies, involving both self-debasing, internalising behaviours, such as anxiety and depression, as well as self-serving, externalising behaviours, including aggression and delinquency (Barriga, Landau, Stinson, Liau, & Gibbs, 2000).

While self-serving cognitive distortions can be useful in that they protect the individual from developing a negative self-concept and blaming herself/himself, they also act to reduce inhibitions towards antisocial behaviour, such as aggression. Sykes and Matza (1957) described the role of self-serving distortions as neutralising guilt and obstructing the development of empathy, while Bandura (1991) suggested that they enable moral disengagement.

Research involving self-serving cognitive distortions has been pivotal in explaining anti-social behaviour (Barriga & Morrison, 2001) and several treatment programmes have been developed with the aim of reducing anti-social behaviours through the process of correcting distorted thinking (Gibbs, Potter, & Goldstein, 1995; Goldstein, 1999; Goldstein, Glick, & Gibbs, 1998; Goldstein & McGinnis, 1997).

Because we believed that the anti-social and aggressive behaviours at the study school may be related to the students' thinking, we chose a survey instrument (see Method below) that would reveal levels of cognitive distortions amongst the students. Although this instrument has no Australian norms, it does have norms established in the United States and we made comparisons between these and the scores obtained by students in our study. Because of specific concerns by senior staff from the school about girls' behaviours, we set out to investigate distorted thinking patterns and how they differed between males and females.

Bullying is a form of aggressive behaviour that is often repetitive and in which there is an imbalance of power between the perpetrator and the victim (Olweus, 1993; Rigby, 2007; Smith & Monks, 2008; Smith, Pepler, & Rigby, 2004). Since Olweus' (1978) pioneering work in Norway, bullying has been studied in many parts of the world (Jimerson, Swearer, & Espelage, 2009; Smith et al., 1999). Over time, new forms of bullying have emerged; for example, cyber-bullying (Dooley, Pyzalski, & Cross, 2009; Slonje & Smith, 2008; Smith et al., 2008; Spears, Owens, Slee, & Johnson, 2008; Spears, Slee, Owens, & Johnson, 2009), and questions have arisen about the amount of more extreme forms of bullying behaviour. In addition, research in the past has considered the extent to which boys and girls differ in their forms of bullying (Bjorkqvist, 1994; Crick & Bigbee, 1998; Galen & Underwood, 1997; James & Owens, 2005; Owens, Daly, & Slee, 2005; Owens, Shute, & Slee, 2000; Owens, Slee, & Shute, 2000). In order to get a contemporary update on these questions, in this study we chose a survey instrument (see Method below) that enabled us to investigate the amounts of a variety of forms of victimisation and bullying by boys and girls – from traditional physical and verbal through cyber to more extreme behaviours including extortion and use of weapons.

The major aim of this study was to investigate the relationship between self-serving cognitive distortions and bullying behaviour. It was anticipated that young people who bullied others would show elevated levels of cognitive distortions when compared with those who were not involved or who were targets of bullying. This is because bullying behaviour is a sub-set of violent or aggressive behaviour (Smith & Monks, 2008) and, as such, self-serving distortions would be an expected characteristic of bullies who engage in aggressive behaviour.

In addition to bullies and victims, Salmivalli, Lagerspetz, Bjorkqvist, Osterman, and Kaukiainen (1996) identified other groups involved in bullying. These are bystanders, who may encourage the bully or defend the victim, bully-victims, who bully others and are themselves also victimised, and a larger group of ‘not involved’ people who play no part in the bullying. Since cognitive distortions are associated with aggressive behaviour it was hypothesised that bullies and bully-victims would be more likely to exhibit cognitive distortions than any of these other groups.

Research questions

• What are the thinking patterns (cognitive distortions) exhibited by the students and how do these compare with the normative US sample?

• How do these thinking patterns differ by sex?

• What is the incidence of the different forms of bullying and victimisation reported by the students? How do these differ by sex?

• What are the relationships between cognitive distortions and bullying?

Method

Ethics approvals

Ethics approval was obtained from the university and the state education department. Parents and students gave informed consent and the usual guarantees of anonymity, confidentiality and the right to withdraw from participation without prejudice were provided.

Participants

The participants were 311 students, 13–16-years-old (mean 14.1 years, standard deviation = 0.82, males = 50.5%), from a school located in a low socio-economic area of metropolitan Adelaide, South Australia, who returned signed parental consent (67% response rate). Based on the Anomalous Responding Scale (see below) 18.9% of the questionnaires were removed before analyses were conducted. These unreliable questionnaires were randomly distributed across age and sex.

Measures

Measurement of Thinking Patterns

Participants completed the How I Think Questionnaire (HIT; Barriga, Gibbs, Potter, & Liau, 2001; Gibbs, Barriga, & Potter, 2001). This is a 54-item self-report instrument, which has a fourth-grade reading level and takes between 5 and 15°min to complete. It primarily measures four types of cognitive distortions (39 items): being self-centred (nine items; e.g. ‘Sometimes you have to get what you want’, ‘Getting what you want is the only important thing’); blaming others (10 items; e.g. ‘If I made a mistake, it's because I got mixed up with the wrong crowd’, ‘If someone leaves a car unlocked, they are asking to have it stolen’); minimising/mislabelling (nine items; e.g. ‘Everybody lies. It's no big deal’, ‘Everybody breaks the law. It's no big deal’); and assuming the worst (11 items; e.g. ‘I can't help losing my temper a lot’, ‘You should hurt people first before they hurt you’). These cognitive distortion items have content relating to four categories of anti-social behaviour derived from the Conduct Disorder and Oppositional Defiant Disorder syndromes listed in the fourth edition of the Diagnostic and Statistical Manual of Mental Health Disorders (American Psychiatric Association, 1994) – opposition/defiance; physical aggression; lying; and stealing. Each cognitive distortion type has at least two and no more than three items of each of the four categories of anti-social behaviour. The four cognitive distortion and four anti-social behaviour categories form eight sub-scales from which three summary scores are derived: the overall (HIT) score is made up of all eight sub-scales; the overt summary score is derived from the opposition/defiance and physical aggression sub-scales; and the covert summary score is derived from the lying and stealing sub-scales.

In addition to the cognitive distortion items, there are eight anomalous responding items (e.g. ‘Sometimes I get bored’, ‘Sometimes I gossip about other people’) and seven positive fillers (e.g. ‘People should try to work on their problems’, ‘Friends should be honest with each other’). The anomalous responding scale is designed to screen for such things as disingenuous responding or impression management. The positive filler items are not scored but are designed to counterbalance the negative content of the distortion items. Students were required to rate on six-point Likert scales (1 = ‘strongly disagree’ to 6 = ‘strongly agree’) the extent to which they agreed with each statement in the questionnaire.

Measurement of victimisation and bullying

Participants completed the Bullying Experiences Questionnaire (Owens & Slee, 2006), which measures bullying and victimisation with 10 global items as follows (note: the examples below show the wording if being victimised in these ways):

• Physical – for example, being punched, hit or kicked.

• Verbal – for example, called names, teased.

• Uninvited sexual behaviour – for example, receiving unwanted sexual comments, jokes, gestures or looks, or were touched, grabbed or pinched in an unwanted sexual way.

• Indirect – for example, left out or excluded, had bad or false things said behind your back (rumours).

• Cyber – for example, received nasty text or SMS messages or nasty emails, bullied through the Internet (e.g. Facebook).

• Destruction of property – for example, your belongings damaged or vandalised.

• Extortion – for example, money or belongings demanded.

• Stealing – for example, items stolen or taken from you.

• Threatening – for example, seriously threatened to be harmed.

• Extreme Violence – for example, had a weapon used against you.

Participants indicated the extent to which they had been victimised by peers and bullied peers on five-point rating scales from 0 = ‘never’ to 4 = ‘very often’. For bullying peers, the wording of the items reflected bullying rather than being victimised; for example, physical included ‘punching, hitting or kicking girls (or boys)’.

Grouping of students into bully groups

Students were classified into victim, bully, bully-victim and not involved groups using the extreme ‘cut-off’ of ‘quite often’ or more. Note that our measures did not enable us to form a separate bystander group.

Students who selected ‘quite often’ (once or twice per week) or ‘very often’ (almost every day) in any of the questions concerning being bullied by a girl/boy or bullying a girl/boy were given a score of one for that question. Scores were then summed so that each student was allocated a score for being bullied by a girl/boy and bullying a girl/boy. These ‘bullying others’ and ‘victimisation’ scores were then used to categorise students into groups:

• Not involved (n = 131, 52.4%) – students with a zero victimisation score and a zero bullying-others score.

• Bully (n = 8, 3.2%) – students with a victimisation score of zero and a bullying-others score of one or more.

• Victim (n = 54, 21.6%) – students with a bullying-others score of zero and a victimisation score of one or more.

• Bully-victim (n = 57, 22.8%) – students with a bullying-others score of one or more and a victimisation score of one or more.

Chi-square analysis was used to test differences in the HIT sub-scales between the bully groups, and Cramer's V was used to calculate the effect size. All analyses were carried out in SPSS 17.0. Cohen's (1992) effect size criteria were used to describe effect sizes as small ( < 0.3), medium ( < 0.5) or large (>0.8).

Results

HIT scores

The overall mean HIT score and the mean scores on all of the summary scores and sub-scales were above those of Barriga et al.'s (2001) North American normative sample (see Table 1).

Table 1 HIT questionnaire summary and scale scores for the study sample compared with Barriga et al.'s (2001) normative sample.

	Study sample		Barriga et al.'s normative sample
Scale	Mean	Standard deviation	Mean	Standard deviation
Overall HIT summary score (HIT)	2.85	0.80	2.39	0.69
Overt summary score (OV)	2.90	0.81	2.44	0.71
Covert summary score (COV)	2.83	0.89	2.34	0.74
Self-centred (SC)	2.72	0.90	2.42	0.74
Blaming others (BO)	2.92	0.87	2.42	0.79
Minimising/mislabelling (MM)	2.76	0.92	2.31	0.78
Assuming the worst (AW)	2.90	0.86	2.35	0.72
Opposition/defiance (OD)	3.00	0.80	2.55	0.72
Physical aggression (PA)	2.79	0.93	2.32	0.78
Lying (L)	3.26	0.97	2.69	0.83
Stealing (S)	2.40	0.95	2.02	0.75

Barriga et al. (2001) examined scores that differentiated referred and non-referred youth in their samples. These cut-off scores were determined to be at the 73rd and 83rd percentiles. Scores between these cut-off values were designated as borderline-clinical, and scores above the 83rd percentile as in the clinical range. These percentile cut-off values coincided well with prevalence rates of disruptive behaviour disorders in the fourth edition of Diagnostic and Statistical Manual of Mental Disorders and the cut-off values for the Achenbach Youth Self Report Externalizing Scale.

The 50th percentile scores of all of the summary and sub-scale scores for the study sample were above those of the normative sample and in Barriga et al.'s (2001) borderline-clinical range (see Table 2).

Table 2 Fiftieth percentile scores of the study sample compared with the normative sample.

	Summary scores			Cognitive distortions				Behavioural referents
	HIT	OV	COV	SC	BO	MM	AW	OD	PA	L	S
Study sample	2.79	2.90	2.77	2.67	2.90	2.67	2.82	2.95	2.80	3.25	2.33
Normative sample	2.36	2.39	2.31	2.33	2.30	2.22	2.27	2.47	2.15	2.59	1.91

Note: HIT, overall HIT summary score; OV, overt summary score; COV, covert summary score; SC, self-centred; BO, blaming others; MM, minimising/mislabelling; AW, assuming the worst; OD, opposition/defiance; PA, physical aggression; L, lying; S, stealing.

For the study sample, the borderline clinical cut-off equated to the 48th percentile, meaning that 52% of the students in the study sample could be described as being in the borderline-clinical range or above. This compares with 26% of the normative sample. The relevant figures for the clinical range are 40% for the study sample compared with 16% for the normative sample.

In relation to sex differences, see Table 3. Males had significantly higher scores than females for all of the summary and scale scores with the exception of self-centred, opposition/defiance and lying.

Table 3 Gender differences on HIT summary and scale scores.

	Gender	Mean	Standard deviation	Effect size (r)	p value
Overall HIT score	Male	2.97	0.73	0.18	< 0.01
	Female	2.68	0.81
Overt summary score	Male	3.02	0.73	0.16	< 0.01
	Female	2.73	0.82
Covert summary score	Male	2.95	0.85	0.18	< 0.01
	Female	2.67	0.89
Self-centred	Male	2.82	0.84	0.14	ns
	Female	2.58	0.92
Blaming others	Male	3.05	0.77	0.18	< 0.01
	Female	2.75	0.90
Minimising/mislabelling	Male	2.91	0.88	0.19	< 0.01
	Female	2.56	0.89
Assuming the worst	Male	3.01	0.81	0.17	< 0.01
	Female	2.74	0.82
Opposition/defiance	Male	3.08	0.76	0.13	ns
	Female	2.88	0.79
Physical aggression	Male	2.96	0.81	0.21	< 0.01
	Female	2.57	0.96
Lying	Male	3.35	0.89	0.11	ns
	Female	3.15	1.03
Stealing	Male	2.55	0.93	0.19	< 0.01
	Female	2.20	0.90

Note: ns, not significant.

Victimisation and bullying

Victimisation

In relation to gender differences in victimisation, see Figure 1. The most common form of victimisation reported to be experienced sometimes or more often by both males and females was verbal and the least experienced were extortion and extreme violence. Males were more likely than females to report experiencing physical victimisation (Mann–Whitney U test, Z = –4.0, p < 0.0001, effect size = 0.26) and this was the second most common type of victimisation experienced by males, but was much less common amongst females (ranked fifth). Destruction of property (Mann–Whitney U test, Z = –2.58, p < 0.01, effect size = 0.17), stealing (Mann–Whitney U test, Z = –3.24, p < 0.001, effect size = 0.21) and extreme violence (Mann–Whitney U test, Z = –3.11, p < 0.002, effect size = 0.21) were much more commonly reported by males than females. There were no significant differences in males' and females' reports of verbal, indirect, threatening, uninvited sexual behaviour, cyber and extortion victimisation.

Figure 1 Types of victimisation experienced ‘sometimes’ or more often by males and females. Note: **p < 0.01, ***p < 0.001, ****p < 0.0001.

Bullying

The most common type of bullying admitted by students to bully peers ‘sometimes’ or ‘more often’ was verbal, followed by physical and then indirect (see Figure 2).

Figure 2 Types of bullying used by males and females ‘sometimes’ or more often to bully other students.Note: *p < 0.05.

The most common form of bullying admitted to by both males and females was verbal. The least reported was extreme violence. Males were more likely than females to report using: threatening behaviour (Mann–Whitney U test, Z = –2.25, p < 0.025, effect size = 0.15); destruction of property (Mann–Whitney U test, Z = –2.56, p < 0.011, effect size = 0.17); stealing (Mann–Whitney U test, Z = –2.38, p < 0.011, effect size = 0.16); extortion (Mann–Whitney U test, Z = –2.17, p < 0.03, effect size = 0.14); and extreme violence (Mann–Whitney U test, Z = –2.5, p < 0.013, effect size = 0.17). Males and females were equally likely to report that they used verbal, physical, indirect, cyber and uninvited sexual behaviour to bully other students.

Relationships between cognitive distortions and bullying status

Percentages of participants who fell within the clinical, borderline-clinical and non-clinical ranges for each of the HIT scale scores were computed for each of the bully status groups (bullies, victims, bully-victims and not involved). The relevant percentages falling within the clinical range are displayed in Table 4.

Table 4 Bully status and percentages in clinical range on HIT questionnaire summary scores and scale scores.

	Bullies (%)	Victims (%)	Bully-victims (%)	Not involved (%)	Significance level	Effect size (Cramer's V)
Overall HIT summary score	62.5	33.3	59.6	27.5	< 0.0001	0.24
Overt summary score	75.0	40.7	63.2	23.7	< 0.0001	0.26
Covert summary score	50.0	25.9	57.9	26.7	< 0.001	0.22
Self-centred	50.0	24.1	49.1	16.8	< 0.0001	0.25
Blaming others	37.5	33.3	56.1	26.0	< 0.01	0.20
Minimising/mislabelling	75.0	31.5	57.9	29.8	< 0.01	0.21
Assuming the worst	75.0	53.7	70.2	29.8	< 0.0001	0.26
Opposition/defiance	62.5	33.3	63.2	21.4	< 0.0001	0.26
Physical aggression	62.5	33.3	59.6	27.5	< 0.001	0.22
Lying	37.5	33.3	54.4	35.1		0.14
Stealing	75	24.1	59.6	31.3	< 0.0001	0.22

Bullies and bully-victims had the highest likelihood of scoring in the clinical range on the overall HIT score, as well as in the clinical range of overt and covert scores. These two groups were also more likely than the victim and not involved groups to score in the clinical range of being self-centred, assuming the worst, opposition/defiance and physical aggression. Bully-victims had the highest likelihood of scoring in the clinical range of blaming others, while bullies had the highest likelihood of scoring in the clinical range of minimising/mislabelling and stealing.

Discussion

This study arose from a request by senior staff of a secondary school who were concerned about levels of anti-social, aggressive and bullying behaviours amongst the students. In response we decided upon a social-cognitive theoretical approach that links behaviour to the ways people think. This approach appears to have been fruitful.

In relation to overall levels of cognitive distortions, it is clear that the participants in our study sample scored consistently much higher than those in the US normative sample. Even though the norms are not Australian, they are from another advanced western country and we would not expect a very large difference in norms between the United States and Australia. The fact that 52% of students in the study sample scored in the borderline range or above and that 40% scored in the clinical range is concerning. Given the relationships previously shown between thinking distortions and aggressive and anti-social behaviours, there is good reason to believe that the behaviours about which the staff members in the study school were concerned are linked to these thinking patterns in their students. In relation to sex differences in the HIT scales, males generally scored significantly higher than did females. If we accept the links between distorted thinking and anti-social and aggressive behaviours, and we know that males have higher levels of these problematic behaviours, then the sex differences in the HIT scales are not surprising.

In relation to overall levels of bullying and victimisation, our results reveal a traditional pattern in which the most frequent forms of bullying and victimisation reported are verbal followed by indirect and physical. These results are consistent with a recent study conducted by one of us in the same city (Skrzypiec, Slee, Murray-Harvey, & Pereira, 2011). We found very low incidences of the more extreme behaviours including use of weapons and extortion. It is important to see such a pattern of results because it tends to dispel the anecdotal (often in the media) reports that children and teenagers are becoming more extreme in their violence. Examination of sex differences reveals a mixed pattern of results depending on whether reports relate to being victimised or to admitting to bullying. Reports of being victimised and admitting to bullying reveal boys exceeding girls in the more extreme behaviours, including destruction of property, stealing, and extreme violence such as use of weapons. Bullying through threatening and extortion were also reported to be exhibited more by boys than girls while physical victimisation was reported to occur more to boys than girls. Note that we found no significant differences between boys and girls for verbal, indirect, sexual and cyber victimisation and bullying. In general, these results fit a pattern in which boys are seen as more directly overt and extreme in their behaviours. This is consistent with recent research (Andreou & Bonoti, 2010; Kim, Kamphaus, Orpinas, & Kelder, 2010; Von Marees & Petermann, 2010). Again this is an important result because it serves to dispel the anecdotal media reports that girls are becoming increasingly extreme in their violent behaviours (e.g., the lads and ladettes television phenomenon). This result also tells the school that, overall, boys report more victimisation and admit to more bullying but there is a range of forms of bullying and victimisation in which there are no significant differences between boys and girls.

The results generally support the hypotheses that bullies and bully-victims would be more likely than those not involved in bullying to exhibit cognitive distortions. The large proportion of bullies and bully-victims who fell into the clinical range was not a surprise given research that has shown the abnormal psycho-social tendencies of bullies, and in particular bully-victims (Ireland & Power, 2004; Nansel et al., 2004; Unnever, 2005).

In summary the results reveal that: the students at the school exhibit worryingly high levels of cognitive distortions; the sorts of victimisation and bullying follow a traditional pattern where verbal, indirect and physical behaviours predominate and there are low levels of the more extreme behaviours; boys exhibit higher levels of the more extreme forms of victimisation and bullying; there are various forms of victimisation and bullying in which there are no significant differences between boys and girls; and there is a direct relationship between distorted thinking and whether or not you are a bully or bully-victim.

These results suggest that interventions at the school which address thinking patterns are needed to try to reduce the troublesome high prevalence of students in the borderline and clinical ranges on the HIT questionnaire. One of the better known social-cognitive programmes for working with troubled youth is EQUIP (Gibbs et al., 1995). This programme uses peer-helping methods and social skills training to help at-risk youth reduce self-serving cognitive distortions in order to assist them to act responsibly. Several studies have shown the effectiveness of the EQUIP programme with juvenile delinquents in reducing self-serving cognitive distortions and promoting a less positive attitude toward anti-social behaviour (Nas, Brugman, & Koops, 2008.); and reducing recidivism and improving conduct (Leeman, Gibbs, & Fuller, 1993; Liau et al., 2004). While the EQUIP programme was developed for older youth and adult offenders, a prevention version of the programme has been developed for use in secondary schools (DiBiase, 2010; DiBiase, Gibbs, & Potter, 2005). The EQUIP for Educators programme (DiBiase et al., 2005) aims to reduce anti-social behaviour by decreasing cognitive distortions (particularly relating to anger management), developing social skills, particularly perspective-taking, and stimulating moral development. The research into the effectiveness of the programme, however, has produced mixed results. While DiBiase (2010) found the programme to be effective in achieving its aims, van der Velden, Brugman, Boom, and Koops (2010) found the programme only partially successful – there was a reduction in attitude toward antisocial behaviour and self-serving cognitive distortions but not of the prevalence of antisocial behaviour, and there was no improvement in moral judgement. In a quasi-experimental study of peer victimisation in secondary schools (van der Meulen, Granizo, & del Barrio, 2010), EQUIP for Educators was found to be partially successful in reducing cognitive distortions and in reducing some types of bullying and social exclusion. Although EQUIP for Educators has to date produced varied results, because of its comprehensiveness and the success of the parent EQUIP programme, we believe that with further modification it could form the basis of a very successful programme for use in secondary schools such as the one we surveyed. In a separate paper, Owens, Skrzypiec, and Wadham (2011) examined literature that reported successful psycho-social interventions with adolescents. For instance, in a comprehensive review Yeager and Walton (2011) reported that successful interventions with youth were those that did not stigmatise participants, minimised resistance and actively engaged students, while Fishbein (2008) emphasised the crucial importance of clearly communicating the messages necessary for behavioural change to adolescents. Jemmott and Jemmott (see Jemmott, Jemmott, Hines, & Fong, 2001; Villarruel, Jemmott, & Jemmott, 2006; Jemmott & Jemmott, 2007) reported that their successful interventions in AIDS education were based around developing agency and self-efficacy in adolescents. Further, Jemmott and Jemmott (2007, p. 248) argued that secondary school students would benefit most from short activities involving active participation, concrete concepts, variation, and repetition. Utilisation of the ideas mentioned by these researchers has the potential to strengthen the EQUIP for Educators programme so that it could be used effectively to reduce cognitive distortions and associated anti-social and aggressive behaviours and specifically bullying.

A limitation of our study is the very low number of bullies in our sample given that we adopted a rigorous criterion cut-off for our bully status groups. We are, however, conducting individual follow-up interviews with students as part of the sociological component of the study. These interviews will enrich our dataset, enabling us to be more confident about our findings despite our small sample size. The interviews will also enable us to determine how social context factors contribute to the patterns of thinking and the behaviours of the students.

Additional information

Notes on contributors

Larry Owens

Larry Owens is a professor in the School of Education at Flinders University. He is the former director of the Flinders Educational Futures Research Institute and current associate head (teaching and learning) of the Faculty of Education, Humanities and Law. His research is concerned with: peer relationships among school students, including gender and age differences in aggression (and particularly indirect aggression among teenage girls); bullying behaviours; and conceptions of popularity and social power among adolescents. He is currently involved in an Australian Research Council linkage grant investigating student behaviour management issues and policies in South Australian schools.

Grace Skrzypiec

Grace Skrzypiec's main research interests centre on adolescent behaviour and well-being. She worked on data collection and analysis of the Australian national KidsMatter and KidsMatter Early Childhood evaluations, and on the data collection and analysis of a bullying intervention in an Australian Research Council-funded study to investigate well-being among school communities in the southern suburbs of Adelaide. Her background includes research on adolescent health with the Commonwealth Scientific and Industrial Research Organisation (CSIRO) and with adolescent offenders at the South Australian Office of Crime Statistics and Research. She is a member of the Student Wellbeing & Prevention of Violence Research Centre at Flinders University.

Ben Wadham

Ben Wadham is a sociologist of education. He co-authored Culture and Education (Pearson Education, 2007) and produced the Young Men Stopping Violence video Fathers and Sons Talk about Violence. Ben researches boys, masculinity and violence in schools, and militarism in Australia. Ben works in the School of Education at Flinders University and has published on issues such as men's health and violence, white masculinities and Aboriginal reconciliation, military culture and rural sustainability. He is a reviewer for the Journal of Men and Masculinities and the Journal of Sociology.