Editorial Credits/ Crédits éditorials

Subscription Rates /Prix d'abonnements

Advertising Rates / Tarifs publicitaires (PDF)

Guest Editorial
Don’t Let Anyone Bully You Into Thinking Bullying Is Not Important!
Ellen L Lipman
(PDF)

In Review
Identifying and Targeting Risk for Involvement in Bullying and Victimization
Wendy M Craig, Debra J Pepler
(PDF)

Consequences of Bullying in Schools
Ken Rigby
(PDF)

Interventions to Reduce School Bullying
Peter K Smith, Katerina Ananiadou, Helen Cowie
(PDF)

Review Paper
Major Depressive Disorder in Adolescence: A Brief Review of the Recent Treatment Literature
Robert Milin, Selena Walker, Joy Chow
(PDF)

Original Research
Behaviour in Children With Language Development Disorders
Ulrike Willinger, Esther Brunner, Gabriele Diendorfer-Radner, Judith Sams, Ulrike Sirsch, Brigitte Eisenwort
(PDF)

An Electrophysiologic Study: Can Semantic Context Processes Be Mobilized in Patients With Thought-Disordered Schizophrenia?
Milena Kostova, Christine Passerieux, Jean-Paul Laurent, Marie-Christine Hardy-Baylé
(PDF)

Brief Communication
Barometric Pressure, Emergency Psychiatric Visits, and Violent Acts
Thomas J Schory, Natasha Piecznski, Sunil Nair, Rif S El-Mallakh
(PDF)

Pilot Study: Access to Fitness Facility and Exercise Levels in Olanzapine-Treated Patients
Suzanne Archie, Jane Hamilton Wilson, Shelley Osborne, Heather Hobbs, Jean McNiven
(PDF)

Behavioural and Substance Use Problems in Rural and Urban Delinquent Youths
Frank J Elgar, John Knight, Graham J Worrall, Gregory Sherman
(PDF)

Book Reviews
(PDF)

A Guide to Asperger Syndrome.
Reviewed by
Jovan Simeon, MD

Effective Leadership.
Reviewed by
Mary Pearson, MA

Handbook of Depression.
Reviewed by
Mostafa Showraki, MD, FRCPC

Letters to the Editor
(PDF)

Re: Evolutionary Perspectives on Schizophrenia

Re: 2002—Defining the 21st Century II

Re: Unfree Associations: Inside Psychoanalytic Institutes

Re: Strategies of Collaboration Between General Practitioners and Psychiatrists: a Survey of Practitioners’ Opinions and Characteristics

Reply to Dr Ungar

Fungal Dermatitis with Olanzapine in Schizophrenia

Re: Canadian Psychiatric Inpatient Religious Commitment

In Review

Interventions to Reduce School Bullying

Peter K Smith, BA, BSc, PhD, FBPS¹, Katerina Ananiadou, BA, MSc, PhD², Helen Cowie, MA, MSc, PhD, Dip Psychotherapy, PGCE, FBPS, Chartered Counselling Psychologist³

Bullying is a form of aggressive behaviour characterized by repeated acts against victims who cannot easily defend themselves (1,2). It can have severe negative consequences, especially for those victimized over a period of time. In the last 2 decades, research and action on school bullying have expanded worldwide (3); there has been widespread circulation of antibullying materials, and some countries legally require schools to have an antibullying policy (4). We review first a range of school-based intervention components to reduce school bullying. Second, we review the outcomes of large-scale intervention projects.

Whole-School Policy

A whole-school policy is a written document that sets out the school’s aims in relation to bullying behaviour, together with a set of strategies to be followed. The document is supported by systems and procedures within the school to ensure that its aims and strategies are effectively implemented, monitored, maintained, and reviewed. The whole-school approach usually emphasizes the democratic involvement of all school members in devising and maintaining the policy and has the advantage of potentially integrating numerous components of an antibullying strategy (5,6). It is usually a core component of the large-scale intervention programs reviewed later in this paper and is best evaluated in that context.

Classroom Climate

One area of debate is the extent to which school antibullying policies and efforts should target bullying directly or focus more generally on improving relationships within the school. Recent work in Norway and Spain has followed the latter approach. A study in Norway found that the quality of classroom management (that is, teacher–pupil relationships) and the social structure of the class (that is, pupil–pupil relationships) substantially predicted rates of reported victimization (7).

Peer-Support Systems

Ways of involving pupils themselves in tackling bullying generally aim to encourage action by the majority of pupils who do not like bullying. Peer-support methods include the following: cooperative group work in the classroom (8); circle time, in which the class addresses relationship issues, including anger, fighting, and bullying (9); circles of friends who provide a support team of peers to work with a vulnerable pupil (10); befriending, in which a pupil or pupils are assigned to “be with” or “befriend” a peer (11); Schoolwatch (12), a program in which pupils elect a committee of peers who propose and implement antibullying activities; conflict resolution or mediation, which is a structured process wherein a neutral third party assists voluntary participants to resolve their dispute; and active-listening or counselling-based approaches, which employ pupil helpers trained and supervised to use active-listening skills to support peers in distress.

These methods hold promise, and evaluations so far find that peer supporters themselves benefit and that the school climate improves generally; however, specific benefits for victims of bullying remain to be proven (13,14). Where peer helpers are not regularly supervised, where there are not enough peer supporters to tackle the problem, or where the problem is particularly severe, the system is less likely to be effective (15). There is a sex divide in recruited peer supporters, with some 80% being girls; the sex balance of teachers in charge of peer-support systems is similarly weighted.

School Tribunals

In a school tribunal or “bully court,” pupils are elected to hear evidence and decide on sanctions or punishments for those involved in bullying. Anecdotal reports indicate that a children’s court set up in a Canadian day school for maladjusted children reduced bullying and increased awareness of the issue (16). After the UK charity Kidscape issued advice and guidance on bully courts in 1990, it claimed that, in 8 schools, reports of being bullied dropped from 70% to 6% (17); however, it has been impossible to verify these claims (18). One case study of a UK secondary school in which a school tribunal was implemented together with a peer-mentoring support scheme did find some evidence for effectiveness (19).

Playground Improvement

A great deal of bullying occurs in playgrounds and school yards, especially in primary schools. Both improved supervision and improved playground design have been advocated. An evaluation of a peer-mediation scheme applied in 3 Canadian primary school playgrounds found that physically aggressive playground behaviour was reduced by 51% to 65% (20).

Curriculum Work

At the classroom level, many programs advocate class discussion to raise awareness, to stimulate ideas, and (or) to develop rules to manage bullying (18,21,22). Various aids are available for classroom teachers, including films and videos, role-play activities, literature, and quality circles (23). There is little evaluation of these activities outside the framework of the large-scale projects reviewed below.

Working With Specific Pupils

Assertiveness training has been recommended as a way to help victims or potential victims of bullying to cope in nonpassive, but also nonaggressive, ways. These techniques can be taught to pupils and appear to help some (18,24).

The Method of Shared Concern is a counselling-based approach (25) for situations wherein a group of pupils have been bullying. This approach focuses on the children doing the bullying as well as on those being bullied. It encourages the bullying children to acknowledge the suffering of the victim and take steps to help the situation. Although used widely in Sweden, little evaluation of the technique exists, and some do not recommend it (1).

A related approach is the No Blame, or Support Group, Approach developed in the UK (26). This problem-solving strategy creates a support group for the bullied pupil comprising both those directly involved in the bullying episode and bystanders. This support group is responsible for solving the problem and for reporting on progress. An evaluation spanning 2 years found immediate success in 80% of cases and delayed success in 14%; in only 6% of cases did the victim report continued bullying (27).

Community Conferencing as Transformative Justice

Community conferencing was developed by Transformative Justice Australia (TJA) and has been extended elsewhere (28). This strategy brings together a community of people in conflict, such as that occurring in school bullying (29). The perpetrator(s) meet the victim(s), each accompanied by friends, family and (or) other supporters. The perpetrator(s) make reparation, and the incident is regarded as finished. When evaluated in Australia, most schools in a pilot program reported high rates of participant satisfaction and high rates of compliance with decisions (30).

Larger-Scale Intervention Projects in Schools With Systematic Evaluation

We now focus on larger-scale projects carried out in more than 1 school and with pre- and posttest evaluation of the outcomes. Bullying rates are estimates of pupils bullying other pupils, and victimization rates are estimates of pupils being bullied by others. We report findings as percentage changes (which we calculated ourselves, where necessary).

Studies have mainly employed 2 types of design. An age-cohort design has time-lagged comparisons between age-equivalent groups (see the Bergen project, described below). This design avoids confound with age but still risks effects of historical period or cohort. Other studies have compared interventions with control groups of participants and schools; in them, reported positive results may be attributable either to decreased bullying in intervention schools or to increased bullying in control schools. This design has some control for historical or cohort effects; however, where all schools may be affected by antibullying initiatives and activity, there remains the question of what constitutes a control school.

Norway
Following the suicides of 3 boys, Norway’s Ministry of Education initiated the first large-scale antibullying campaign in fall 1983. The Ministry distributed a booklet for teachers and an information folder for parents to schools nationwide and also made available a video showing role plays of bullying situations. To establish baseline bullying rates before the start of the campaign, all primary and junior high schools were invited to give the Olweus Bully–Victim questionnaire to pupils. The effects of the program were systematically evaluated in 2 samples from schools in Bergen and Rogaland, respectively.

We describe first the Bergen sample. This comprised 2500 students from 42 primary and secondary schools, aged 11 to 14 years at the initial evaluation. In addition to the Olweus questionnaire, students completed questionnaires on anti- social behaviour and classroom climate (1,21,31). Because the antibullying campaign was nationwide, it was not possible to compare intervention schools with control schools. The study employed an age-cohort design involving 4 age (grade) cohorts tested at Time 1 (May 1983), Time 2 (May 1984), and Time 3 (May 1985). Over the course of the program, this approach allowed 5 time-lagged comparisons between age-equivalent groups (for example, baseline data from Grade 5 children at Time 1 were compared with baseline data of Grade 4 children at Time 2, at which time the latter children had reached Grade 5 age but had also experienced 8 months of intervention).

The 8-month evaluation assessed the effects of the nationwide circulation of material. Approximately 15 months after the start of the program, researchers met for 2 hours with Bergen school staff providing individual feedback information (31). It is not clear whether the program implemented in Bergen at the time amounted to the full Olweus program. The current Olweus antibullying program was developed “in connection with the nationwide campaign” (1, p 38). By introducing clear and firm rules and creating a warm and positive school environment, it aims to restructure the school setting to remove positive, and increase negative, consequences of bullying behaviour. In this program, antibullying measures work at school, classroom, and individual levels.

Using unspecified composite measures, the follow-up evaluation found reduced reports of direct bullying (that is, physical and verbal bullying) and of having been bullied. Effects were stronger after 20 months of intervention. For victimization, reductions after 8 months averaged 48% for boys and 58% for girls; after 20 months, they averaged 52% for boys and 62% for girls. For bullying others, reductions after 8 months averaged 16% for boys and 30% for girls; after 20 months, they averaged 35% for boys and 74% for girls. Antisocial behaviour also lessened, and school climate improved.

The second follow-up survey took place in October 1986, in Rogaland (32,33). This sample comprised 7000 students aged 8 to 16 years, in 37 primary and secondary schools. The fullest report of results describes direct victimization, social exclusion, and direct bullying in grades 4 to 9 only (ages 11 to 16 years; sample size not reported) (34).

In Rogaland, reported victimization increased by 44% for boys and decreased by 12.5% for girls. Social exclusion increased by 12.5% for boys and decreased by 8% for girls. Notably, however, bullying increased by 24% for boys and by 14% for girls. Summing over sex, outcomes are clearly negative on all measures.

Schools’ involvement was measured. For victimization, there was a linear relation between degree of involvement and positive changes, especially in primary schools (33). For bullying, there was a similar trend for primary schools only.

Although based on the same nationwide intervention campaign, the Bergen and Rogaland outcomes contrast strongly. As Roland emphasizes, the research team in Rogaland did not intervene in the schools between baseline and follow-up (in contrast to the Bergen study) (35). Olweus states that “the studies in Bergen and Rogaland were completely different in terms of planning, data quality, times of measurement, and contact with the schools, and accordingly, also in terms of expected results” (1, p 39).

The Bergen findings received more publicity in the 1990s, and a full version of the Olweus antibullying program became available. Since then, another large-scale intervention was implemented in Bergen, involving 14 intervention and 16 comparison schools (a total of 3200 students aged 11 to 15 years) over a period of 6 to 7 months (late 1997 to June 1998) (36). Preliminary findings suggest that being bullied and bullying others have been reduced by 25% to 30% in intervention schools.

Following the success of Bergen’s antibullying program, other countries developed and implemented several intervention programs with similar aims and methods.

England
From 1991 to 1993, the Sheffield Anti-Bullying project worked with 23 schools (16 primary and 7 secondary) and 6500 students aged 8 to 16 years (18). Although inspired by the Bergen project, this project was designed independently. Its core component was a whole-school policy against bullying. Schools could also choose to implement the following additional interventions: curriculum work; playground interventions; and work at the individual level with bullies, victims, or the peer group. The effort that schools put into these core and optional program components was monitored and varied considerably.

A preintervention survey was carried out at the end of 1990, followed by a second survey 18 months after the start of the program. According to the survey, victimization rates decreased by 14% in primary schools and by 7% in secondary schools. Bullying rates decreased by 12% in primary schools and by 12% in secondary schools. The proportion of pupils who told a teacher about bullying increased by 6% in primary schools and by 32% in secondary schools; the proportion of bullies reporting that a teacher had talked to them about it increased by 5% in primary schools and by 38% in secondary schools.

The project included 4 comparison schools (1 primary and 3 secondary), of which 2 did worse than most project schools, and 2 did not. All UK schools experienced some national circulation of antibullying materials during this period, and 1 of the comparison secondary schools did a lot of antibullying work. When data from all these schools were combined and action-outcome relations examined, positive correlations were found between measures of action taken and outcome measures such as changes in frequency of being bullied (in primary schools) or of telling a teacher (in secondary schools).

Another intervention program took place between 1991 and 1993 (37). Two schools in London and 2 schools in Liverpool took part (1 primary and 1 secondary school in each city; n = 1284 pupils). In all schools, a staff–student antibullying working party was set up to design and implement an antibullying policy. Video and peer-support programs were used in the primary schools, and assertiveness training and peer-mediation skills in the secondary schools. Decreased bullying levels were found in both primary schools (that is, the number of those never bullied increased on average by 10% after the first year and by 61% in the second year) but less consistently in both secondary schools (that is, although the number of those never bullied increased on average by 4%, the number decreased in the London secondary school, which was affected at the time by increased racial tension in the surrounding community).

Canada
An antibullying program in Toronto (38,39) worked with approximately 1000 children, aged 8 to 14 years, from 4 elementary schools. Modelled after the Bergen program, the intervention worked at school, classroom, and individual levels. Program components included developing a code of behaviour; increasing supervision of playgrounds and corridors; establishing class rules; and introducing various classroom activities, such as drama and language work with bullying themes. Three schools implemented a peer-level conflict-mediation program.

Preintervention assessment and postintervention assessment after 18 months showed a nonsignificant increase (20%) in children bullied more than once or twice a term; however, assessment also showed a significant decrease (18%) in children reporting having been bullied at least once in the last 5 days. There was a significant increase (22%) in bullying others more than once or twice weekly and an increase (24%) in bullying others at least once during the last 5 days.

Germany
From 1994 to 1996, an intervention took place in Schleswig-Holstein (40,41). In it, 37 primary and secondary schools (n = 10 600 students, aged 8 to 18 years) participated in the preintervention survey and a follow-up 1 or 2 years later. The program used an age-cohort design and followed the Bergen model, including core intervention measures at school, classroom, and individual levels.

Outcomes varied with age. Up to grade 10 (approximately age 16 years), having been directly bullied decreased on average by 17% across age groups; however, in grades 11 and 12 (ages 17 to 18 years), it increased by 44%. Social exclusion increased on average by 1.6% up to grade 10, and by 25% for grades 11 and 12. Bullying others decreased on average by 5% up to grade 10; however, in grades 11 and 12, it increased on average by 40.5%.

US
In South Carolina’s rural communities, 39 schools participated in an intervention from 1995 to 1997 (42). Here, 6250 students in grades 4 to 6 (ages 9 to 11 years) were followed over 2 years. During the first year, 11 schools received the intervention, and 28 schools acted as control schools. During the second year, 7 of the original 28 control schools also started receiving the intervention. The results refer to differences between the intervention and control schools after the first year and also to differences between year 1 and year 2 of the program. Data were collected once before and twice after the program began. The project was similar to the Olweus program, including core interventions at all 3 levels; however, it offered additional support materials for school staff and involved members of the local community.

Victimization rates did not change significantly for intervention schools in year 1 or year 2, compared with control schools. In intervention schools, bullying rates decreased significantly (by 20%) after 1 year; for control schools, they increased slightly (by 9%). After 2 years of intervention, no further significant differences were found between the 2 groups of schools, and both intervention and control groups showed slightly increased rates of victimization. With regard to general antisocial behaviour, neither the overall program effects nor the second-year program effects were significant: at best, the program showed trends in the intervention schools for slower rates of increase in certain antisocial behaviours, compared with control schools. No differences were found between intervention and control schools in attitudes toward bullying, either at year 1 or at year 2. No program effects were found on any measures for students in the 7 control schools that received the intervention in year 2.

Belgium (Flanders)
A project carried out from 1995 to 1997 within 18 primary and secondary schools in the Flemish part of Belgium (n = 1104 students, aged 10 to 16 years) used an experimental design (43). In this project, 3 primary and 3 secondary schools were randomly assigned to each of 3 conditions: treatment with support, treatment without support, and control. Schools in either of the 2 treatment conditions took part in the intervention program, which was developed along the principles of the Bergen and Sheffield programs. To test the importance of external support, school staff in the treatment with support condition also undertook 25 hours of training and feedback on implementing antibullying measures.

Assessments were carried out before the start of the intervention program and at posttests 8 and 20 months later. Results are reported in terms of 3 outcome scales: victim, bully, and positive interactions. For victimization, there was no significant condition × time effect; both treatment and control schools showed a small (1% to 3%) decrease in victimization scores, which was more consistent in primary than in secondary schools. For bullying others, there was a significant condition × time effect. After 20 months, primary schools receiving treatment with support showed no change, primary schools receiving treatment without support showed a decrease of 2%, and primary control schools showed an increase of 5%. Secondary schools receiving treatment with support showed, notably, a 1% increase, secondary schools receiving treatment without support showed a 1% decrease, and secondary control schools showed no change (that is, differences were not significant). For positive interactions, there was also little difference among conditions: all schools reported increases of around 3% to 9% over time; however, the treatment without support schools reported smaller increases than did the control schools (at least in primary schools).

One study finding was that the intervention had some differential impact on primary and secondary school students in regard to bullying behaviour. The study also found no significant differences between outcomes for treatment with support and treatment without support.

Spain
The Sevilla Anti-Violencia Escolar (SAVE) project took place in Seville from 1995 to 1996 and from 1999 to 2000 (22,44). The intervention program was inspired by the Sheffield and Bergen projects but developed autonomously. It had 4 parts: the democratic management of interpersonal relationships; cooperative group work and the curriculum; training in emotions, attitudes and values; and direct interventions for pupils at risk or involved in bullying. Ten schools participated in the intervention program, with pupils aged 8 to 18 years. Five of these schools (n = 910 students) took part in a postintervention survey 4 years after the initial survey. Three different schools (n = 751 students) participated in the posttest as control schools; however, these schools did not have pretest data, and this is therefore not a full experimental design.

In the intervention schools, the number of victimized pupils decreased by 57%, and the number of bullying pupils decreased by 16%. The proportion of bullies to victims (that is, those taking part in bullying and also being bullied) decreased by 57% and that of bystanders increased by 7%. The posttest results showed a significantly lower incidence of bullying problems at intervention schools, compared with control schools. The number of children reporting positive relationships with their peers increased by 16%, and long-term victims (those reporting being victimized for a year or more) decreased by 41%.

Switzerland
In Bern, a bullying prevention program was undertaken in kindergarten from January through May, 1998 (45). This program had 8 intervention schools (n = 152 children) and 8 control schools (n = 167 children). The children ranged in age from 4 to 7 years. The program focused on enhancing teachers’ capacity to handle bully–victim problems and included sensitization, observing behaviour, practising consistent behaviour and appropriate use of positive and negative sanctions, and examining the role and responsibility of noninvolved children. Data were collected from peer nominations of victims and bullies and from teacher ratings before, and 6 months after, the start of the program.

In the intervention group, peer nominations of children as either bullies or victims did not differ significantly between pre- and posttest assessments (specifically, victim nominations decreased by 4% and bully nominations by 12%). However, in the control group, victim nominations increased significantly, by 71%; bully nominations also increased significantly, by 22%. For reported bullying, teacher ratings showed no significant differences in overall outcomes. For reported victimization, teacher ratings showed some decline in physical and indirect victimization (that is, exclusion) in the intervention group, but increased verbal bullying. In the control group, teacher-rated victimization increased for property-related bullying and remained stable for other types of bullying.

Overview of Large-Scale Interventions

The large-scale antibullying programs have had mixed results. The first implementation, in Bergen, has been the most successful, with subsequent attempts yielding much more modest outcomes—and at times, even negatives ones. Only the Seville program has produced outcomes matching the Bergen study. Several factors may explain these variable outcomes.

Type of Intervention
Most programs have been multilevelled or multifaceted. There is no clear evidence as to which components have been most important. The Sheffield study attempted to assess program components independently but emerged with a general factor, “amount of total intervention” (18); a follow-up of primary schools 1 year later did suggest that keeping the whole-school policy active was important for continuing positive outcomes (46). The Seville project attempted some pupil assessment of program component effectiveness. Again, although some differences were found, all components registered positive assessments.

Some projects have added components to the Olweus program (for example, peer conflict mediation in Toronto and additional support materials and community involvement in South Carolina). These projects had less positive outcomes than the Bergen project, and there is therefore no evidence that the added components had positive effects.

Length of Program
Intervention length has varied, with most projects lasting 1 to 2 years. The Seville project had the longest intervention (4 years), and is one of the most successful, suggesting that program length is an important variable. The Bergen project found that outcomes for bullying were better after 20 months than after 8 months but that outcomes for victimization were only slightly better. The South Carolina project found no beneficial effects from a second year of intervention, nor was it clearly found to be beneficial in the Flanders project.

Support by Researchers
The extra support provided for Bergen schools offers an explanation for the different outcomes in Bergen and Rogaland: the research team’s continuing interest in school efforts, feedback on survey results, and further encouragement and support in antibullying work may lead to more effective work in the school. However, the Flanders project provided a direct experimental test of this hypothesis, and further support to schools from the researchers had no additional effect (43). Perhaps what is important is not external support but how much time and effort schools themselves decide to invest in the program.

Effort Invested by Schools
Various researchers have reported a positive relation between outcomes obtained and the effort put into the program by each school (18,21,40). In the Sheffield project, schools that put more effort into the program (for example, by implementing more optional components or requesting more support from the research team) obtained better outcomes (18). This positive dosage effect was also reported in the Rogaland study (33), though more clearly for primary than for secondary schools.

Age of Students
Some programs had a stronger positive effect on primary school pupils, compared with secondary school pupils. This was found in the Sheffield and Flanders studies, as well as in the London–Liverpool project. In the Schleswig-Holstein project, reported victimization (for direct bullying) decreased less for older children after the intervention and, in fact, increased for adolescents aged 17 to 18 years.

Developmental characteristics of older children and organizational features of secondary schools have been suggested as explanations (43). Younger children more willingly accept teacher authority and curriculum activities and school policies that reflect teacher influence. Older children—especially those involved in bullying and other antisocial activities—may explicitly reject teacher influence and values advocated by the school. The general peer climate and attitudes toward victims also become somewhat more negative in adolescence, particularly among boys (47). Further, secondary schools are larger and organized by year-group rather than by class, making whole-school processes more difficult to promote effectively (48).

Student Sex
Although many projects did not report sex differences, both Norwegian studies indicate that girls are more receptive to antibullying interventions. Girls generally have more favourable attitudes to victims, especially in adolescence (47,49). Girls are also significantly more willing to play an active part in challenging school bullying (for example, through participation in peer-support interventions) (13).

Program Comprehensiveness
The Bergen and Rogaland projects were part of a nationwide campaign, and the Schleswig-Holstein and South Carolina studies were part of large area campaigns. Conversely, the Toronto project and the London–Liverpool project focused on a few schools in a local area. Large-scale comprehensive programs can benefit from more support and resources and from associated media publicity. Nevertheless, while the success of the Bergen project may have been influenced by participation in a nationwide campaign, the Rogaland study, which had overall negative outcomes, was part of the same campaign. Similarly, although the Schleswig-Holstein and South Carolina studies were large-scale, they had less successful outcomes.

Suggestions for Future Research

Measurement Issues
One difficulty in comparing outcomes of different projects is that, often, different measures are used. However, the presence of various outcome measures does have benefits. Informed comparisons are helped by the use of various standard outcome measures (for example, victimization, bullying, attitudes, coping strategies, and school climate) to ascertain which are affected by an intervention, and in what ways. It would be preferable to supplement self-report data with peer-nomination and teacher data, given limited concordance between these different kinds of assessment (50,51).

Some researchers attribute the rather small postintervention changes in victimization levels to sensitization effects (18,43). Intervention programs raise pupil awareness of bullying; this may indeed be one of the program aims—or even a general prerequisite, as with the Olweus program. Making children more aware of bullying problems increases the likelihood that they will identify and report bullying. For example, they may realize that rumour spreading or social exclusion are forms of bullying, which they had not previously understood. This could increase victimization responses at posttest, quite independently of any actual changes in behaviours. Raised awareness may affect not only self-report data but also peer nomination data and teacher reports. Only observational methods appear potentially able to exclude the sensitization issue. Observational methods are costly and time-consuming, but they are possible to use (52) and should be considered as a supplementary measure in future projects, perhaps on some subsample of schools or classes.

Developmental Issues
The pattern of lesser program effects with older students indicates that intervention programs need to start earlier (perhaps at preschool age) or be tailored more to particular age groups. Research on the stability of victim and bully status suggests that few pupils enter into stable victim roles before ages 8 to 9 years (53,54). Future intervention programs might focus on children younger than age 8 to 9 years to prevent vulnerable children from being systematically targeted and stereotyped into a stable victim role from which it may be difficult to escape. The Bern antibullying program in kindergarten schools found some evidence for program effects in reducing victim nominations (45).

Sex and Types of Bullying
Do we tackle indirect bullying and girls’ bullying effectively? Boys and girls use and experience different types of bullying: boys’ bullying is more physical; girls’ bullying is more indirect or relational and includes social exclusion. The Sheffield project follow-up concluded that girls’ bullying, while less frequent than boys’, may be more difficult to tackle: boys’ bullying continued to fall in all 4 schools surveyed, but girls’ bullying only in 1 (46). There are difficulties in tackling relational bullying among adolescent girls (55). Policies and curriculum interventions need to target indirect and relational bullying as well as direct bullying.

Incentive Issues
Do any effects last after the intervention stops? Legal requirements plus periodic inspections may help this process. Other incentives to encourage schools to maintain antibullying work are the efforts of parents, school councils, or governing bodies; the possibility of adverse media publicity; and the possibility of legal action by former pupils or their parents.

School and Community Issues
The projects discussed have essentially focused on school-based interventions. It has been argued that models encapsulating the issue as solely a school-based problem are doomed to partial success, at best (56). Neglect of wider issues—deprivation in sections of society; positive presentation of violence in the mass media; and opportunities and reinforcement for violence, often in venues outside school— means that we cannot expect substantial positive outcomes to school-based initiatives, especially in the inner-city areas where social deprivation is most obvious and deeply embedded.

It could be argued that this analysis is less applicable to the European social democracies and Canada, which have less obvious social inequality than is seen in inner-city regions of the US. Nevertheless, the variable success of school-based antibullying interventions may suggest that, while worthwhile, they will have limited impact while we neglect such factors as the quality of parenting, parent training, and support for parents; the socioeconomic conditions of families and the opportunity benefits for crime and antisocial behaviour; the ways in which abuses of power are presented, and at times shown positively, in the mass media; and the ways in which abuses of behaviour are tolerated and rewarded in the wider society.

Summary

In our view, schools should continue with antibullying interventions. The cost of continued victimization is known to be considerable (57), and it is a moral imperative for schools to do what they can, within reason, to prevent this. Some form of school policy or class rules, plus curriculum work on awareness and other focused interventions, need not require a lot of time and effort and can have positive effects, especially in primary schools and especially when schools are fully committed to their antibullying work. However, more research and evaluation is needed to maximize the effectiveness of such endeavours.

Funding and Support

We acknowledge the financial support of the TMR project Causes and Nature of Bullying (contract ERB-FMRX-CT97-0139 from the European Commission).

References

1. Olweus D. Sweden. In: Smith PK, Morita Y, Junger-Tas J, Olweus D, Catalano R, Slee P, editors. The nature of school bullying: a cross-national perspective. London, New York: Routledge; 1999. p 7–27.

2. Smith PK, Brain P. Bullying in schools: Lessons from two decades of research. Aggressive Behavior 2000;26:1–9.

3. Smith PK, Morita Y, Junger-Tas J, Olweus D, Catalano R, Slee P, editors. The nature of school bullying: a cross-national perspective. London, New York: Routledge; 1999.

4. Ananiadou K, Smith PK. Legal requirements and nationally circulated materials against school bullying in European countries. Criminal Justice 2002;2:471–91.

5. Thompson D, Sharp S. Improving schools: establishing and integrating whole school behaviour policies. London: David Fulton; 1999.

6. Suckling A, Temple C. Bullying: a whole-school approach. London, Philadelphia: Jessica Kingsley; 2002.

7. Roland E, Galloway D. Classroom influences on bullying. Educational Research 2002;44:299–312.

8. Cowie H, Smith PK, Boulton M, Laver R. Co-operation in the multi-ethnic classroom. London: David Fulton; 1994.

9. Mosley J. Quality circle time in the primary classroom. Trowbridge (UK): Learning Development Aids; 1996.

10. Newton C, Wilson D. Circles of friends. Dunstable: Folens; 1999.

11. Cowie H, Sharp S. Peer counselling in schools: a time to listen. London: David Fulton; 1996.

12. Warton K, Barry S. Schoolwatch: an evaluation. London: Home Office Research, Development and Statistics Directorate; 1999.

13. Cowie H. Bystanding or standing by: gender issues in coping with bullying in English schools. Aggressive Behavior 2000;26:85–97.

14. Cowie H, Wallace P. Peer support into action. London: Sage; 2000.

15. Cowie H, Olafsson R. The role of peer support in helping the victims of bullying in a school with high levels of aggression. School Psychology International 2000;21:79–95.

16. Laslett R. Bullies: a children’s court in a day school for maladjusted children. J Special Education 1980:4:391–7.

17. Elliott M, editor. Bullying—a practical guide to coping for schools. Harlow: Longman; 1991.

18. Smith PK, Sharp S, editors. School bullying: insights and perspectives. London: Routledge; 1994.

19. Mahdavi J, Smith PK. The operation of a bully court and perceptions of its success: a case study. School Psychology International 2002;23:327–41.

20. Cunningham CE, Cunningham LJ, Martorelli V, Tran A, Young J, Zacharias R. The effects of primary division, student-mediated conflict resolution programs on playground aggression. J Child Psychol Psychiatry 1998;39:653–62.

21. Olweus D. Bullying at school: what we know and what we can do. Oxford: Blackwell; 1993.

22. Ortega R, Lera MJ. The Seville anti-bullying project. Aggressive Behavior 2000;26:113–23.

23. Rigby K. Stop the bullying: a handbook for schools. London: Jessica Kingsley; 2002.

24. Ross D. Childhood bullying and teasing: what school personnel, other professionals and parents can do. 2nd ed. Alexandria (VA): American Counseling Association; 2003.

25. Pikas A. A pure concept of mobbing gives the best results for treatment. School Psychology International 1989;10:95–104.

26. Robinson G, Maines B. Crying for help: the No Blame approach to bullying. Bristol (UK): Lucky Duck Publishing; 1997.

27. Young S. The support group approach to bullying in schools. Educational Psychology in Practice 1998;14:32–9.

28. Braithwaite J, Mugford S. Conditions of successful reintegration ceremonies: dealing with young offenders. British Journal of Criminology 1994;32:2.

29. Moore D. Pride, shame and empathy: community conferencing as transformative justice in education, In: Slee PT, Rigby K, editors. Children’s peer relations. London: Routledge; 1998. p 254–71.

30. Hyndman M, Thorsborne M, Wood S. Community accountability conferencing: trial report. Brisbane: Department of Education; 1996.

31. Olweus D. Bully/victim problems among schoolchildren: basic facts and effects of a school-based intervention program. In: Pepler DJ, Rubin KH, editors. The development and treatment of childhood aggression. Hillsdale (NJ): Erlbaum; 1991. p 411–48.

32. Roland E. Bullying: the Scandinavian research tradition. In: Tattum DP, Lane DA, editors. Bullying in schools. Stoke-on-Trent: Trentham Books; 1989. p 21–32.

33. Roland E. Bullying: a developing tradition of research and management. In: Tattum D, editor. Understanding and managing bullying. Oxford: Heinemann Educational; 1993. p15-–0.

34. Munthe E. Bullying in Scandinavia. In: Roland E, Munthe E, editors. Bullying: an international perspective. London: David Fulton; 1989. p 66–78.

35. Roland E. Bullying in school: three national innovations in Norwegian schools in 15 years. Aggressive Behavior 2000;26:135–43.

36. Olweus D, Limber S. The bullying prevention program. In: Elliott DS, series editor. Blueprints for violence prevention. Volume 9. Boulder (CO): University of Colorado; 1999.

37. Pitts J, Smith P. Preventing school bullying. London: Home Office Police Research Group; 1995.

38. Pepler D, Craig W, Ziegler S, Charach A. A school-based anti-bullying intervention: preliminary evaluation. In: Tattum D, editor. Understanding and managing bullying. Oxford: Heinemann Educational; 1993. p 76–91.

39. Pepler DJ, Craig WM, Ziegler S, Charach A. An evaluation of an anti-bullying intervention in Toronto schools. Can J Community Mental Health 1994;13:95–110.

40. Hanewinkel R, Eichler D. Ergebnisse einer interventionsstudie zur pr@vention schulischer gewalt. In: Schäfer M, Frey D, editors. Aggression und gewalt unter kindern und jugendlichen. Göttingen: Hogrefe-Verlag; 1999. p 245–64.

41. Hanewinkel R, Knaack, R. Mobbing: gewaltprävention in schule in Schleswig-Holstein. Report. Kronshagen: Landesinstitut Schleswig-Holstein fhr Praxis und Theorie der Schule; 1997.

42. Melton GB, Limber SP, Cunningham P, Osgood DW, Chambers J, Flerx V, Henggeler S, Nation M. Violence among rural youth. Final report to the Office of Juvenile Justice and Delinquency Prevention.Washington (DC): OJJDP; 1998.

43. Stevens V, de Bourdeaudhuij I, van Oost P. Bullying in Flemish schools: an evaluation of anti-bullying intervention in primary and secondary schools. British J Educational Psychology 2000;70:195–210.

44. Ortega R, Del Rey R. Aciertos y desaciertos del proyecto Sevilla anti-violencia escolar (SAVE). Revista de Educaci#n 2001;324:253–70.

45. Alsaker FD, Valkanover S. Early diagnosis and prevention of victimization in kindergarten. In: Juvonen J, Graham S, editors. Peer harassment in school. The plight of the vulnerable and the victimised. New York: Guilford; 2001. p 175–95.

46. Eslea M, Smith PK. The long-term effectiveness of anti-bullying work in primary schools. Educational Research 1998;40:203–18.

47. Olweus D, Endresen IM. The importance of sex-of-stimulus object: age trends and sex differences in empathic responsiveness. Social Development 1998;3:370–88.

48. Arora CMJ. Is there any point in trying to reduce bullying in secondary schools? Educational Psychology in Practice 1994;10:155–62.

49. Menesini E, Eslea M, Smith PK, Genta ML, Giannetti E, Fonzi A, and others. A cross-national comparison of children’s attitudes towards bully/victim problems in school. Aggressive Behavior 1997;23:245–57.

50. Pellegrini AD, Bartini M. An empirical comparison of methods of sampling aggression and victimization in school settings. J Educational Psychology 2000;92:360–66.

51. Ladd GW, Kochenderfer-Ladd B. Identifying victims of peer aggression from early to middle childhood: analysis of cross-informant data for concordance, estimation of relational adjustment, prevalence of victimization, and characteristics of identified victims. Psychological Assessment 2002;14:74–96.

52. Pepler DJ, Craig WM. A peek behind the fence: naturalistic observations of aggressive children with remote audiovisual recording. Dev Psychol 1995;31:548–53.

53. Kochenderfer BJ, Ladd GW. Peer victimization: cause or consequence of school maladjustment? Child Dev 1996;67:1305–17.

54. Monks C, Smith PK, Swettenham J. Aggressors, victims and defenders in preschool: peer, self and teacher reports. Merrill-Palmer Quarterly. Forthcoming.

55. Owens L, Shute R, Slee P. “Guess what I just heard!” Indirect aggression among teenage girls in Australia. Aggressive Behavior 2000;26:67–83.

56. Devine J, Lawson HA. The complexity of school violence: commentary from the US. In: Smith PK, editor. Violence in schools: the response in Europe. London, New York: RoutledgeFalmer; 2002. p 332–50.

57. Hawker DSJ, Boulton MJ. Twenty years research on peer victimization and psychosocial maladjustment: a meta-analytic review of cross-sectional studies. J Child Psychol Psychiatry 2000;41:441–55.

Author(s)

Manuscript received and accepted July 2003.

1. Professor of Psychology and Head, Unit of School and Family Studies, Goldsmiths College, University of London, London.

2. Research Officer, Institute of Education, University of London, London.

3. Professor of Mental Health, Head of Postgraduate Education, University of Surrey, Guildford, Surrey.

Address for correspondence: Professor PK Smith, Department of Psychology, Goldsmiths College, New Cross, London SE14 6NW, UK

email: p.smith@gold.ac.uk

1 | 2