Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

This chapter provides an overview of the international research on bullying in schools with a particular focus on the negative mental health effects. We recognize 3 main groups of students: students who bully others, students who are bullied by others, and students who do both. We begin by defining bullying as it occurs in schools and note challenges in defining bullying, especially given the emergence of cyber bullying (which itself has implications for schools). Next, we review the research literature for each of the above 3 student groups in terms of the rates of these behaviours, the factors that predict whether a student engages in these behaviours, and the effects of these behaviours with a focus on mental health. Along the way, we comment on the methodological limitations of the research to date. Next, theoretical frameworks for bullying in school, and in particular, the links between bullying and mental health are described and it is noted that this is an area for further development in the literature. We then review prevention and early intervention programs which are designed to promote mental health and reduce bullying. In the concluding section, we summarize the issues confronting the bullying in schools research literature and highlight some areas where future research is likely to be particularly illuminating.

Defining Bullying

The most accepted and widely used definition of bullying in schools includes 3 main features:

  • Aggressive behaviour that is intentional.

  • The aggressive behaviour is repeated.

  • There is a power imbalance between the student who does the bullying and the student who is bullied, with victims often being unable to easily defend themselves from perpetrators (Olweus 1993).

However, it is increasingly recognized that the conceptualization and measurement of bullying can be difficult, particularly when trying to capture the repetitious nature and power imbalances reflected in current definitions (Dooley et al. 2009; Grigg 2010). Bullying can be covert (e.g., exclusion, spreading rumors) or overt (e.g., verbal and physical abuse). Cyber bullying is often described as an extension of bullying in schools or so-called “traditional” bullying, with similar defining features to bullying in schools. Cyber bullying is also often considered a specific form of covert bullying that involves the use of electronic devices to carry out bullying but can also be overt (e.g., deliberate cyber stalking, sending derogatory or hate mail) (Spears et al. 2009). Electronic media such as computers, mobile telephones and tablets are used by young people to bully, embarrass, exclude or humiliate others, via methods such as email, chat-rooms, social networking sites, instant messaging, websites, telephone calls, video and text messaging (Cross et al. 2009; Smith et al. 2008). Although cyber bullying may not occur in schools, it may still have implications for schools when students bring into school events that began on the internet.

The emergence of cyber bullying has presented a number of challenges for the prevailing definitions of bullying in schools. For example, how is a power imbalance defined over the internet? Is it necessary for cyber bullying to occur more than once given that bullying through the internet may have a profound impact even if it occurs only once? The effects of that one instance of cyber bullying can be far greater than prolonged off-line bullying. These are challenges researchers are currently trying to address. Much work remains to be done to clarify the key components of bullying in schools given the advent of cyber bullying.

As outlined in the Introduction, in this chapter we recognise different groups of students who engage in bullying behaviour or are recipients of such behaviour.

  • Students who bully only (perpetrators only).

  • Students who are bullied only (victims only).

  • Students who are both bullied and bully others (both perpetrators and victims).

Increasingly, another group is recognised, that of bystanders. These students do not engage in bullying nor are they the receivers of bullying, however, they are present when bullying occurs and often choose not to stop the bullying or in some cases encourage the student who is bullying others. While we acknowledge the existence of this group of students and their role in bullying situations, they will not be a focus of our review in this chapter.

As highlighted bullying, remains a social problem, however the extent to which it proves problematic is uncertain, given the complexities associated with the measurement of bullying. In addition, a range of factors influence our ability to compare the incidence of bullying across studies. The way in which bullying is measured has been known to differ across studies (Solberg and Olweus 2003), particularly in terms of the varied definitional features used in the measurement of bullying. In addition, the time frame by which respondents are asked to report bullying also differs across bullying instruments, as do the behavioural indicators used to constitute bullying in such surveys. There is further diversity in the way in which bullying scales are presented. Some researchers choose to clearly define bullying before measuring the bullying acts, whilst others refrain from defining (or at times even mentioning the term) bullying, but rather operationalize bullying by asking participants to rate their involvement in a predetermined list of aggressive behaviours. In addition, the timing of the administration of the bullying survey has been considered important in influencing incidence rates of bullying (Smith 2011). Given that such disparities exist in the measurement and definition of bullying across studies, drawing conclusions about the prevalence rates of bullying over time is difficult. Likewise, the extent to which young people are perpetrators and victims of cyber bullying is questionable, as the calculation and measurement of prevalence rates of cyber bullying presents further challenges. As technology is constantly evolving, cyber bullying continues to manifest in many different ways. Therefore, it is important to consider that the calculation and comparison of prevalence rates of cyber bullying across time and across studies is difficult, due to the historical changes in the measurement and definitions of cyber bullying (Rivers et al. 2011). Nonetheless, researchers have reported rates of bullying; however these studies have been based on cross-sectional studies. Results from such research confirm that bullying is an international phenomenon, with rates differing considerably across countries (Due et al. 2005). Rates of bullying will be highlighted in the remainder of the chapter as they relate to the 3 bullying subgroups: bullies (perpetrators), victims, and bully-victims.

Bullying-Perpetrator

A bully is someone who repeatedly attacks another individual who does not fight back (Berger 2007, p. 96).

Rates and Patterns of Bullying Perpetration

Nansel et al. (2001) surveyed over 15,000 American students across grades 6–10 (ages 12–16). Of the 29.9 % of the sample who indicated some form of engagement within the bully situation, 13.0 % appeared to take on the bully role. Further, the frequency of bullying increased through grades 6–8 compared to students in grades 9 and 10.

The Health and Behaviour in School-age Children (HBSC) 2001/2002 international report found that overall, approximately 35.0 % of young people were involved in bullying others at least once during the previous couple of months. Of this group, 14.0 % engaged in fighting, 9.0 % engaged in fighting and bullying, while 8.0 % engaged in bullying alone.

These rates vary substantially by age, countries and regions: 9.0–54.0 % for 11 years, 17.0–71.0 % for 13 year olds and 19.0–73.0 % for 15 year olds. The mean percentages for the 3 age groups are 30.0 %, 38.0 % and 36.0 % respectively. Austria, Estonia, Germany, Latvia, Lithuania, Switzerland and Ukraine are consistently in the top quartile across all age groups, and the Czech Republic, Ireland, Scotland, Slovenia, Sweden, The Former Yugoslav Republic of Macedonia and Wales are in the lowest quartile.

In all countries and regions and all age groups, boys report bullying others more often than girls. In most countries and regions, the higher increase in reported bullying occurs between ages 11 and 13. Further, across all age groups, 10 countries and regions show an increase with age; 3 show similar rates (Belgium (Flemish), Belgium (French) and Greenland) and 2, a decrease with age (Israel and Norway).

In 2005, Due et al. (2005) also demonstrate from their survey across 28 European and North American countries that bullying is an international phenomenon, where rates vary considerably across countries. Prevalence of bullying was reported to range between 5.1 % (girls) – 6.3 % (boys) in Sweden, to 38.2 % (girls) and 41.4 % (boys) in Lithuania. This appears consistent with the HBSC 2001/2002 research.

Regarding the prevalence of cyber bullying, in their national US study of 10- to 17-year olds, Ybarra and Mitchell (2004) reported that 15.0 % of their sample had engaged in online harassment behaviour. In Cross et al. (2009) national Australian study of 8- to 14-year-olds, it was revealed that for all forms of bullying, including cyber bullying, 11.0 % of boys and 7.0 % of girls had participated. Rates for cyber bullying alone were 3.8 % for boys and 3.3 % for girls.

More recently, Hemphill et al. (2011, 2012b) present one of the first 3-year longitudinal studies reporting on differential bullying and victimisation rates within a sample of students in mid- to later- adolescence. The results showed that the most common form of bullying in grades 9 through grades 11 was relational aggression, with up to 72.0 % of boys and 65.0 % of girls in grade 9 reporting that they engaged in relational aggression. In grades 9–11, rates of traditional bullying perpetration and relational aggression were higher in boys than girls. Across time, gender differences in rates of traditional bullying perpetration increased in boys and girls from grades 9 to 11, while relational aggression decreased over time in both girls and boys. The most common combination of bullying subtypes across grades 9–11 was traditional bullying and relational aggression across the entire sample. Such rates increased over time.

Interestingly, Hemphill et al. (2011, 2012b) found a high rate of relational aggression amongst boys. This was contrary to expectation as relational aggression was originally conceptualised as a form of aggression characterised by behaviours engaged in mostly by girls (Crick and Grotpeter 1995). However, it was found that 72.0 % of boys and 56.0 % girls in grade 9 reported relational aggression engagement (Hemphill et al. 2011, 2012b). This difference in findings requires further analysis that accounts for age, school level and investigative methodologies (e.g., peer nominations vs. survey).

The existing overt bully/aggressor literature suggests an inverse relationship between age and bully perpetrator behaviour. However, Hemphill et al. (2011, 2012b) found that the role of the bully within the traditional context increased in both sexes from grades 9 to 11. Again, further investigation accounting for methodological differences is required. For example, such an outcome may reflect the lack of research cohesion around the operationalization of the bullying construct.

Predictors of Bullying Perpetration

Risk factors of bullying perpetration have been extensively studied. A common research finding is that having a history of involvement in bullying or aggressive/antisocial behavior increases the likelihood of engagement in subsequent future bullying perpetration. For instance, prior engagement in relational aggression, which in itself is considered a subtype of bullying (Van der Wal et al. 2003), is predictive of both traditional and cyber bullying perpetration when measured 2 years later in year 9 (Hemphill et al. 2012a). Furthermore, being the victim or perpetrator of traditional bullying in year 7 predicted bullying perpetration in year 9 (Hemphill et al. 2012a). Results from a meta-analysis across 153 studies exploring bullying in childhood and adolescence found that externalizing behaviors (“deviant, aggressive, disruptive and noncompliant responses”) were the strongest individual level predictors of being a perpetrator of bullying (Cook et al. 2010). Strong associations have also been noted between bullying perpetration amongst students and violent-related behaviors, such as weapon carrying, frequent fighting and being injured in a fight (Nansel et al. 2003). The odds of being a bully and also engaging in these behaviors are especially high when bullying occurs away from the school environment (Nansel et al. 2003).

Although externalizing problem behaviours have been found to predict being a bully, there is also evidence that internalizing problems such as depressive symptoms are associated with bullying perpetration (Slee 1995). This relationship may be particularly pertinent to boys, given that the association between bullying others and levels of depression differs for boys and girls (Austin and Joseph 1996; Slee 1995). The identification of depression as a predictor of bullying is supported by longitudinal research by Sourander and collaborators (2000) who found that experiencing high levels of depressive symptoms at age 8 predicted bullying perpetration at age 16, as did being male (Sourander et al. 2000).

At the school level researchers have found that attending a school with a positive climate and being connected to school is associated with a lower risk of involvement in bullying perpetration (Williams and Guerra 2007). School suspension may also be related to bullying perpetration as it has been shown to increase the likelihood of other violent (Hemphill et al. 2009) and antisocial behavior (Hemphill et al. 2006) at 12 month follow-up, independently of other established risk factors. Researchers have also highlighted a link between poor scholastic achievement and problem behaviors including bullying at school, with academic failure (at Year 7) found to predict traditional bullying perpetration (at Year 9) (Hemphill et al. 2012a, b).

Family conflict is also an established predictor of youth violence and physical aggression (Hawkins et al. 2000; Hemphill et al. 2009; Herrenkohl et al. 2000). In a study by Hemphill et al., family conflict at year 7 was considered a predictor of traditional bullying perpetration 2 years later (Hemphill et al. 2012a, b). Such findings suggest that young people living in a home environment characterized by conflict may themselves engage in problem behaviour, including bullying, in other contexts (Farrington and Ttofi 2011; Hawkins et al. 2000; Hemphill et al. 2009; Herrenkohl et al. 2000). This is further supported by studies which have indicated that being exposed to domestic violence predicts being a bully, particularly when mothers are violent towards fathers (Baldry 2003). Furthermore, in the family context high parental support is negatively related to physical, verbal, relational and cyber bullying perpetration (Wang et al. 2009), whereas having a poor emotional bond with a caregiver increases the likelihood of being involved in online bullying perpetration (Ybarra and Mitchell 2004). Poor family management (reflected by lack of clear rules and monitoring of students) is also an established risk factor for violent and antisocial behaviors (Hawkins et al. 2000; Hemphill et al. 2009; Herrenkohl et al. 2000).

The impact of peers on bullying have also been studied and found to be influential on bullying. In particular, peer influences which impact on the adjustment of young people (e.g. associating with antisocial friends and being involved in prosocial group activities) have been noted as strong predictors of bullying perpetration amongst children and adolescents (Cook et al. 2010).

Outcomes of Bullying Perpetration

Bullying others has been linked to a number of adverse outcomes, with behavioural and psychosocial outcomes commonly cited. Much research is available which suggests that being involved in bullying at school increases the likelihood of future engagement in externalising behaviours such as antisocial behaviour, violent or offending behaviour. In a study by Hemphill and collaborators (2012) which explored the short-term longitudinal psychosocial consequences of traditional bullying among Victorian students in the mid to late secondary school years, it was found that traditional bullying perpetration in year 10 predicted multiple future psychosocial outcomes such as theft, violent behaviour and binge drinking when assessed a year later (Hemphill et al. 2012). In addition, child and adolescent perpetrators of bullying are at heightened risk for later criminality (Sourander et al. 2006) and recidivist criminality in later years (Olweus 1993, 1999). In a longitudinal study by Olweus, it was found that over half of young people who were classified as bullies in grades 6–9 were later convicted of crime by age 24, with 35–40.0 % of bullies being convicted of multiple criminal offences (Olweus 1993, 1999). Cross-sectional studies have also highlighted that perpetration of bullying by young people is associated with increased risk of medically treated injuries when assessed across multiple European countries, the USA and Canada (Pickett et al. 2002). Furthermore, bullying perpetrators at school have been found to experience health problems, emotional adjustment difficulties, social and school adjustment problems, as well as use of alcohol to a greater extent than their non-involved peers; a finding which was consistently found in various European countries, the USA and Canada (Nansel et al. 2004). There is further evidence that frequent bullying perpetration at school is associated with an increased risk of having self-harming thoughts, particularly around suicidal ideation (Kaltiala-Heino et al. 1999). As these later studies are based on cross-sectional data, we are limited to drawing causal inferences about the direction of such relationships.

Traditional bullying perpetration has also been associated with a number of other long term mental health adversities, in the form of internalising problems similar to those of bullying victimisation. Gibb and collaborators (2011) noted long term consequences of bullying perpetration at age 13–15 years, including mental health and adjustment problems at age 16–30 years (Gibb et al. 2011). Higher rates of internalising problems (major depression, suicidal ideation, suicide attempt), alcohol and illicit substance dependence, as well as externalising problems (violent offending, property offending and arrests/convictions) at ages 16–30 were noted for those who bullied others in adolescence, as compared to those who did not (Gibb et al. 2011). Furthermore, young people engaging in bullying perpetration at school are more likely to experience increased aggression over time (Kim et al. 2006).

Bullying Victimization

According to Olweus (1994) a student is being bullied or victimized when he or she is exposed, repeatedly and over time, to negative actions on the part of one or more other students (Olweus 1994, p. 1171). As bullying can be both overt and covert nature, it is possible that young people may initially be oblivious to being the victim of bullying. This is especially true when it comes to cyber bullying.

Rates and Patterns of Bullying Victimization

Rates of bullying victimisation have been found to vary across countries and settings. In a cross national sample of schools across Europe and North America, a large variation in the rates of victimisation were reported. Students in Sweden reported the lowest rates of bullying victimisation (5.1 % for girls and 6.3 % for boys) as compared to students from Lithuania who reported the highest rates (38.2 % for girls and 41.4 % for girls) (Due et al. 2005). A US national survey indicated that approximately 32.0 % of students aged 12–18 reported being victimized at school via traditional bullying methods, with rates of cyber bullying being much lower (3.7 %) (National Center for Educational Statistics: Institute of Education Sciences 2011). In another study reporting rates of bullying victimization amongst a national representative sample of US students, it was found that up to 8.4 % of the sample were frequently bullied (weekly) (Nansel et al. 2001). A national Australian study of 8- to 14-year-olds found that rates of being bullied ranged from nearly 24.0 to 29.0 %. Males and females reported similar rates of victimization, with 27.0 % and 26.0 % of males and females respectively reporting being victimized (Cross et al. 2009). In a longitudinal study which surveyed approximately 800 students from Victoria, Australia, rates of traditional bullying victimization were reported to be 28.0 % in year 9. These rates rose to almost 39.0 % 2 years later as measured in year 11. However, rates of cyber bullying victimization remained relatively stable from year 9 (14.0 %) to year 11 (13.4 %) (Hemphill et al. 2012a, b).

Results from various studies highlight a general trend for the rates of bullying victimization (including frequent victimization) to decline with age of the victimized person (Due et al. 2005; Nansel et al. 2001; Olweus 1994). The persistence of bullying victimization has also been explored with results from studies suggesting that being the victim of bullying can extend over multiple years. For instance, being bullied at age 8 was associated with bullying victimization 8 years later, with approximately 90.0 % of boys who reported being victims at age 16 also being victims at age 8, as compared to about 50.0 % of females who reported being a victim of bullying at age 16 also victims at age 8 (Sourander et al. 2000). In a study by Kumpulainen and collaborators (1999) it was noted that although the number of students involved in bullying declined from age 8 to age 12, many children shifted their bullying status over this time. For instance, 9.4 % of children who were bullied at age 8 were then bullying other children at age 12, whereas 7.2 % of those bullied at age 8 experienced both perpetration and victimization at age 12 (Kumpulainen et al. 1999). Such information illustrates the important interplay which operates between the different bullying subtypes over time.

Various studies have also provided evidence that rates of bullying victimization are higher for males than females, however at times gender differences have been minimal (Due et al. 2005; Kumpulainen et al. 1999). Males have also been found to be bullied more frequently than females (Nansel et al. 2001).

These noted trends in the rates of bullying victimization may in fact be specific to the bullying subtype investigated. For instance, research suggests that there are higher portions of females being bullied by technology (7.7 %) as opposed to males (5.2 %) (Cross et al. 2009). Likewise, it has been found that rates of cyber bullying victimisation consistently increase from year 4 to year 9 (Cross et al. 2009). These results suggest that males are more likely to be subjected to traditional forms of bullying victimisation, whilst females may be exposed to cyber bullying victimisation more so than males. Given that such disparities exist, it is important for such differences to be reflected in the measurement of bullying.

Predictors of Bullying Victimization

A range of factors predict bully victimisation amongst youth. These span across individual, family peer and school domains. Mental health factors such as experiencing internalising problems and displaying externalising behaviours have often been cited as predictors of bullying victimisation. In particular young people who experience internalising problems such as emotional problems are often considered more likely to experience bullying victimisation. A study by Hodges and Perry (1999) found that internalising problems (i.e. withdrawal and anxiety-depression) amongst children and early adolescents contributed to bullying victimisation over a 1 year period (Hodges and Perry 1999). Likewise, students who were identified as victims of bullying at age 5–7 experienced more internalizing problems prior to bullying as compared to their peers (Arseneault et al. 2006). Internalising problems such as high levels of depressive symptomology at age 8 have also been associated with later bullying victimisation over an 8 year period (Sourander et al. 2000). Finally, having a low self-regard was also found to predict later victimisation by peers when assessed longitudinally (Egan and Perry 1998).

Externalising behaviours have also been linked to bullying victimisation. Physical aggression in preschool (early childhood aged 17 months) has been found to predict childhood peer victimisation trajectories (Barker et al. 2008). Furthermore, young people who became victims of bullying between the ages of 5 and 7 experienced more externalising problems (aggression and delinquency) than their peers. This association was only true for girls (Arseneault et al. 2006). The results of a cross-sectional study also revealed that being bullied in school and away from school was associated with violent behaviour such as weapon carrying, fighting and injuries sustained from physical fights (Nansel et al. 2003). Given that these findings are based on cross-sectional data the temporal ordering of these factors is questionable.

Family influences are also known to shape young people’s behaviours and experiences, including their experiences of bullying. Therefore, understanding the role that families play in predicting bullying victimisation is crucial to developing a systemic and holistic approach to address bullying. Children residing in home environments characterised by violence and marital conflict (i.e. are exposed to domestic/interparental violence) (Baldry 2003; Beran and Violato 2004) and children who are maltreated at home (Shields and Cicchetti 2001) are at greater risk of being victimised by their peers at school. As noted, mental health problems encountered by young people are predictors of bullying victimisation. Likewise, parental mental health problems such as depression (Beran and Violato 2004) have also been linked to school bullying victimisation amongst youth. Again, these associations between family level factors and bullying victimization are based on cross-sectional data.

In addition to individual and family influences, experiences at school have long been studied in relation to bullying as the school environment is a prominent context in which bullying occurs. School factors have been identified as predicting being a victim of bullying. School climate in particular predicted bullying victimisation amongst children and adolescents (Cook et al. 2010). It has been reported that young people who are victims of bullying are generally more unhappy at school in the early years as compared to their peers (Arseneault et al. 2006).

It is not surprising that peer factors are associated with bullying victimisation, given that bullying is often facilitated through social interaction with peers and that bullying is often conceptualised as a relationship problem (Pepler et al. 2004; Spears et al. 2009). Furthermore, as bullying often occurs in the school context amongst peers, it is expected that peer-relational factors impact on bullying experiences. In a meta analysis of 13 commonly referred to individual and contextual predictors of bullying across 153 studies, it was found that peer status (quality of relationships children and adolescents have with their peers p. 67) had the strongest effect in predicting victimization status amongst children and adolescents (Cook et al. 2010). Peer rejection has also been found to be associated with peer victimisation, as has being disliked by peers (Beran and Violato 2004; Hodges and Perry 1999). In the meta-analysis reported by Cook and collaborators (2010) “social competence” also had a relatively large effect size in relation to being victimised (Cook et al. 2010). Similarly, Egan and Perry (1998) found that young people with poor social skills are at greater risk of experiencing peer bullying victimisation, particularly when they also had a low self-regard (Egan and Perry 1998).

Although individual, family, school and peers factors have been noted to put young people at increased risk of bullying victimization, it is also possible that victimization may contribute to young people further experiencing these same adjustment difficulties as a consequence of being the victim of bullying. For instance, in a study conducted by Egan and Perry (1998) self-regard (peer social competence) was found to predict bullying victimisation, but victimisation was also found to predict later self-perceived peer social competence (Egan and Perry 1998). A second study noted similar findings. Hodges and Perry (1999) found that internalising problems, physical weakness and peer rejection predicted victimisation 1 year later. Furthermore, victimisation also predicted internalising problems and peer rejection 1 year later (Hodges and Perry 1999). These findings illustrate that young people who experience adjustment issues early in life may be immersed in a vicious cycle incorporating early adjustment problems which lead to being bullied, which then results in experiencing further adjustment difficulties as the result of being bullied. Few studies have empirically tested such reciprocal relationships over time as the longitudinal data needed to test these relationships are often scarce. Nonetheless it important to consider both the antecedents and consequences of bullying to truly understand the complex nature of this social problem that occurs in schools. The consequences of being bullied will now be discussed.

Outcomes of Bullying Victimization

The impact of being bullied often affects the physical, social and emotional wellbeing of those being victimised. Bullying can be severely violent in nature, and so, bullying can have detrimental physical health consequences on the victims of bullying. There is also vast evidence to support that young people who are victims of bullying also experience adjustment issues and mental health problems in the short and long term. Findings from both cross-sectional and longitudinal studies have highlighted the association between bullying victimisation in childhood and adolescence and a range of mental health problems, of which self-harming behaviour, violence and psychotic symptoms have been included (Arseneault et al. 2010). However, the temporal ordering of these relationships is sometimes questionable when reliant on cross-sectional studies. We focus on describing findings from longitudinal studies which have highlighted the association between bullying victimization and experiencing later mental health problems.

There is a tendency for those who are bullied to be at increased risk of experiencing future internalising problems (e.g. being withdrawn, somatic complaints, anxious/depressed) (Arseneault et al. 2006; Hodges and Perry 1999). In a study conducted by Bond et al. (2001) it was found that being victimised in year 8 (age 13) was associated with anxiety and depressive symptomatology in the preceding year (Bond et al. 2001). This longitudinal association has also been examined in the later school years. Being victimised (via traditional bullying methods) in year 10 has been associated with depressive symptomology in year 11 (Hemphill et al. 2012). Likewise, Tfoti and Farrington (2011) found that depression may be considered a long term consequence of being bullied at school, even when assessed up to 36 years post bullying victimization and controlling for a range of other childhood risk factors (Ttofi and collaborators 2011). There is further evidence that being bullied at school (as assessed retrospectively) is associated with a greater likelihood of having a diagnoses of depression when assessed much later in life (31–51 years) (Lund et al. 2008).

Higher rates of both internalising and externalising problems have been noted between the ages of 16–30 amongst students who were bullied at age 13–15, of which major depression, anxiety disorder, alcohol dependence, illicit drug dependence, conduct disorder/antisocial personality disorder, violent offending, property offending, arrest/conviction have been included (Gibb et al. 2011). In addition to the link between bullying victimisation and later internalising problems, bullying victimisation has also been associated with increased risk of experiencing externalising problems, particularly for girls (Arseneault et al. 2006).

Given the social nature of bullying, research exists which suggests that those being bullied also suffer socially, with many at increased risk of experiencing peer rejection over time (Hodges and Perry 1999). According to Kim and collaborators (2006), young people who are bullied in 7th grade are also at greater risk of experiencing social problems 10 months later (Kim et al. 2006). Likewise, it has been suggested that students who are victims of bullying between the ages of 5 and 7 exhibit less prosocial behaviour at age 7 (Arseneault et al. 2006). Furthermore school experiences, such as happiness with school and with peers are affected by bullying victimisation (Arseneault et al. 2006).

Long term exposure to bullying victimization may also be fatal, as evidenced by reports around the world of young people committing suicide as the result of being bullied (Berger 2007). Furthermore, higher rates of suicidal ideation and suicide attempts have been noted amongst students who were bullied between the ages of 13–15 compared to students who did not report being bullied, when these outcomes were measured amongst participants aged 16–30 (Gibb et al. 2011).

These very serious consequences of being bullied emphasise the importance for schools to have practices and services in place to offer support and to educate young people about how to productively cope with bullying, in order to minimise the impact of the bullying experience.

Bullying Perpetration and Victimization

As the name suggests, a bully-victim is someone who is simultaneously a perpetrator of bullying but is also a victim of bullying (Berger 2007).

Rates and Patterns of Bullying Perpetration and Victimization

Nansel et al. (2001) surveyed over 15,000 American students across grades 6–10 (ages 12–16). Of the 29.9 % of the sample who indicated some form of engagement within the bully situation, 6.3 % reported concurrently being both the bully and the victim. Further, males were more likely than females to occupy this dual role.

The Health and Behaviour in School-age Children (HBSC) 2001/2002 international report found overall, that 35.0 % of their sample indicated no aggressive behaviours. Interestingly, 24.0 % of the sample engaged in fighting or bullying and victimisation concurrently.

More recently, Hemphill et al. (2011, 2012b) present one of the first 3-year longitudinal studies reporting on differential bullying and victimisation rates within a sample of students in mid- to later- adolescence. It was found that 10.0 % of the sample fell within the bully/victim role, within both the traditional and cyber contexts. However, for the traditional bully/victim category this percentage rose to 27.0 % for those in grade 11 (Hemphill et al. 2012).

Predictors of Bullying-Victimization

Family and peer factors have also been studied as correlates of bully-victim status. It has been found that parenting characteristics such as being rejected by parents, as evidenced by displays of hostility and punishment impact on being a bully-victim at school, compared to young people who neither bully or are bullied (Veenstra et al. 2005). Also parental vulnerability to psychopathology, particularly externalizing disorders was associated with being a bully-victim at school (Veenstra et al. 2005). Peer factors such as peer status and peer influences have also been identified as strong contextual predictors of being a bully-victim in childhood and adolescence (Cook et al. 2010). Furthermore young people reported higher levels of avoidance behaviour toward bully-victims peers (as compared to bullies, victims, uninvolved or other youth) as indicated by peer nomination (Juvonen et al. 2003).

Bully-victims also present with higher rates of school refusal behaviour (Kumpulainen et al. 1998) and school disengagement (Juvonen et al. 2003) compared to peers. They also experience interpersonal problems (including “fighting fatigue, loneliness, reduced social interest, somatic concerns”) and ineffectiveness (including “pessimism, self-depreciation, school-work difficulty, self-blame, indecisiveness”) to a greater degree than their peers (Kumpulainen et al. 1998). These results from multiple research studies suggest that bully-victims are confronted with an array of psychosocial difficulties. As such, engaging in the role of a bully-victim may have serious ramifications for such young people. These will now be discussed.

Outcomes of Bullying Perpetration and Victimization

It has been suggested that young people who maintain the dual role of bullying others and are also victims themselves experience greater levels of adjustment difficulties as compared to peers who maintain a single role in the bullying dynamic (Arseneault et al. 2010). A study by Kumulainen and Rasanen (2000) revealed that students who were bully-victims at age 8 had higher levels of psychiatric symptoms and were at heightened risk of being deviant at age 15, as compared to non-involved students (Kumpulainen and Rasanen 2000). Consistent with this, Kim and collaborators (2006) also found bully-victims at Grades 7 and 8 were at much greater risk of experiencing aggression and externalizing problems 10 months later, with odds ratios of 4.9 and 4.6 noted respectively for these consequences (Kim et al. 2006). Long term consequences of being a bully-victim in childhood include increased likelihood of being involved in multiple criminal offenses at age 16–20, with traffic offenses most likely amongst this group (Sourander et al. 2007).

Young people classified as bully-victims in childhood (age 5–7) also experienced more externalising problems and exhibited poorer prosocial behaviours as compared to victims and peers not involved in bullying (Arseneault et al. 2006).

Although externalising behavioural problems have commonly been cited as consequences of being a bully-victim, this group have also been linked to internalising problems in the short and long term. Bully-victims in early adolescence (age 13) have been found to experience higher levels of internalizing problems such as depression, low self-esteem, failure expectations and self-harm compared to peers who were not involved as a bully or a victim, when assessed up to 2 year later (Ozdemir and Stattin 2011). Bully-victims are at increased risk of persistent suicidal ideation as compared to non-involved peers (Kim et al. 2009).

Bully-victims have been found to encounter school adjustment difficulties. For instance lower academic performance and reading efficiency have been noted amongst bully-victims at age 7 compared to students who were considered victims only, or alternatively not involved at bullying at all. In addition, bully-victims also seem less happy at school as compared to peers (Arseneault et al. 2006).

Theoretical Framework to Explain Bullying and Links to Mental Health

Despite the extensive research literature on bullying in schools, there is no one established theory to explain the development of bullying and in particular how it impacts on mental health. To date, research on bullying in schools has focused on describing the phenomena rather than explaining it. Hence, there is no one accepted theory of bullying in schools, rather researchers tend to draw on one of many different psychological theories, such as Social Learning Theory, Social Cognitive Theory, Evolutionary Psychology and the list goes on. Like most complex and similar behaviours (e.g., violence) it is likely that a tripartite model that recognises there are multiple influences on the development of bullying including (1) biological/genetic; (2) psychological characteristics of the student; and (3) social environment to which the student is exposed. Much more research remains to be done to adequately develop a theoretical understanding of bullying in schools. The book chapter by Dixon and Smith (2011) is a novel attempt to provide an integrated theory of bullying in schools. Further theorising and research of this nature is needed in this field of research, particularly in relation to how bullying in schools relates to mental health.

What Can Be Done to Reduce Bullying? What Can Schools Do?

There exists an abundance of prevention and intervention strategies and programs reported in the research literature that aim to reduce bullying perpetration, victimisation and wider antisocial behaviours. These range from parental education and training programs to improving community safety and the school environment. The most common types of intervention include school anti-bully policies, school programs targeting families, teachers and peer groups about the impact of bullying on an individual’s emotional and psychological health, and individual social skills and assertiveness training. Schools are also increasing their attention on the school climate. Teachers are playing a greater role in recognising when bullying occurs through supervision of social/play areas where much of the bullying can take place. Other schools have introduced peer support groups that teach students how to stand up to bullies. This is particularly targeted for victims of bullying, providing those affected with skills in assertiveness and social interaction.

Many anti-bullying interventions have taken place in Norway (Olweus programme), UK, Australia, Belgium, Canada, Germany, Finland, Ireland, Italy, Spain, Switzerland and the USA (Monks and Coyne 2011; Smith et al. 2004). Much of the literature focuses on prevention and intervention programs utilising indirect and direct approaches. A whole school approach has provided encouraging results in reducing bullying; however is limited by the paucity of strong evidence of its overall effectiveness and sustainability (Tangen and Campbell 2010; Swearer et al. 2010). The whole school approach takes into account different facets of bullying prevention and intervention including the quality and style of leadership and management in schools and its curriculum, school activities, the quality of teacher supervision, and anti-bullying policies. Researchers posit that this indirect approach is limited by the exclusion of direct intervention for the perpetrators, who require strategies to engage in more prosocial behaviours (Smith et al. 2004; Swearer et al. 2010; Vreeman and Carroll 2007).

Schools who adopt a direct approach to managing bullying often require that teachers are skilled in detecting incidents of bullying and deal directly with the perpetrator(s) and victim(s). However, several issues evolve from the direct approach such as the teacher’s skills and training in managing bullying behaviour as well as their moral standing in response to the bullying. Another view supports that bullying should not only be addressed directly with those involved, but should be considered in the wider social context. This view is supported by Luiselli et al. (2005) who identified the need for social skills training, promotion of social competency (by teaching students how to interact more effectively with peers and adults through enhanced conflict resolution, developing problem solving skills), developing skills in negotiation, and developing friendship-building abilities.

Although much of the literature has investigated the nature and impact of traditional forms of bullying in the schoolyard and in the classroom, many students are now also exposed to cyber bullying. Young people engage in cyber bullying to harm others repeatedly via technology including social networking sites (e.g., facebook, twitter), websites, web cameras and mobile phones. This has presented a new set of challenges for schools: students can readily contact each other through this technology allowing victims to be reached anytime and anyplace. Secondly, cyber bullying is often covert and the perpetrator can remain anonymous reducing the likelihood of being detected or reprimanded by teachers and parents. Therefore, many bullying incidents go unreported. Furthermore, bullies can spread rumours or insults quickly to a large amount of people in a short amount of time. This type of bullying is difficult for the victim to defend themselves and is almost impossible for the school to contain.

Even though there have been various school-based interventions that have been evaluated and contributed to the reduction of bullying in schools, bullying continues to remain a significant issue for schools impacting on young people and their school experience. So what can schools do? There is some consensus that bullying prevention and intervention need to be introduced in the early school years and should be reviewed in on-going and consistent fashion, regularly practiced in schools, facilitated by teachers and supported by family or significant others (Swearer et al. 2010). There is not one prevention or intervention strategy that has been found to be effective in schools, however there is growing evidence to support a multi-factorial approach to ameliorating the problem. Bullying prevention and interventions should address the systems directly affecting children and adolescents including families, schools, peer groups, teacher–student relationships, parent–child relationships, parent–school relationships, neighbourhoods, and cultural expectations (Swearer et al. 2010). Although this holistic perspective has been studied in such areas as school violence, the application of this framework in the bullying area is in its infancy. Future research is warranted to investigate whether this can significantly reduce attitudes and perceptions supportive of bullying and also can create meaningful and sustainable behaviour change amongst young people.

Conclusion

An extensive literature on bullying in schools has been built through over 40 years of research. This research has shown us that bullying is common, that there are a number of factors in a student’s life that predict who does the bullying and who is bullied, and that it is associated with poor mental health, as well as a range of costs to students, families, schools, and the broader community.

Despite this research, there are a number of areas that require further research. First, the advent of cyber bullying has strengthened questions about the conceptualisation and definitions of bullying. For example, how is a power imbalance defined over the internet? Is it necessary for cyber bullying to occur more than once given that bullying through the internet may have a profound impact even if it occurs only at one time. These are important questions that the field of bullying research will need to address in coming years.

Clarifying the definition and conceptualisation of bullying in schools will no doubt improve the measurement of bullying and this will assist researchers to calculate more accurate prevalence rates for bullying. Currently, it is difficult to compare the rates found in different studies because they have used different measures and therefore often report very different rates of bullying in schools. Developing psychometrically sound measures of bullying in schools that are grounded in conceptual and theoretical frameworks is likely to result in more accurate estimates of the occurrence of bullying in schools.

Third, much of the research on bullying in schools to date has operated in a theoretical vacuum. There is no one accepted theory of bullying in schools, rather researchers tend to draw on one of many different psychological theories, such as Social Learning Theory, Social Cognitive Theory, Evolutionary Psychology and the list goes on. Like most complex and similar behaviours (e.g., violence) it is likely that there multiple influences on the development of bullying ranging from biological/genetic to psychological characteristics of the individual to the social environment to which the student is exposed. Much more research remains to be done to adequately develop a theoretical understanding of bullying in schools. The book chapter by Dixon and Smith (2011) is a novel attempt to provide an integrated theory of bullying in schools that recognises bullying occurs in a social system (school). Further theorizing and research of this nature is needed in this field of research, particularly in relation to how bullying in schools relates to mental health.

Within the bullying in schools literature, there are increasingly descriptions of 4 main groups:

  • Students who do not bully and are not bullied,

  • Students who bully only,

  • Students who are bullied only,

  • Students who are both bullied and bully others.

There is more research required to better understand these different groups. What are the mental health outcomes for students who have been bullied and also bully others? It seems likely that they will be worse than those students who bully only or have been bullied only. Is this group of students who both bully and have been bullied qualitatively different to the other groups? Under what circumstances do students end up in this group?

Cyber bullying has presented many challenges for schools. If bullying occurs outside school grounds particularly over the internet, should schools become involved? This is a question researchers and educators are still grappling with today. Increasingly the answer seems to be that if what happens outside school is likely to have consequences within the school environment, then schools need to know what is going on and how to address it. So, yes, schools do need to make cyber bullying their business. This will be an important area of consideration in future years.

The extensive literature on bullying in schools includes reports on the effectiveness of various prevention and early intervention programs – generally with the aim of reducing bullying but also improving the mental health of students more broadly. However, there remain challenges in this area; a key one is to ensure that knowledge about effective prevention and early intervention programs is disseminated to schools. To date, anecdotally, schools continue to programs that do not have an evidence base. There need to be improved ways of helping schools to make good choices about the programs they use. They also need to monitor and evaluate how the program works in their school – since what works for one school may not work for another.

In conclusion, research on bullying in schools shows that this is a common behaviour, many factors influence whether or not students bully or are bullied, and there are serious mental health effects of bullying on students – many of these long-lasting. More consideration of conceptual and theoretical issues is required in this field. Further, more research is needed to give accurate estimates of the rates of bullying in schools, how students come to engage in bullying, be bullied, or both, and how to minimise the negative mental health effects of bullying on students in the short and long term. By successfully attending to all of this, the rates of bullying in schools and the associated mental health effects can be reduced.