Cyberbullying A Bully Centered Intervention Program Psychology Essay

Published Date: 23 Mar 2015

Cyberbullying, an emerging trend of bullying in its unprecedented technological form, has fast become a social issue that warrants immediate intervention. However, current literature has offered mostly intervention strategies that focus on the treatment of the victims, rather than the rehabilitation of the cyberbullies. Such victim-centric intervention program, while helpful, does not appear to be an effective strategy in tackling cyberbullying. This review provides an overview of the cyberbullying problem by focusing on its (a) definition; (b) modus operandi; (c) perpetrators and (d) theories. These findings form the foundation for the recommendations of the bully-centric intervention program. The recommended intervention program serves to complement existing victim-centric interventions to effectively manage the cyberbullying problem. Finally, this review concludes by highlighting plausible limitations and future directions of the recommended intervention program.

Introduction

Traditionally, bullying involves the harassment of others using physical means (Guerin & Hennessy, 2002). However, over the decade, traditional bullying has rapidly evolved into a new form through technological means. Cyberbullying, a term coined by Bill Belsey (2004), has fast become a global phenomenon that warrants immediate attention. Studies put together have evidenced that cyberbullying is increasingly prevalent, especially in schools, as close to 25% of students have been found cyberbullying (Dehue, Bolman & Vollink, 2008; Li, 2006; Hinduja & Patchin, 2008). Worryingly, this trend is expected to worsen with increasing ubiquity of technological devices among students.

The adverse consequences of cyberbullying on victims have been well documented by many researchers. In fact, an emerging body of research has proposed that the impact of cyberbullying is even more detrimental than traditional bullying (Belsey, 2004; Scheneider, Gruman & Coutts, 2012). This is because researchers have found that cyber-victims, compared to traditional victims, have a significantly higher risk of drug overdose and suicidal tendency (Dooley, Pyzalski & Cross, 2009; Holt, Finkelhor & Kantor, 2007a). Considering the severity of the cyberbullying, an effective intervention program is not only necessary, but critical in the management of this prickly problem.

Accumulated findings have revealed that current cyberbullying intervention programs are predominantly victim-centric, which focus largely on the treatment process for cyber-victims, and how individuals can protect themselves from being cyber-bullied. Examples of such intervention programs include "Don't Suffer in Silence strategy" (DfES, 2002), "Prevention Curriculum for Cyberbullying" (Kowalski & Agatston, 2009), and "Four step process: Stop, Save, Block, and Tell" (Snakenborg, Van Acker, and Gable, 2011).

While the abovementioned victim-centric programs have its merits, its effectiveness in eradicating cyberbullying is largely limited by its failure to focus on the root of the problem, which is the cyberbullies themselves. Therefore, this highlights the need for a bully-centric intervention program that can effectively tackle cyberbullying head-on. Such bully-centric approach serves to complement existing victim-centric intervention programs in the effective management of the cyberbullying problem. More importantly, it also serves to treat the underlying causes of the cyberbullying behaviours, which could otherwise lead to other problematic behaviours as well.

Therefore, the objective of this review is to recommend a bully-centric cyberbullying intervention program, which focuses on both the preventive and treatment aspects. However, before the recommendations, it is imperative to have an in-depth understanding of the cyberbullying. Therefore, this review also aims to provide an overview of the cyberbullying problem, by focusing on four main aspects (a) definition of cyberbullying; (b) modus operandi of cyberbullying; (c) profiles of cyberbullies and (d) explanations of cyberbullying behaviours. These findings serve as the foundation for the recommendations of the intervention program.

Cyberbullying defined

To date, a standardised operational definition of cyberbullying has not been agreed on. Nonetheless, most of the existing studies have identified three major conditions to be met in order to define an action as cyberbullying. First - the malicious intent to harm. Second - impact on victims is more psychological and emotional in nature. Third - repetition of online behaviours. However, the 'repetition criterion' is debatable considering the 'permanence' nature of some types of technological means, such as website postings, that may be online for a long period of time. As such, even a single aggressive action, posting on the website, could be devastating enough to the victims due to the 'permanence' of the action. Besides, if the 'repetition criterion' is defined by the frequency the postings have been viewed or forwarded, then viewers are equally guilty of meeting the 'repetition criterion'.

Cyberbullying has been proposed to be an evolved version of traditional bullying (Li, 2005). However, it is presumptuous to assume that cyberbullying is a direct replacement of traditional bullying. This is because other researchers (e.g., Raskauskas & Stoltz, 2007; Ybarra & Mitchell, 2004a; Ybarra et al. 2007) have provided substantial evidence to suggest that cyberbullying can be more accurately seen as an extension of traditional bullying. In other words, technological means simply become an extension of 'traditional schoolyard' for bullying occurrences. As Li (2007) has aptly summed it "Cyberbullying is another form of traditional bullying using 21st century technologies".

Since cyberbullying has been argued to be an extension of traditional bullying, this suggests that the core characteristics of traditional and cyberbullying could display some similarities. Indeed, comparisons between studies have revealed that the intentions remain the same, as both forms of bullying are intended to humiliate, threaten, and to cause helplessness in victims (Strom & Strom, 2005). Furthermore, like traditional bullying, power imbalance also exists between cyber-victims and cyberbullies. However, power exists in different forms for both types of bullying. For traditional bullying, victims are usually physically weaker than their perpetrators. For cyberbullying, physicality plays a lesser role. Instead, victims are usually less technologically gifted than their perpetrators.

Modus operandi of cyberbullying

Tools

Cyberbullying behaviours are carried out via technological tools. The technological tools may include mobile phones, e-mail, website postings, instant messaging (IM), internet chat rooms, social networking sites and blogs. These tools, while seemingly innocuous, can be potentially misused for cyberbullying purposes. For instance, mobile phones and e-mail can be used to send threatening messages to harass the victims. Phishing can occur in internet chat rooms whereby victims can be tricked into sharing sensitive information, which then leads to the exploitation of that information (Englander, 2011). Social networking sites can similarly be misused as a platform for posting humiliating photos and videos of victims.

However, with technological advancement over the decade, it is likely that some technological tools are more popular for use than the others. Indeed, findings from Slonje and Smith (2008) have confirmed that cyberbullies are increasing switching to social networking sites and website postings, rendering tools like mobile phones and e-mail obsolete. It is evident that cyberbullies are getting increasingly sophisticated in their choice of technological tools, as social networking sites and website postings can comparatively lead to more psychological damage due to its permanence and publicly viewable contents.

Cyber-bystanders

Even though, cyberbullies are primarily responsible for the engagement of cyberbullying behaviours, the cyber-bystanders deserve mention as well. Technically, any individuals who witness the cyberbullying behaviours are considered cyber-bystanders (Cross, Monks, Campbell, Spears & Slee, 2011). Often, the mere presence of the cyber-bystanders is enough to exert some degree of emotional blow to the victims. This is mainly due to victim's awareness that their humiliation has been witnessed by the cyber-bystanders. Victims in their adolescence are especially vulnerable to such psychological traumatisation due to their heightened self-consciousness and preoccupation of 'imaginary audience' witnessing their victimisation (Elkind & Ginsberg, 2007).

Indifferent attitude among the cyber-bystanders also indirectly leads to the continuation of cyberbullying. The cyber-bystanders' indifference could be aptly explained by the bystander effect (Latane & Darley, 1968). Essentially, this social psychological phenomenon refers to situations where individuals' likelihood to help is in inverse relation to the number of bystanders. Explanatorily, the wide breadth of audience in the cyberspace leads to diffusion of responsibilities, and eventual indifference among the cyber-bystanders. While indifferent bystanders can play an indirect role in the continuation of cyberbullying, other bystanders could take on a more directive role by forwarding the malicious content to other people, even with the full understanding of its detrimental impact on the victims. Considering their direct contribution to the victims' harassment, the latter group of cyber-bystanders may also be considered cyberbullies.

Profiles of cyberbullies

This section aims to provide descriptions of the cyberbullies, in terms of their age, gender and typology. The profiling is useful to identify individuals who are at risk of engaging in cyberbullying behaviours. With this understanding, a targeted intervention can be developed to prevent and reduce the occurrences of cyberbullying behaviours.

Age

Based on existing literature, age is an important factor in cyberbullying behaviours (Kowalski & Limber, 2007; Moody, 2001). However, current studies have revealed mixed findings, and have yet to reach a consensus on which age group is most vulnerable to develop cyberbullying behaviours. Slonje and Smith (2008) have proposed that upper primary school students are most at risk to cyberbully. Tokunaga (2010) further extended on their finding, and proposed that middle school students are equally at risk to engage in cyberbullying behaviours. Contrastingly, Beale and Hall (2007) provided anecdotal evidence to suggest that high school students are most likely to cyberbully. Several reasons that can be attributed to explain these mixed findings include differences in sample population, size and research design. Despite the mixed findings, one shared conclusion among the researchers is that students, in their adolescence, are most at risk to develop cyberbullying behaviours.

There are several explanations why students are most vulnerable to develop cyberbullying behaviours, they are as follows. First - students in their adolescence are typically less socially mature than adults (Frisen, Jonsson, Persson, 2007). As a result of their lack in social maturity, they have a higher tendency to engage in socially irresponsible behaviours, like cyberbullying. Second - young students tend draw a less distinctive line between aggressive and non-aggressive behaviours (Monks & Smith, 2006). Thus what is considered as cyberbullying behaviours may be deemed otherwise. This offers a possible explanation for the Inadvertent's cyberbullying behaviours. Third - cyberbullying behaviours are possible mechanisms for students to explore the relatively unchartered concept of social hierarchy in schools (Pellegrini & Bartini, 2002). Exploitation of the digital tools, which allows students to dominate others, strengthens their quest to climb up the social hierarchy in schools. Fourth - students in their adolescence are especially susceptible to peer influence (Lounsbury, 2000). Explanatorily, students who are acquainted cyberbullies are more likely to be influenced into cyberbullying as well. As such, students' cyberbullying behaviours can be due to negative peer influence. Last - young students are particularly prone to the 'invulnerability fallacy' (Sternberg, 2003). As a result, they tend to exaggerate their self-importance, and believe that they can do whatever they want, and get away with it. This could explain the students' tendency to engage in reckless and self-asserting acts, like cyberbullying.

Gender

Current literature, that attempts to investigate which gender is more at risk developing cyberbullying behaviours, have revealed mixed results. Empirical studies put together have presented two main themes of findings.

One group of researchers (e.g., Blair, 2003; Kowalski, Limber & Agatston, 2008; Mason, 2008; Thorp, 2004) have postulated that cyberbullying behaviours are more commonly observed in females. This could be because females are more inclined to engage in covert (Graig, Pepler & Atlas, 2000) and psychological form of bullying (Huang & Chou, 2010). Similarly, cyberbullying allows females to operate covertly and to inflict psychological pain on their victims. Besides, females may feel more comfortable executing their bullying behaviours via online mediums than males, as they tend to communicate more regularly on mediums like internet chat-rooms and e-mail (Thorp, 2004).

On the contrary, another group of researchers (e.g., Bhat, 2008; Dilmac, 2009; Li, 2005; Wang, Iannotti & Nansel, 2009) proposed that males are more likely to display cyberbullying behaviours. Since traditional bullying, which has been predominated by males, is widely supported to be the precursor to cyberbullying (Raskauskas & Stoltz, 2007), it would be logical to assume the predominance of males in cyberbullying as well. Furthermore, males' preference in covert tools like the internet for their bullying behaviours can be explained by their preferred communication style, which is typically less direct (Tannen, 1990).

Based on the above findings and explanations, it is clear that cyberbullying behaviours are neither observed only in males or females. Rather, it is evident that both genders are equally capable of engaging in cyberbullying. Therefore, it is suggested that both males and females are equally at risk of developing cyberbullying behaviours. Interestingly, the finding's revelation of an equal involvement of females in cyberbullying is in sharp contrast to existing bullying literature that evidenced the insignificance of female involvement (Olweus, 1993; Whitney & Smith, 1993).

Typology

Aftab (2008) has categorised the cyberbullies into four main types based on their motives and characteristics. They are the (a) Inadvertent; (b) Vengeful Angel; (c) Mean Girls and (d) Power Hungry. The categorisation of the cyberbullies is useful to understand the motivational and operational differences among the cyberbullies (Trolley, Hanel & Shields, 2006). With this understanding, intervention program can be tailored for different types of cyberbullies based on their specific needs. However, Aftab's typology should be used with caution as it is more anecdotally than empirically supported.

The inadvertent. This group of cyberbullies is also commonly known as the indirect cyberbullies (Deno, 2003). Aftab argued that this group is the most innocuous among the cyberbullies because they usually lack the malicious intent to victimise others, and their online behaviours are more reactive than proactive. As a result, they usually do not perceive themselves as cyberbullies (Aftab, 2008; Trolley, Hanel & Shields, 2006).

The Vengeful Angels. This group usually consists of individuals who are past victims of traditional and/or cyberbullying (Sontag, Clemans, Graber & Lyndon, 2011). As the name suggests, the Vengeful Angels are typically vengeful as they seek to exert revenge for their past victimisation. This may explain the reactive nature of their online behaviours. Their past victim status are often due to their lack in physicality and popularity. Consequently, their offline revenge plan may be hindered by these disadvantages. However, the irrelevance of such disadvantages in the cyberspace setting allows the Vengeful Angel to carry out their revenge plan with ease.

Mean Girls. As the name suggests, this group of cyberbullies operate primarily in an all female ensemble. They typically engage in cyberbullying for 'fun', and feed by admiration from group members and the audience (Aftab, 2008).

Power Hungry. Due to their advanced technological skills, the Power Hungry's unfulfilled sense of power is often satisfied and effectuated by their technological ability to manipulate and exploit the cyberspace, like creating of online hate polls and hacking into their victim's online account. As their sense of self-importance depends on the positive reactions of other people, the Power Hungry tend to brag about their cyberbullying involvement. Similar to Mean Girls, this group is proactively aggressive and their aggressive online behaviours are typically pre-mediated and controlled. Other than cyberbullying, this group usually engages in traditional bullying to satiate their incessant need for power. Cyberbullies who also engage in traditional bullying are commonly referred to as combined aggressor (Smith, Mahdavi, Carvalho & Tippett, 2006).

Explanations of cyberbullying behaviours

The rather unexplored scholarly field of cyberbullying has offered very limited explanations for cyberbullying behaviours. The current cyberbullying explanations include (a) succorance (b) empathy (c) general strain theory. However, it is presumptuous to assume that there are no other explanations for cyberbullying behaviours. Due to the close similarities of both traditional and cyberbullying, in terms of the motivation and intention (Hinduja & Patchin, 2008; Strom & Strom, 2005), existing bullying theories may offer valuable insights and contributions in the understanding of cyberbullying behaviours as well. Four plausible cyberbullying explanations include (a) biopsychological theory; (b) differential association theory; (c) social learning theory and (d) labelling theory. This section serves to discuss and evaluate both the current and plausible explanations of cyberbullying behaviours.

Current explanations

Succorance. Succorance refers to the search for care, support and affection from other people (Brody, 1997). It was proposed by Dilmac (2009) to be a psychological need that cyberbullies crave. A separate study conducted by Walker, Sockman and Koehn (2011) has further confirmed the positive relationship between succorance and cyberbullying behaviours. Despite the empirical evidence, both research studies failed to fully explain the relationship between succorance and cyberbullying. One possible explanation is that individuals whose affection needs are not met offline turn to cyberbullying in order to garner support and attention from the cyber-bystanders. In other words, these individuals may turn to cyber-bystanders for attention and support via their cyberbullying behaviours.

However, the above proposed explanation contradicts the finding of Vandebosch and van Cleemput (2008) who concluded that attention from other people will, in fact, dissuade cyberbullying behaviours. This is because the cyberbullies would prefer to carry out their acts in an anonymous manner without attracting attention. Considering this contradictory finding, it appears that the succorance explanation does not apply to all types of cyberbullies. Since the Power Hungry and Mean Girls' cyberbullying behaviours are mainly motivated by the positive reactions from the audience (Ybarra & Mitchell, 2004), the succorance explanation is seemingly more applicable for these two types of cyberbullies.

Empathy. Broadly, empathy refers to the comprehension and sharing of other people's feelings (Epstein, 1972). Current research studies and meta-analyses have strongly supported the negative relationship between cyberbullying and empathy level (Ang & Goh, 2010; Nickerson, Mele & Princiotta, 2008; Steffgen, Konig, Pfetsch & Melzer, 2011).

A separate study conducted by Aoyama, Brak and Talbert (2011) extended on this finding by proposing that de-individuation could also explain cyberbullying behaviours. De-individuation is defined as the lessening of normative behaviours and cognitive decision making which lead to undesirable outcome (Festinger, Pepitone & Newcomb, 1952). Due to the covert and distant nature of the cyberbullying medium where users' abilities to witness the impact of their action are compromised, users may grossly underestimate the adverse impact of their behaviours (Suler, 2001), which eventually results in deregulated behaviours, like cyberbullying. Englander and Muldowney (2007) have aptly described this phenomenon as the "online disinhibition effect".

The Reduced Social Cues (RSC) model proposed by Siegel and McGuire (1984) complements the explanation suggested by Aoyama et al. (2011). The RSC model put forth that technological use result in less direct feedback of the user's online behaviours. This could then result in the deficiency in empathy and sensitivity level of the users, and eventually causing deregulated online behaviours.

In sum, the explanation offered by Aoyama et al. (2011) complemented by the RSC model, assumes that people unintentionally cyberbully due to their attenuated empathy level caused by the lack of insight of their online behaviours. However, a study conducted by Juvonen and Gross (2008) contradicted this explanation. In their study, close to two thirds of the cyberbullies understood what they were doing, and the consequences of their actions. Their finding clearly shows that cyberbullying behaviours can be intentionally executed with the full understanding of the detrimental effects. The researchers further explained that people with low empathy level engage in cyberbullying behaviours due to the anonymity afforded by the cyberspace. Evidently, the 'lack of insight' explanation by Aoyama et al is not applicable to Juvonen and Gross's finding.

The abovementioned empathy explanations offer two distinctive views on why people cyberbully. Perhaps, the explanation offered by Aoyama et al. (2011) is more appropriate to explain the cyberbullying behaviours of the Inadvertent. This is because, unlike other cyberbullies, the Inadvertent often lack the malicious intent, and their unintentional behaviours could be most aptly explained by their lack of insight. The second explanation offered by Juvonen and Gross could be more useful in explaining the behaviours of the Vengeful Angels. Unlike the other cyberbullies, this group is most likely to understand the ramifications of their cyberbullying behaviours due to their own victimisation experience. The covert and anonymous nature of the cyberspace makes it especially enticing for this group to execute their revenge plan.

Jolliffe and Farrington (2004) has further suggested the differential role that cognitive and affective empathy play in cyberbullying. Briefly, cognitive and affective empathy refers to the ability to understand (Hogan, 1969), and experience (Epstein, 1972) other people's emotions respectively. Studies conducted by Feshbach (1987) and Shechtman (2002) have displayed strong evidence to suggest that affective empathy plays a more influential role in traditional bullying. On the contrary, studies conducted by Crick (1995) and Joinsen (1998) may have suggested that cognitive empathy play a more important role in cyberbullying. This is because their findings have revealed that people with lower level of cognitive empathy are more likely to engage in relational and indirect bullying. Since cyberbullying can also be classified as relational or indirect aggression (Berger, 2007), it could be argued that lower cognitive empathy contributes more significantly to cyberbullying behaviours. However, this argument is contested by Almedia, Marinho, Esteves, Gomes and Correia (2008) as they suggested that both types of empathy play an equally important role in cyberbullying.

Ang and Goh (2010) took these issues into consideration. Their results supported the suggestion put forth by Jolliffe and Farrington (2004) that both cognitive and affective empathy play a differential role in cyberbullying behaviours. However, the argument that low cognitive empathy is more influential in explaining cyberbullying behaviours is only partially supported. This is because the argument is found to be only applicable for the male subjects. In their study, more cyberbullying behaviours were observed in male subjects with low level of cognitive empathy, regardless of their affective empathy level. By the same token, less cyberbullying behaviours were observed in male subjects with high level of cognitive empathy, also regardless of their affective empathy level. Clearly, low cognitive empathy plays an influential role in explaining cyberbullying behaviours in male subjects. However, for female subjects, low level of cognitive empathy does not always lead to more cyberbullying behaviours. This is because the repercussion of low level of cognitive empathy can be buffered by their high level of affective empathy. The female subjects are more likely to display cyberbullying behaviours only when their affective empathy level is low. Clearly, low affective empathy plays a more significant role in explaining females' cyberbullying behaviours.

General strain theory. The general strain theory was proposed by Agnew (1992) to explain deviant behaviours in youths. Even though this theory was not originally developed to explain cyberbullying, several researchers (e.g., Hay, Carter & Meldrum, 2010; Moon, Morash, McCluskey & Hwang, 2009; Rivituso, 2012) have attempted to adopt the general strain theory as a theoretical framework to explain cyberbullying behaviours. According to Agnew (1992), three types of strains that an individual may confront include (a) failure to attain positive goals; (b) removal of positive stimuli and (c) exposure to negative stimuli. These strains subsequently give rise to negative emotions, and if unresolved, lead to deviant behaviours. Put another way, negative emotions mediate the positive relationship between strains and deviancy. Patchin and Hinjuja (2011) extended on Agnew's proposition by suggesting that cyberbullying behaviours could reduce the negative emotions of individuals confronted with strains. As a result, individuals faced with strains may find cyberbullying appealing as such behaviours could provide some form of psychological relief for their frustrations (Ma, 2001).

In a study conducted by Hay et al. (2010) to investigate the applicability of general strain theory in explaining cyberbullying behaviours, it was found that the general strain theory is more applicable in explaining males' cyberbullying behaviours. This is because unlike females who tend to exhibit internalising response to strains, like suicidal ideation (Agnew & Brezina, 1997; Hay, 2003; Piquero & Sealock, 2004), males are more likely to display externalising responses. Furthermore, males tend to possess less positive coping skills and social resources compared to their female counterparts. As a result, males are more likely to respond to their strains in a maladaptive manner, like cyberbullying.

In another similar investigation study conducted by Patchin and Hinduja (2011), the results of their study revealed that subjects who reported strain were more likely to display cyberbullying behaviours. Additionally, subjects who reported negative emotions like angriness and/or frustration were also more likely to cyberbully. However, unlike Agnew's proposition, their study revealed that the negative emotions did not play any mediating role in the positive relationship between strain and cyberbullying behaviours. Instead, both strain and negative emotions, independent of each other, directly influence cyberbullying behaviours. One possible reason why Patchin and Hinduja failed to demonstrate the mediating role of negative emotions could be due to the cross-sectional nature of their data collection, where the variables of interest were all measured at the same time. Individual differences are harder to control for in a cross-sectional study (Mann, 2003). As a result, the desired relationship between the variables of interest may not materialise as expected. Similar to the findings of Patchin and Hinduja, an earlier study conducted by Mazerolle (1998) also failed to demonstrate the mediating role of anger in the relationship between strain and deviant behaviours. Instead, both strain and anger were found to be independently related to the deviant behaviours.

In a separate but related study, not only did Moon et al. (2009) fail to demonstrate the mediating role of negative emotions in the relationship between strain and cyberbullying behaviours, they also failed to find any significant relationship between negative emotions and cyberbullying behaviours. One possible explanation could be because the negative emotions measured in their study were trait-based rather than situational-based. Compared to situational-based emotions, trait-based emotions play a significantly less directive role in an individual's externalising behaviours (Mazerolle, Piquero & Capowich, 2003). As such, this could account for the insignificance in the relationship between the trait-based negative emotions and the cyberbullying behaviours.

In sum, Patchin and Hinduja (2011) have demonstrated that negative emotions, similar to strains, can independently explain cyberbullying behaviours. These findings were similar to the findings of Mazerolle (1998). Contrarily, Moon et al. (2009) have evidenced that negative emotions play no part in explaining cyberbullying behaviours at all. Based on the above findings, it is clear that strains do contribute to cyberbullying behaviours. However, the role of negative emotions is less clear. Therefore, strains, rather than general strain theory, could be a better explanation of cyberbullying behaviours.

Even though Patchin and Hinduja (2011) have demonstrated that cyberbullying is a result of strains, they have also conducted other similar studies that demonstrated that cyberbullying could also be a source of strain by itself. In their earlier studies (Patchin &Hinduja, 2007, 2010), they found that students who had been cyber-bullied were significantly more likely to display cyberbullying behaviours, particularly towards their perpetrators. Other researchers like Hoobler (2010) and Rivituso (2012) had revealed similar findings. Ybarra and Mitchell (2004) proposed that cyber-victims who turn into cyberbullies, commonly known as Vengeful Angel, tend to have stronger needs to sanctions. Their retaliatory responses, in the form of cyberbullying behaviours, are ways for them to right the "wrongs" they suffered. Moreover, the protection and anonymity offered by the cyberbullying medium make it especially enticing for them to carry out their revenge plan.

However, Dombeck (2007) argued that not all individuals who experience strain of cyberbullying will turn into cyberbullies. Dombeck proposed the social maturity theory to support this argument. According to socially maturity theory, socially mature individuals, usually adults, deal with confronted strains in a socially responsible manner that is accepted by the society. As such, socially mature individuals who are cyberbullied are more likely to respond in a socially acceptable manner, like reporting to the authority. Suggestively, social maturity can be regarded as a protective factor against cyberbullying behaviours. This may also explain why cyberbullying behaviours are usually engaged by youths, who are usually less socially mature (Kowalski & Limber, 2007; Slonje & Smith, 2007).

Plausible explanations

Due to the limitedness of current literature in offering explanations of cyberbullying behaviours, this section serves to introduce plausible explanations of cyberbullying behaviours. Four major theories, which have been commonly applied in the explanations of traditionally bullying, will be adopted in the explanations of cyberbullying behaviours. These four theories include the biopsychological theory, differential association theory, social learning theory and labelling theory. The theories' applicability in explaining the cyberbullying behaviours will also be discussed and evaluated.

Biopsychological theory. According to Gray's biopsychological theory (1981), bullying behaviours can be explained from a biopsychological perspective. Broadly, Gray's theory hypothesises that individuals who display bullying behaviours have a higher Behavioural Activation System (BAS), characterised by lower level of cortical arousal and autonomic reactivity. Individuals with higher BAS favour activities that are more arousing and exciting in order to meet their higher behavioural activation threshold. Similarly, cyberbullies could be argued to have a higher BAS than non-cyberbullies. Cyberbullying could be seen as behaviours for them to satisfy their higher behavioural activation threshold. Gray (1991) further found that individuals with high BAS tend to be more impulsive and display less inhibition, characteristics which are closely similar to the behaviours exhibited by the cyberbullies (Steffgen et al., 2011).

In an extension to Gray's earlier work, an empirical study conducted by Keltner, Gruenfeld and Anderson (2003) revealed that both indirect and direct bullies displayed poor response to punishment. Upon further investigation, the researchers concluded that bullies have a lower Behavioural Inhibition System (BIS), characterised by lower level of inhibition and lower sensitivity to punishment. Analogously, cyberbullies' uninhibited behaviours (Bauman, 2009) and the ineffectiveness of punishment on them (Steffgen & Russon, 2002) could also imply their lower BIS.

While Gray's biopsychological theory is useful in explaining cyberbullying behaviours in terms of biopsychological traits like high BAS and/or low BIS, the theory does not offer a holistic explanation as it lacks a 'social' perspective. Using only the biological and psychological perspectives in the explanation leads to the inherent assumption that cyberbullying tendencies are inborn, and may have discounted the importance of external influence which could be equally important in the explanation of cyberbullying behaviours. Furthermore, Gray's theory also does not account for how cyberbullying behaviours are acquired. These limitations may be addressed by the differential association theory.

Differential association theory. The differential association theory developed by Sutherland (1947), which primarily focuses on the influences of external social context, proposes deviant behaviours to be products of social interactions. Sutherland's theory has been well-supported by several empirical studies (e.g. Espelage, Bosworth, & Simon, 2000; Pepler, Jiang, Craig & Connolly, 2008) in explaining bullying behaviours. Results from these studies have evidenced that individuals who are associated with bullies are significantly more likely to bully as well. Similarly, it is possible that individuals who are associated with cyberbullies are also more likely to cyberbully. In an extension to Sutherland's theory, Matsueda and Anderson (1998) further found that individuals, who are associated with intimates who have positive attitudes towards bullying, can be influenced into bullying also. This may indicate that even mere association with intimates who attitudinally endorse cyberbullying, but not necessarily cyberbullying, can lead to an individual's cyberbullying behaviours.

The differential association theory addresses two issues which were not touched on by Gray's biopsychological theory - the role of external influence in cyberbullying behaviours, and how cyberbullying behaviours are acquired. However, the differential association theory is also not without its limitations. This is largely because the theory reduces an individual to a mere object who is not capable of making an independent decision, but only reacts to stimuli presented. Notably, individual differences are not accounted for as the theory does not explain why some people do not turn into cyberbullies despite being closely associated with cyberbullies or intimates who endorse cyberbullying. Furthermore, the motivations of the cyberbullying behaviours are not explored. These limitations may be compensated by the social learning theory.

Social learning theory. The social learning theory was developed by Akers and Burgess (1966) as a revision to Sutherland's differential association theory. Like Sutherland, Akers and Burgess posited that individuals learn bullying behaviours from their social environment. However, unlike Sutherland's theory, the social learning theory does not assume that individuals automatically acquire bullying behaviours by mere association. Instead, Akers and Burgess argued that bullying behaviours are learned via operant conditioning. One particular tool of operant conditioning is positive reinforcement. In the case of cyberbullying, positive reinforcement may include bestowment of status and attention by peers (Campbell, 2005). Peer acceptance is another form of positive reinforcement suggested by Snakenborg et al. (2011). Based on their findings, it was revealed that that close to half of their subjects started and continued their cyberbullying behaviours out of fear of being isolated by their peers. Cyberbullying behaviours simply become a mean for these subjects to gain peer acceptance. Put another way, peer acceptance serves as a positive reinforcer for the initiation and maintenance of an individual's cyberbullying behaviours.

As seen, the social learning theory addresses the limitations of the differential association theory. Unlike differential association theory, the theory put forth by Akers and Burgess has demonstrated that positive reinforcement, rather than mere association, is necessary for the learning of cyberbullying behaviours. Importantly, the social learning theory has also illustrated that a person is capable of making an independent decision to cyberbully. The self-made decision to cyberbully may be motivated by its rewarding consequences, like peer acceptance and attention.

While the social learning theory is useful in explaining how positive reinforcement, like peer acceptance, leads to the initiation of cyberbullying behaviours. The theory does not fully account for other ways individuals can be inducted into cyberbullying. For example, some individuals, particularly the traditional bullies, may turn to cyberbullying as a result of discriminatory labels attached on them. The labelling theory serves to elaborate on this example.

Labelling theory. Labelling theory, also known as social reaction theory, was developed by Tanneubaum (1938) to explain how labels tagged to individuals can lead to deviant behaviours. Similarly, Tanneubaum's theory may explain how labels tagged to traditional bullies can lead to cyberbullying behaviours. Since bullying labels are often attached to the individuals rather than their bullying behaviours (Cassidy, Jackson & Brown, 2005). This may cause the labelled individuals to believe that the bullying problems originate with them, which eventually results in self-fulfilling prophecy (Collins, 1991; Polakow, 1993; Sohbat, 2003). As a result, the individuals behave in manners consistent to the bullying labels. The ubiquitous cyberbullying medium simply becomes another convenient platform for these labelled individuals to assume their labels. Moreover, the labels attached to bullies, often permanent, may lessen their will to alleviate their bullying behaviours (Corrado, Cohen & Odgers, 2003). This could consequently exacerbate their bullying behaviours and give rise lead to cyberbullying behaviours as they become increasingly resistant to positive changes.

Findings from Twemlow, Fonagy, Sacco and Brethour (2006) may have offered another possible explanation why traditional bullies turn into cyberbullies. In their longitudinal study, it was found that most of the participants who exhibited direct bullying behaviours during the initial measurement were equally likely to display indirect bullying behaviours five years later. This could suggest that bullying, in any forms, is an innate attitudinal characteristic, which entraps individuals in the bullying role. The cyber-setting merely becomes another platform for these individuals to thrive in their bullying roles.

Recommendations of intervention program

With a clearer understanding of the cyberbullying problem, this section aims to recommend a bully-centric intervention program that can effectively tackle cyberbullying. A natural approach is to target the cyberbullies in order to treat their cyberbullying behaviours. Such approach, considered the tertiary intervention, is important, but should not be the only focus of the intervention program. This is because the tertiary intervention approach only focuses on the treatment aspect, but not the prevention aspect, which is equally, if not more important in the long run. Thus, this highlights the need for the implementation of primary and secondary interventions that also focus on the prevention aspect. Since majority of the cyberbullies are students, and cyberbullying behaviours are often seen as an extension of schoolyard bullying (Li, 2006; Raskauskas & Stoltz, 2007), schools are natural places for the implementation of the intervention work. Therefore, this section aims to recommend an effective intervention program that not only to treat, but also prevent cyberbullying behaviours in schools, through the implementation of primary, secondary and tertiary intervention work. Such multipronged approach may prove to be effective in the management of cyberbullying behaviours (Stevens et al., 2000; Vreeman & Carroll, 2007).

Primary and secondary intervention

The primary and secondary intervention approach will be applied to all students, inclusive of both the cyberbullies and non-cyberbullies. The main objective of this approach is to prevent the precipitation of cyberbullying behaviours in the first place among the non-cyberbullies. However, the cyberbullies benefit as well. This is because such preventive approach complements and enhances the efficacy of their treatment programs, by containing the recidivism. In addition, this intervention approach will be equally applied to both male and female students since accumulated findings have not empirically concluded the predominance of either gender in cyberbullying. Broadly, the primary and secondary intervention focuses on school regulations, practices, curriculum and teachers.

School regulations. Since cyberbullying behaviours, like traditional bullying behaviours, are often detrimental in nature, a natural tendency for schools would be to implement harsh regulations to tackle cyberbullying. An example of such regulations includes the banning of technological devices in school. However such punitive regulation is only a stopgap measure, at best. This is because students are still able to engage in cyberbullying outside school.

When students feel restrained by the punitive regulations, psychological reactance can even result (Brehm & Brehm, 1981) whereby they may intentionally engage in the discouraged behaviours, like cyberbullying, so as to demonstrate their independence. Pelaez, Field, Pickens and Hart (2008) have also demonstrated the deleterious effect of an environment governed by harsh regulations. Their study has revealed that children who are brought up in a punitive environment, characterised by high control and low warmth, tend to be more proactively aggressive, and their rebellion often takes the form of voracious acting-out behaviour. Evidently, school environments that are operated punitively and harshly can counterproductively breed cyberbullying behaviours in students.

Considering the ineffectiveness and detrimental effect of punitive school regulations, schools should consider adopting a non-punitive approach in the preventive effort against cyberbullying, for instance, the encouragement of responsible usage of technological devices. Such non-punitive approach is advantageous as it lessens the probability of psychological reactance in students. Besides, in a non-punitive school environment where fear of punishment can be put to rest (Jackson, 1964), cyberbullies are more likely to disclose their cyberbullying involvement, which is perhaps the most important step in the initial treatment of the cyberbullies.

School practices. However, even in a non-punitive environment, voluntary admission of cyberbullying involvement may be difficult considering the grave nature of most cyberbullying offences. As such, identification of the cyberbullies is a key process in the intervention effort against cyberbullying. Labelling is a useful method in the identifying the cyberbullies. Despite its usefulness, current literature has suggested that labelling could be potentially detrimental as well. As explained earlier by the labelling theory, labelling of the cyberbullies may potentially lead to the continuation of their cyberbullying behaviours as they become less motivated for positive changes. In addition, labelled cyberbullies are more likely to be discriminated by teachers and other students which could further exacerbate their initial cyberbullying behaviours.

Perhaps, rather than labelling the individuals, labelling of their cyberbullying behaviours may be significantly less impactful. This is because attention would be paid more on the problematic behaviours of the individual, rather than the "problematic" individual. This may consequently lead to less discrimination, which then strengthens the individual's will to overcome their problematic cyberbullying behaviours. However, the labelling of an individual's cyberbullying behaviours is not without its problems. The absence of a standardised operational measurement of cyberbullying behaviours has made labelling of the behaviours difficult. Without a standardised measurement, teachers may inevitably display individual differences in their interpretation of cyberbullying behaviours. As a result, labels may not be uniformly applied to all students, which may then raise the question of fairness.

As explained, labelling of either the cyberbullies or their cyberbullying behaviours can be equally problematic. But it appears that labels can be more detrimental when they are applied to the cyberbullies themselves. Despite the potential adverse impact of labels, labelling of the students' cyberbullying behaviours is still relatively useful to sieve out students who may need more targeted intervention work than others. Therefore, this highlights the need for schools to develop a standardised tool for detection of cyberbullying behaviours in students. With a standardised measurement method, labelling of the students' cyberbullying behaviours would be fairer.

Punishment. Upon identification of the cyberbullies, schools will then have to decide on the necessity of punishment. However, even if punishment is deemed necessary, enforcement of punishment can be tricky for three reasons. First - schools may not have the jurisdiction to enforce punishment for cyberbullying that are conducted away from schools. Second - punishment may not be fair to some cyberbullies, like the Inadvertent, whose cyberbullying behaviours are unintentional. Third - even if the cyberbullies have the malicious intent, the lack of physical wounds in their victims may make punishment unjustified.

Furthermore, punishment may not necessarily be the most ideal form of response to cyberbullying behaviours. As Gray's biopsychological theory have suggested, cyberbullies are usually less sensitive to punishment due to their lower BIS. As such, punishment does little to deter their cyberbullying behaviours. Even though harsh punishment may lead to immediate compliance of the cyberbullies, its effect in permanently eliminating their cyberbullying behaviours is limited. This is because harsh punishment usually does not result in long-term attitudinal change which is the key to long-term behavioural change (Aronson & Carlsmith, 1963). Besides, harsh punishment via coercive means, like confiscating of cyberbullies' technological devices, does not allow cyberbullies the chance to autonomously change their cyberbullying behaviours, and could even lead to several adverse impacts, like anxiety disorder and low self-esteem (Horn, Joseph & Cheng, 2004; Roland, 2002).

Considering the ineffectiveness and detrimental impact of harsh punishment, milder form of punishment is suggested instead. The effectiveness of mild punishment has been well acknowledged by many researchers (e.g., (Friedman & Brinker, 2001; Meares, Kahan & Katyal, 2004; Pfeffer & Lawler, 1980). Examples of mild punishment include verbal warnings and time-out. Like harsh punishment, mild punishment can equally lead to the immediate compliance of the cyberbullies. Unlike harsh punishment, mild punishment can potentially lead to more lasting attitudinal and behavioural changes. As Aronson and Carlsmith (1963) explained, in the absence of tangible external justification, like avoidance of harsh punishment, cyberbullies who receive mild punishment will resort to internal justification to account for their immediate compliance. Internal justification can be in the form of positive attitudinal change, which then results in long-term behavioural change. Evidently, mild punishment can be even more effective than harsh punishment in the eradication of problematic cyberbullying behaviours in the long run.

Curriculum. Educational curriculum can be an important component in the intervention effort against cyberbullying behaviours. Importantly, the educational curriculum should aim to inculcate an anti-cyberbullying attitude in students. This is because, as mentioned previously, Matsueda and Anderson (1998) have provided strong evidence to suggest that even attitudinal endorsement of cyberbullying can eventually lead to actual cyberbullying behaviours.

Educational materials could touch on digital citizenship and netiquette. Such "information quotient" curriculum serves to overcome barriers of ignorance and also raises awareness of the destructive effects of cyberbullying. This may be especially relevant to the Inadvertent as their cyberbullying behaviours, often unintentional, are resulted from their lack of insight of their online behaviours (Aoyama et al., 2011). Therefore, it is important for the educational curriculum to reduce the students' tendency to trivialise their potentially destructive behaviours. Other than information providing, feedback could also be provided for the students' usage behaviour of their devices. Such information cum feedback approach may lead to deeper processing of the educational materials, which then leads to a higher probability of actual attitudinal and behavioural changes (Wakefield, 2010).

Other than the Inadvertent, another group of students that the curriculum should focus on is the cyber-bystanders. Indifferent cyber-bystanders are no less culpable than cyberbullies. This is because, to certain extent, their nonchalant attitude contributes to the continuation of the cyberbullying behaviours. Since the bystander effect, which was discussed earlier, serves to explain the cyber-bystanders' indifference, the curriculum should also emphasise on how cyber-bystanders can overcome the bystander effect.

One plausible strategy that could be imparted in the curriculum is the "befriending" intervention strategy which was developed by Stevens et al. (2000). This strategy focuses on enhancing the bystanders' understanding of the victim's traumatic experience, and recommends several concrete steps that bystanders can take to help the victims. Two key measures recommended are the reporting of the cyberbullies, and stopping the dissemination of derogatory materials.

However, identifying the cyber-bystanders would be difficult, and almost unpractical. This is because any students who witness or are knowledgeable of cyberbullying behaviours can be technically considered cyber-bystanders. The identification of the cyber-bystanders is further complicated by the covert nature of the cyberbullying medium. Perhaps, a more pragmatic approach would be to impart the "befriending" strategy to all students since all of them are potential cyber-bystanders. Despite the strong endorsement of the "befriending" strategy by current bullying intervention programs (see Cowie & Hutson, 2005; Crothers, Kolbert & Barker, 2006; McElearney, Roosmale, Scott & Stephenson, 2008), the lack of literature to support its efficacy in counteracting the cyber-bystanders' bystander effect has made future empirical testing necessary.

Extracurricular activities. In schools, strains that students are confronted with include academic stress and peer conflict. These strains then produce urges for "corrective behaviours", and negative emotions like frustrations (Agnew, 1992). For some students, the "corrective behaviours" may be in the form of cyberbullying. However, such deviant behaviours are avoidable if there are positive avenues in schools where they can engage in their "corrective behaviours". An example of such positive avenue is the extracurricular activities where students can, at least, temporarily disengage from the strains, and vent their frustrations in a constructive manner. Besides, the extracurricular activities serve to occupy students' time after school that could otherwise be spent on cyberbullying activities.While some students cyberbully in response to strains, some may also cyberbully because of their high BAS. To the latter group of students, cyberbullying serves to satisfy their high behavioural activation threshold. Since these cyberbullies are primarily motivated by the "exciting" nature of cyberbullying, extracurricular activities that are equally or more "exciting" could serve to replace the students' cyberbullying behaviours.

Social control theory. The social control theory provides an additional theoretical framework to explain how the extracurricular activities can prevent cyberbullying behaviours in students. Broadly, the social control theory (Hirschi, 1969) hypothesizes that deviancy can be prevented if adolescents are closely bonded to pro-social groups. Suggestively, cyberbullying behaviours can be prevented if students are tightly bonded to their extracurricular activities groups that promote pro-social behaviours. An example of such group is the Community Involvement Program (CIP), where students gather to make positive contribution to their community.

In order to ensure that students are tightly bonded to their pro-social groups, CIP could be made compulsory in schools. The mandatory nature of CIP also ensures the students' commitment to the pro-social activities. Ideally, the bonds built between students over the pro-social activities reduce the probability of them cyberbullying each other. Furthermore, the resulted camaraderie between the students could lead to stronger identification with their pro-social groups. This has two positive implications. One - positive reinforcement of pro-social behaviours. Students who strongly identify with their groups are more sensitive and susceptible to reinforcements given by their pro-social groups (Kiesner, 2002). Therefore, students are more receptive of the pro-social behaviours reinforced by their fellow group members. Since prosociality is negatively correlated with aggression (Barber & Olsen, 1997), modeling of pro-social behaviours, similarly, can reduce cyberbullying behaviours. Two - positive conformity. Students who are closely bonded to groups that subscribe to pro-social activities are less likely to engage in antisocial activities, like cyberbullying, that are against their groups' pro-social values and could potentially strain their relationships with their identified group members. As a result, students are more likely to conform to the pro-social activities, which naturally exclude cyberbullying. This could potentially lead to long-term belief and endorsement of pro-social values. Plausibly, attitudinal endorsement of positive values serves as protective factor against cyberbullying behaviours. The two positive implications highlight the advantages of extracurricular activities, like CIP, in buffering against cyberbullying behaviours among students.

Teachers. Traditionally, teachers play a central role in bullying intervention programs (Crothers, Kolbert, & Barker, 2006). This is primarily due to their physical proximity with students and active presence in schools. However, in the case of cyberbullying, the roles that teachers can play are comparatively more restricted due to the discreet nature of cyberbullying behaviours, whereby the physical proximity of teachers with their students lends little advantage in the detection of cyberbullying behaviours. Moreover, the cyberbullying problem may be beyond the comprehension of teachers due to its youthful and intricate nature. As a result, teachers may be clueless in handling cyberbullying situations. Even if the teachers are aware of the cyberbullying problem, the absence of direct guidelines for teachers may result in teachers' inaction even when cyberbullying behaviours are detected or reported. This consequently leads to students' unwillingness in reporting of cyberbullying, which then further contributes to the exacerbation of the cyberbullying problem.

Despite the constraints teachers face, the importance of teachers in the intervention work against cyberbullying should not be discounted. There are two ways teachers can contribute to the intervention effort - satisfying the students' need for attention and support, and paying more attention to past-victims.

Results from Sourander et al. (2010) have clearly concluded that inattentive and unsupportive teachers tend to elicit cyberbullying behaviours in students. Using Dilmac's succorance theory as a basis for elaboration, when students' need for attention and support are not satisfied by their teachers, the students are more likely to turn to cyberspace to meet their needs. Attention and support rendered by the cyber-bystanders may eventually give rise to cyberbullying behaviours. Therefore, this leads to the suggestion that attention and support given by the teachers can lessen the probability of students fulfilling their psychological needs online. This reduces the possibly of cyberbullying behaviours.

While some students' cyberbullying behaviour are primarily influenced by the cyber-bystanders, other students' cyberbullying behaviours may be largely driven by their past experience of being traditionally and/or cyber-bullied. The latter group of students is commonly known as the Vengeful Angel (Aftab, 2008). Teachers can play a considerable role in preventing past-victims turning into Vengeful Angels. The first step that teachers should take is to identify these past-victims, and refer them for professional counselling. However, the roles of teachers do not stop there. Considering their advantageous physical proximity with these students, the involvement of teachers during and after the counselling sessions is equally important, as continued supervision on the teachers' part may help prevent recidivism in these students.

Tertiary intervention

The focus of the tertiary intervention is on the treatment and rehabilitation of the cyberbullies. Due to the targeted approach of the tertiary intervention, teachers may not necessarily have the relevant knowledge for its implementation. Instead, counsellors would be more appropriate for the implementation of the treatment due to their expertise. Two treatment approaches will be suggested. They are empathy training and anger management. However, before the treatment, motivational interviewing is useful in enhancing the cyberbullies' readiness for positive changes.

Motivational interviewing. Motivational interviewing (MI) is a widely endorsed counselling technique that helps to "motivate" people before the actual treatment process (Amrhein et al., 2003). MI is especially beneficial for cyberbullies who are unmotivated for treatment. Coercing of unmotivated cyberbullies to undergo the treatment programs would be unproductive due to their unready mental state. MI could psyche the cyberbullies for the treatment programs by helping them understand the adverse consequences of their behaviours, and pointing out the inconsistencies in their self-view with their cyberbullying behaviours (Britt, Hudson & Blampied, 2003). Treatment programs that target cyberbullies who are motivated to change would have a higher probability of achieving its full efficacy.

Empathy training. Broadly, empathy training helps to build the bullies' empathy level. It is achieved by instilling a general sense of positive feeling, coupled with an enhanced understanding of their victims' grievances (Milsom & Gallo, 2006). Such training has been proven to be successful in reducing bullying behaviours (Bjorkqvist, Osterman & Kaukiainen, 2000). However, its efficacy in reducing cyberbullying behaviours is less clear due to the lack of empirical testing. Nonetheless, the empirically supported negative relationship between empathy level and cyberbullying (Ang & Goh, 2010; Nickerson et al., 2008; Steffgen et al., 2011) suggests that empathy training can also reduce cyberbullying behaviours. Such training is most likely to benefit cyberbullies with low empathy level.

As discussed earlier, findings of Ang and Goh (2010) have evidenced that male's cyberbullying behaviours can be explained more by their low cognitive empathy level. The low cognitive explanation for male's cyberbullying behaviours can be further elaborated by Silfver and Helkama (2007) who have empirically concluded that male's moral judgement is directly dependent on their cognitive empathy level. In other words, males with lower empathy level are significantly more likely to have a lower moral judgement. Since morality is in inverse relationship with deviancy (Blackburn, 1988), individuals with lower moral judgement are more likely to engage in cyberbullying behaviours. Implicationally, empathy training for male cyberbullies should emphasize on cognitive empathy, which serves to enhance their moral judgement as well. With a higher cognitive empathy level and moral judgement, males' cyberbullying behaviours can be greatly reduced.

On the contrary, Ang and Goh (2010) failed to fully demonstrate the relationship between low cognitive empathy level and females' cyberbullying behaviours. This is because they have found that that low cognitive empathy level does not necessarily lead to cyberbullying behaviours in females. Ang and Goh's findings can be further explained by Lennon and Eisenberg (1987) who posited that females have a significantly higher affective empathy level than males. Consequentially, females' higher affective empathy level can advantageously buffer against their low cognitive empathy level. This illustrates why females' cyberbullying behaviours can be explained more by their lower affective empathy level. Implicationally, emphasis should be placed on building of affective empathy level for female cyberbullies. With a higher affective empathy level, females are less likely to display cyberbullying behaviours.

In sum, empathy training appears to be a promising treatment program for the cyberbullies. Existing findings have provided strong evidence that the efficacy of the empathy training could be further enhanced by placing extra emphasis on cog