Cyberbullying: A Systematic Literature Review on the Definitional Criteria

Nowadays, there are a lot of research being done regarding cyberbullying issues. However, there are debates among researchers due to inconsistencies in defining cyberbullying. The development of cyberbullying definition is difficult to establish because there are no agreements among the researchers on the similarities and differences between traditional bullying and cyberbullying. Cyberbullying tends to borrow the definition from traditional bullying and extends the definition by adding criterion which suits the cyber world, where it happens through the use of electronic communications. The definitional criteria of cyberbullying is important to establish the definition of cyberbullying that may help to understand the concept of cyberbullying in depth and differ the cyberbullying from traditional bullying in an aspect of its definition. The aim of this study is to classify the criteria of cyberbullying using the conceptual definitions developed by researchers from all around the world. This study is based on systematic literature review by using PRISMA flowchart and a table of analysis on the criteria of cyberbullying. The findings identified on the criteria of cyberbullying are intentional, repetition, power imbalance, anonymity, publicity, aggression, and electronic devices and media. As the technology development is become more advanced in a short time, the research on the criteria of cyberbullying should be updated by time in order to align with the technology advancement. Besides, the research on other areas in cyberbullying may be developed as this paper may help in depth understanding on the criteria of cyberbullying to other researchers.


Introduction
Cyberbullying definitions are generally based on traditional bullying definitions (Dehue et al., 2008;Kowalski & Limber, 2007;Menesini, 2012;Smith et al., 2008). In many cases, cyberbullying is defined based on Olweus's original traditional bullying definition Smith et al., 2008), whereby the author takes three main points to make 1) intent to harm, 2) repetition, and 3) interpersonal relationship with power imbalance (Olweus and Limber, 2010). Through broadened definition of traditional bullying, cyberbullying has been described as "an aggressive act or behaviour that is carried out using electronic means by a group or an individual repeatedly and over time against a victim who cannot easily defend him or herself" . From this perspective, cyberbullying is a deliberate misuse of power that happens by the use of information and communication technology (ICTs). Despite the definition of cyberbullying by Smith et al (2008) (or related ones) is commonly used in the sense of cyberbullying, a few of its definitional aspects are argued. There are two criteria that distinguish bullying from more general aggression such as intent to cause harm, which are repetition and power imbalance. These criteria can be seen as equally straightforward in terms of traditional bullying but more complex to apply in cyberbullying. Furthermore, several scholars have expressed concern that using traditional bullying concepts to describe cyberbullying could not result in adequate representations of cyberbullying (Slonje et al., 2013;Tokunaga, 2010). According to Corcoran, Guckin, and Prentice (2015), since cyber-based networking has its own distinct essence, it is impossible to apply traditional bullying criteria to cyber-based violence. Yet, in order to provide a useful and coherent body of information, it is critical to reach a consensus on the description of the phenomenon as an empirical concept and to make an attempt to evaluate cyberbullying in a bullying sense (Olweus & Limber, 2018). Other researchers with the same intention have strongly debated regarding the definitions, operationalisation of concept and measurements of cyberbullying (Corcoran, Guckin, & Prentice, 2015;Thomas et al., 2014;Waasdorp & Bradshaw, 2015;Zych, Ortega-Ruiz, & Del Rey, 2015b). Tran et al (2018) stated that the development of conceptual definitions is a challenge for all research fields, especially in context where researchers are exploring a new scientific problem such as cyberbullying. According to Slonje, Smith & Frisen, (2013) due to various difficulties in the criteria of repetition and power imbalance, identifying cyberbullying can be more complicated than defining traditional bullying. As a result, it is important for the study to return to the basic tenets for identifying and conceptualising cyberbullying behaviours in order to step into a new paradigm of bullying research. In addition, conceptualising cyberbullying is considered essential because cyberbullying can be present in a variety of ways and through a variety of networks (Kowalski, Giumetti, Schroeder, & Lattanner, 2014). Another ensuing issue here would be the cultural aspects. Since there are variations across different countries in the meaning and definition of cyberbullying , it is vital to reach a consensus about the definitional criteria for cyberbullying.

PRISMA Flowchart
The researcher used PRISMA flowchart and analysis of systematic review on the definitions of cyberbullying in order to come out with the criteria of cyberbullying for the methodology in this study (Figure 1).

Literature Search
The articles searched were published between 2010 until December 2020. The searching was performed using various electronic databases such as Scopus, Academia, Google Scholar, and Science Direct. The keywords used in the literature search were "cyberbullying", "online bullying", "internet bullying", "definitional", "definition", "criteria", "intention", "repetitive", "power imbalance", "aggression", "anonymity", "publicity", "electronic media" and "electronic device". The search string used by the researcher (("cyberbullying" OR "online bullying" OR "internet bullying") AND ("definitional" OR "definition" "criteria" OR "intention*" OR "repetitive" OR "repetition" OR "power imbalance" OR "aggression" OR "anonymity" OR "publicity" OR "electronic device" OR "electronic media")). The keywords are based on the search engines of each database and the terms used in the combinations in order to locate the maximum number of studies. The terms included with reference to the context (cyberbullying, traditional bullying, online bullying, internet bullying, internet aggression, electronic aggression, definitional, definition, criteria, intention, repetitive, power imbalance, aggression, anonymity, publicity), study population (child, adolescents and youth or young adults), and methodology (qualitative and quantitative), combined with the Boolean AND and OR-operator to search in the title or abstract.

Inclusion and Exclusion Criteria Inclusion
• Articles were written in English.Papers were published between 2010 to December 2020.
• The study mentioned the definition of cyberbullying used for the study.
• The articles mentioned the issues on definitional criteria of cyberbullying and discussed how each criteria of cyberbullying can develop the definition of cyberbullying. • The study must mention clearly the definitional criteria of cyberbullying for the researcher to record the criteria that has been used by the authors in study.

Exclusion
• Studies on cyberbullying in working environment or at workplace.
• Studies not matching quality criteria.

Search Result
The results from the search retrieved 2660 publications from the electronic databases. There were 450 records of publications screened where 303 records were excluded. After the abstracts were screened, 147 records of publications were retrieved and reviewed for eligibility. In the eligibility stage, there were 87 records of publications excluded because of stated reasons. 60 records of publications were included in this review and met the inclusion criteria for this study. Figure 1 provides a detailed information regarding the study selection.

Coding Strategy
After the process of identification of articles using the PRISMA flowchart, the method of this study then continued with the process of analysing the criteria of cyberbullying. 60 records of publications included in this study were analysed and coded in each review and the definitional criteria of cyberbullying. The researcher used a table to list, code and gather the definitions used by the previous papers and listed out the criteria of cyberbullying from the definitions that had been selected or included. As the researcher had listed, coded and gathered all the criteria of cyberbullying through the use of definitions, then the criteria were analysed to come out with the finding for this study. The researcher listed 60 papers that mentioned clearly the definitions used in the studies. By using the definitions, the researcher points out the criteria of cyberbullying that had been mentioned in the definitions. The researcher adapted the study from Naruskov et al. (2012) which used the five criteria of cyberbullying, which are intention, repetitive, power imbalance, anonymity, and publicity. The researcher added on two more criteria which are aggression and electronic device and media as the value added for this study. Thus, there are seven total criteria of cyberbullying mentioned or identified in this study. Figure 2 shows the flow of analysing the definitional criteria of cyberbullying that done by the researcher in order to get the result for this study.

Results and Discussion
The findings from this systematic literature analyses the definitional criteria which found that there are seven definitional criteria of cyberbullying. The definitional criteria of cyberbullying are extensions from traditional bullying which are intention, repetitive, and power imbalance.
The definitional criteria of cyberbullying as found from this study are intention, repetitive, power imbalance, anonymity, publicity, aggression, and electronic device and media. The researcher adapted the five criteria of cyberbullying from Naruskov et al. (2012) which are intention, repetitive, power imbalance, anonymity, and publicity. Through this study, the researcher added two more criteria of cyberbullying which are aggression and electronic device and media, according to the systematic literature review analysis on the definition of cyberbullying as stated by authors from previous studies.

Intentionality
In the cyber context, Naruskov et al (2012) find intentionality even harder to specify due to the extreme essence of contact throughout this situation (Kowalski et al., 2008;Menesini and Nocentini, 2009). Nocentini et al (2010) examined a sample of youths from three different European countries: Italy, Germany, and Spain. In terms of intentionality, respondents agreed that it is significant, but it cannot be perceived independently from power imbalance, since the effect on the victim and interpretation of the action should be more important than the justification for the perpetrator's actions in assessing the extent and meaning of the behaviour. Tran et al. (2018) all groups agreed that a central characteristic of bullying is intentional and purposeful intent to harm. The intent to harm is one, if not the central characteristic of the definition of cyberbullying. For instance, according to a teacher group, cyberbullying "must cause some loss for the victim" and according to psychologists, cyberbullying occurs when "someone transmits something that makes the other person feel uncomfortable, affects his or her dignity, or their mental or physical health." According to the professionals, cyberbullying is an intentionally harmful behaviour but different from harassment. "If an Internet user posts something that does not have a specific target such as "I hate you", it still can be cyberbullying. Individuals who read the posts may be specifically targeted and experience being bullied." Naruskov et al (2012) in this respect, with a number of scholars relate to the irreversible and public nature of social interaction: even though it is shared to the Internet by the perpetrator once it could be viewed by several people, and then it becomes very impossible to track and erase where the content may have been downloaded, stored and/or distributed to a broader audience and thus not to the public (Dooley et al., 2009;Kowalski et al., 2008;Shariff 2008;Slonje and Smith, 2008). In fact, the victim and other visitors can browse one-off posts twice and more (Kowalski et al., 2008). According to Dooley et al (2009) there continues to be gaps between the perpetrator and the victim in terms of determining how often these aggressive events arise and the possible consequences. However, in the cyber world, such actions do not have to be replicated or repeated as often to inflict harm, and this makes repetition in cyberbullying distinct from the repetition in traditional bullying (Dooley et al., 2009). From research focus group interviews, Nocentini et al (2010) confirmed that repetition was considered a very powerful cyberbullying criteria because it helps differentiate a prank from an intentional attack. In addition, respondents from Italy and Germany referred to the association between repetition and publicity in digital environment, where if a single event of harassment was sent or shown just once to a few users, it was still considered a repeated act by youths. Tran et al. (2018) stated that the majority of focus group discussions covered discussion of frequency or repetition, which is not an important factor in deciding whether a behaviour is cyberbullying. According to the expert community, "frequency is not the problem, but the degree of harm to which everything is exposed." According to the teachers, "only one action can be considered bullying if it is strong enough." Psychologists have suggested that repetition was not central to the definition, for example, it is possible to attack an individual violently only once, but it still have a long-term impact on the person being bullied.

Repetitive
Power Imbalance Naruskov et al (2012) stated that in scope of cyberbullying, the power imbalance resides in conditions where the victims "cannot easily defend themselves" due to the offensive conduct of the perpetrator (Grigg, 2010). Furthermore, Nocentini et al (2010) discovered that power imbalance and intent to harm must be perceived in relation to one another, and the participants claimed that while the victim is impaired by the acts, the behaviour must be classified as cyberbullying. In a simulated world, the power imbalance might well be exacerbated by 'real world' power criteria, for instance, physical size or age, or an enhanced computing capability of the perpetrator (Dooley et al., 2009;Hinduja and Patchin, 2008;Kowalski et al., 2008;Vandebosch and Cleemput, 2008). Grigg (2010) argues that power imbalances will also rely on the perpetrator's situational benefits, such as elevated social position in those social networks, with the aid of other group members. Moreover, the discreet and public nature of the incident and the constant accessibility of the victim will generate inequalities in power structures (Dooley et al., 2009;Kowalski et al., 2008;Slonje and Smith, 2008). Tran et al. (2018) stated that cyberbullying includes behaviours where an individual or a group of individuals are stronger than the intended victim in the opinion of the majority of the audience. This was the consensus of all the groups. A group of students stated that "usually when someone is being bullied, they will feel isolated, as nobody is on their side of the network and that they are weak." According to the expert group, cyberbullying involves "the intimidation of one person/group by another person/group that is more powerful than the other group, so the behaviour is morally intimidating" (Tran et al., 2018).

Anonymity
Naruskov et al (2012) (Kowalski et al., 2008). Protected account names and fake identities protect perpetrators, allowing them to act freely, and making it more difficult for the victim to deal with them (Kowalski et al., 2008;Shariff 2008). Vandebosch and Cleemput (2008) reported that in focus group interviews with Belgian students that the victim's feeling of disempowerment and indignation could be caused by the perpetrator's anonymity. Furthermore, the study discovered that the majority of the participants who had been harassed by electronic media had received anonymous assaults. Mishna et al (2008) conducted a focus group of 38 students in grades 5 and 8. They indicated that the respondents thought of cyberbullying as a major issue because of the anonymity that encourages people to behave in ways they might never do in real life. However, the research showed that the students' own experiences on cyberbullying were not as obscured as they were in the form of student groups and interactions. Nocentini et al (2010) reported that for all the research countries which are Italy, Germany and Spain, the requirements for anonymity were not perceived to be specified criteria by the participants, but they indicated that it could induce vulnerability and insecurity in the victims. Tran et al (2018) stated that according to a group of experts, in cyberbullying the perpetrator "can be anyone." According to the psychologists and teachers, cyber perpetrator can be anonymous or can pretend to be someone else to cause harm to the third person. According to the teachers, in certain cases the cyber perpetrator and cyber victim are not even acquainted with each other "they can hurt an unfamiliar person . . . they talk in public chat group, then fight with each other without knowing the other person" (Tran et al., 2018).

Publicity
According to Naruskov et al (2012), publicity in the cyber world is the complete opposite of a private mode of contact involving only two people, as reported by (Nocentini et al., 2010). In brief, the public essence of Internet communication necessitates interaction with a large audience. For instance, once a picture or a video is posted on a social media platform, an unlimited proportion of people will view and repost and share it with the world. Nocentini et al (2010) claimed publicity criteria are not necessary for labelling and functioning as cyberbullying, but publicity criteria are still valid as it may represent the severity of the harassment. Moreover, adolescents from Italy, Germany and Spain have suggested that the presence of audience makes cyberbullying visible. Accordingly, studies by Slonje and Smith (2008) indicated the use of photo or video recordings was acknowledged to be more extreme over other types of cyberbullying, primarily because a potentially huge viewer and targets could be detected. Allowing private data to be transparent was the most widely used method of cyberbullying in the study by Lenhart (2007). Equally, Patchin and Hinduja (2010) claimed the most extensively recorded form of cyberbullying is posting something online about someone to make others laugh.

Aggression
Cyberbullying may be either direct or relational, constructive or reactive, and manifest in a multitude of ways (Skye Wingate, Minney & Guadagno, 2013). Relational violence (Crick & Grotpeter, 1995) can also involve behaviours such as delivering offensive messages, uploading hurtful comments on social networks, and sharing humiliating details such as images or private writings (Law et al., 2012;Smith, Mahdavi et al., 2006). Since the perpetrator targets the same victim frequently, this activity can be categorised as cyberbullying. Relational aggression is more likely to entail a power imbalance within the social system than a power discrepancy that is dependent on physical characteristics. Reactive aggression is the response of the perpetrator to an actual or potential threat, while proactive aggression is intentionally incited by the perpetrator with a view to obtaining profit or reaching objectives (Crick & Dodge, 1996;Dodge & Coie, 1987) varying through humour or excitement to the extortion of personal information (Law et al., 2012). The virtual world has an inadequacy tangible signals and other social impulses, therefore communication over the Internet can be used in an undefined context, and hence communicators can assign aggressive intentions to neutral stimuli. Sharing intimate communications with unwanted recipients or making embarrassing photos public are examples of cyberbullying behaviours that are typically called reactive aggression (Law et al., 2012). Proactive cyberbullying behaviour might include developing aggressive networks solely for the intention of harming another user or creating a false virtual entity to elicit possibly hazardous information from the victim (Law et al., 2012). Current evidence suggests that cyberbullying probably comes from proactive aggression rather than reactive aggression (Calvete, Orue, Estevez, Villardon, & Padilla, 2010). Calvete et al. (2010) also proposed that cyberbullying can be graded as passive and indirect violence compared to other aggressive behaviours in young people. According to J. Pyalski (2012), networking structures such as the Internet and cell phones can be used as tools for aggressive behaviour against others. Cyber perpetrators may use such actions as the word 'electronic aggression' understood as any kind of aggression or harassment, which includes mocking, lying, trying to make fun of, posting rude or cruel comments, making accusations, threatening or violent comments via e-mail, web portal, social networking, group chat or messaging apps (David-Ferdon & Feldman Hertz, 2007). Aggressive acts on a few types of electronic communication, such as those featuring visual imagery depicting a victim, are more harmful than others (Slonje & Smith, 2008). A few reports have focused on the way in which acts of aggression are carried out, defining a particular set of technological acts of harassment as cyberbullying is based on the existence of criteria relevant to school abuse (Cassidy et al., 2011). According to Mehari et al. (2014), recent cyberbullying research is primarily focused on the perception that it is a different type of aggression, a complement to physical, verbal, and social abuse. As a result, scholars have generally not distinguished between physical, verbal, and relational aggression in their definitions and measures of cyberbullying. This is predicated on the assumption that the aggressivity medium is more important than the shape. Nevertheless, it is probable that media constitutes an extra element in which violence can be graded. In certain circumstances, the same violent action can be carried out across the internet as rumours can be shared in individual or by messaging apps. In other situations, aggression by means of electronic media systems may be interpreted differently. For instance, the sharing of inappropriate photos is targeted at adolescents' image, credibility, and relationships. As a result, this paradigm is known to be relational cyberbullying.

Electronic Device and Media
This criteria is important to be included in the definition of the cyberbullying because technically, cyberbullying exists in the rapid development of technology and technology includes electronic devices such as smartphones, laptops, computers, tablets and electronic medias such as social networking sites, websites, e-mail, and other forms of electronic communication. Trolley and Hanel (2010) reported that there is a wide range of electronic media available. As a number of easy-to-access electronic devices and media with constant availability, possible misuse by user will lead to an increased number of cyberbullying cases. Mobile phones or smartphones are popular electronic devices used for communication with a wide range of applications and as useful gadgets, as it is a compact device that can be carried around and has many features and functions that are useful for users. The advantages of mobile phones or smartphones to take images and videos of others make it possible for users to take corrupted or offensive photos, then send the photos or videos to other people, or posting it to a cyberbully on social networking sites with the purpose of revealing or humiliating the cyber victim (Barlett & Gentile, 2012;Rivers & Noret, 2010;Trolley & Hanel, 2010).
The development of technology and the internet simplify access to social media as easily as at one's fingertips. Beng & Hua (2019) stated that users of social media or social networking sites can easily post a status on Facebook, share a personal photo on Twitter, and comment on others' status on social networking sites. On the Internet, people can create dummy accounts to leave an anonymous message without fearing the repercussions of their actions. In addition, people can comment on whatever they like on social networking sites. Table 1 below shows the electronic medias that are used as a platform to cyberbully others. Social networking platforms include Facebook, Instagram, Twitter, WhatsApp, LinkedIn, Google+ and more. It is used to create friendships, staying connected, sharing data or knowledge with someone else and post new images and clips (Snakenborg et al., 2011;Trolley & Hanel, 2010). Cyber perpetrator can post violent, humiliating, and hurtful messages, images, and videos, and may be viewed by millions of users around the world (Shaw and Cross, 2012;Stauffer et al. 2012). In addition, online media platforms divert people from reality, weaken real relationships, promote isolation, boost suicidal thoughts, and distribute confidential data to an anonymous public (Balog et al. 2013). Instant messaging: Similar to text messaging, but instant messaging allows a person to reply directly in terms of a "live chat" (Snakenborg et al., 2011). Users of these instant messaging applications will freely gain advantage of emerging technologies and using that to deliver aggressive, intimidating, and offensive responses to each other (McElligott, 2006). E-mail: Electronic messages sent to an individual using a specific e-mail address. It can be used to send the same message to multiple receivers at once (Barlett and Gentile, 2012;Smith and Slonje, 2010;Snakenborg et al., 2011). The message sent could be transmitted to numerous people who did not intend to get the message, thereby invasive to the privacy of the person (Snakenborg et al., 2011). Chat rooms: A portal that helps users to meet and communicate with others about their common interest in real-time. Chat rooms should be beneficial and is a practical way for young people to stay connected with others. However, chat rooms offer a convenience for online fraud and cyberbullying, where the step in identifying the cyber perpetrator is challenging (Jacobs et al., 2012). Websites: Personal websites are intended to provide information about a person, community or institution, but people may also create websites that are defamatory, hostile, and offensive to others or targeted individuals (Franks, 2015). It reveals the cyber victim and encourages others to access the website and comment on the post, hence adding to the victimisation of the potential victim (Barlett and Gentile, 2012;Smith and Slonje, 2010;Twyman et al., 2010). Wong et al. (2013) confirmed that there are reports of websites explicitly designed to encourage people to commit suicide. Video hosting sites: Youtube is an instance of video sharing platform that encourages users to post and watch interesting content (Snakenborg et al., 2011). Vicious, embarrassing, and abusive videos can be posted, downloaded, and commented on by millions of users around the globe, as the internet is open and easy to reach. Online video games: Many theorists believe that social domination or control is the primary motive for aggression across contexts, particularly cyberbullying (Espelage et al., 2013;Kowalski et al., 2012;Law et al., 2012). Ross and Weaver (2012) discovered evidence of an observational learning analysis of depressive behaviour "grieving" when playing video games. Yang (2012) investigated cyberbullying among Taiwanese youth and discovered that a preference for violent games predicted hostility, which predicted cyberbullying and cyber-victimisation.
Cyberbullying is one of the forms of harassment that involves actions or acts intended to harm victims who are unable to defend themselves easily by the use of information and communication technologies (ICTs) and electronic devices such as smartphones or mobile phones, computers or laptops, and tablets (Slone, Smith and Frisén, 2013;Smith, del Barrio and Tokunaga, 2013). There are several versions of the concept of cyberbullying, that is consistently used by . The definition of cyberbullying is expanded from the definition of traditional bullying, as reported by Smith et al (2008) "an aggressive act or behaviour that is carried out using electronic means by a group or an individual repeatedly and over time against a victim who cannot easily defend him or herself." Blumenfeld (2013) has acknowledged that cyberbullying is a type of current action that has emerged alongside rapid growth of technology advancement. Brody and Vangelisti (2017) stated that it is important to define the concept of cyberbullying to describe cyberbullying. This is because the definition of cyberbullying is needed in solving the social problem issues in the aspect of risk factor, prevention strategy, and others. There are arguments regarding the definition of cyberbullying by previous authors. Selkie, Fales & Moreno (2016) stated that the application of cyberbullying's definition is unclear, specifically, whether online activities may be called aggressive in the absence of severe in-person psychology stimuli such as voice tone and facial expressions. Olweus (2012) and Smith (2015) have argued that cyberbullying is a behavioural issue that occurs online or in the virtual environment, described as an intentional violence that is repeated over time and is conducted using electronic devices. Moreover, conceptualising cyberbullying is regarded as critical due to the fact that cyberbullying can take several different forms and exist in a variety of settings (Kowalski et al. 2014). There are many forms or types of cyberbullying that existed which each one of the forms has different description and the conceptualising of cyberbullying might help for a detail description of the forms. Another issue that arises as a result of this is within the cultural aspect. There are few studies that have been carried out in few countries in defining cyberbullying towards their own countries. Since the meaning and concept of cyberbullying vary across countries , it is critical to find consensus on the definitional criteria for cyberbullying.

Conclusions
The aim of this research is to determine the definitional criteria of cyberbullying in order to establish a definition of cyberbullying. Through a systematic review of previous researches, this paper sought to clarify the definitional criteria of cyberbullying. From the findings of this study, it may help scholars to have an in-depth understanding of the definitional criteria of cyberbullying. The findings from this study also may help scholars to further the research on the conceptualisation of cyberbullying. The conceptualisation of cyberbullying is important because when people clearly understand the concept of cyberbullying, many cyber-based social problems can be prevented or solved. In addition, it may help the government, policy makers, and legislators in developing a better concept and definition on cyberbullying in order to develop or establish the new rules or laws on cyberbullying.