cyberbullying research paper 2021

SYSTEMATIC REVIEW article

Cyberbullying among adolescents and children: a comprehensive review of the global situation, risk factors, and preventive measures.

• 1 School of Political Science and Public Administration, Wuhan University, Wuhan, China
• 2 School of Medicine and Health Management, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
• 3 College of Engineering, Design and Physical Sciences, Brunel University London, Uxbridge, United Kingdom

Background: Cyberbullying is well-recognized as a severe public health issue which affects both adolescents and children. Most extant studies have focused on national and regional effects of cyberbullying, with few examining the global perspective of cyberbullying. This systematic review comprehensively examines the global situation, risk factors, and preventive measures taken worldwide to fight cyberbullying among adolescents and children.

Methods: A systematic review of available literature was completed following PRISMA guidelines using the search themes “cyberbullying” and “adolescent or children”; the time frame was from January 1st, 2015 to December 31st, 2019. Eight academic databases pertaining to public health, and communication and psychology were consulted, namely: Web of Science, Science Direct, PubMed, Google Scholar, ProQuest, Communication & Mass Media Complete, CINAHL, and PsycArticles. Additional records identified through other sources included the references of reviews and two websites, Cyberbullying Research Center and United Nations Children's Fund. A total of 63 studies out of 2070 were included in our final review focusing on cyberbullying prevalence and risk factors.

Results: The prevalence rates of cyberbullying preparation ranged from 6.0 to 46.3%, while the rates of cyberbullying victimization ranged from 13.99 to 57.5%, based on 63 references. Verbal violence was the most common type of cyberbullying. Fourteen risk factors and three protective factors were revealed in this study. At the personal level, variables associated with cyberbullying including age, gender, online behavior, race, health condition, past experience of victimization, and impulsiveness were reviewed as risk factors. Likewise, at the situational level, parent-child relationship, interpersonal relationships, and geographical location were also reviewed in relation to cyberbullying. As for protective factors, empathy and emotional intelligence, parent-child relationship, and school climate were frequently mentioned.

Conclusion: The prevalence rate of cyberbullying has increased significantly in the observed 5-year period, and it is imperative that researchers from low and middle income countries focus sufficient attention on cyberbullying of children and adolescents. Despite a lack of scientific intervention research on cyberbullying, the review also identified several promising strategies for its prevention from the perspectives of youths, parents and schools. More research on cyberbullying is needed, especially on the issue of cross-national cyberbullying. International cooperation, multi-pronged and systematic approaches are highly encouraged to deal with cyberbullying.

Introduction

Childhood and adolescence are not only periods of growth, but also of emerging risk taking. Young people during these periods are particularly vulnerable and cannot fully understand the connection between behaviors and consequences ( 1 ). With peer pressures, the heat of passion, children and adolescents usually perform worse than adults when people are required to maintain self-discipline to achieve good results in unfamiliar situations. Impulsiveness, sensation seeking, thrill seeking, and other individual differences cause adolescents to risk rejecting standardized risk interventions ( 2 ).

About one-third of Internet users in the world are children and adolescents under the age of 18 ( 3 ). Digital technology provide a new form of interpersonal communication ( 4 ). However, surveys and news reports also show another picture in the Internet Age. The dark side of young people's internet usage is that they may bully or suffer from others' bullying in cyberspace. This behavior is also acknowledged as cyberbullying ( 5 ). Based on Olweus's definition, cyberbullying is usually regarded as bullying implemented through electronic media ( 6 , 7 ). Specifically, cyberbullying among children and adolescents can be summarized as the intentional and repeated harm from one or more peers that occurs in cyberspace caused by the use of computers, smartphones and other devices ( 4 , 8 – 12 ). In recent years, new forms of cyberbullying behaviors have emerged, such as cyberstalking and online dating abuse ( 13 – 15 ).

Although cyberbullying is still a relatively new field of research, cyberbullying among adolescents is considered to be a serious public health issue that is closely related to adolescents' behavior, mental health and development ( 16 , 17 ). The increasing rate of Internet adoption worldwide and the popularity of social media platforms among the young people have worsened this situation with most children and adolescents experiencing cyberbullying or online victimization during their lives. The confines of space and time are alleviated for bullies in virtual environments, creating new venues for cyberbullying with no geographical boundaries ( 6 ). Cyberbullying exerts negative effects on many aspects of young people's lives, including personal privacy invasion and psychological disorders. The influence of cyberbullying may be worse than traditional bullying as perpetrators can act anonymously and connect easily with children and adolescents at any time ( 18 ). In comparison with traditional victims, those bullied online show greater levels of depression, anxiety and loneliness ( 19 ). Self-esteem problems and school absenteeism have also proven to be related to cyberbullying ( 20 ).

Due to changes in use and behavioral patterns among the youth on social media, the manifestations and risk factors of cyberbullying have faced significant transformation. Further, as the boundaries of cyberbullying are not limited by geography, cyberbullying may not be a problem contained within a single country. In this sense, cyberbullying is a global problem and tackling it requires greater international collaboration. The adverse effects caused by cyberbullying, including reduced safety, lower educational attainment, poorer mental health and greater unhappiness, led UNICEF to state that “no child is absolutely safe in the digital world” ( 3 ).

Extant research has examined the prevalence and risk factors of cyberbullying to unravel the complexity of cyberbullying across different countries and their corresponding causes. However, due to variations in cyberbullying measurement and methodologies, no consistent conclusions have been drawn ( 21 ). Studies into inconsistencies in prevalence rates of cyberbullying, measured in the same country during the same time period, occur frequently. Selkie et al. systematically reviewed cyberbullying among American middle and high school students aged 10–19 years old in 2015, and revealed that the prevalence of cyberbullying victimization ranged from 3 to 72%, while perpetration ranged from 1 to 41% ( 22 ). Risk and protective factors have also been broadly studied, but confirmation is still needed of those factors which have more significant effects on cyberbullying among young people. Clarification of these issues would be useful to allow further research to recognize cyberbullying more accurately.

This review aims to extend prior contributions and provide a comprehensive review of cyberbullying of children and adolescents from a global perspective, with the focus being on prevalence, associated risk factors and protective factors across countries. It is necessary to provide a global panorama based on research syntheses to fill the gaps in knowledge on this topic.

Search Strategies

This study strictly employed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We consulted eight academic databases pertaining to public health, and communication and psychology, namely: Web of Science, Science Direct, PubMed, Google Scholar, ProQuest, Communication & Mass Media Complete, CINAHL, and PsycArticles. Additional records identified through other sources included the references of reviews and two websites, Cyberbullying Research Center and United Nations Children's Fund. With regard to the duration of our review, since most studies on cyberbullying arose around 2015 ( 9 , 21 ), this study highlights the complementary aspects of the available information about cyberbullying during the recent 5 year period from January 1st, 2015 to December 31st, 2019.

One researcher extracted keywords and two researchers proposed modifications. We used two sets of subject terms to review articles, “cyberbullying” and “child OR adolescent.” Some keywords that refer to cyberbullying behaviors and young people are also included, such as threat, harass, intimidate, abuse, insult, humiliate, condemn, isolate, embarrass, forgery, slander, flame, stalk, manhunt, as well as teen, youth, young people and student. The search formula is (cyberbullying OR cyber-bullying OR cyber-aggression OR ((cyber OR online OR electronic OR Internet) AND (bully * OR aggres * OR violence OR perpetrat * OR victim * OR threat * OR harass * OR intimidat * OR * OR insult * OR humiliate * OR condemn * OR isolate * OR embarrass * OR forgery OR slander * OR flame OR stalk * OR manhunt))) AND (adolescen * OR child OR children OR teen? OR teenager? OR youth? OR “young people” OR “elementary school student * ” OR “middle school student * ” OR “high school student * ”). The main search approach is title search. Search strategies varied according to the database consulted, and we did not limit the type of literature for inclusion. Journals, conference papers and dissertations are all available.

Specifically, the inclusion criteria for our study were as follows: (a). reported or evaluated the prevalence and possible risk factors associated with cyberbullying, (b). respondents were students under the age of 18 or in primary, junior or senior high schools, and (c). studies were written in English. Exclusion criteria were: (a). respondents came from specific groups, such as clinical samples, children with disabilities, sexual minorities, specific ethnic groups, specific faith groups or samples with cross-national background, (b). review studies, qualitative studies, conceptual studies, book reviews, news reports or abstracts of meetings, and (c). studies focused solely on preventive measures that were usually meta-analytic and qualitative in nature. Figure 1 presents the details of the employed screening process, showing that a total of 63 studies out of 2070 were included in our final review.

Figure 1 . PRISMA flow chart diagram showing the process of study selection for inclusion in the systematic review on children and adolescents cyberbullying.

Meta-analysis was not conducted as the limited research published within the 5 years revealed little research which reported odds ratio. On the other hand, due to the inconsistency of concepts, measuring instruments and recall periods, considerable variation could be found in research quality ( 23 ). Meta-analysis is not a preferred method.

Coding Scheme

For coding, we created a comprehensive code scheme to include the characteristics. For cyberbullying, we coded five types proposed by Willard ( 24 – 26 ), which included verbal violence, group violence, visual violence, impersonating and account forgery, and other behaviors. Among them, verbal violence is considered one of the most common types of cyberbullying and refers to the behavior of offensive responses, insults, mocking, threats, slander, and harassment. Group violence is associated with preventing others from joining certain groups or isolating others, forcing others to leave the group. Visual violence relates to the release and sharing of embarrassing photos and information without the owners' consent. Impersonating and account forgery refers to identity theft, stealing passwords, violating accounts and the creation of fake accounts to fraudulently present the behavior of others. Other behaviors include disclosure of privacy, sexual harassment, and cyberstalking. To comprehensively examine cyberbullying, we coded cyberbullying behaviors from both the perspectives of cyberbullying perpetrators and victims, if mentioned in the studies.

In relation to risk factors, we drew insights from the general aggression model, which contributes to the understanding of personal and situational factors in the cyberbullying of children and adolescents. We chose the general aggression model because (a) it contains more situational factors than other models (e.g., social ecological models) - such as school climate ( 9 ), and (b) we believe that the general aggression model is more suitable for helping researchers conduct a systematic review of cyberbullying risk and protective factors. This model provides a comprehensive framework that integrates domain specific theories of aggression, and has been widely applied in cyberbullying research ( 27 ). For instance, Kowalski and colleagues proposed a cyberbullying encounter through the general aggression model to understand the formation and development process of youth cyberbullying related to both victimization and perpetration ( 9 ). Victims and perpetrators enter the cyberbullying encounter with various individual characteristics, experiences, attitudes, desires, personalities, and motives that intersect to determine the course of the interaction. Correspondingly, the antecedents pertaining to cyberbullying are divided into two broad categories, personal factors and situational factors. Personal factors refer to individual characteristics, such as gender, age, motivation, personality, psychological states, socioeconomic status and technology use, values and perceptions, and other maladaptive behaviors. Situational factors focus on the provocation/support, parental involvement, school climate, and perceived anonymity. Consequently, our coders related to risk factors consisting of personal factors and situational factors from the perspectives of both cyberbullying perpetrators and victims.

We extracted information relating to individual papers and sample characteristics, including authors, year of publication, country, article type, sampling procedures, sample characteristics, measures of cyberbullying, and prevalence and risk factors from both cyberbullying perpetration and victimization perspectives. The key words extraction and coding work were performed twice by two trained research assistants in health informatics. The consistency test results are as follows: the Kappa value with “personal factors” was 0.932, and the Kappa value with “situational factors” was 0.807. The result shows that the coding consistency was high enough and acceptable. Disagreements were resolved through discussion with other authors.

Quality Assessment of Studies

The quality assessment of the studies is based on the recommended tool for assessing risk of bias, Cochrane Collaboration. This quality assessment tool focused on seven items: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other sources of bias ( 28 ). We assessed each item as “low risk,” “high risk,” and “unclear” for included studies. A study is considered of “high quality” when it meets three or more “low risk” requirements. When one or more main flaw of a study may affect the research results, the study is considered as “low quality.” When a lack of information leads to a difficult judgement, the quality is considered to be “unclear.” Please refer to Appendix 1 for more details.

This comprehensive systematic review comprised a total of 63 studies. Appendices 2 , 3 show the descriptive information of the studies included. Among them, 58 (92%) studies measured two or more cyberbullying behavior types. The sample sizes of the youths range from several hundred to tens of thousands, with one thousand to five thousand being the most common. As for study distribution, the United States of America, Spain and China were most frequently mentioned. Table 1 presents the detail.

Table 1 . Descriptive information of studies included (2015–2019).

Prevalence of Global Cyberbullying

Prevalence across countries.

Among the 63 studies included, 22 studies reported on cyberbullying prevalence and 20 studies reported on prevalence from victimization and perpetration perspectives, respectively. Among the 20 studies, 11 national studies indicated that the prevalence of cyberbullying victimization and cyberbullying perpetration ranged from 14.6 to 52.2% and 6.3 to 32%, respectively. These studies were conducted in the United States of America ( N = 4) ( 29 – 32 ), South Korea ( N = 3) ( 33 – 35 ), Singapore ( N = 1) ( 36 ), Malaysia ( N = 1) ( 37 ), Israel ( N = 1) ( 38 ), and Canada ( N = 1) ( 39 ). Only one of these 11 national studies is from an upper middle income country, and the rest are from highincome countries identified by the World Bank ( 40 ). By combining regional and community-level studies, the prevalence of cyberbullying victimization and cyberbullying perpetration ranged from 13.99 to 57.5% and 6.0 to 46.3%, respectively. Spain reported the highest prevalence of cyberbullying victimization (57.5%) ( 41 ), followed by Malaysia (52.2%) ( 37 ), Israel (45%) ( 42 ), and China (44.5%) ( 43 ). The lowest reported victim rates were observed in Canada (13.99%) and South Korea (14.6%) ( 34 , 39 ). The reported prevalence of cyberbullying victimization in the United States of America ranged from 15.5 to 31.4% ( 29 , 44 ), while in Israel, rates ranged from 30 to 45% ( 26 , 42 ). In China, rates ranged from 6 to 46.3% with the country showing the highest prevalence of cyberbullying perpetration (46.30%) ( 15 , 43 , 45 , 46 ). Canadian and South Korean studies reported the lowest prevalence of cyberbullying perpetration at 7.99 and 6.3%, respectively ( 34 , 39 ).

A total of 10 studies were assessed as high quality studies. Among them, six studies came from high income countries, including Canada, Germany, Italy, Portugal, and South Korea ( 13 , 34 , 39 , 46 – 48 ). Three studies were from upper middle income countries, including Malaysia and China ( 37 , 43 ) and one from a lower middle income country, Nigeria ( 49 ). Figures 2 , 3 describe the prevalence of cyberbullying victimization and perpetration respectively among high quality studies.

Figure 2 . The prevalence of cyberbullying victimization of high quality studies.

Figure 3 . The prevalence of cyberbullying perpetration of high quality studies.

Prevalence of Various Cyberbullying Behaviors

For the prevalence of cyberbullying victimization and perpetration, the data were reported in 18 and 14 studies, respectively. Figure 4 shows the distribution characteristics of the estimated value of prevalence of different cyberbullying behaviors with box plots. The longer the box, the greater the degree of variation of the numerical data and vice versa. The rate of victimization and crime of verbal violence, as well as the rate of victimization of other behaviors, such as cyberstalking and digital dating abuse, has a large degree of variation. Among the four specified types of cyberbullying behaviors, verbal violence was regarded as the most commonly reported behaviors in both perpetration and victimization rates, with a wide range of prevalence, ranging from 5 to 18%. Fewer studies reported the prevalence data for visual violence and group violence. Studies also showed that the prevalence of impersonation and account forgery were within a comparatively small scale. Specific results were as follows.

Figure 4 . Cyberbullying prevalence across types (2015–2019).

Verbal Violence

A total of 13 studies reported verbal violence prevalence data ( 15 , 26 , 34 , 37 – 39 , 42 , 43 , 47 , 48 , 50 , 51 ). Ten studies reported the prevalence of verbal violence victimization ranging from 2.8 to 47.5%, while seven studies claimed perpetration prevalence ranging from 1.5 to 31.8%. Malaysia reported the highest prevalence of verbal violence victimization (47.5%) ( 37 ), followed by China (32%) ( 43 ). China reported that the prevalence of verbal violence victimization ranged from 5.1 to 32% ( 15 , 43 ). Israel reported that the prevalence of verbal violence victimization ranged from 3.4 to 18% ( 26 , 38 , 42 ). For perpetration rate, Malaysia reported the highest level at 31.8% ( 37 ), while a study for Spain reported the lowest, ranging from 3.2 to 6.4% ( 51 ).

Group Violence

The prevalence of group violence victimization was explored within 4 studies and ranged from 5 to 17.8% ( 26 , 34 , 42 , 43 ), while perpetration prevalence was reported in three studies, ranging from 10.1 to 19.07% ( 34 , 43 , 47 ). An Israeli study suggested that 9.8% of respondents had been excluded from the Internet, while 8.9% had been refused entry to a group or team ( 26 ). A study in South Korea argued that the perpetration prevalence of group violence was 10.1% ( 34 ), while a study in Italy reported that the rate of online group violence against others was 19.07% ( 47 ).

Visual Violence

The prevalence of visual violence victimization was explored within three studies and ranged from 2.6 to 12.1% ( 26 , 34 , 43 ), while the perpetration prevalence reported in four studies ranged from 1.7 to 6% ( 34 , 43 , 47 , 48 ). For victimization prevalence, a South Korean study found that 12.1% of respondents reported that their personal information was leaked online ( 34 ). An Israel study reported that the prevalence of outing the picture was 2.6% ( 26 ). For perpetration prevalence, a South Korean study found that 1.7% of respondents had reported that they had disclosed someone's personal information online ( 34 ). A German study reported that 6% of respondents had written a message (e.g., an email) to somebody using a fake identity ( 48 ).

Impersonating and Account Forgery

Four studies reported on the victimization prevalence of impersonating and account forgery, ranging from 1.1 to 10% ( 15 , 42 , 43 ), while five studies reported on perpetration prevalence, with the range being from 1.3 to 9.31% ( 15 , 43 , 47 , 48 , 51 ). In a Spanish study, 10% of respondents reported that their accounts had been infringed by others or that they could not access their account due to stolen passwords. In contrast, 4.5% of respondents reported that they had infringed other people's accounts or stolen passwords, with 2.5% stating that they had forged other people's accounts ( 51 ). An Israeli study reported that the prevalence of being impersonated was 7% ( 42 ), while in China, a study reported this to be 8.6% ( 43 ). Another study from China found that 1.1% of respondents had been impersonated to send dating-for-money messages ( 15 ).

Other Behaviors

The prevalence of disclosure of privacy, sexual harassment, and cyberstalking were also explored by scholars. Six studies reported the victimization prevalence of other cyberbullying behaviors ( 13 , 15 , 34 , 37 , 42 , 43 ), and four studies reported on perpetration prevalence ( 34 , 37 , 43 , 48 ). A study in China found that 1.2% of respondents reported that their privacy had been compromised without permission due to disputes ( 15 ). A study from China reported the prevalence of cyberstalking victimization was 11.9% ( 43 ), while a Portuguese study reported that this was 62% ( 13 ). In terms of perpetration prevalence, a Malaysian study reported 2.7% for sexual harassment ( 37 ).

Risk and Protective Factors of Cyberbullying

In terms of the risk factors associated with cyberbullying among children and adolescents, this comprehensive review highlighted both personal and situational factors. Personal factors referred to age, gender, online behavior, race, health conditions, past experiences of victimization, and impulsiveness, while situational factors consisted of parent-child relationship, interpersonal relationships, and geographical location. In addition, protective factors against cyberbullying included: empathy and emotional intelligence, parent-child relationship, and school climate. Table 2 shows the risk and protective factors for child and adolescent cyberbullying.

Table 2 . Risk and protective factors of cyberbullying among children and adolescents.

In terms of the risk factors associated with cyberbullying victimization at the personal level, many studies evidenced that females were more likely to be cyberbullied than males ( 13 , 26 , 29 , 38 , 43 , 52 , 54 , 55 , 58 ). Meanwhile, adolescents with mental health problems ( 61 ), such as depression ( 33 , 62 ), borderline personality disorder ( 63 ), eating disorders ( 41 ), sleep deprivation ( 56 ), and suicidal thoughts and suicide plans ( 64 ), were more likely to be associated with cyberbullying victimization. As for Internet usage, researchers agreed that youth victims were probably those that spent more time online than their counterparts ( 32 , 36 , 43 , 45 , 48 , 49 , 60 ). For situational risk factors, some studies have proven the relationship between cyberbullying victims and parental abuse, parental neglect, family dysfunction, inadequate monitoring, and parents' inconsistency in mediation, as well as communication issues ( 33 , 64 , 68 , 73 ). In terms of geographical location, some studies have reported that youths residing in city locations are more likely to be victims of cyberbullying than their peers from suburban areas ( 61 ).

Regarding the risk factors of cyberbullying perpetration at the personal level, it is generally believed that older teenagers, especially those aged over 15 years, are at greater risk of becoming cyberbullying perpetrators ( 55 , 67 ). When considering prior cyberbullying experiences, evidence showed that individuals who had experienced cyberbullying or face-to-face bullying tended to be aggressors in cyberbullying ( 35 , 42 , 49 , 51 , 55 ); in addition, the relationship between impulsiveness and cyberbullying perpetration was also explored by several pioneering scholars ( 55 , 72 , 80 ). The situational factors highlight the role of parents and teachers in cyberbullying experiences. For example, over-control and authoritarian parenting styles, as well as inharmonious teacher-student relationships ( 61 ) are perceived to lead to cyberbullying behaviors ( 74 , 75 ). In terms of differences in geographical locations, students residing in cities have a higher rate of online harassment than students living in more rural locations ( 49 ).

In terms of the protective factors in child and adolescent cyberbullying, scholars have focused on youths who have limited experiences of cyberbullying. At the personal level, high emotional intelligence, an ability for emotional self-control and empathy, such as cognitive empathy ability ( 44 , 55 ), were associated with lower rates of cyberbullying ( 57 ). At the situational level, a parent's role is seen as critical. For example, intimate parent-child relationships ( 46 ) and open active communication ( 19 ) were demonstrated to be related to lower experiences of cyberbullying and perpetration. Some scholars argued that parental supervision and monitoring of children's online activities can reduce their tendency to participate in some negative activities associated with cyberbullying ( 31 , 46 , 73 ). They further claimed that an authoritative parental style protects youths against cyberbullying ( 43 ). Conversely, another string of studies evidenced that parents' supervision of Internet usage was meaningless ( 45 ). In addition to conflicting roles of parental supervision, researchers have also looked into the role of schools, and posited that positive school climates contribute to less cyberbullying experiences ( 61 , 79 ).

Some risk factors may be protective factors under another condition. Some studies suggest that parental aggressive communication is related to severe cyberbullying victims, while open communication is a potential protective factor ( 19 ). Parental neglect, parental abuse, parental inconsistency in supervision of adolescents' online behavior, and family dysfunction are related to the direct or indirect harm of cyberbullying ( 33 , 68 ). Parental participation, a good parental-children relationship, communication and dialogue can enhance children's school adaptability and prevent cyberbullying behaviors ( 31 , 74 ). When parental monitoring reaches a balance between control and openness, it could become a protective factor against cyberbullying, and it could be a risk factor, if parental monitoring is too low or over-controlled ( 47 ).

Despite frequent discussion about the risk factors associated with cyberbullying among children and adolescents, some are still deemed controversial factors, such as age, race, gender, and the frequency of suffering on the internet. For cyberbullying victims, some studies claim that older teenagers are more vulnerable to cyberbullying ( 15 , 38 , 52 , 53 ), while other studies found conflicting results ( 26 , 33 ). As for student race, Alhajji et al. argued that non-white students were less likely to report cyberbullying ( 29 ), while Morin et al. observed no significant correlation between race and cyberbullying ( 52 ). For cyberbullying perpetration, Alvarez-Garcia found that gender differences may have indirect effects on cyberbullying perpetration ( 55 ), while others disagreed ( 42 , 61 , 68 – 70 ). Specifically, some studies revealed that males were more likely to become cyberbullying perpetrators ( 34 , 39 , 56 ), while Khurana et al. presented an opposite point of view, proposing that females were more likely to attack others ( 71 ). In terms of time spent on the Internet, some claimed that students who frequently surf the Internet had a higher chance of becoming perpetrators ( 49 ), while others stated that there was no clear and direct association between Internet usage and cyberbullying perpetration ( 55 ).

In addition to personal and situational factors, scholars have also explored other specific factors pertaining to cyberbullying risk and protection. For instance, mindfulness and depression were found to be significantly related to cyber perpetration ( 76 ), while eating disorder psychopathology in adolescents was associated with cyber victimization ( 41 ). For males who were familiar with their victims, such as family members, friends and acquaintances, they were more likely to be cyberstalking perpetrators than females or strangers, while pursuing desired closer relationships ( 13 ). In the school context, a lower social likability in class was identified as an indirect factor for cyberbullying ( 48 ).

This comprehensive review has established that the prevalence of global childhood and adolescent victimization from cyberbullying ranges from 13.99 to 57.5%, and that the perpetration prevalence ranges from 6.0 to 46.3%. Across the studies included in our research, verbal violence is observed as one of the most common acts of cyberbullying, including verbal offensive responses, insults, mocking, threats, slander, and harassment. The victimization prevalence of verbal violence is reported to be between 5 and 47.5%, and the perpetration prevalence is between 3.2 and 26.1%. Personal factors, such as gender, frequent use of social media platforms, depression, borderline personality disorder, eating disorders, sleep deprivation, and suicidal tendencies, were generally considered to be related to becoming a cyberbullying victim. Personal factors, such as high school students, past experiences, impulse, improperly controlled family education, poor teacher-student relationships, and the urban environment, were considered risk factors for cyberbullying perpetration. Situational factors, including parental abuse and neglect, improper monitoring, communication barriers between parents and children, as well as the urban environment, were also seen to potentially contribute to higher risks of both cyberbullying victimization and perpetration.

Increasing Prevalence of Global Cyberbullying With Changing Social Media Landscape and Measurement Alterations

This comprehensive review suggests that global cyberbullying rates, in terms of victimization and perpetration, were on the rise during the 5 year period, from 2015 to 2019. For example, in an earlier study conducted by Modecki et al. the average cyberbullying involvement rate was 15% ( 81 ). Similar observations were made by Hamm et al. who found that the median rates of youth having experienced bullying or who had bullied others online, was 23 and 15.2%, respectively ( 82 ). However, our systematic review summarized global children and adolescents cyberbullying in the last 5 years and revealed an average cyberbullying perpetration rate of 25.03%, ranging from 6.0 to 46.3%, while the average victimization was 33.08%, ranging from 13.99 to 57.5%. The underlying reason for increases may be attributed to the rapid changing landscape of social media and, in recent years, the drastic increase in Internet penetration rates. With the rise in Internet access, youths have greater opportunities to participate in online activities, provided by emerging social media platforms.

Although our review aims to provide a broader picture of cyberbullying, it is well-noted in extant research that difficulties exist in accurately estimating variations in prevalence in different countries ( 23 , 83 ). Many reasons exist to explain this. The first largely relates poor or unclear definition of the term cyberbullying; this hinders the determination of cyberbullying victimization and perpetration ( 84 ). Although traditional bullying behavior is well-defined, the definition cannot directly be applied to the virtual environment due to the complexity in changing online interactions. Without consensus on definitions, measurement and cyberbullying types may vary noticeably ( 83 , 85 ). Secondly, the estimation of prevalence of cyberbullying is heavily affected by research methods, such as recall period (lifetime, last year, last 6 months, last month, or last week etc.), demographic characteristics of the survey sample (age, gender, race, etc.), perspectives of cyberbullying experiences (victims, perpetrators, or both victim and perpetrator), and instruments (scales, study-specific questions) ( 23 , 84 , 86 ). The variety in research tools and instruments used to assess the prevalence of cyberbullying can cause confusion on this issue ( 84 ). Thirdly, variations in economic development, cultural backgrounds, human values, internet penetration rates, and frequency of using social media may lead to different conclusions across countries ( 87 ).

Acknowledging the Conflicting Role of the Identified Risk Factors With More Research Needed to Establish the Causality

Although this review has identified many personal and situational factors associated with cyberbullying, the majority of studies adopted a cross-sectional design and failed to reveal the causality ( 21 ). Nevertheless, knowledge on these correlational relationships provide valuable insights for understanding and preventing cyberbullying incidents. In terms of gender differences, females are believed to be at a higher risk of cyberbullying victimization compared to males. Two reasons may help to explain this. First, the preferred violence behaviors between two genders. females prefer indirect harassment, such as the spreading of rumors, while males tend toward direct bullying (e.g., assault) ( 29 ) and second, the cultural factors. From the traditional gender perspective, females tended to perceive a greater risk of communicating with others on the Internet, while males were more reluctant to express fear, vulnerability and insecurity when asked about their cyberbullying experiences ( 46 ). Females were more intolerant when experiencing cyberstalking and were more likely to report victimization experiences than males ( 13 ). Meanwhile, many researchers suggested that females are frequent users of emerging digital communication platforms, which increases their risk of unpleasant interpersonal contact and violence. From the perspective of cultural norms and masculinity, the reporting of cyberbullying is also widely acknowledged ( 37 ). For example, in addition, engaging in online activities is also regarded as a critical predictor for cyberbullying victimization. Enabled by the Internet, youths can easily find potential victims and start harassment at any time ( 49 ). Participating in online activities directly increases the chance of experiencing cyberbullying victimization and the possibility of becoming a victim ( 36 , 45 ). As for age, earlier involvement on social media and instant messaging tools may increase the chances of experiencing cyberbullying. For example, in Spain, these tools cannot be used without parental permission before the age of 14 ( 55 ). Besides, senior students were more likely to be more impulsive and less sympathetic. They may portray more aggressive and anti-social behaviors ( 55 , 72 ); hence senior students and students with higher impulsivity were usually more likely to become cyberbullying perpetrators.

Past experiences of victimization and family-related factors are another risk for cyberbullying crime. As for past experiences, one possible explanation is that young people who had experienced online or traditional school bullying may commit cyberbullying using e-mails, instant messages, and text messages for revenge, self-protection, or improving their social status ( 35 , 42 , 49 , 55 ). In becoming a cyberbullying perpetrator, the student may feel more powerful and superior, externalizing angry feelings and relieving the feelings of helplessness and sadness produced by past victimization experiences ( 51 ). As for family related factors, parenting styles are proven to be highly correlated to cyberbullying. In authoritative families, parents focus on rational behavioral control with clear rules and a high component of supervision and parental warmth, which have beneficial effects on children's lifestyles ( 43 ). Conversely, in indulgent families, children's behaviors are not heavily restricted and parents guide and encourage their children to adapt to society. The characteristics of this indulgent style, including parental support, positive communication, low imposition, and emotional expressiveness, possibly contribute to more parent-child trust and less misunderstanding ( 75 ). The protective role of warmth/affection and appropriate supervision, which are common features of authoritative or indulgent parenting styles, mitigate youth engagement in cyberbullying. On the contrary, authoritarian and neglectful styles, whether with excessive or insufficient control, are both proven to be risk factors for being a target of cyberbullying ( 33 , 76 ). In terms of geographical location, although several studies found that children residing in urban areas were more likely to be cyberbullying victims than those living in rural or suburban areas, we cannot draw a quick conclusion here, since whether this difference attributes to macro-level differences, such as community safety or socioeconomic status, or micro-level differences, such as teacher intervention in the classroom, courses provided, teacher-student ratio, is unclear across studies ( 61 ). An alternative explanation for this is the higher internet usage rate in urban areas ( 49 ).

Regarding health conditions, especially mental health, some scholars believe that young people with health problems are more likely to be identified as victims than people without health problems. They perceive health condition as a risk factor for cyberbullying ( 61 , 63 ). On the other hand, another group of scholars believe that cyberbullying has an important impact on the mental health of adolescents which can cause psychological distress consequences, such as post-traumatic stress mental disorder, depression, suicidal ideation, and drug abuse ( 70 , 87 ). It is highly possible that mental health could be risk factors, consequences of cyberbullying or both. Mental health cannot be used as standards, requirements, or decisive responses in cyberbullying research ( 13 ).

The Joint Effort Between Youth, Parents, Schools, and Communities to Form a Cyberbullying-Free Environment

This comprehensive review suggests that protecting children and adolescents from cyberbullying requires joint efforts between individuals, parents, schools, and communities, to form a cyberbullying-free environment. For individuals, young people are expected to improve their digital technology capabilities, especially in the use of social media platforms and instant messaging tools ( 55 ). To reduce the number of cyberbullying perpetrators, it is necessary to cultivate emotional self-regulation ability through appropriate emotional management training. Moreover, teachers, counselors, and parents are required to be armed with sufficient knowledge of emotional management and to develop emotional management capabilities and skills. In this way, they can be alert to the aggressive or angry emotions expressed by young people, and help them mediate any negative emotions ( 45 ), and avoid further anti-social behaviors ( 57 ).

For parents, styles of parenting involving a high level of parental involvement, care and support, are desirable in reducing the possibility of children's engagement in cyberbullying ( 74 , 75 ). If difficulties are encountered, open communication can contribute to enhancing the sense of security ( 73 ). In this vein, parents should be aware of the importance of caring, communicating and supervising their children, and participate actively in their children's lives ( 71 ). In order to keep a balance between control and openness ( 47 ), parents can engage in unbiased open communication with their children, and reach an agreement on the usage of computers and smart phones ( 34 , 35 , 55 ). Similarly, it is of vital importance to establish a positive communication channel with children ( 19 ).

For schools, a higher priority is needed to create a safe and positive campus environment, providing students with learning opportunities and ensuring that every student is treated equally. With a youth-friendly environment, students are able to focus more on their academic performance and develop a strong sense of belonging to the school ( 79 ). For countries recognizing collectivist cultural values, such as China and India, emphasizing peer attachment and a sense of collectivism can reduce the risk of cyberbullying perpetration and victimization ( 78 ). Besides, schools can cooperate with mental health agencies and neighboring communities to develop preventive programs, such as extracurricular activities and training ( 44 , 53 , 62 ). Specifically, school-based preventive measures against cyberbullying are expected to be sensitive to the characteristics of young people at different ages, and the intersection of race and school diversity ( 29 , 76 ). It is recommended that school policies that aim to embrace diversity and embody mutual respect among students are created ( 26 ). Considering the high prevalence of cyberbullying and a series of serious consequences, it is suggested that intervention against cyberbullying starts from an early stage, at about 10 years old ( 54 ). Schools can organize seminars to strengthen communication between teachers and students so that they can better understand the needs of students ( 61 ). In addition, schools should encourage cyberbullying victims to seek help and provide students with opportunities to report cyberbullying behaviors, such as creating online anonymous calls.

Conclusions and Limitations

The comprehensive study has reviewed related research on children and adolescents cyberbullying across different countries and regions, providing a positive understanding of the current situation of cyberbullying. The number of studies on cyberbullying has surged in the last 5 years, especially those related to risk factors and protective factors of cyberbullying. However, research on effective prevention is insufficient and evaluation of policy tools for cyberbullying intervention is a nascent research field. Our comprehensive review concludes with possible strategies for cyberbullying prevention, including personal emotion management, digital ability training, policy applicability, and interpersonal skills. We highlight the important role of parental control in cyberbullying prevention. As for the role of parental control, it depends on whether children believe their parents are capable of adequately supporting them, rather than simply interfering in their lives, restricting their online behavior, and controlling or removing their devices ( 50 ). In general, cyberbullying is on the rise, with the effectiveness of interventions to meet this problem still requiring further development and exploration ( 83 ).

Considering the overlaps between cyberbullying and traditional offline bullying, future research can explore the unique risk and protective factors that are distinguishable from traditional bullying ( 86 ). To further reveal the variations, researchers can compare the outcomes of interventions conducted in cyberbullying and traditional bullying preventions simultaneously, and the same interventions only targeting cyberbullying ( 88 ). In addition, cyberbullying also reflects a series of other social issues, such as personal privacy and security, public opinion monitoring, multinational perpetration and group crimes. To address this problem, efforts from multiple disciplines and novel analytical methods in the digital era are required. As the Internet provides enormous opportunities to connect young people from all over the world, cyberbullying perpetrators may come from transnational networks. Hence, cyberbullying of children and adolescents, involving multiple countries, is worth further attention.

Our study has several limitations. First, national representative studies are scarce, while few studies from middle and low income countries were included in our research due to language restrictions. Many of the studies included were conducted in schools, communities, provinces, and cities in high income countries. Meanwhile, our review only focused on victimization and perpetration. Future studies should consider more perspectives, such as bystanders and those with the dual identity of victim/perpetrator, to comprehensively analyze the risk and protective factors of cyberbullying.

Data Availability Statement

The original contributions presented in the study are included in the article/ Supplementary Material , further inquiries can be directed to the corresponding author/s.

Author Contributions

SH, CZ, RE, and WZ conceived the study and developed the design. WZ analyzed the result and supervised the study. CZ and SH wrote the first draft. All authors contributed to the article and approved the submitted version.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2021.634909/full#supplementary-material

1. Ang RP. Adolescent cyberbullying: a review of characteristics, prevention and intervention strategies. Aggress Violent Behav. (2015) 25:35–42. doi: 10.1016/j.avb.2015.07.011

CrossRef Full Text | Google Scholar

2. Reyna VF, Farley F. Risk and rationality in adolescent decision making: implications for theory, practice, and public policy. Psychol Sci Public Interest. (2006) 7:1–44. doi: 10.1111/j.1529-1006.2006.00026.x

PubMed Abstract | CrossRef Full Text | Google Scholar

3. UNICEF ed. Children in a Digital World . New York, NY: UNICEF (2017).

Google Scholar

4. Thomas HJ, Connor JP, Scott JG. Integrating traditional bullying and cyberbullying: challenges of definition and measurement in adolescents - a review. Educ Psychol Rev. (2015) 27:135–52. doi: 10.1007/s10648-014-9261-7

5. Baldry AC, Farrington DP, Sorrentino A. “Am I at risk of cyberbullying”? A narrative review and conceptual framework for research on risk of cyberbullying and cybervictimization: the risk and needs assessment approach. Aggress Violent Behav. (2015) 23:36–51. doi: 10.1016/j.avb.2015.05.014

6. Olweus D. Bullying at School: What We Know and What We Can Do . Oxford; Cambridge, MA: Blackwell (1993).

PubMed Abstract | Google Scholar

7. Dooley JJ, Pyzalski J, Cross D. Cyberbullying versus face-to-face bullying: a theoretical and conceptual review. J Psychol. (2009) 217:182–8. doi: 10.1027/0044-3409.217.4.182

8. Smith PK, Mahdavi J, Carvalho M, Fisher S, Russell S, Tippett N. Cyberbullying: its nature and impact in secondary school pupils. J Child Psychol Psychiatry. (2008) 49:376–85. doi: 10.1111/j.1469-7610.2007.01846.x

9. Kowalski RM, Giumetti GW, Schroeder AN, Lattanner MR. Bullying in the digital age: a critical review and meta-analysis of cyberbullying research among youth. Psychol Bull. (2014) 140:1073–137. doi: 10.1037/a0035618

10. León Vicente I. Cybervictimization by cyberbullying: children at risk and children as risk (dissertation). University of the Basque Country, Leioa, Spain (2016).

11. Hinduja S, Patchin JW. Identification, Prevention, and Response. (2020).

12. Jadambaa A, Thomas HJ, Scott JG, Graves N, Brain D, Pacella R. Prevalence of traditional bullying and cyberbullying among children and adolescents in Australia: a systematic review and meta-analysis. Aust N Z J Psychiatry. (2019) 53:878–88. doi: 10.1177/0004867419846393

13. Pereira F, Matos M. Cyber-stalking victimization: what predicts fear among Portuguese adolescents? Eur J Crim Policy Res. (2016) 22:253–70. doi: 10.1007/s10610-015-9285-7

14. Reed LA, Ward LM, Tolman RM, Lippman JR, Seabrook RC. The association between stereotypical gender and dating beliefs and digital dating abuse perpetration in adolescent dating relationships. J Interpers Violence . (2018). doi: 10.1177/0886260518801933

15. Huang CL, Yang SC, Hsieh LS. The cyberbullying behavior of Taiwanese adolescents in an online gaming environment. Children Youth Serv Rev. (2019) 106:104461. doi: 10.1016/j.childyouth.2019.104461

16. Raskauskas J, Huynh A. The process of coping with cyberbullying: a systematic review. Aggress Violent Behav. (2015) 23:118–25. doi: 10.1016/j.avb.2015.05.019

17. Bradshaw J, Crous G, Rees G, Turner N. Comparing children's experiences of schools-based bullying across countries. Children Youth Serv Rev. (2017) 80:171–80. doi: 10.1016/j.childyouth.2017.06.060

18. Hutson E, Kelly S, Militello LK. Systematic review of cyberbullying interventions for youth and parents with implications for evidence-based practice. Worldviews Evid Based Nurs. (2018) 15:72–9. doi: 10.1111/wvn.12257

19. Larranaga E, Yubero S, Ovejero A, Navarro R. Loneliness, parent-child communication and cyberbullying victimization among Spanish youths. Comp Hum Behav. (2016) 65:1–8. doi: 10.1016/j.chb.2016.08.015

20. van Geel M, Vedder P, Tanilon J. Relationship between peer victimization, cyberbullying, and suicide in children and adolescents: a meta-analysis. JAMA Pediatr. (2014) 168:435–42. doi: 10.1001/jamapediatrics.2013.4143

21. Zych I, Ortega-Ruiz R, Del Rey R. Systematic review of theoretical studies on bullying and cyberbullying: facts, knowledge, prevention, and intervention. Aggress Violent Behav. (2015) 23:1–21. doi: 10.1016/j.avb.2015.10.001

22. Selkie EM, Kota R, Chan Y-F, Moreno M. Cyberbullying, depression, and problem alcohol use in female college students: a multisite study. Cyberpsychol Behav Soc Netw. (2015) 18:79–86. doi: 10.1089/cyber.2014.0371

23. Brochado S, Soares S, Fraga S. A scoping review on studies of cyberbullying prevalence among adolescents. Trauma Violence Abus. (2017) 18:523–31. doi: 10.1177/1524838016641668

24. Nocentini A, Calmaestra J, Schultze-Krumbholz A, Scheithauer H, Ortega R, Menesini E. Cyberbullying: labels, behaviours and definition in three European Countries. Aust J Guid Couns. (2010) 20:129–42. doi: 10.1375/ajgc.20.2.129

25. Willard NE. Cyberbullying and Cyberthreats: Responding to the Challenge of Online Social Aggression, Threats, and Distress . Champaign: Research Press (2007).

26. Aizenkot D, Kashy-Rosenbaum G. Cyberbullying victimization in whatsapp classmate groups among Israeli Elementary, Middle, and High School Students. J Interpers Violence . (2019). doi: 10.1177/0886260519842860

27. Anderson CA, Bushman BJ. Human aggression. Ann Rev Psychol. (2002) 53:27–51. doi: 10.1146/annurev.psych.53.100901.135231

28. Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions . (2011). Available online at: http://handbook-5-1.cochrane.org/ (accessed January 17, 2021).

29. Alhajji M, Bass S, Dai T. Cyberbullying, mental health, and violence in adolescents and associations with sex and race: data from the 2015 youth risk behavior survey. Global Pediatr Health . (2019) 6:2333794X19868887. doi: 10.1177/2333794X19868887

30. Grinshteyn E, Yang YT. The association between electronic bullying and school absenteeism among high school students in the United States. J School Health. (2017) 87:142–9. doi: 10.1111/josh.12476

31. Mesch GS. Parent-child connections on social networking sites and cyberbullying. Youth Soc. (2018) 50:1145–62. doi: 10.1177/0044118X16659685

32. Sam J, Wisniewski P, Xu H, Rosson MB, Carroll JM. How are social capital and parental mediation associated with cyberbullying and cybervictimization among youth in the United States? In: Stephanidis C, editor. HCI International 2017 – Posters' Extended Abstracts Communications in Computer Information Science Cham: Springer International Publishing. p. 638–644.

33. Hong JS, Kim DH, Thornberg R, Kang JH, Morgan JT. Correlates of direct and indirect forms of cyberbullying victimization involving South Korean adolescents: an ecological perspective. Comput Hum Behav. (2018) 87:327–36. doi: 10.1016/j.chb.2018.06.010

34. Lee C, Shin N. Prevalence of cyberbullying and predictors of cyberbullying perpetration among Korean adolescents. Comp Hum Behav. (2017) 68:352–8. doi: 10.1016/j.chb.2016.11.047

35. You S, Lim SA. Longitudinal predictors of cyberbullying perpetration: evidence from Korean middle school students. Person Ind Differ. (2016) 89:172–6. doi: 10.1016/j.paid.2015.10.019

36. Holt TJ, Fitzgerald S, Bossler AM, Chee G, Ng E. Assessing the risk factors of cyber and mobile phone bullying victimization in a nationally representative sample of Singapore Youth. Int J Offend Ther Comp Criminol. (2016) 60:598–615. doi: 10.1177/0306624X14554852

37. Marret MJ, Choo WY. Factors associated with online victimisation among Malaysian adolescents who use social networking sites: a cross-sectional study. BMJ Open. (2017) 7:e014959. doi: 10.1136/bmjopen-2016-014959

38. Tesler R, Nissanholtz-Gannot R, Zigdon A, Harel-Fisch Y. The association of cyber-bullying and adolescents in religious and secular schools in Israel. J Relig Health. (2019) 58:2095–109. doi: 10.1007/s10943-019-00938-z

39. Beran T, Mishna F, McInroy LB, Shariff S. Children's experiences of cyberbullying: a Canadian National Study. Child School. (2015) 37:207–14. doi: 10.1093/cs/cdv024

40. World Bank Country and Lending Groups – World Bank Data Help Desk. Available online at: https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups [accessed January 25, 2021).

41. Marco JH, Tormo-Irun P. Cyber victimization is associated with eating disorder psychopathology in adolescents. Front Psychol. (2018) 9:987. doi: 10.3389/fpsyg.2018.00987

42. Olenik-Shemesh D, Heiman T. Cyberbullying victimization in adolescents as related to body esteem, social support, and social self-efficacy. J Genet Psychol. (2017) 178:28–43. doi: 10.1080/00221325.2016.1195331

43. Rao J, Wang H, Pang M, Yang J, Zhang J, Ye Y, et al. Cyberbullying perpetration and victimisation among junior and senior high school students in Guangzhou, China. Inj Prev. (2019) 25:13–9. doi: 10.1136/injuryprev-2016-042210

44. Lee C. Weak Commitment to School, Deviant Peers, and Cyberbullying Victimization-Strain in Adolescent Cyberbullying . (2017). Available online at: https://search.proquest.com/pqdtglobal/docview/1937915933/abstract/9D85437600564444PQ/30 (accessed June 17, 2020).

45. Lin MT. Risk factors associated with cyberbullying victimization and perpetration among Taiwan children (dissertation). The University of Texas, Austin, TX, United States (2019).

46. Pieschl S, Porsch T. The complex relationship between cyberbullying and trust. Int J Dev Sci. (2017) 11:9–17. doi: 10.3233/DEV-160208

47. Brighi A, Menin D, Skrzypiec G, Guarini A. Young, bullying, and connected. Common pathways to cyberbullying and problematic internet use in adolescence. Front Psychol. (2019) 10:1467. doi: 10.3389/fpsyg.2019.01467

48. Festl R. Perpetrators on the internet: analyzing individual and structural explanation factors of cyberbullying in school context. Comp Hum Behav. (2016) 59:237–48. doi: 10.1016/j.chb.2016.02.017

49. Olumide AO, Adams P, Amodu OK. Prevalence and correlates of the perpetration of cyberbullying among in-school adolescents in Oyo State, Nigeria. Int J Adolesc Med Health. (2016) 28:183–91. doi: 10.1515/ijamh-2015-0009

50. Baldry AC, Sorrentino A, Farrington DP. Cyberbullying and cybervictimization versus parental supervision, monitoring and control of adolescents' online activities. Child Youth Serv Rev. (2019) 96:302–7. doi: 10.1016/j.childyouth.2018.11.058

51. Garaigordobil M. Cyberbullying in adolescents and youth in the Basque Country: prevalence of cybervictims, cyberaggressors, and cyberobservers. J Youth Stud. (2015) 18:569–82. doi: 10.1080/13676261.2014.992324

52. Morin HK, Bradshaw CP, Kush JM. Adjustment outcomes of victims of cyberbullying: the role of personal and contextual factors. J School Psychol. (2018) 70:74–88. doi: 10.1016/j.jsp.2018.07.002

53. Baraldsnes D. The prevalence of cyberbullying and the views of 5-12 grade pupils and teachers on cyberbullying prevention in Lithuanian Schools. Uinv J Educ Res. (2015) 3:949–59. doi: 10.13189/ujer.2015.031201

54. Razjouyan K, Mobarake AH, Sadr SS, Ardestani SMS, Yaseri M. The relationship between emotional intelligence and the different roles in cyberbullying among high school students in Tehran. Iran J Psychiatry Behav Sci. (2018) 12:UNSP e11560. doi: 10.5812/ijpbs.11560

55. Alvarez-Garcia D, Carlos Nunez J, Garcia T, Barreiro-Collazo A. Individual, family, and community predictors of cyber-aggression among adolescents. Eur J Psychol Appl Legal Context. (2018) 10:79–88. doi: 10.5093/ejpalc2018a8

56. Horzum MB, Ayas T, Randler C, Dusunceli B. The effects of empathy and circadian preference on cyberbullying of adolescents in Turkey. Biol Rhythm Res . (2019). doi: 10.1080/09291016.2019.1603839

57. Carmen Martinez-Monteagudo M, Delgado B, Manuel Garcia-Fernandez J, Rubio E. Cyberbullying, aggressiveness, and emotional intelligence in adolescence. Int J Environ Res Public Health. (2019) 16:5079. doi: 10.3390/ijerph16245079

58. Sasson H, Mesch G. The role of parental mediation and peer norms on the likelihood of cyberbullying. J Genet Psychol. (2017) 178:15–27. doi: 10.1080/00221325.2016.1195330

59. Wang X, Lei L, Liu D, Hu H. Moderating effects of moral reasoning and gender on the relation between moral disengagement and cyberbullying in adolescents. Person Ind Differ. (2016) 98:244–9. doi: 10.1016/j.paid.2016.04.056

60. Simsek N, Sahin D, Evli M. Internet addiction, cyberbullying, and victimization relationship in adolescents a sample from Turkey. J Addict Nurs. (2019) 30:201–10. doi: 10.1097/JAN.0000000000000296

61. McQuillan BE. Ecological Factors Associated with Middle School Students' Experiences of Cyberbullying . (2016). Available online at: https://search.proquest.com/pqdtglobal/docview/1794167537/abstract/9D85437600564444PQ/4 (accessed June 17, 2020).

62. Rose CA, Tynes BM. Longitudinal associations between cybervictimization and mental health among U.S. adolescents. J Adolesc Health. (2015) 57:305–12. doi: 10.1016/j.jadohealth.2015.05.002

63. Stockdale LA, Coyne SM, Nelson DA, Erickson DH. Borderline personality disorder features, jealousy, and cyberbullying in adolescence. Pers Individ Differ. (2015) 83:148–53. doi: 10.1016/j.paid.2015.04.003

64. Chen Q, Lo Camilla KM, Yuhong Z, Anne C, Ling CK, Patrick I. Family poly-victimization and cyberbullying among adolescents in a Chinese school sample. Child Abuse Negl. (2018) 77:180–7. doi: 10.1016/j.chiabu.2018.01.015

65. Landoll RR, La Greca AM, Lai BS, Chan SF, Herge WM. Cyber victimization by peers: prospective associations with adolescent social anxiety and depressive symptoms. J Adolesc. (2015) 42:77–86. doi: 10.1016/j.adolescence.2015.04.002

66. Iranzo B, Buelga S, Cava M-J, Ortega-Baron J. Cyberbullying, psychosocial adjustment, and suicidal ideation in adolescence. Psychosoc Interv. (2019) 28:75–81. doi: 10.5093/pi2019a5

67. Buelga S, Cava MJ, Musitu G, Torralba E. Cyberbullying aggressors among Spanish secondary education students: an exploratory study. Interact Tech Smart Ed. (2015) 12:100–15. doi: 10.1108/ITSE-08-2014-0025

68. Katz I, Lemish D, Cohen R, Arden A. When parents are inconsistent: parenting style and adolescents' involvement in cyberbullying. J Adolesc. (2019) 74:1–2. doi: 10.1016/j.adolescence.2019.04.006

69. Cénat JM, Blais M, Lavoie F, Caron P-O, Hébert M. Cyberbullying victimization and substance use among Quebec high schools students: the mediating role of psychological distress. Comp Hum Behav. (2018) 89:207–12. doi: 10.1016/j.chb.2018.08.014

70. Hoareau N, Bages C, Allaire M, Guerrien A. The role of psychopathic traits and moral disengagement in cyberbullying among adolescents. Crim Behav Ment Health. (2019) 29:321–31. doi: 10.1002/cbm.2135

71. Khurana A, Bleakley A, Jordan A, Romer D. The protective effects of parental monitoring and internet restriction on adolescents' risk of online harassment. J Youth Adolesc. (2015) 44:1039–47. doi: 10.1007/s10964-014-0242-4

72. Martínez I, Murgui S, Garcia OF, Garcia F. Parenting in the digital era: protective and risk parenting styles for traditional bullying and cyberbullying victimization. Comp Hum Behav. (2019) 90:84–92. doi: 10.1016/j.chb.2018.08.036

73. Yusuf S, Salleh H, Bahaman A, Shamsul M, Ramli N, Ramli AN, et al. Parental attachment and cyberbullying experiences among Malaysian children. Pertanika J Scholarly Res Rev . (2018) 4:67–80.

74. Martinez-Ferrer B, Leon-Moreno C, Musitu-Ferrer D, Romero-Abrio A, Callejas-Jeronimo JE, Musitu-Ochoa G. Parental socialization, school adjustment and cyber-aggression among adolescents. Int J Environ Res Public Health. (2019) 16:4005. doi: 10.3390/ijerph16204005

75. Moreno–Ruiz D, Martínez–Ferrer B, García–Bacete F. Parenting styles, cyberaggression, and cybervictimization among adolescents. Comp Hum Behav. (2019) 93:252–9. doi: 10.1016/j.chb.2018.12.031

76. Ho SS, Chen L, Ng APY. Comparing cyberbullying perpetration on social media between primary and secondary school students. Comp Educ. (2017) 109:74–84. doi: 10.1016/j.compedu.2017.02.004

77. Gómez-Ortiz O, Romera EM, Ortega-Ruiz R, Del Rey R. Parenting practices as risk or preventive factors for adolescent involvement in cyberbullying: contribution of children and parent gender. Int J Environ Res Public Health. (2018) 15:2664. doi: 10.3390/ijerph15122664

78. Wright MF, Kamble SV, Soudi SP. Indian adolescents' cyber aggression involvement and cultural values: the moderation of peer attachment. Sch Psychol Int. (2015) 36:410–27. doi: 10.1177/0143034315584696

79. Holfeld B, Leadbeater BJ. Concurrent and longitudinal associations between early adolescents' experiences of school climate and cyber victimization. Comput Hum Behav. (2017) 76:321–8. doi: 10.1016/j.chb.2017.07.037

80. Álvarez-García D, Núñez JC, González-Castro P, Rodríguez C, Cerezo R. The effect of parental control on cyber-victimization in adolescence: the mediating role of impulsivity and high-risk behaviors. Front Psychol. (2019) 10:1159. doi: 10.3389/fpsyg.2019.01159

81. Modecki KL, Minchin J, Harbaugh AG, Guerra NG, Runions KC. Bullying prevalence across contexts: a meta-analysis measuring cyber and traditional bullying. J Adolesc Health. (2014) 55:602–11. doi: 10.1016/j.jadohealth.2014.06.007

82. Hamm MP, Newton AS, Chisholm A, Shulhan J, Milne A, Sundar P, et al. Prevalence and effect of cyberbullying on children and young people: a scoping review of social media studies. JAMA Pediatr. (2015) 169:770. doi: 10.1001/jamapediatrics.2015.0944

83. Gaffney H, Farrington DP, Espelage DL, Ttofi MM. Are cyberbullying intervention and prevention programs effective? A systematic and meta-analytical review. Aggress Violent Behav. (2019) 45:134–53. doi: 10.1016/j.avb.2018.07.002

84. Selkie EM, Fales JL, Moreno MA. Cyberbullying prevalence among US middle and high school-aged adolescents: a systematic review and quality assessment. J Adolesc Health. (2016) 58:125–33. doi: 10.1016/j.jadohealth.2015.09.026

85. Ybarra ML, Boyd D, Korchmaros JD, Oppenheim J. Defining and measuring cyberbullying within the larger context of bullying victimization. J Adolesc Health. (2012) 51:53–8. doi: 10.1016/j.jadohealth.2011.12.031

86. Kowalski RM, Limber SP, McCord A. A developmental approach to cyberbullying: prevalence and protective factors. Aggress Violent Behav. (2019) 45:20–32. doi: 10.1016/j.avb.2018.02.009

87. Dilmac B, Yurt E, Aydin M, Kasarci I. Predictive relationship between humane values of adolescents cyberbullying and cyberbullying sensibility. Electron J Res Educ Psychol. (2016) 14:3–22. doi: 10.14204/ejrep.38.14123

88. Reed KP, Cooper RL, Nugent WR, Russell K. Cyberbullying: a literature review of its relationship to adolescent depression and current intervention strategies. J Hum Behav Soc Environ. (2016) 26:37–45. doi: 10.1080/10911359.2015.1059165

Keywords: cyberbullying, children, adolescents, globalization, risk factors, preventive measures

Citation: Zhu C, Huang S, Evans R and Zhang W (2021) Cyberbullying Among Adolescents and Children: A Comprehensive Review of the Global Situation, Risk Factors, and Preventive Measures. Front. Public Health 9:634909. doi: 10.3389/fpubh.2021.634909

Received: 29 November 2020; Accepted: 10 February 2021; Published: 11 March 2021.

Reviewed by:

Copyright © 2021 Zhu, Huang, Evans and Zhang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

*Correspondence: Wei Zhang, weizhanghust@hust.edu.cn

† These authors have contributed equally to this work and share first authorship

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Cyberbullying detection and machine learning: a systematic literature review

• Published: 24 July 2023
• Volume 56 , pages 1375–1416, ( 2023 )

Cite this article

• Vimala Balakrisnan 1 &
• Mohammed Kaity 1  

778 Accesses

Explore all metrics

The rise in research work focusing on detection of cyberbullying incidents on social media platforms particularly reflect how dire cyberbullying consequences are, regardless of age, gender or location. This paper examines scholarly publications (i.e., 2011–2022) on cyberbullying detection using machine learning through a systematic literature review approach. Specifically, articles were sought from six academic databases (Web of Science, ScienceDirect, IEEE Xplore, Association for Computing Machinery, Scopus, and Google Scholar), resulting in the identification of 4126 articles. A redundancy check followed by eligibility screening and quality assessment resulted in 68 articles included in this review. This review focused on three key aspects, namely, machine learning algorithms used to detect cyberbullying, features, and performance measures, and further supported with classification roles, language of study, data source and type of media. The findings are discussed, and research challenges and future directions are provided for researchers to explore.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

A survey on sentiment analysis methods, applications, and challenges

Mayur Wankhade, Annavarapu Chandra Sekhara Rao & Chaitanya Kulkarni

The disaster of misinformation: a review of research in social media

Sadiq Muhammed T & Saji K. Mathew

Artificial Intelligence and Fraud Detection

https://trends.google.com/trends/explore?date=2011-01-01%202022-12-31&q=deep%20learning&hl=en .

Agrawal S, Awekar A (2018) Deep learning for detecting cyberbullying across multiple social media platforms. In European Conference on Information Retrieval (pp. 141–153). Springer, Cham

Aizenkot D, Kashy-Rosenbaum G (2018) Cyberbullying in WhatsApp classmates’ groups: evaluation of an intervention program implemented in israeli elementary and middle schools. New Media & Society 20(12):4709–4727

Article   Google Scholar  

Akhter MP, Zheng JB, Naqvi IR, Abdelmajeed M, Sadiq MT (2020) Automatic Detection of Offensive Language for Urdu and Roman Urdu. IEEE Access 8:91213–91226.

Aldhyani TH, Al-Adhaileh MH, Alsubari SN (2022) Cyberbullying identification system based deep learning algorithms. Electronics 11(20):3273

Al-Garadi MA, Hussain MR, Khan N, Murtaza G, Nweke HF, Ali I, …, Gani A (2019) Predicting cyberbullying on social media in the big data era using machine learning algorithms: review of literature and open challenges. IEEE Access 7:70701–70718

Al-garadi MA, Varathan KD, Ravana SD (2016) Cybercrime detection in online communications: the experimental case of cyberbullying detection in the Twitter network. Comput Hum Behav 63:433–443

Al-Harigy LM, Al-Nuaim HA, Moradpoor N, Tan Z (2022) Building towards Automated Cyberbullying Detection: A Comparative Analysis. Computational Intelligence and Neuroscience, 2022

Alom Z, Carminati B, Ferrari E (2020) A deep learning model for Twitter spam detection. Online Social Networks and Media 18:100079

Alpaydin E (2010) Introduction to machine learning, 2nd edn. MIT Press

Ates EC, Bostanci E, Guzel MS (2021) Comparative performance of machine learning algorithms in cyberbullying detection: using turkish language preprocessing techniques. arXiv preprint arXiv :2101.12718

Ayo FE, Folorunso O, Ibharalu FT, Osinuga IA (2020) Machine learning techniques for hate speech classification of twitter data: state-of-the-art, future challenges and research directions. Comput Sci Rev 38:100311

Balakrishnan V (2015) Cyberbullying among young adults in Malaysia: the roles of gender, age and internet frequency. Comput Hum Behav 46:149–157

Balakrishnan V, Khan S, Arabnia HR (2020a) Improving cyberbullying detection using Twitter users’ psychological features and machine learning. Computers & Security 90:101710

Balakrishnan V, Khan S, Arabnia HR (2020b) Improving cyberbullying detection using Twitter users’ psychological features and machine learning. Computers & Security 90:101710

Balakrishnan V, Khan S, Fernandez T, Arabnia HR (2019) Cyberbullying detection on twitter using Big Five and Dark Triad features. Pers Individ Differ 141, 252–257.

Bretschneider U, Wöhner T, Peters R (2014) Detecting online harassment in social networks.

Buan TA, Ramachandra R (2020) Automated Cyberbullying Detection in Social Media Using an SVM Activated Stacked Convolution LSTM Network. In Proceedings of the 2020 the 4th International Conference on Compute and Data Analysis (pp. 170–174)

Camerini AL, Marciano L, Carrara A, Schulz PJ (2020) Cyberbullying perpetration and victimization among children and adolescents: a systematic review of longitudinal studies. Telematics Inform 49:101362

Chatzakou D, Kourtellis N, Blackburn J, De Cristofaro E, Stringhini G, Vakali A (2017) Mean birds: Detecting aggression and bullying on twitter. In Proceedings of the 2017 ACM on web science conference (pp. 13–22)

Chavan VS, Shylaja SS (2015) Machine learning approach for detection of cyber-aggressive comments by peers on social media network. In 2015 International Conference on Advances in Computing, Communications and Informatics (ICACCI) (pp. 2354–2358). IEEE

Cheng L, Guo R, Silva YN, Hall D, Liu H (2021) Modeling temporal patterns of cyberbullying detection with hierarchical attention networks. ACM/IMS Trans Data Sci 2(2):1–23

Cheng L, Li J, Silva YN, Hall DL, Liu H (2019) Xbully: Cyberbullying detection within a multi-modal context. In Proceedings of the Twelfth ACM International Conference on Web Search and Data Mining (pp. 339–347)

Chen Y, Zhou Y, Zhu S, Xu H (2012) Detecting offensive language in social media to protect adolescent online safety. In 2012 International Conference on Privacy, Security, Risk and Trust and 2012 International Confernece on Social Computing (pp. 71–80). IEEE

Dadvar M, De Jong F (2012) Cyberbullying detection: a step toward a safer internet yard. In Proceedings of the 21st International Conference on World Wide Web (pp. 121–126)

Dadvar M, Jong FD, Ordelman R, Trieschnigg D (2012) Improved cyberbullying detection using gender information. In Proceedings of the Twelfth Dutch-Belgian Information Retrieval Workshop (DIR 2012) . University of Ghent

Dadvar M, Trieschnigg D, Ordelman R, de Jong F (2013) Improving cyberbullying detection with user context. In European Conference on Information Retrieval (pp. 693–696). Springer, Berlin, Heidelberg

Dey R, Bag S, Sarkar RR (2021) Identification of stable housekeeping genes for normalization of qPCR data in a pathogenic fungus. J Microbiol Methods 180:106106

Google Scholar  

Dinakar K, Picard R, Lieberman H (2015) Common sense reasoning for detection, prevention, and mitigation of cyberbullying. In IJCAI International Joint Conference on Artificial Intelligence .

Dinakar K, Reichart R, Lieberman H (2011) Modeling the detection of textual cyberbullying. In Proceedings of the International Conference on Weblog and Social Media 2011

Divyashree VH, Deepashree NS (2016) An effective approach for cyberbullying detection and avoidance. International Journal of Innovative Research in Computer and Communication Engineering , 14

Djuraskovic O, Cyberbullying Statistics F (2020) and Trends with Charts: First Site Guide; 2020. Available from: https://firstsiteguide.com/cyberbullying-stats/

Elmezain M, Malki A, Gad I, Atlam ES (2022) Hybrid deep learning model–based prediction of images related to Cyberbullying. Int J Appl Math Comput Sci 32(2):323–334

MATH   Google Scholar  

Fahrnberger G, Nayak D, Martha VS, Ramaswamy S (2014) SafeChat: A tool to shield children’s communication from explicit messages. In 2014 14th International Conference on Innovations for Community Services (I4CS) (pp. 80–86). IEEE

Fang Y, Yang S, Zhao B, Huang C (2021) Cyberbullying detection in social networks using bi-gru with self-attention mechanism. Information 12(4):171

Foong YJ, Oussalah M (2017), September Cyberbullying system detection and analysis. In 2017 European Intelligence and Security Informatics Conference (EISIC) (pp. 40–46). IEEE

Galán-García P, Puerta JGDL, Gómez CL, Santos I, Bringas PG (2016) Supervised machine learning for the detection of troll profiles in twitter social network: application to a real case of cyberbullying. Log J IGPL 24(1):42–53

MathSciNet   Google Scholar  

García-Recuero Á (2016) Discouraging abusive behavior in privacy-preserving online social networking applications. In Proceedings of the 25th International Conference Companion on World Wide Web (pp. 305–309)

Ge S, Cheng L, Liu H (2021) Improving cyberbullying detection with user interaction. In Proceedings of the Web Conference 2021 (pp. 496–506)

Goodboy AK, Martin MM (2015) The personality profile of a cyberbully: examining the Dark Triad. Comput Hum Behav 49:1–4

Goodfellow I, Bengio Y, Courville A (2016) Deep learning. MIT Press

Haidar B, Chamoun M, Serhrouchni A (2017a) Multilingual cyberbullying detection system: Detecting cyberbullying in Arabic content. In 2017 1st Cyber Security in Networking Conference (CSNet) (pp. 1–8). IEEE

Haidar B, Chamoun M, Serhrouchni A (2017b) A multilingual system for cyberbullying detection: arabic content detection using machine learning. Adv Sci Technol Eng Syst J 2(6):275–284

Hani J, Nashaat M, Ahmed M, Emad Z, Amer E, Mohammed A (2019) Social media cyberbullying detection using machine learning. Int J Adv Comput Sci Appl 10(5):703–707

Hinduja S, Patchin JW (2010) Bullying, cyberbullying, and suicide. Archives of suicide research 14(3):206–221

Hosmer DW Jr, Lemeshow S, Sturdivant RX (2013) Applied Logistic Regression. Wiley

Hosseinmardi H, Mattson SA, Rafiq RI, Han R, Lv Q, Mishra S (2015b) Detection of cyberbullying incidents on the instagram social network. arXiv preprint arXiv:1503.03909

Hosseinmardi H, Mattson SA, Rafiq RI, Han R, Lv Q, Mishr S (2015a) Prediction of cyberbullying incidents on the instagram social network. arXiv preprint arXiv:1508.06257

Hosseinmardi H, Rafiq RI, Han R, Lv Q, Mishra S (2016) Prediction of cyberbullying incidents in a media-based social network. In 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) (pp. 186–192). IEEE

Huang Q, Singh VK, Atrey PK (2014) Cyber bullying detection using social and textual analysis. In Proceedings of the 3rd International Workshop on Socially-Aware Multimedia (pp. 3–6)

Hutter F, Kotthoff L, Vanschoren J (2019) Automated machine learning: methods, systems, challenges. Springer

Kaity M, Balakrishnan V (2019) An automatic non-english sentiment lexicon builder using unannotated corpus. J Supercomputing 75(4):2243–2268

Kelleher JD, Tierney B, Tierney B (2018) Data science: an introduction. CRC Press

Kitchenham B, Charters S (2007) Guidelines for performing systematic literature reviews in software engineering. Tech Rep EBSE 1:1–57

Koutsou A, Tjortjis C (2018) Predicting hospital readmissions using random forests. IEEE J Biomedical Health Inf 22(1):122–130

Kumar A, Nayak S, Chandra N (2019) Empirical analysis of supervised machine learning techniques for Cyberbullying detection. In International Conference on Innovative Computing and Communications (pp. 223–230). Springer, Singapore

Kumar A, Sachdeva N (2020) Multi-input integrative learning using deep neural networks and transfer learning for cyberbullying detection in real-time code-mix data. Multimedia Systems

Kumar A, Sachdeva N (2021) Multimodal cyberbullying detection using capsule network with dynamic routing and deep convolutional neural network. Multimedia Syst, 1–10

LeCun Y, Bengio Y, Hinton G (2015) Deep Learn Nat 521(7553):436–444

Li W, Li X (2021) Cyberbullying among college students: the roles of individual, familial, and cultural factors. Int J Environ Res Public Health 18(11):1–17

López-Vizcaíno MF, Nóvoa FJ, Carneiro V, Cacheda F (2021) Early detection of cyberbullying on social media networks. Future Generation Computer Systems 118:219–229

Lu N, Wu G, Zhang Z, Zheng Y, Ren Y, Choo KKR (2020) Cyberbullying detection in social media text based on character-level convolutional neural network with shortcuts. Concurrency and Computation: Practice and Experience, e5627

Maity K, Sen T, Saha S, Bhattacharyya P (2022) MTBullyGNN: a graph neural network-based Multitask Framework for Cyberbullying Detection. IEEE Transactions on Computational Social Systems

Malik CI, Radwan RB (2020) Adolescent victims of cyberbullying in Bangladesh- prevalence and relationship with psychiatric disorders. Asian J Psychiatr 48:101893

Mangaonkar A, Hayrapetian A, Raje R (2015) Collaborative detection of cyberbullying behavior in Twitter data. In 2015 IEEE international conference on electro/information technology (EIT) (pp. 611–616). IEEE

Manning CD, Raghavan P, Schütze H (2008) Introduction to Information Retrieval. Cambridge University Press

McEvoy MP, Williams MT (2021) Quality Assessment of systematic reviews and Meta-analyses of physical therapy interventions: a systematic review. Phys Ther 101(4):pzaa226

Mercado RNM, Chuctaya HFC, Gutierrez EGC (2018) Automatic cyberbullying detection in spanish-language social networks using sentiment analysis techniques. Int J Adv Comput Sci Appl 9(7):228–235

Monteiro RP, Santana MC, Santos RM, Pereira FC (2022) Cyberbullying victimization and mental health in higher education students: the mediating role of perceived social support. J interpers Violence, 1–23

Nahar V, Al-Maskari S, Li X, Pang C (2014) Semi-supervised learning for cyberbullying detection in social networks. In Australasian Database Conference (pp. 160–171). Springer, Cham

Nahar V, Unankard S, Li X, Pang C (2012) Sentiment analysis for effective detection of cyber bullying. Asia-Pacific Web Conference

Nandhini BS, Sheeba JI (2015) Online social network bullying detection using intelligence techniques. Procedia Comput Sci 45:485–492

Niu M, Yu L, Tian S, Wang X, Zhang Q (2020) Personal-bullying detection based on Multi-Attention and Cognitive Feature. Autom Control Comput Sci 54(1):52–61

Noviantho, Isa SM, Ashianti L (2018) Cyberbullying classification using text mining. In Proceedings - 2017 1st International Conference on Informatics and Computational Sciences , ICICoS 2017

Patil C, Salmalge S, Nartam P (2020) Cyberbullying detection on multiple SMPs using modular neural network. Advances in Cybernetics, Cognition, and machine learning for Communication Technologies. Springer, Singapore, pp 181–188

Chapter   Google Scholar  

Pawar R, Raje RR (2019) Multilingual Cyberbullying Detection System. In 2019 IEEE International Conference on Electro Information Technology (EIT) (pp. 040–044). IEEE

Pires TM, Nunes IL (2019) Support vector machine for human activity recognition: a comprehensive review. Artif Intell Rev 52(3):1925–1962

Pradhan A, Yatam VM, Bera P (2020) Self-Attention for Cyberbullying Detection. In 2020 International Conference on Cyber Situational Awareness, Data Analytics and Assessment (CyberSA) (pp. 1–6). IEEE

Pérez PJC, Valdez CJL, Ortiz MDGC, Barrera JPS, Pérez PF (2012) MISAAC: Instant messaging tool for cyberbullying detection. In Proceedings of the 2012 International Conference on Artificial Intelligence , ICAI 2012 (pp. 1049–1052)

Rafiq RI, Hosseinmardi H, Han R, Lv Q, Mishra S (2018) Scalable and timely detection of cyberbullying in online social networks. In Proceedings of the 33rd Annual ACM Symposium on Applied Computing (pp. 1738–1747)

Raisi E, Huang B (2018) Weakly supervised cyberbullying detection with participant-vocabulary consistency. Social Netw Anal Min 8(1):38

Reynolds K, Kontostathis A, Edwards L (2011) Using machine learning to detect cyberbullying. In 2011 10th International Conference on Machine learning and applications and workshops (Vol. 2, pp. 241–244). IEEE

Rosa H, Matos D, Ribeiro R, Coheur L, Carvalho JP (2018) A “deeper” look at detecting cyberbullying in social networks. In 2018 International Joint Conference on Neural Networks (IJCNN) (pp. 1–8). IEEE

Rosa H, Pereira N, Ribeiro R, Ferreira PC, Carvalho JP, Oliveira S, Coheur L, Paulino P, Veiga Simão AM, Trancoso I (2019) Automatic cyberbullying detection: A systematic review. Computers in Human Behavior, 93, 333–345

Salawu S, He Y, Lumsden J (2017) Approaches to automated detection of cyberbullying: a survey. IEEE Trans Affect Comput.

Sanchez H, Kumar S (2011) Twitter bullying detection. ser. NSDI , 12 (2011), 15

Shah N, Maqbool A, Abbasi AF (2021) Predictive modeling for cyberbullying detection in social media. J Ambient Intell Humaniz Comput 12(6):5579–5594

Singh A, Kaur, M (2020) Intelligent content-based cybercrime detection in online social networks using cuckoo search metaheuristic approach [Article]. J Supercomput 76(7):5402–5424

Singh VK, Ghosh S, Jose C (2017) Toward multimodal cyberbullying detection. In Proceedings of the 2017 CHI Conference Extended Abstracts on Human Factors in Computing Systems (pp. 2090–2099)

Soni D, Singh VK (2018) See no evil, hear no evil: Audio-visual-textual cyberbullying detection. Proceedings of the ACM on Human-Computer Interaction , 2 (CSCW), 1–26

Squicciarini A, Rajtmajer S, Liu Y, Griffin C (2015) Identification and characterization of cyberbullying dynamics in an online social network. In Proceedings of the 2015 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining 2015 (pp. 280–285)

Sugandhi R, Pande A, Agrawal A, Bhagat H (2016) Automatic monitoring and prevention of cyberbullying. Int J Comput Appl 8:17–19

Tahmasbi N, Rastegari E (2018) A socio-contextual approach in automated detection of public cyberbullying on Twitter. ACM Trans Social Comput 1(4):1–22

Tan SH, Zou W, Zhang J, Zhou Y (2020) Evaluation of machine learning algorithms for prediction of ground-level PM2.5 concentration using satellite-derived aerosol optical depth over China. Environ Sci Pollut Res 27(29):36155–36170

Tarwani S, Jethanandani M, Kant V (2019) Cyberbullying Detection in Hindi-English Code-Mixed Language Using Sentiment Classification. In International Conference on Advances in Computing and Data Sciences (pp. 543–551). Springer, Singapore

Tomkins S, Getoor L, Chen Y, Zhang Y (2018) A socio-linguistic model for cyberbullying detection. In 2018 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) (pp. 53–60). IEEE

van Geel M, Goemans A, Toprak F, Vedder P (2017) Which personality traits are related to traditional bullying and cyberbullying? A study with the big five, Dark Triad and sadism. Pers Indiv Differ 106:231–235

Van Hee C, Jacobs G, Emmery C, Desmet B, Lefever E, Verhoeven B, …, Hoste V (2018) Automatic detection of cyberbullying in social media text. PLoS ONE, 13(10), e0203794

Van Hee C, Lefever E, Verhoeven B, Mennes J, Desmet B, De Pauw G, …, Hoste V (2015) Detection and fine-grained classification of cyberbullying events. In International Conference Recent Advances in Natural Language Processing (RANLP) (pp. 672–680)

Wang W, Xie X, Wang X, Lei L, Hu Q, Jiang S (2019) Cyberbullying and depression among chinese college students: a moderated mediation model of social anxiety and neuroticism. J Affect Disord 256:54–61

Whiting P, Savović J, Higgins JP et al (2016) ROBIS: a new tool to assess risk of bias in systematic reviews was developed. J Clin Epidemiol 69:225–234

Witten IH, Frank E, Hall MA (2016) Data Mining: practical machine learning tools and techniques, 4th edn. Morgan Kaufmann Publishers

Wright MF (2017) Cyberbullying in cultural context. J Cross-Cult Psychol 48(8):1136–1137

Wu J, Wen M, Lu R, Li B, Li J (2020) Toward efficient and effective bullying detection in online social network. Peer-to-Peer Netw Appl, 1–10

Wu T, Wen S, Xiang Y, Zhou W (2018) Twitter spam detection: survey of new approaches and comparative study. Computers & Security 76:265–284

Yin D, Xue Z, Hong L, Davison BD, Kontostathis A, Edwards L (2009) Detection of harassment on web 2.0. Proceedings of the Content Analysis in the WEB , 2 , 1–7

Zhang X, Tong J, Vishwamitra N, Whittaker E, Mazer JP, Kowalski R, Hu H, Luo F, Macbeth J, Dillon E (2017) Cyberbullying detection with a pronunciation based convolutional neural network. In Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016

Zhao R, Mao K (2017) Cyberbullying detection based on semantic-enhanced marginalized denoising auto-encoder. IEEE Trans Affect Comput 8(3), 328–339. Article 7412690

Zhao R, Zhou A, Mao K (2016) Automatic detection of cyberbullying on social networks based on bullying features. In Proceedings of the 17th international conference on distributed computing and networking (pp. 1–6)

Zhong H, Li H, Squicciarini AC, Rajtmajer SM, Griffin C, Miller DJ, Caragea C (2016) Content-Driven Detection of Cyberbullying on the Instagram Social Network. In IJCAI (pp. 3952–3958)

Download references

Author information

Authors and affiliations.

Faculty of Computer Science and Information Systems, Universiti Malaya, Kuala Lumpur, 50603, Malaysia

Vimala Balakrisnan & Mohammed Kaity

You can also search for this author in PubMed   Google Scholar

Contributions

VB wrote the original draft; performed analysis; revised the article; MK performed data collection; performed analysis; revised the article; All authors reviewed the manuscript.

Corresponding author

Correspondence to Vimala Balakrisnan .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Balakrisnan, V., Kaity, M. Cyberbullying detection and machine learning: a systematic literature review. Artif Intell Rev 56 (Suppl 1), 1375–1416 (2023). https://doi.org/10.1007/s10462-023-10553-w

Download citation

Accepted : 08 July 2023

Published : 24 July 2023

Issue Date : October 2023

DOI : https://doi.org/10.1007/s10462-023-10553-w

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Cyberbullying
• Machine learning
• Systematic literature review
• Find a journal
• Publish with us
• Track your research

Mitigating Bias in Session-based Cyberbullying Detection: A Non-Compromising Approach

Lu Cheng , Ahmadreza Mosallanezhad , Yasin Silva , Deborah Hall , Huan Liu

Export citation

• Preformatted

Markdown (Informal)

[Mitigating Bias in Session-based Cyberbullying Detection: A Non-Compromising Approach](https://aclanthology.org/2021.acl-long.168) (Cheng et al., ACL-IJCNLP 2021)

• Mitigating Bias in Session-based Cyberbullying Detection: A Non-Compromising Approach (Cheng et al., ACL-IJCNLP 2021)
• Lu Cheng, Ahmadreza Mosallanezhad, Yasin Silva, Deborah Hall, and Huan Liu. 2021. Mitigating Bias in Session-based Cyberbullying Detection: A Non-Compromising Approach . In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) , pages 2158–2168, Online. Association for Computational Linguistics.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Child Health Nurs Res
• v.27(1); 2021 Jan

Empathy, cyberbullying, and cybervictimization among Filipino adolescents

Tadena shannen.

1 Graduate Student, School of Nursing, Hallym University, Chuncheon, Korea

Shin-Jeong Kim

2 Professor, School of Nursing, Research Institute of Nursing Science, Hallym University, Chuncheon, Korea

Jungmin Lee

3 Lecturer, School of Nursing, Hallym University, Chuncheon, Korea

The purpose of this study was to gain insights into empathy, cyberbullying, and cybervictimization among Filipino adolescents.

The participants were 168 junior high school students in the 7th to 10th grades at a public high school in Cavite Province, Philippines. Data were collected on demographic characteristics, the Basic Empathy Scale, and Revised Cyber Bullying Inventory-II scores. The data were analyzed in SPSS using descriptive statistics (frequency and mean), the independent t-test, bivariate correlation, and one-way analysis of variance.

The degree of empathy did not show a statistically significant relationship with cyberbullying (r=−.07, p =.359) but did show a significant relationship with cybervictimization (r=.18, p =.025). Furthermore, cyberbullying had a statistically significant association with cybervictimization (r=.60, p <.001).

Although empathy does not necessarily affect cyberbullying, higher levels of empathy were found among cyberbullying victims. However, an alarming result of this study is the possibility that victims may become cyberbullies. Therefore, we should highlight empathy as part of efforts to prevent cyberbullying and to solve various cyber-related problems. Since cyberbullying and cybervictimization are closely related, it is important to focus on this relationship and to make multilateral efforts to ensure that cyberbullying does not lead to other negative issues.

INTRODUCTION

Cyberbullying is more harmful than other types of violence because a negative post or comment can reach a limit-less number of social media (SM) users, thereby increasing the opportunity of prolonged exposure and permanency [ 1 , 2 ]. Furthermore, younger generations currently tend to embrace the use of SM [ 1 ]. A term related to “cyberbullying” is “cybervictimization’, which refers to the experience of being victimized through the use of electronic information on the internet [ 2 ]. Cyberbullying is considered to be a form of violent behavior; therefore, it is expected that cyberbullying causes detrimental effects for both the victim and the perpetrator.

The use of the internet for SM has steadily grown over time. As of January 2020, the estimated number of active SM users worldwide was 3.8 billion, reflecting an increase from the number of 3.4 billion in January 2019 [ 3 ]. In particular, during the coronavirus disease 2019 (COVID-19) pandemic, there has been an increase in the number of digital platform users [ 4 ]. Individuals are now actively posting their new achievements during the pandemic lockdown and expressing their opinions more vocally than before, as SM is now the most frequently used communication method for most individuals [ 4 ]. One study showed that as SM use increases, students who are prone to bullying are more likely to face cyberbullying [ 4 ]. There can be a limitless exchange of information among the SM users, and this information can sometimes be used in negative ways that may cause harm or make someone feel harassed. The act of harassing or harming someone via the use of SM is called cyberbullying. Cyberbullying can happen by sending harassing messages, posting derogatory comments on SM sites, posting humiliating pictures, and threatening someone electronically [ 5 ].

Victims of cyberbullying usually report feelings of depression, anger, frustration, and even suicidal thoughts. Cyberbullying has been found to be linked to low self-esteem, family problems, academic problems, violence, and delinquent behaviors [ 1 , 4 ]. In general, the estimated number of adolescents experiencing cyberbullying is around 10%-40%, although the percentage depends on age and the definition of cyberbullying used in the study [ 1 , 2 ]. In the United States, among a sample of 5,700 middle to high school students, 33.8% admitted to being cyberbullied, while 11.5% admitted to being perpetrators of cyberbullying [ 1 ]. In South Korea, 10.8% of 50,000 elementary to high school students experienced cyberbullying [ 6 ]. In a survey conducted in the Philippines, it was found that 80% of 1,143 respondents aged 13-16 years experienced cyberbullying [ 7 ].

A psychological factor related to the development of violent behavior is having a low empathy level. Empathy can be defined using two dimensions: cognitive and affective. Cognitive empathy is the ability to communicate, tolerate, recognize, and perceive emotions, while affective empathy refers to the ability to perceive and share both the positive and negative emotions of other people [ 8 ]. A study showed that lower levels of empathy increased the development of violent or aggressive behaviors, while a higher level of empathy decreased the development of violent or aggressive behaviors [ 9 ]. Since cyberbullying is considered to be a violent behavior, it is possible that empathy influences cyberbullying.

Adolescence is an important developmental stage when adolescents transition from childhood to adulthood. The changes experienced by adolescents also affect their roles in their family, with friends, and in their communities. Thus, it is vital to pay special attention to adolescents [ 10 ]. The parents of Filipino adolescents set the expectation that they are to be aware of their actions and the consequences corresponding to those actions [ 11 ]. Camaraderie is a factor valued by Filipino adolescents, which is why it is natural for them to be friendly, resulting in the expectation that they know how to empathize with someone [ 11 ]. Nonetheless, little research has discussed Filipino adolescents’ characteristics, and although favorable characteristics have been mentioned, there has been an increase in cases of cyberbullying in the Philippines [ 7 ].

It has been reported that inhabitants of the Philippines spend an average of 4 hours a day using SM, exceeding the values reported for other countries [ 3 ]. Since prolonged exposure to SM can increase victimization, Filipinos may be exposed to the harmful effects of SM for a longer time than others [ 1 ]. Although there has been an increase in the number of cases of cyberbullying in the Philippines, little research has tackled this issue. Because of the detrimental effects of cyberbullying, it is important for adolescents to be aware of its seriousness and the importance of putting an end to this violent behavior. Since adolescence is a vital stage of life, it is important to properly guide adolescents throughout this period. To the researchers’ knowledge, no published study has yet investigated empathy among Filipino adolescents and its possible influence on cyberbullying and cybervictimization.

The purpose of this study was to investigate empathy, cyberbullying, and cybervictimization among Filipino adolescents, with the following specific goals:

• To identify the degree of empathy, cyberbullying, and cybervictimization among junior high school students.
• To identify differences in empathy, cyberbullying, and cybervictimization according to participants’ general characteristics.
• To identify the relationships among empathy, cyberbullying, and degree of cybervictimization.

2. Theoretical Framework

Two theoretical frameworks were used to support the study: the social learning theory developed by Albert Bandura [ 12 ] and the general strain theory developed by Robert Agnew [ 13 ] ( Figure 1 ).

Empathy, cyberbullying, and cybervictimization framework.

Social learning theory [ 12 ] reflects the understanding that a person learns and develops attitudes, behaviors, and emotional reactions by observing others’ actions. It combines cognitive and behavioral learning theory. Cyberbullying incidents and SM may influence ones’ empathy due to what a person observes or experiences through SM. Along with other factors that may also affect someone’s empathy, a person might, or might not, engage in cyberbullying.

General strain theory [ 13 ] supports the assumption that negative experiences can lead to strain or stress. Victimization is considered a stressful experience, and according to this theory, experiences of stress or strain may cause a person to perform a positive or negative action to decrease the stress or strain. Within this framework, the study used general strain theory to investigate whether cybervictimization increased victims’ likelihood to engage in cyberbullying as a response to the strain that they experienced.

1. Study Design

This research is a descriptive study using a questionnaire to investigate empathy, cyberbullying, and cybervictimization among Filipino adolescents.

2. Participants

The participants were junior high school students who were currently enrolled in a public secondary school in Cavite Province, Philippines. Junior high school is the term used in the Philippines to describe grade 7 to 10 students, whereas in South Korea, the term middle or high school is used. The participants were students from grades 7 to 10 who were aged 12 to 19 years old and used SM applications. The school was selected using convenience sampling. The sample size for the correlation was calculated using G*Power 3.1.9.2. The results indicated that 164 participants were needed for an effect size of .25, a significance level of .05, and a test power of .95. Anticipating an 80% response rate, 201 participants were recruited. Responses from 168 participants were analyzed after excluding uncompleted questionnaires.

3. Measurements

The Basic Empathy Scale was developed by Jolliffe and Farrington [ 14 ] and validated among 363 adolescents (grade 10) in the United Kingdom. The scale is a 20-item instrument with two subscales (cognitive empathy with 9 items, Cronbach’s α = .79; affective empathy with 11 items, Cronbach’s α = .85). It is a 5-point Likert-type scale (strongly disagree=1, disagree=2, neutral=3, agree=4, and strongly agree=5). Eight items are reverse-scored. A higher score indicates a higher empathy level [ 14 ]. For the current study, four items were omitted to obtain a more acceptable Cronbach’s α. Each of the eight items was used to measure cognitive empathy and affective empathy. Cronbach’s α values of .67 for the complete empathy scale, .51 for cognitive empathy, and .52 for affective empathy were recorded. Cronbach’s α values ranging from .50 to .75 are generally accepted because these values indicate moderate reliability [ 15 ]. The researcher conducted a pilot study without omitting the four questions and recorded a Cronbach’s α of .79 for the complete empathy scale, .67 for cognitive empathy, and .65 for affective empathy. After omitting the four questions, Cronbach’s α values of .81 for the complete empathy scale, .80 for cognitive empathy, and .66 for affective empathy were recorded.

2) Cyberbullying and cybervictimization

The Revised Cyber Bullying (Cyber Victim) Inventory-II is the second revision of the tool developed by Topcu and Erdur-Baker [ 16 ] and validated among 1,803 high school students attending public high school in Ankara, Turkey. The purpose of the revision was to update and eliminate specific technology names. Consisting of 10 items, participants rate each item twice (once for reporting cyberbullying experience and once for reporting cybervictimization) on a 4-point rating scale (1=never, 2=once, 3=twice or three times, 4=more than three times). Scores are added to achieve the total score in each form. A total score of 10 means being not involved in cyberbullying or cybervictimization, respectively, whereas a higher score indicates more frequent engagement in cyberbullying and cybervictimization. The Cronbach’s α coefficients were .80 for the cybervictimization part and .79 for the cyberbullying part [ 17 ]. In this study, Cronbach’s α values of .80 for cyberbullying and .75 for cybervictimization were recorded. A pilot study recorded a Cronbach’s α of .88 for the entire questionnaire, .57 for the cyberbullying subscale, and .92 for the cybervictimization subscale.

3) Content validity index

The modified questionnaire used in this study to measure empathy was reviewed and approved by a group of experts, including two professors from the child and adolescent department and three nurses who had both worked in a pediatric unit at a university hospital for over 5 years and had doctoral degrees in nursing. The content validity index was .91 using the rating of item relevance.

4. Data Collection Method

After receiving approval to conduct the study from the Institutional Review Board of Hallym University, the researcher coordinated with a teacher at a public secondary school in the Philippines. The researcher was endorsed by the school principal, and a letter describing the timetable for gathering data, the research tools, the purpose of the study, and permission to conduct the study was personally given. After receiving permission to conduct the study, endorsement letters, student and parental consent forms were distributed to the students and were signed by both the student and parents as proof that they agreed to participate in the study. After receiving the consent forms, a pilot study was conducted among two students from each grade level to assess the validity of the tool. After validation, the researcher collected data by distributing the questionnaires, giving instructions, and remaining in the classroom for the duration of questionnaire completion be able to respond to any questions or clarification from the participants. It took around 15 minutes for each class to complete the questionnaires. The collected questionnaires were kept and locked in the cabinet of the corresponding author’s research room and were not accessible to any other researchers.

5. Ethical Considerations

The two instruments used in this study were approved for use by the original authors through email. In addition, the study was conducted after receiving approval from the Institutional Review Board of Hallym University (HIRB-2019-092). The study was conducted at a public secondary school. The researcher explained the purpose and procedure of the research to the participants who agreed to participate by completing an informed consent form. Assurance was given that the results of the study would only be used for the current study and that the anonymity of the participants would be maintained. Participants were provided snacks worth about $1 per person as compensation.

6. Statistical Analysis

Data were analyzed using SPSS for Windows version 25 (IBM Corp., Armonk, NY, USA). Participants’ demographic characteristics were analyzed in terms of percentage, frequency, and mean±standard deviation. The Shapiro-Wilk hypothesis test and the Levene test for equality of variance were performed to confirm the normality of the distribution and within-group normality. The t-test and analysis of variance were conducted to identify differences in empathy, cyberbullying, and cybervictimization according to participants’ characteristics. The Duncan multiple range test was used to determine specific differences between pairs of means. Bivariate correlation analysis was conducted to assess the relationships among empathy, cyberbullying, and cybervictimization

1. Descriptive Characteristics of the Participants

The descriptive characteristics of the participants are shown in Table 1 . Of the 168 participants, 44.0%(n=74) were male and 56.0%(n=94) were female, and their ages were 12-19 years old, with a mean of 14.3±1.4 years. Forty-four (26.2%) of the participants were in the 10th grade, followed by 25.6 % (n=43) in eighth grade, 24.4% (n=41) in seventh grade, and 23.8% (n=40) in the ninth grade. The most commonly utilized SM platform by the participants was Facebook (98.8%; n= 166). The average time of SM usage per day was 296.1±266.1 minutes (4.93 hours or almost 5 hours). Among the 168 participants, the majority used SM applications more than seven times per day (n=57, 34.3%), followed by three to four times (n=46, 27.7%), five to six times (n=39, 23.5%), and one to two times (n=24, 14.5%). The cell phone was the primary device used to check SM platforms (n=154, 91.7%). The predominant purpose of using SM was to watch videos (n=121, 72.0%), while the least common purpose was to read articles (n=62, 36.9%).

Demographic Characteristics of Participants ( N =168)

SM, social media; SNS, social networking site.

2. Degree of Empathy, Cyberbullying, and Cybervictimization

The content, mean score, and standard deviation of the Basic Empathy Scale and Revised Cyber Bullying Inventory-II are presented in Table 2 . The total mean score for empathy was 3.4±0.4 out of 5, indicating a high level of empathy among the participants. The item with the highest mean score was “I can understand my friend’s happiness when she/he does well at something” (3.7±1.1). Followed by “When someone is feeling ‘down’ I can usually understand how they feel” (3.6±1.2), “I often become sad when watching sad things on or in films” (3.6±1.3), “I can often understand how people are feeling even before they tell” (3.7±1.1), and “I can usually work out when people are cheerful” (3.6±1.1). The item with the lowest mean score was “I find it hard to know when my friends are frightened” (2.8±1.1). The average score was divided by 10 which is the number of items. Thus, for cyberbullying was 1.6±0.5, and that for cybervictimization was 1.7±0.6 out of 4, which indicated the existence of cyberbullying and cybervictimization, with a slightly higher score for the latter. “Insulting someone [on the internet]” was the item with the highest average score (1.9±1.0 for cyberbullying and 2.1±1.1 for cybervictimization), followed by “taking over the password of someone’s account [on the internet]”(1.8±0.9 for cyberbullying and 1.8±0.9 for cybervictimization). “Creating a humiliating website [on the internet]” showed the lowest mean score (1.1±0.5 for cyberbullying and 1.2±0.6 for cybervictimization).

Empathy, Cyberbullying, and Cybervictimization of the Participants ( N =168)

3. Differences in the Level of Empathy, Cyberbullying, and Cybervictimization According to Participants’ General Characteristics

Patterns in empathy, cyberbullying and cybervictimization according to participants’ general characteristics are shown in Table 3 . Total empathy, cognitive empathy, and affective empathy were significantly higher among female participants than among male participants (t=3.82, p <.001, t=2.19, p =.030, and t=4.05, p =.001, respectively). However, there were no significant differences in cyberbullying and cybervictimization by gender (t=0.20, p =.838, and t=0.35, p =.724, respectively).

Differences in Participants’ Empathy, Cyberbullying, and Cybervictimization According to Their Characteristics ( N =168)

There was a statistically significant difference in total empathy and both cognitive and affective empathy according to grade level (F=3.90, p =.010, F=3.80, p =.012, and F=3.45, p = .010, respectively). Empathy was significantly higher among grade 10 students than among students of the other grade levels, but the Duncan post hoc test showed no significant difference in the mean scores. In addition, participants’ grade showed a statistically significant association with cyberbullying (F=3.03, p =.031). Cyberbullying was significantly more common among students in grade 8 than among students of the other grade levels, but the Duncan post hoc test showed no significance. In addition, cybervictimization showed a statistically significant difference according to grade level (F=7.66, p <.001), and was also higher in grade 8 participants, but the Duncan post hoc test showed no significance.

Statistically significant differences in total empathy and both cognitive and affective empathy were found according to participants’ age (t=2.86, p =.005, t=2.67, p =.008, and t=2.51, p =.013, respectively). There were no significant differences in cyberbullying or cybervictimization by age (t=0.59, p =.558, and t=1.13, p =.262, respectively). Finally, there were no significant differences in any dependent variables according to the daily use of SM.

4. Correlations among Empathy, Cyberbullying, and Cybervictimization

Cybervictimization was significantly correlated with participants’ total empathy (r=.18, p =.025) and cognitive empathy (r=.17, p =.030), but not with affective empathy (r=.12, p =.114). Total empathy was not significantly correlated with cyberbullying (r=−.07, p =.359). Cyberbullying was significantly correlated with cybervictimization (r=.60, p <.001) ( Table 4 ).

Correlations among Empathy, Cyberbullying, and Cybervictimization ( N =168)

The Philippines has a total population of 108 million, of whom 73 million are active SM users [ 3 , 18 ]. In the current study, Facebook was the most used SM application used by adolescents. In 2020, Facebook had approximately of 2.4 billion monthly users worldwide, thus making it the most popular SM application [ 3 ]. In the Philippines, Facebook is currently the most popular SM application, with at least 70 million users, and most SM users use cellphones to access SM; this trend was also found in a study conducted among Filipino adolescents [ 18 ]. The Philippines have had the longest average daily time of SM use for 4 consecutive years, which reached 4 hours in 2020 [ 3 ]. This finding is in line with the current study, which recorded an average of almost 5 hours of SM use. Most respondents used SM to watch videos (n=121) and for communication (n=106). Watching videos and communication were also the top 2 purposes of using SM in an earlier study [ 18 ].

Filipinos have been observed to value camaraderie and find it natural to be friendly to others, for which reason they place considerable importance on interpersonal relationships [ 11 ]. The importance placed on friendship may be interpreted as reflecting a high level of empathy among Filipino adolescents. The existence of cyberbullying and cybervictimization was evident according to the findings of the study. Even with the increase in cyberbullying in the Philippines, some undocumented cases may exist due to victims’ fear of their situation being publicized. Some schools also tend to under-report cyberbullying due to fears of tarnishing the school’s reputation [ 19 ].

Demographic characteristics are believed to possibly influence empathy, cyberbullying, and cybervictimization. A significant relationship was found between empathy and gender in the current study, with female adolescents scoring higher for empathy than male adolescents. This may result from gender expectations in Filipino culture between the two genders. Female adolescents are expected to help their mothers with chores and to take care of their siblings, which fosters nurturing characteristics and helps them to understand different emotions. In contrast, male adolescents are trained to be logical and independent, and to keep their emotions to themselves [ 20 ]. This may explain why the female participants in this study had a higher degree of empathy than the male participants. However, no significant relationship was found between cyberbullying or cybervictimization and gender, similar to another study conducted in the Philippines [ 21 ]. These results show that when considering factors influencing empathy, other demographic characteristics such as age, educational level, and economic level must also be considered, as well as gender. Thus, further research is needed to predict the variables that affect Filipino adolescents’ empathy regarding cyberbullying and cybervictimization.

The current study established a significant relationship between age and empathy. Another study also reported a similar result, and this relationship reflects the development of empathy, which grows with age until it reaches full development in adolescence [ 22 ]. However, some studies have reported a decrease in cognitive empathy and an increase in affective empathy in the elderly. A decrease in cognitive empathy may be related to reduced activity in brain areas associated with cognitive function [ 23 ].

No significant relationship was found between age and cybervictimization or cyberbullying in this study. This finding is similar to a study conducted among Filipino university students, where it was found that cyberbullying and cybervictimization experiences of students were not significantly related to their age [ 24 ].

There is scarce evidence regarding the relationships among grade level, empathy, cyberbullying, and cybervictimization. The current study identified a significant relationship between grade level and empathy, as well as with cyberbullying and cybervictimization. A study among grade 7-12 students found that the cognitive empathy of victims of cyberbullying was high [ 25 ]. However, in a comparison of cognitive empathy in cyberbully-victims (a term for victims who also engaged in cyberbullying) and bystanders, lower grade levels had higher levels of cognitive empathy than other groups, while higher grade levels had higher levels of cognitive empathy only among cyberbully-victims, but not bystanders. Furthermore, it was found that victimization increased after grade 7 and continued through grade 12 [ 25 ].

Empathy has consistently shown a negative relationship with antisocial behavior [ 15 ], implying that empathy may have a possible influence on cyberbullying. A study investigating the relationship between cyberbullying and empathy found out that a low level of both cognitive and affective empathy predicted cyberbullying [ 26 ]. The current study found no significant relationship between empathy and cyberbullying. The same result was found in another study, and a possible explanation for this result is that even though the participants had a high empathy level, which may make them sensitive to others’ emotions, they may have viewed cyberbullying as a normal phenomenon [ 27 ]. As such, it is crucial to provide Filipino adolescents with the necessary information to protect themselves against cyberbullying. In addition, active interest and participation of parents and teachers are needed for the development of programs, practices, and policies to protect and support this population from this problem.

Few studies have reported a relationship between empathy and cybervictimization. It was found that cybervictims scored higher than cyberbullies and uninvolved students or bystanders in both cognitive and affective empathy [ 28 ]. One study reported higher affective empathy among cybervictims [ 29 ]. Another study reported significant relationships of both cognitive and affective empathy with cybervictimization [ 30 ]. In the current study, total empathy and cognitive empathy, but not affective empathy, had significant relationships with cybervictimization. This finding of distinct relationships of cognitive and affective empathy with cybervictimization was not found in previous research. The significant relationship between cybervictimization and cognitive empathy may be explained by cybervictims’ capability to understand the perspective of others, while still having problematic social relationships, as shown by the lack of a significant association between affective empathy and cybervictimization. Furthermore, affective empathy is important in establishing social relationships [ 27 ].

Taking revenge is a possible response of a cybervictim to his or her bullies. The desire to take revenge may lead a cybervictim to engage in cyberbullying [ 30 ]. Another explanation is the application of general strain theory, which was developed by Agnew [ 13 ]. Greater exposure or experiences of cyberbullying may increase the likelihood that cybervictims will engage in cyberbullying to release their stress. This is an example of a negative response to strain. A study conducted in the Philippines and Turkey reported a positive relationship between cyberbullying and cybervictimization [ 21 ]. The current study yielded a significant relationship between cyberbullying and cybervictimization.

Several limitations were found in this study. The first is the length of the duration of data collection, which resulted in a small sample of participants from a single school. The second is that the researcher used convenience sampling to choose where to conduct the study and the participants to be included. Participants were picked by the teachers due to their availability and knowledge regarding SM use. The third is the use of a self-report survey. There is a possibility that students over-reported or misreported some data (e.g., SM time usage per day and the reasons for SM usage). The questionnaires used may have made the students feel discomfort, anxiety, and fear due to the questions. The fourth is the scarcity of articles about empathy, cyberbullying, cybervictimization, and characteristics of Filipino adolescents. The fifth is the reliability of the Basic Empathy Scale. This study recorded Cronbach’s α values of .67 for the total Basic Empathy Scale, .51 for the cognitive empathy subscale, and .52 for the affective empathy subscale. This is considered a low reliability score, but according to a book by Hinton et al. [ 15 ], a Cronbach’s α value ranging from .50 to .75 is accepted because it indicates moderate reliability.

Despite the limitations of this study, it still made a pertinent contribution to the study of empathy, cyberbullying, and cybervictimization.

The findings of this study shed important light on the issues of cyberbullying and cybervictimization, which could become more problematic in the COVID-19 pandemic. Empathy was not found to be negatively associated with cyberbullying. This can be explained by a lack of self-awareness, because some SM users post or send whatever they want without thinking about whether the content can harm others [ 1 ]. It is also possible that cyberbullying was viewed as a normal phenomenon, especially as SM use has increased over time. Cognitive empathy, but not affective empathy, was significantly associated with cybervictimization. It is possible for cybervictims to understand the emotions of others but still have difficulty in establishing social relationships or feeling the emotions of other people. The finding of this study that cyberbullying was significantly associated with cybervictimization is alarming. This relationship may be explained by the stress that cybervictims experience and their desire to take revenge on the perpetrators, especially since SM can be anonymously.

We expect that the findings of this study may contribute to the prevention of cyber-related problems among adolescents as it offers basic guidelines. Since adolescence can be a confusing period due to changes and transitions from childhood to adulthood, adolescents deserve special attention and guidance from people surrounding them. It is vital for parents, teachers, and even guardians to establish open and good communication with adolescents to properly guide them towards betterment. At this time, with one’s capacity to empathize, it is possible to become a bystander or a defender of a victim, for which reason we should highlight empathy in efforts to prevent and solve various types of cyber-related problems. We also suggest that as cyberbullying and cybervictimization are closely related, it is important to focus on this relationship and to make multilateral efforts to ensure that cyberbullying does not lead to other negative problems.

Acknowledgments

Conflict of interest

No existing or potential conflict of interest relevant to this article was reported.

Data availability

Please contact the corresponding author for data availability.