This 

document is downloaded from DR‑NTU (https://dr.ntu.edu.sg)
Nanyang Technological University, Singapore.

A meta‑analysis of factors predicting
cyberbullying perpetration and victimization :
from the social cognitive and media effects
approach

Chen, Liang; Ho, Shirley S.; Lwin, May Oo

2016

Chen, L., Ho, S. S., & Lwin, M. O. (2017). A meta‑analysis of factors predicting cyberbullying
perpetration and victimization : from the social cognitive and media effects approach. New
Media & Society, 19(8), 1194‑1213. doi:10.1177/1461444816634037

https://hdl.handle.net/10356/142038

https://doi.org/10.1177/1461444816634037

© 2016 The Author(s). All rights reserved. This paper was published by SAGE Publications in
New Media & Society and is made available with permission of The Author(s).

Downloaded on 22 Mar 2023 02:56:36 SGT

 1

A Meta-Analysis of Factors Predicting Cyberbullying Perpetration and Victimization:

From the Social Cognitive and Media Effects Approach

Liang Chen

Shirley S. Ho

May O. Lwin

Nanyang Technological University, Singapore

Author Note
Correspondence concerning this article should be addressed to Liang Chen, Nanyang
Technological University, WKWSCI Building Level 5, 31 Nanyang Link, 637718 Singapore.
Email: [email protected]

This is the final version of a manuscript that appears in New Media & Society. The APA
citation for the published article is:

Chen, L., Ho, S. S., & Lwin, M. O. (2017). A meta-analysis of factors predicting
cyberbullying perpetration and victimization: From the social cognitive and media effects
approach. New Media & Society, 19(8), 1194-1213. DOI: 10.1177/1461444816634037

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship,
and/or publication of this article: This work was funded by the Singapore’s Ministry of
Education Tier 1 Grant (Grant No.: M4011117).
 2

Abstract

Cyberbullying has become a critical social issue, which severely threatens children

and adolescents’ physical and psychological health. The current research systematically

examined the predictors of cyberbullying from the social cognitive and media effects

approach. Specifically, this study identified 16 predictors of cyberbullying perpetration and

victimization and examined the magnitude of the effects of these predictors by meta-

analyzing 81 empirical studies, which represented a total sample of 99,741 participants and

yielded 259 independent correlations. The results revealed that risky information and

communications technology (ICT) use, moral disengagement, depression, social norms, and

traditional bullying perpetration were the main predictors of cyberbullying perpetration,

while risky ICT use and traditional bullying victimization were the major contributors of

cyberbullying victimization. According to the moderator analyses, country of the sample,

sampling method, age, and media platform were significant moderators of the relationships

between some specific predictors and cyberbullying perpetration and victimization.

Implications for future cyberbullying research were discussed.

Keywords: Cyberbullying perpetration, cyberbullying victimization, media effects, meta-

analysis, social cognition

 3

A Meta-Analysis of Factors Predicting Cyberbullying Perpetration and Victimization:

From the Social Cognitive and Media Effects Approach

With the rise of information and communication technologies (ICT), individuals,

especially children and adolescents, are able to access a large amount of information and

interact with others. This brings many benefits to their social and emotional development

(Kowalski et al., 2008). However, these communication technologies have also exposed

children and adolescents to cyberbullying, which threatens their psychological and physical

health (Kowalski et al., 2008).

Cyberbullying refers to “an aggressive, intentional act carried out by a group or

individual, using electronic forms of contact, repeatedly and over time against a victim who

cannot easily defend him or herself” (Smith et al., 2008: 376). In recent years, with the

proliferation of social media, cyberbullying has become one of the most common social

issues worldwide (Görzig and Frumkin, 2013). In the United States, approximately 30% of

middle school students reported having experienced cyberbullying (Patchin and Hinduja,

2010). The National Children’s Home of United Kingdom (NCH, 2005) has found that 20%

of the children aged between 7 and 11 years in the United Kingdom have been cyberbullied.

Besides, Lwin et al. (2012) examined online harassment among 537 teenagers in Singapore

and found that 51% of the respondents claimed to have been cyberbullied.

Furthermore, the large number of cyberbullying incidents, coupled with the high

severity of the incidents, has threatened the mental and physical health of children and

adolescents, with victims experiencing harmful consequences such as psychosomatic

symptoms (Kowalski et al., 2008; Patchin and Hinduja, 2006), anti-social behaviors (Low

and Espelage, 2013), and suicidal behaviors (Hinduja and Patchin, 2010).

In order to curb cyberbullying, many researchers have started to investigate this issue.

Numerous descriptive studies exploring the prevalence of cyberbullying have been conducted
 4

earlier (Dehue et al., 2008; Slonje and Smith, 2008). Tokunaga (2010) conducted a meta-

synthesis to list different definitions of cyberbullying and to examine the prevalence of

cyberbullying and the relationships between demographic factors and cyberbullying. In more

recent years, researchers have conducted numerous studies to examine the predictors of

cyberbullying from different perspectives, including media exposure (Akbulut et al., 2010;

Erdur-Baker, 2010), personal factors (Grabe and Holfeld, 2012; Patchin and Hinduja, 2010),

and environmental factors (Festl and Quandt, 2013; Xiao and Wong, 2013). Furthermore,

Kowalski et al. (2014) conducted a meta-analysis to explore predictors and outcomes of

cyberbullying among adolescents based on empirical studies from 2002 to early 2013.

Although comprehensive, this meta-analysis had a number of limitations. First, their study

focused only on adolescents, without taking into consideration other vulnerable groups such

as children and emerging adults. Next, although they analyzed studies in North America and

Europe, they did not include studies from Asia. It is important to include Asian samples into

the meta-analysis as recent reports have shown that the highest incidences of cyberbullying

are occurring in countries such as China and India (Microsoft, 2012). Finally, due to the fact

that the general aggression model was used as a theoretical framework in their meta-analysis,

they highlighted certain predictors but overlooked several other important factors, such as

social norms and popularity. Many studies have used the social dominance theory and the

theory of planned behavior to explain the phenomenon of cyberbullying (Espelage et al.,

2013; Heirman and Walrave, 2008; Lazuras et al., 2013). As popularity and social norms are

key factors in these theories, they should be important in exploring cyberbullying. These

research gaps, coupled with the fact that many new empirical studies on cyberbullying have

sprung up over the last 2 years, gave us the impetus to examine cyberbullying with a different

theoretical perspective.

In order to clarify the research gaps in the domain of cyberbullying and to offer
 5

directions for future cyberbullying research, the current research applies the social cognitive

and media effects approach to systematically meta-analyze the entire range of empirical

studies across all age groups and countries from 2004 to 2015. Specifically, this study seeks

to analyze 16 predictors of cyberbullying and to examine the magnitude of the effects of

these predictors across 81 empirical studies. Besides, we select and analyze four potential

moderators: country, sampling method, age, and media platform to explain the variance in the

magnitude of the effects of these predictors across different studies.

Theoretical framework

Social cognitive theory suggests human functioning as an interaction of personal,

behavioral, and environmental influences and posits that both personal factors and

environmental factors can affect human behavior (Bandura, 1986). The theory has been

widely applied in traditional bullying (TB) (Hymel et al., 2005; Mouttapa et al., 2004) to

understand the complexity and social nature of TB. Like TB, cyberbullying is heralded as

individuals’ social behavior, which can also be explained by personal and environmental

factors (Xiao and Wong, 2013). In addition, the media effects model suggests that media can

affect audiences’ thoughts and behaviors, especially aggressive behaviors (Ball-Rokeach and

DeFleur, 1976; Felson, 1996). Many studies have indicated that how individuals use media

could affect their involvement in cyberbullying (Festl et al., 2013; Mark and Ratliffe, 2011).

Based on these considerations, this study integrates the social cognitive theory and media

effects model as a theoretical framework to identify three main categories of factors in order

to explain cyberbullying: media exposure, personal factors, and environmental factors.

Media exposure

Based on the media effects model, media have an influence on individuals’ behaviors

(Valkenburg and Peter, 2013). In the context of cyberbullying, many studies have found that

individuals who spend more time using ICT are more likely to be involved in cyberbullying
 6

(Akbulut et al., 2010; Bauman, 2009; Zhou et al., 2013). This is consistent with the online

disinhibition effect (Suler, 2004). The absence of rich emotive cues and immediate feedback

in online interactions create a disinhibition effect, which diminishes self-censorship. In the

online environment, individuals generally feel less restrained and are likely to express their

thoughts more freely (Ho and McLeod, 2008; Suler, 2004). This encourages aggressive

behavior toward others (Espelage et al., 2013).

Additionally, risky ICT use has been considered as an important factor of

cyberbullying victimization. Many studies have indicated that engagement in risky Internet

activities, such as sharing personal information or photos online (Kwan and Skoric, 2013;

Mesch, 2009) and adding strangers as online friends (Görzig and Ólafsson, 2013; Kwan and

Skoric, 2013), accounts for a higher likelihood of becoming victims of cyberbullying.

Therefore, frequency of ICT use and risky ICT use are positively associated with both

cyberbullying perpetration and victimization.

Personal factors

According to the social cognitive theory (Bandura, 1986), personal and environmental

factors are the main determinants of individuals’ behaviors. Personal factors could include

personality traits, belief, skill level, available effort, and experiences (Gifford and Nilsson,

2014; Gist and Mitchell, 1992). In the context of cyberbullying, 10 and 8 personal factors

have been identified to predict cyberbullying perpetration and victimization, respectively.

Specifically, extant empirical studies on cyberbullying have documented that moral

disengagement (Bussey et al., 2015; Perren and Sticca, 2011), narcissism (Braithwaite et al.,

2009), depression (Kowalski et al., 2008), and self-efficacy (Bussey et al., 2015;) are

positively associated with cyberbullying perpetration. On the other hand, self-esteem

(Pyżalski, 2012; Yang et al., 2006), emotional management (Kellerman et al., 2013),

popularity (Buelga et al., 2015), and school commitment (Hemphill et al., 2014) are
 7

negatively associated with cyberbullying perpetration.

Additionally, research examining cyberbullying victimization has demonstrated that

moral disengagement (Almeida et al., 2012) and depression (Didden et al., 2009; Low and

Espelage, 2013) are positive predictors of cyberbullying victimization. Conversely, self-

esteem (Brewer and Kerslake, 2015; Yang et al., 2006), emotional management (Hemphill et

al., 2014), and self-efficacy in defending (Chang et al., 2015a; Depaolis and Williford, 2015)

are negatively associated with cyberbullying victimization.

In addition, TB perpetration and victimization have been considered as crucial

personal factors, which are positively associated with both cyberbullying perpetration and

victimization (Hemphill et al., 2012; Sticca et al., 2013; Ybarra and Mitchell, 2004).

Therefore, this study explores the degree to which these personal factors are associated with

cyberbullying perpetration and victimization.

Environmental factors

According to the social cognitive theory, besides personal factors, environmental

factors are another category of determinants of individuals’ behaviors, which could promote

or inhibit violent behaviors (Bandura, 1986). Environmental factors refer to the roles played

by parents, peers, and other persons who are important to an individual (Bruning, 1999). The

values, attitudes, and feedback provided by these referent groups could drive individuals’

behaviors (Ho et al., 2014). In terms of cyberbullying, extant studies have revealed that

general parental interaction (Cappadocia et al., 2013) and parental mediation of child’s ICT

use (Roberto et al., 2014; Shapka and Law, 2013) are negatively associated with both

cyberbullying perpetration and victimization. Besides, some studies have found that social

norms, including subjective, descriptive, and injunctive norms, are also positively related to

cyberbullying perpetration (Bastiaensens et al., 2015; Heirman and Walrave, 2008).

Therefore, this study examines the extent to which these environmental factors are associated
 8

with cyberbullying perpetration and victimization.

Research questions

According to the literature review in the preceding section, the following research

questions are proposed:

RQ1. To what degree are the following 14 factors—(a) frequency of ICT use, (b) risky

ICT use, (c) moral disengagement, (d) narcissism, (e) depression, (f) self-efficacy, (g)

self-esteem, (h) emotional management, (i) school commitment, (j) TB perpetration, (k)

TB victimization, (l) social norms, (m) parental interaction, and (n) parental

mediation—associated with cyberbullying perpetration?

RQ2. To what degree are the following 11 factors—(a) frequency of ICT use, (b) risky

ICT use, (c) self-esteem, (d) moral disengagement, (e) depression, (f) self-efficacy in

defending, (g) emotional management, (h) TB perpetration, (i) TB victimization, (j)

parental interaction, and (k) parental mediation—associated with cyberbullying

victimization?

Method

Literature search

To identify relevant studies, the first step was to locate studies using the following

key words: cyberbullying, cyberbullying perpetration, cyberbullying victimization, cyber-

victimization, cyberbully, cybervictim, online bullying, online victimization, Internet

bullying, Internet victimization, online harassment, Internet harassment, online aggression,

and Internet aggression. The following databases were searched: PsycINFO, MEDLINE,

Communication & Mass Media Complete, EBSCO, Educational Resources Information

Center (ERIC), Web of Science Direct, PubMed, and Nursing & Allied Health Source. To be

thorough, this study reviewed the reference lists of the collected publications for additional
 9

studies. We also reviewed non-peer-reviewed papers, such as book chapters and conference

papers, to reduce the potential influence of publication bias (Sohn, 1996).

Filter process

Studies were included in this meta-analysis based on the following inclusion and

exclusion criteria:

1. The article should focus on predictors of cyberbullying which is defined as an

aggressive, intentional act carried out using electronic forms of contact (Smith et

al., 2008). As a result, studies focusing on online harassment or aggression were

also included in this meta-analysis.

2. The quantitative information for meta-analytical purposes that could be computed

or converted into effect size estimates must be available in the selected studies.

Essential data include sample sizes, test statistics, means, correlations, and odds

ratios.

3. If one study used two or more data sets, it was considered as two or more studies.

The original search process identified 1037 related studies. After going through the

abstracts of all 1037 articles, only 139 examined predictors of cyberbullying. Furthermore, 80

of the 139 studies provided the quantitative information for calculation of effect sizes, but

one of these studies used two data sets. Therefore, after the filtering process, the final sample

for this meta-analysis was 81 studies, which represented a total sample of 99,741 participants

and yielded 259 independent correlations.

Coding procedures

Prior to coding, two coders were trained based on the specific coding criteria.

Following this, they were asked to practice coding. Any discrepancies were resolved, and the

decision rules were established to avoid ambiguities. Following this, the two coders

individually coded all 81 studies. Specific study characteristics, predictors, moderators, and
 10

statistical indexes in each study were coded. The inter-rater agreement (Cohen’s κ) ranged

from .80 to 1.00, with the average κ = .83.

Calculation of effect sizes

To compute effect sizes, we took the statistics of comparable formats as reported in

each study. The data formats mainly included correlations, odds ratios, and means with

standard deviations. We entered all the data to the Comprehensive Meta-Analysis (CMA)

software program and then converted the data into one uniform form of Pearson product-

moment correlation (Pearson’s r). According to Cohen’s guidelines of effect size for social

science (Cohen, 1988, 1992), small effect size is indicated by Pearson’s r of .10−.23; medium

effect size is indicated by Pearson’s r of .24−.36; and large effect size is indicated by

Pearson’s r of .37 or higher. The total effect sizes of each predictor and dependent variables

(cyberbullying perpetration or victimization) were averaged Pearson’s r across studies,

weighting each observed correlation according to the study’s sample size and standard error

(Hedges and Olkinm, 1985).

With respect to the homogeneity test, the Q statistic was used in this meta-analysis.

According to the guideline presented by Hedges and Vevea (2007), we used the random

effects model because most Q statistics were significant.

Results

Description of the studies

First, we provide a summary description of all the 81 studies included in this meta-

analysis. Among the 81 studies, 9 (11.11%) were published from 2004 to 2009, while the

remaining 72 studies (88.89%) were published between 2010 and 2015. This revealed that

there was a noticeable increase in cyberbullying research since the term “cyberbullying” was

coined by Belsey in 2004. Of the 81 studies, 67 (82.72%) examined the predictors of

cyberbullying perpetration, while 56 (69.14%) explored the predictors of cyberbullying

 11

victimization. Most studies employed samples from North America and Europe: 34 studies

(41.98%) used samples from North America (the United States and Canada); 31 studies

(38.27%) were from European countries, such as the United Kingdom, Belgium, Greece, and

the Netherlands; and the remaining 16 studies (19.75%) were from the Asia-Pacific region,

such as Australia, China, and Singapore. Most studies used samples of adolescents (85.19%).

Only nine studies (11.11%) examined cyberbullying among adults and three (3.70%) studies

focused on children. Finally, 47 (58.02%) studies used non-random samples, whereas 34

(41.98%) studies applied the random sampling method.

Weighted correlations

Cyberbullying perpetration. Weighted correlations between predictors and

cyberbullying perpetration are presented in Table 1. For media exposure, the predictive

power of risky ICT use (r = .29, p < .05) was stronger than frequency of ICT use (r = .20, p <

.05) when predicting cyberbullying perpetration. According to Cohen’s guidelines of effect

size (Cohen, 1988, 1992), the association between frequency of ICT use and cyberbullying

perpetration was small, while the association between risky ICT use and cyberbullying

perpetration was medium. Both predictors’ weighted effect sizes were significantly

distinguished from zero because their 95% confidence intervals did not cross zero.

With regard to personal factors, the strongest positive predictor was TB perpetration

(r = .39, p < .05). The effect size was large. Moral disengagement (r = .28, p < .05),

depression (r = .25, p < .05), narcissism (r = .23, p < .05), and self-efficacy (r = .21, p < .05)

were also positively associated with cyberbullying perpetration, but their effect sizes were

medium or small and weaker than those of TB perpetration. The weakest effect size was

noted for TB victimization (r = .19, p < .05). On the other hand, self-esteem (r = −.13, p <

.05), school commitment (r = −.10, p < .05), and emotional management (r = −.09, p < .05)

had weak negative associations with cyberbullying perpetration. All the above effect sizes
 12

were significantly different from zero. However, the results showed that popularity was not a

significant predictor because its 95% confidence intervals crossed zero. In terms of

environmental factors, social norms were positively associated with cyberbullying (r = .27, p

< .05). The effect size was at a medium level. Besides, parental interaction (r = −.17, p < .05)

and parental mediation (r = −.07, p < .05) were weak negative predictors. These effect sizes

were significantly different from zero.

Cyberbullying victimization. Weighted correlations between predictors and

cyberbullying victimization are shown in Table 2. For media exposure, risky ICT use (r =

.26, p < .05) had a larger effect size than frequency of ICT use (r = .19, p < .05) with respect

to being cybervictims. Based on Cohen’s guidelines (Cohen, 1988, 1992), the former’s effect

size was medium, but the latter’s effect size was small. Both effect sizes were significantly

different from zero. Among personal factors, TB victimization (r = .32, p < .05) had

moderate positive correlations with cyberbullying victimization. TB perpetration (r = .21, p <

.05), depression (r = .20, p < .05), and moral disengagement (r = .14, p < .05) were also

positive predictors, but their effect sizes were small. Conversely, self-esteem (r = −.22, p <

.05), emotional management (r = −.20, p < .05), and self-efficacy in defending (r = −.03, p <

.05) had weak negative effects on cyberbullying victimization. All these effect sizes were

significantly different from zero.

In terms of environmental factors, both parental interaction (r = −.09, p < .05) and

parental mediation (r = −.07, p < .05) were negatively associated with cyberbullying

victimization. Their effect sizes were small, but significantly different from zero.

Moderator analyses

Based on the recommendations of many scholars (e.g. Cook et al., 2010; Rosenthal,

1991), we conducted moderator analyses for factors with heterogeneous effect sizes across

studies because a significant Q statistic indicates significant heterogeneity of effect sizes

 13

across different studies. Specifically, we employed country of the sample, sampling method,

and age as potential moderators of all the factors with heterogeneous effect sizes across

studies and cyberbullying. In addition, media platform was used as a potential moderator of

media exposure factors and cyberbullying. The selections were based on findings from

previous studies as well as recommendations from scholars. Previous studies have indicated

cross-cultural differences in cyberbullying, in which adolescents in Asia were more likely to

be involved in cyberbullying than their counterparts in the West (Microsoft, 2012; Perren et

al., 2010). Besides this, some scholars suggested that the varying prevalence rates of

cyberbullying across previous studies might be due to the use of different sampling methods

and the assessment of participants from different age groups (Bauman, 2013; Kowalski et al.,

2012). Moreover, Brandtzæg et al. (2009) found that children’s experience of cyberbullying

can differ when they use different media platforms. Therefore, based on these considerations,

this study employed these four factors as potential moderators.

According to the moderator analyses, first, country of the sample significantly

moderated the association of parental interaction (QB = 100.76, p < .05) with cyberbullying

perpetration (see Table 3). Second, sampling method significantly moderated the associations

of frequency of ICT use (QB = 4.19, p < .05) and parental interaction (QB = 6.43, p < .05)

with cyberbullying perpetration. Moreover, it was a significant moderator of frequency of

ICT use and cyberbullying victimization (QB = 10.70, p < .05) (see Table 4). Third, age was a

significant moderator, which can explain the differences among effect sizes of TB

perpetration (QB = 80.10, p < .05) and TB victimization (QB = 21.41, p < .05) when assessed

for cyberbullying perpetration. Moreover, it significantly moderated the associations of

frequency of ICT use (QB = 12.95, p < .05) and TB victimization (QB = 13.13, p < .05) with

cyberbullying victimization (see Table 5). Finally, media platform significantly moderated

the relationships between frequency of ICT use and cyberbullying perpetration (QB = 12.14, p
 14

< .05) and cyberbullying victimization (QB = 23.12, p < .05) (see Table 6).

Discussion

This meta-analysis not only identifies the main predictors of cyberbullying

perpetration and victimization but also examines the magnitude of the effects of these

predictors. Primarily, the results revealed that all the proposed factors from the social

cognitive and media effects approach, with the exception of popularity, significantly

predicted cyberbullying perpetration or victimization. This indicates that integrating the

social cognitive theory and media effects model provides a promising theoretical framework

to explore the phenomenon of cyberbullying.

Furthermore, the results show that risky ICT use, moral disengagement, depression,

social norms, and TB perpetration were more strongly related to cyberbullying perpetration

than other predictors. Their aggregated effect sizes were large or medium based on Cohen’s

guidelines of effect size (Cohen, 1988, 1992). According to our findings as well as empirical

evidence from recent studies (e.g. Akbulut, 2014), it is worthwhile for educators to conduct

workshops or seminars to guide students on how to use ICT safely in order to curb children

and adolescents’ aggressive behavior online. Moreover, educators should also advocate moral

engagement by developing children’s abilities to employ both substantive knowledge and

moral reasoning when evaluating the events (Colby and Ehrlich, 2003) and mitigate the onset

of depression in children by understanding children’s emotional situation, providing a

relaxing environment, and encouraging their autonomy (Beyondblue, 2011). As social norms

have a relatively strong effect on cyberbullying perpetration, correcting individuals’

misperception about the prevalence of cyberbullying among their peers could also be

important. Besides, educators and parents should pay more attention to offenders of TB and

increase their awareness of the negative outcomes of cyberbullying. Additionally, the results

revealed that only small associations were established between frequency of ICT use, self-
 15

esteem, narcissism, self-efficacy, emotional management, school commitment, TB

victimization, and cyberbullying perpetration. Moreover, popularity is not a significant

predictor of cyberbullying perpetration. Despite this, we cannot simply conclude that these

factors are not important. One reason for the weak or null relationships might be due to the

fact that only a small number of studies have focused on these predictors. Thus, the effect

sizes were calculated based on limited studies. Another reason might be due to the fact that

some predictors have a long-term effect on cyberbullying, but cross-sectional studies are not

able to capture it.

In addition, this study examined the relationships between 11 proposed predictors and

cyberbullying victimization. Similar to cyberbullying perpetration, risky ICT use was a

relatively strong predictor of cyberbullying victimization. Besides, TB victimization was

another major contributor of cyberbullying victimization. These two predictors’ effect sizes

were at medium level. Thus, educators and parents should pay more attention to TB victims

to increase their awareness of cyber safety in order to protect them from cyberbullying

victimization. Additionally, weak associations were found between frequency of ICT use,

self-esteem, depression, moral disengagement, self-efficacy in defending, emotional

management, parental interaction, parental mediation, TB perpetration, and cyberbullying

victimization.

According to the results of the moderator analyses, country of the sample significantly

moderated the negative relationship between parental interaction and cyberbullying

perpetration. Moreover, the negative relationship was stronger in Western countries than in

Asian countries. It is plausible that Western parents are more likely to employ autonomy-

supportive style instead of punitive parenting style when interacting with their children than

Asian parents (Kim and Wong, 2002). Children with supportive parental behaviors tend to be

less aggressive (Finkenauer et al., 2005). Thus, children who interact frequently with their
 16

parents in Western countries are less likely to cyberbully others than children in Asian

countries.

Second, sampling method significantly moderated the relationships between

frequency of ICT use and cyberbullying perpetration and victimization. Moreover, it was a

significant moderator of parental interaction and cyberbullying perpetration. Specifically, the

associations between frequency of ICT use and cyberbullying perpetration and victimization

were stronger in studies that used non-random samples compared to studies that used random

samples. In addition, the relationship between parental interactions and cyberbullying

perpetration was smaller in studies that used non-random samples than in studies that used

random samples. One possibility for these differences might be due to the fact that most non-

random samples were also respondents in niche environments (Paul et al., 2000). For

example, one of the studies (Bauman, 2009) included in our meta-analysis examined the

associations of frequency of ICT use and parental interaction with cyberbullying based on

rural intermediate school students. In another study included in our meta-analysis, Festl et al.

(2013) explored the influence of frequency of ICT use on cyberbullying by focusing on

academic track school students. Compared with regular schools, rural intermediate schools

and academic track schools might have less opportunities and resources to provide workshops

or seminars regarding cyberbullying for their students. As such, students in these niche

schools who use ICT frequently might be more likely to be involved in cyberbullying than

regular school students. Besides, parents living in rural areas are relatively less educated and

are less likely to use supportive parenting practices when interacting with their children

(Scott et al., 2012). Since children who interact frequently with their parents exhibit lower

levels of aggression (Clauss-Ehlers, 2010), parental interaction in rural areas might, therefore,

be less effective than in urban areas in protecting children from cyberbullying. Based on

these considerations, effect sizes varying across different studies may be due to the fact that
 17

non-random samples tend to be students in niche schools or environments.

Third, age significantly moderated the relationships between two TB factors and

cyberbullying perpetration and the association between TB victimization and cyberbullying

victimization. Moreover, the relationships were stronger when studies focused on the older

people. One possible explanation is that young people tend to solve physical bullying by

fighting, whereas older individuals were more likely to extend offline bullying to online

(Perren et al., 2010). Besides, age significantly moderated the association of frequency of

ICT use with cyberbullying victimization. Moreover, the association was stronger when

studies examined younger people than adults. It is plausible that adults are better able to

manage their online behaviors and protect themselves when using ICT compared to

adolescents and children.

Finally, media platform significantly moderated the associations between frequency

of ICT use and cyberbullying perpetration and victimization. It is not surprising that

frequency of social media use has a stronger effect on cyberbullying perpetration and

victimization than other media platforms. Indeed, many studies have documented that

cyberbullying is most rampant on social media (Smith et al., 2008). One plausible

explanation is that social media are extremely popular among children and adolescents, thus

becoming an essential part of their social life. Moreover, social media integrate online

discussion forum, chat room, blog, and e-mail, which allow users to send private messages,

comment on others’ postings, post personal information, and upload photos. Thus,

cyberbullying is more likely to manifest in social media than in other media (Kwan and

Skoric, 2013).

Conclusion

The findings of this meta-analysis present some practical implications for educators

and parents. For example, according to our findings, educators and parents could remind
 18

students to refrain from adding strangers as online friends and disclosing personal

information online. Educators and parents should also encourage students to use ICT in

appropriate ways in order to protect themselves from cyberbullying. In addition to media

exposure, the results indicate that parent–child interactions and parental mediation have a

negative association with cyberbullying. Thus, parents could increase the interactions with

children, be more involved in their lives, and manage children’s ICT use to prevent

cyberbullying.

Additionally, this meta-analysis provides guidelines for the direction of future

cyberbullying research. Our moderator analysis shows that the frequency of social media use

has a stronger effect on cyberbullying than other media platforms. Moreover, research has

indicated that cyberbullying is most rampant on social media (Smith et al., 2008). However,

only limited studies have examined the predictors of cyberbullying on social media (Kwan

and Skoric, 2013). Future studies should focus on examining cyberbullying on social media.

Next, while many children have become offenders and victims of cyberbullying, only a few

studies examined predictors of cyberbullying perpetration or victimization in childhood. Our

moderator analysis reveals that age could significantly moderate the relationships between

many predictors and cyberbullying perpetration and victimization. Thus, future studies should

explore cyberbullying among primary or elementary school students and compare

cyberbullying across different age groups. Finally, most existing studies on cyberbullying

were from North America and Europe, but limited studies used Asian samples. As pointed

out, Asia has the highest level of cyberbullying (Microsoft, 2012). Our moderator analysis

indicates that cultural difference exists in the associations between parental interaction and

cyberbullying perpetration. Moreover, previous research has indicated that cyberbullying

varies across different cultures (Li, 2007; Microsoft, 2012). Thus, it is necessary to

investigate cyberbullying in Asia in future studies.

 19

References

References marked with an asterisk indicate studies included in the meta-analysis.

*Akbulut Y, Sahin YL and Eristi B (2010) Cyberbullying victimization among Turkish

online social utility members. Educational Technology & Society 13(4): 192–201.

Akbulut Y (2014) Effect of case–based video support on cyberbullying awareness. Australian

Educational Computing, 29(1): 1–12.

*Almeida A, Correia I, Marinho S, et al. (2012) Virtual but not less real: a study of

cyberbullying and its relations to moral disengagement and empathy. In: Li Q, Cross

D and Smith PK (eds) Cyberbullying in the Global Playground: Research from

International Perspectives. Chichester: Wiley-Blackwell, pp. 223–244.

*Ang RP, Tan KA and Mansor AT (2010) Normative beliefs about aggression as a mediator

of narcissistic exploitativeness and cyberbullying. Journal of Interpersonal Violence

26(13): 2619–2634.

Ball-Rokeach SJ and DeFleur ML (1976) A dependency model of mass-media effects.

Communication Research 3(1): 3–21.

Bandura A (1986) Social Foundations of Thought and Action: A Social Cognitive Theory.

Englewood Cliffs, NJ: Prentice Hall.

*Barlett CP and Gentile DA (2012) Attacking others online: the formation of cyberbullying

in late adolescence. Psychology of Popular Media Culture 1(2): 123–135.

*Baroncelli A and Ciucci E (2014) Unique effects of different components of trait emotional

intelligence in traditional bullying and cyberbullying. Journal of Adolescence 37(6):

807–815.

*Bastiaensens S, Pabian S, Vandebosch H, et al. (2015) From normative influence to social

pressure: how relevant others affect whether bystanders join in cyberbullying. Social

Development 25(1): 193–211.

 20

*Bauman S (2009) Cyberbullying in a rural intermediate school: an exploratory study.

Journal of Early Adolescence 30(6): 803–833.

Bauman S (2013) Cyberbullying: What does research tell us?. Theory Into Practice 52(4):

249–256.

*Bauman S and Pero H (2010) Bullying and cyberbullying among deaf students and their

hearing peers: an exploratory study. Journal of Deaf Stud Deaf Education 16(2): 236–

253.

Belsey B (2004) Cyber-bullying: an emerging threat to the “always on” generation. Available

at: http://www.bullying.org.html (accessed 10 March 2014).

Beyondblue (2011) Depression in Adolescents and Young Adults. Available at: https://www.

nhmrc.gov.au/_files_nhmrc/publications/attachments/ext0007_cp_guideline_depressi

on_ adolescents_young_beyondblue.pdf (accessed 10 August 2015).

*Bossler AM, Holt TJ and May DC (2012) Predicting online harassment victimization among

a juvenile population. Youth & Society 44(4): 500–523.

Braithwaite V, Ahmed E and Braithwaite J (2009) Workplace bullying and victimization: the

influence of organizational context, shame and pride. International Journal of

Organisational Behaviour 13(2): 71–94.

Brandtzæg PB, Staksrud E, Hagen I, et al. (2015) Norwegian children’s experiences of

cyberbullying when using different technological platforms. Journal of Children and

Media 3(4): 349–365.

*Brewer G and Kerslake J (2015) Cyberbullying, self-esteem, empathy and loneliness.

Computers in Human Behavior 48: 255–260.

*Buelga S, Iranzo B, Cava MJ, et al. (2015) Psychological profile of adolescent

cyberbullying aggressors. Revista de Psicología Social 30(2): 382–406.

Bruning SD (1999) Relationships between organizations and publics: Development of a

 21

multidimensional organization-public relationship scale. Public Relations Review

25(2): 157–170.

*Bussey K, Fitzpatrick S and Raman A (2015) The role of moral disengagement and self-

efficacy in cyberbullying. Journal of School Violence 14(1): 30–46.

*Cappadocia MC, Craig WM and Pepler D (2013) Cyberbullying prevalence, stability, and

risk factors during adolescence. Canadian Journal of School Psychology 28(2): 171–

192.

*Chang FC, Chiu CH, Miao NF, et al. (2015a) Online gaming and risks predict cyberbullying

perpetration and victimization in adolescents. International Journal of Public Health

60(2): 257–266.

*Chang FC, Chiu CH, Miao NF, et al. (2015b) The relationship between parental mediation

and Internet addiction among adolescents, and the association with cyberbullying and

depression. Comprehensive Psychiatry 57: 21–28.

Clauss-Ehlers CS (2010) Encyclopedia of Cross-Cultural School Psychology. New York:

Springer Science & Business Media.

Cohen J (1988) Statistical Power Analysis for the Behavioral Sciences. Mahwah, NJ:

Lawrence Erlbaum Associates.

Cohen J (1992) Statistical power analysis. Current Directions in Psychological Science 1:

98–101.

Colby A and Ehrlich T (2003) Undergraduate Education and the Development of Moral and

Civic Responsibility. Available at: http://www.gwu.edu/~ccps/Colby.html (accessed

10 August 2015).

*Connell NM, Schell-Busey NM, Pearce AN, et al. (2013) Badgrlz? Exploring sex

differences in cyberbullying behaviors. Youth Violence and Juvenile Justice 12(3):

209–228.
 22

Cook CR, Williams KR and Guerra NG (2010) Predictors of bullying and victimization in

childhood and adolescence: a meta-analytic investigation. School Psychology

Quarterly 25(2): 65–83.

*Del Rey R, Elipe P and Ortega-Ruiz R (2012) Bullying and cyberbullying: overlapping and

predictive value of the co-occurrence. Psicothema 24(4): 608–613.

*DePaolis K and Williford A (2015) The nature and prevalence of cyber victimization among

elementary school children. Child & Youth Care Forum 44(3): 377–393.

*Didden R, Sigafoos J and Korzilius H (2009) Form and function of communicative

behaviours in individuals with Angelman syndrome. Journal of Applied Research in

Intellectual Disabilities 22(6): 526–537.

Dehue E, Bolman C and Vollink T (2008) Cyberbullying: youngsters’ experiences and

parental perception. CyberPsychology & Behavior 11(2): 217–223.

*Erdur-Baker Ö (2010) Cyberbullying and its correlation to traditional bullying, gender, and

frequent and risky usage of internet mediated communication tools. New Media &

Society 12(1): 109–126.

Espelage DL, Rao MA and Craven R (2013) Relevant theories for cyberbullying research. In:

Auman S, Walker J and Cross D (eds) Principles of Cyberbullying Research:

Definition, Methods, and Measures. New York and London: Routledge, pp. 49–67.

Felson RB (1996) Mass media effects on violent behavior. Annual Review of Sociology 22:

103–128.

Festl R and Quandt T (2013) Social relations and cyberbullying: the influence of individual

and structural attributes on victimization and perpetration via the Internet. Human

Communication Research 39(1): 101–126.

*Festl R, Scharkow M and Quandt T (2013) Peer influence, internet use and cyberbullying: a

comparison of different context effects among German adolescents. Journal of

 23

Children and Media 7(4): 446–462.

Finkenauer C, Engels RCME and Baumeister RF (2005) Parenting behaviour and adolescent

behavioural and emotional problems: the role of self-control. International Journal of

Behavioral Development 29(1): 58–69.

*Floros GD, Siomos KE, Fisoun V, et al. (2013) Adolescent online cyberbullying in Greece:

the impact of parental online security practices, bonding, and online impulsiveness.

Journal of School Health 83(6): 445–453.

*Foley C, May D, Blevins KR, et al. (2014) An exploratory analysis of cyber-harassment of

K-12 teachers by parents in public school setting. Educational Policy 1: 1–24.

*Gamez-Guadix M, Orue I, Smith PK, et al. (2013) Longitudinal and reciprocal relations of

cyberbullying with depression, substance use, and problematic internet use among

adolescents. The Journal of Adolescent Health 53(4): 446–452.

Gifford R and Nilsson A (2014) Personal and social factors that influence pro environmental

concern and behaviour: a review. International Journal of Psychology 49(3): 141–

157.

Gist ME and Mitchell TR (1992) Self-efficacy: a theoretical analysis of its determinants and

malleability. Academy of Management Review 17(2): 183–211.

*Görzig A and Frumkin L (2013) Cyberbullying experiences on-the-go: when social media

can become distressing. Journal of Psychosocial Research on Cyberspace 7(1): article

1.

*Görzig A and Ólafsson K (2013) What makes a bully a cyberbully? Unravelling the

characteristics of cyberbullies across twenty-five European countries. Journal of

Children and Media 7(1): 9–27.

*Grabe M and Holfeld B (2012) Middle school students’ perceptions of and responses to

cyberbullying. Journal of Educational Computing Research 46(4): 395–413.

 24

*Hase CN, Goldberg SB, Smith D, et al. (2015) Impacts of traditional bullying and

cyberbullying on the mental health of Middle school and High school students.

Psychology in the Schools 52(6): 607–617.

Hedges LV and Olkinm I (1985) Statistical Methods for Meta-Analysis. Prlando, FL:

Academic Press.

Hedges LV and Vevea JL (2007) Fixed- and random-effects models in meta-analysis.

Psychological Methods 3(4): 486–504.

*Heirman W and Walrave M (2008) Assessing concerns and issues about the mediation of

technology in cyberbullying. Journal of Psychosocial Research on Cyberspace 2(2):

1–12.

*Hemphill SA, Kotevski A and Tollit M (2012) Longitudinal predictors of cyber and

traditional bullying perpetration in Australian secondary school students. The Journal

of Adolescent Health 51(1): 59–65.

*Hemphill SA, Tollit M, Kotevski A, et al. (2014) Predictors of traditional and cyber-

bullying victimization a longitudinal study of Australian secondary school students.

Journal of Interpersonal Violence 30(15): 2567–2590.

Hinduja S and Patchin JW (2010) Bullying, cyberbullying, and suicide. Archives of suicide

research 14(3): 206-221.

Ho SS and McLeod DM (2008) Social-psychological influences on opinion expression in

face-to-face and computer-mediated communication. Communication Research 35(2):

190–207.

Ho SS, Poorisat T, Neo RL, et al. (2014) Examining how presumed media influence affects

social norms and adolescents’ attitudes and drinking behavior intentions in rural

Thailand. Journal of Health Communication 19(3): 282–330.

*Holfeld B and Grabe M (2012) An examination of the history, prevalence, characteristics,

 25

and reporting of cyberbullying in the United States. In: Li Q, Cross D and Smith PK

(eds) Cyberbullying in the Global Playground: Research from International

Perspectives. Chichester: Wiley-Blackwell, pp. 117–142.

*Holfeld B and Leadbeater BJ (2015) The nature and frequency of cyber bullying behaviors

and victimization experiences in young Canadian children. Canadian Journal of

School Psychology 30(2): 116–135.

*Holt TJ, Fitzgerald S, Bossler AM, et al. (2014) Assessing the risk factors of cyber and

mobile phone bullying victimization in a nationally representative sample of

Singapore youth. International Journal of Offender Therapy and Comparative

Criminology 1: 1–18.

Hymel S, Rocke-Henderson N and Bonanno RA (2005) Moral disengagement: a framework

for understanding bullying among adolescents. Journal of Social Sciences 8(1): 1–11.

*Kellerman I, Margolin G, Borofsky LA, et al. (2013) Electronic aggression among emerging

adults motivations and contextual factors. Emerging Adulthood 1(4): 293–304.

Kim SY and Wong VY (2002) Assessing Asian and Asian American parenting: a review of

the literature. In: Kurasaki K, Okazaki S and Sue S (eds) Asian American Mental

Health: Assessment Theories and Methods. New York: Kluwer, pp. 85–201.

Kowalski RM, Giumetti GW, Schroeder AN, et al. (2012) Cyber bullying among college

students: evidence from multiple domains of college life. In: Wankel LA and Wankel

C (eds) Misbehavior Online in Higher Education. Bingley: Emerald Group Publishing

Limited, pp. 293–321.

Kowalski RM, Giumetti GW, Schroeder AN, et al. (2014) Bullying in the digital age: a

critical review and meta-analysis of cyberbullying research among youth.

Psychological Bulletin 140(4): 1073–1139.

Kowalski RM, Limber SP and Agatston PW (2008) Cyberbullying: Bullying in the Digital
 26

Age. Malden, MA: Blackwell Publishers.

*Kowalski RM, Morgan C and Limber S (2012) Traditional bullying as a potential warning

sign of cyberbullying. School Psychology International 33(5): 505–519.

*Kwan GCE and Skoric MM (2013) Facebook bullying: an extension of battles in school.

Computers in Human Behavior 29(1): 16–25.

*Lazuras L, Barkoukis V, Ourda D, et al. (2013) A process model of cyberbullying in

adolescence. Computers in Human Behavior 29(3): 881–887.

*Lester L, Cross D and Shaw T (2012) Problem behaviours, traditional bullying and

cyberbullying among adolescents: longitudinal analyses. Emotional and Behavioural

Difficulties 17(3–4): 435–447.

*Li Q (2007) New bottle but old wine: a research of cyberbullying in schools. Computers in

Human Behavior 23(4): 1777–1791.

*Low S and Espelage D (2013) Differentiating cyber bullying perpetration from non-physical

bullying: commonalities across race, individual, and family predictors. Psychology of

Violence 3(1): 39–52.

Lwin MO, Li B and Ang RP (2012) Stop bugging me: an examination of adolescents’

protection behavior against online harassment. Journal of Adolescence 35(1): 31–41.

*MacDonald CD and Roberts-Pittman B (2010) Cyberbullying among college students:

prevalence and demographic differences. Procedia – Social and Behavioral Sciences

9: 2003–2009.

*Marcum CD, Higgins GE and Ricketts ML (2010) Potential factors of online victimization

of youth: an examination of adolescent online behaviors utilizing routine activity

theory. Deviant Behavior 31(5): 381–410.

*Mark L and Ratliffe KT (2011) Cyber worlds: new playgrounds for bullying. Computers in

the Schools 28(2): 92–116.

 27

*Mesch GS (2009) Parental mediation, online activities, and cyberbullying. Cyberpsychology

& Behavior 12(4): 387–393.

Microsoft (2012) Online Bullying among Youth 8–17 Years Old. Available at:

http://download. microsoft.com/download/E/8/4/E84BEEAB-7B92-4CF8-B5C7-

7CC20D92B4F9/WW%20Online%20Bullying%20Survey%20%20Executive%20Su

mmary%20-%20China_Final.pdf (accessed 10 December 2013).

*Mishna F, Khoury-Kassabri M, Gadalla T, et al. (2012) Risk factors for involvement in

cyber bullying: victims, bullies and bully–victims. Children and Youth Services

Review 34(1): 63–70.

*Mitchell KJ, Finkelhor D, Wolak J, et al. (2011) Youth Internet victimization in a broader

victimization context. The Journal of Adolescent Health 48(2): 128–134.

*Mitchell KJ, Wolak J and Finkelhor D (2008) Are blogs putting youth at risk for online

sexual solicitation or harassment? Child Abuse & Neglect 32(2): 277–294.

Mouttapa M, Valente T, Gallaher P, et al. (2004) Social network predictors of bullying and

victimization. Adolescence 39(154): 315–335.

*Navarro JN and Jasinski JL (2012) Going cyber: using routine activities theory to predict

cyberbullying experiences. Sociological Spectrum 32(1): 81–94.

*Navarro R, Serna C, Martínez V, et al. (2013) The role of Internet use and parental

mediation on cyberbullying victimization among Spanish children from rural public

schools. European Journal of Psychology of Education 28(3): 725–745.

NCH (2005) Putting U in the Picture: Mobile Bullying Survey. Available at:

http://www.avapro-ject.org.uk/media/28482/mobile_bullying_report.pdf (accessed 10

December 2013).

*Pabian S and Vandebosch H (2014a) Developmental trajectories of (cyber) bullying

perpetration and social intelligence during early adolescence. The Journal of Early
 28

Adolescence 36(2): 145–170.

*Pabian S and Vandebosch H (2014b) Using the theory of planned behaviour to understand

cyberbullying: the importance of beliefs for developing interventions. European

Journal of Developmental Psychology 11(4): 463–477.

*Pabian S and Vandebosch H (2015a) An Investigation of Short-Term Longitudinal

Associations between Social Anxiety and Victimization and Perpetration of

Traditional Bullying and Cyberbullying 45(2): 328–339.

*Pabian S and Vandebosch H (2015b) Short-term longitudinal relationships between

adolescents’(cyber) bullying perpetration and bonding to school and teachers.

International Journal of Behavioral Development 1: 1–11.

Patchin JW and Hinduja S (2006) Bullies move beyond the schoolyard a preliminary look at

cyberbullying. Youth violence and juvenile justice 4(2): 148–169.

Patchin JW and Hinduja S (2010) Cyberbullying and self-esteem. Journal of School Health

80(12): 614–621.

Paul B, Salwen MB and Dupagne M (2000) The third-person effect: a meta-analysis of the

perceptual hypothesis. Mass Communication & Society 3(1): 57–85.

*Perren S and Gutzwiller-Helfenfinger E (2012) Cyberbullying and traditional bullying in

adolescence: differential roles of moral disengagement, moral emotions, and moral

values. European Journal of Developmental Psychology 9(2): 195–209.

*Perren S and Sticca F (2011) Bullying and morality: are there differences between

traditional bullies and cyberbullies? The biennial SRCD meeting, Montréal, QC,

Canada, March.

Perren S, Dooley J, Shaw T, et al. (2010) Bullying in school and cyberspace: associations

with depressive symptoms in Swiss and Australian adolescents. Child and Adolescent

Psychiatry and Mental Health 4(28): 1–10.

 29

*Pornari CD and Wood J (2010) Peer and cyber aggression in secondary school students: the

role of moral disengagement, hostile attribution bias, and outcome expectancies.

Aggressive Behavior 36(2): 81–94.

*Price M, Chin MA, Higa-McMillan C, et al. (2013) Prevalence and internalizing problems

of ethnoracially diverse victims of traditional and cyber bullying. School Mental

Health 5(4): 183–191.

*Pyżalski J (2012) From cyberbullying to electronic aggression: typology of the

phenomenon. Emotional and Behavioural Difficulties 17(3–4): 305–317.

*Reyns BW, Burek MW, Henson B, et al. (2013) The unintended consequences of digital

technology: exploring the relationship between sexting and cybervictimization.

Journal of Crime and Justice 36(1): 1–17.

*Roberto AJ, Eden J, Savage MW, et al. (2014) Prevalence and predictors of cyberbullying

perpetration by high school seniors. Communication Quarterly 62(1): 97–114.

*Robson C and Witenberg RT (2013) The influence of moral disengagement, morally based

self-esteem, age, and gender on traditional bullying and cyberbullying. Journal of

School Violence 12(2): 211–231.

Rosenthal R (1991) Meta-Analytic Procedures for Social Research. Newbury Park, CA:

SAGE.

Scott S, Doolan M, Beckett C, et al. (2012) How Is Parenting Style Related to Child

Antisocial Behaviour? Preliminary Findings from the Helping Children Achieve

Study. London: UK Department of Education.

*Shapka JD and Law DM (2013) Does one size fit all? Ethnic differences in parenting

behaviors and motivations for adolescent engagement in cyberbullying. Journal of

Youth and Adolescence 42(5): 723–738.

Slonje R and Smith PK (2008) Cyberbullying: another main type of bullying? Scandinavian
 30

Journal of Psychology 49(2): 147–154.

Smith PK, Mahdavi J, Carvalho M, et al. (2008) Cyberbullying: Its nature and impact in

secondary school pupils. Journal of child psychology and psychiatry 49(4): 376–385.

Sohn D (1996) Publication bias and the evaluation of psychotherapy efficacy in reviews of

the research literature. Clinic Psycholology Review 16(2): 147–156.

*Sticca F, Ruggieri S, Alsaker F, et al. (2013) Longitudinal risk factors for cyberbullying in

adolescence. Journal of Community & Applied Social Psychology 23(1): 52–67.

Suler J (2004) The online disinhibition effect. Cyberspace Behavior 7(3): 321–326.

*Tarablus T, Heiman T and Olenik-Shemesh D (2015) Cyber bullying among teenagers in

Israel: an examination of cyber bullying, traditional bullying, and socioemotional

functioning. Journal of Aggression, Maltreatment & Trauma 24(6): 707–720.

Tokunaga RS (2010) Following you home from school: a critical review and synthesis of

research on cyberbullying victimization. Computers in Human Behavior 26(3): 277–

287.

*Twyman K, Saylor C, Taylor LA, et al. (2010) Comparing children and adolescents engaged

in cyberbullying to matched peers. Cyberpsychology, Behavior & Social Networking

13(2): 155–199.

Valkenburg PM and Peter J (2013) The differential susceptibility to media effects model.

Journal of Communication 63(2): 221–243.

*Vazsonyi AT, Machackova H, Sevcikova A, et al. (2012) Cyberbullying in context: direct

and indirect effects by low self-control across 25 European countries. European

Journal of Developmental Psychology 9(2): 210–227.

*Walrave M and Heirman W (2011) Cyberbullying: predicting victimisation and

perpetration. Children & Society 25(1): 59–72.

*Wegge D, Pabian S, Vandebosch H, et al. (2015) Popularity through online harm: the
 31

longitudinal associations between cyberbullying and sociometric status in early

adolescence. The Journal of Early Adolescence 31(1): 86–107.

*Williford A, Elledge LC, Boulton AJ, et al. (2013) Effects of the KiVa antibullying program

on cyberbullying and cybervictimization frequency among Finnish youth. Journal of

Clinical Child and Adolescent Psychology 42(6): 820–833.

*Wolak J, Mitchell KJ and Finkelhor D (2007) Does online harassment constitute bullying?

An exploration of online harassment by known peers and online-only contacts. The

Journal of Adolescent Health 41(6): S51–S58.

*Workman M (2010) A behaviorist perspective on corporate harassment online: validation of

a theoretical model of psychological motives. Computers & Security 29(8): 831–839.

*Wright MF and Li Y (2013) Normative beliefs about aggression and cyber aggression

among young adults: a longitudinal investigation. Aggressive Behavior 39(3): 161–

170.

*Wright MF, Kamble SV and Soudi SP (2015) Indian adolescents’ cyber aggression

involvement and cultural values: the moderation of peer attachment. School

Psychology International 1: 1–18.

*Xiao B and Wong YM (2013) Cyber-bullying among university students: an empirical

investigation from social cognitive perspective. International Journal of Business and

Information 8(1): 34–69.

*Yang S, Kim J, Kim S, et al. (2006) Bullying and victimization behaviours in boys and girls

at South Korean primary schools. Journal of the American Academy of Child and

Adolescent Psychiatry 45(1): 69–77.

*Ybarra ML and Mitchell KJ (2004) Youth engaging in online harassment: associations with

caregiver-child relationships, Internet use, and personal characteristics. Journal of

Adolescence 27(3): 319–336.

 32

*Zhou Z, Tang H, Tian Y, et al. (2013) Cyberbullying and its risk factors among Chinese

high school students. School Psychology International 34(6): 630–647.

 33

Tables

Table 1.
Meta-analysis of predictors of cyberbullying perpetration.
95% CI
Correlates k N Effect Low High Q
Size
Media Exposure
Frequency of ICT use 18 21,035 .20 .14 .25 303.98***
Risky ICT use 9 15,083 .29 .23 .36 139.50***
Personal Factors
Self-esteem 6 6489 -.13 -.19 -.07 2.38
Moral disengagement 10 5854 .28 .20 .36 86.97***
Narcissism 2 764 .23 .16 .29 0.33
Depression 7 6262 .25 .16 .35 75.24***
Self-efficacy 5 4233 .21 .07 .33 74.74***
Emotional management 4 2546 -.09 -.15 -.03 5.29
Popularity 3 1307 -.26 -.21 .63 112,73***
School commitment 3 8066 -.10 -.20 -.01 35.72***
TB perpetration 32 75,396 .39 .35 .43 1062.38***
TB victimization 22 66,586 .19 .14 .24 650.93***
Environmental Factors
Social norms 6 6167 .27 .17 .37 82.62***
Parental interaction 6 8202 -.17 -.29 -.04 161.95***
Parental mediation 6 7362 -.07 -.14 -.01 43.94***
CI: confidence interval; ICT: information and communication technology; TB: traditional
bullying; k: number of studies; N: total sample size for all studies combined; effect size:
Pearson’s r; 95% CI: lower and upper limits of 95% CI for effect size.
Effect size calculations were based on the random effects model.
*** p < .001.
 34

Table 2.
Meta-analysis of predictors of cyberbullying victimization.
95% CI
Correlates k N Effect Low High Q
Size
Media Exposure
Frequency of ICT use 23 50,523 .19 .14 .23 442.40***
Risky ICT use 11 14,837 .26 .19 .32 141.06***
Personal Factors
Self-esteem 5 2766 -.22 -.28 -.15 10.78
Moral disengagement 6 4592 .14 .06 .23 39.12***
Depression 9 33,443 .20 .13 .28 169.58***
Self-efficacy in 2 2975 -.03 -.06 -.01 0.90
defending
Emotional management 4 4044 -.20 -.34 -.08 1.66
TB perpetration 22 67,637 .21 .18 .24 189.71***
TB victimization 24 71,474 .32 .30 .36 721.19***
Environmental Factors
Parental interaction 6 6907 -.09 -.12 -.06 7.17
Parental mediation 8 8331 -.07 -.13 -.0` 43.93***
CI: confidence interval; ICT: information and communications technology; TB: traditional
bullying; k: number of studies; N: total sample size for all studies combined; effect size:
Pearson’s r; 95% CI: lower and upper limits of 95% CI for effect size.
Effect size calculations were based on the random effects model.
***p < .001.
 35

Table 3.
Moderator analysis: country
DV Predictors Categories k Effect Q
size
Cyberbullying perpetration Parental interaction Asia-Pacific 2 -.07 100.76***
Europe 1 -.40
North America 3 -.28
DV: Dependent variable; k: number of studies; effect size: Pearson’s r. Effect size
calculations were based on the random effects model.
***p < .001.
 36

Table 4.
Moderator analysis: sampling
DV Predictors Categories k Effect size Q
Cyberbullying Frequency of ICT use Non-random 11 .25 4.19*
perpetration Random 7 .13
Parental interaction Non-random 1 .05 6.43*
Random 5 .20
Cyberbullying Frequency of ICT use Non-random 10 .27 10.70**
victimization Random 13 .10
DV: Dependent variable; ICT: information and communications technology; k: number of
studies; effect size: Pearson’s r. Effect size calculations were based on the random effects
model.
*p < .05; **p < .01.
 37

Table 5.
Moderator analysis: age
DV Predictors Categories k Effect Q
size
Cyberbullying TB perpetration Adult 3 .50 80.10***
perpetration Teen 27 .40
Child 2 .14
TB victimization Adult 1 .33 21.41***
Teen 18 .20
Child 3 .01
Cyberbullying Frequency of ICT use Adult 4 .05 12.95**
victimization Teen 18 .17
Child 1 .14
TB victimization Adult 1 .48 13.13**
Teen 20 .33
Child 3 .24
DV: Dependent variable; TB: traditional bullying; ICT: information and communications
technology; k: number of studies; Effect size: Pearson’s r. Effect size calculations were based
on the random effects model.
**p < .01; ***p < .001.
 38

Table 6.
Moderator analysis: media platform
DV Predictors Categories k Effect Q
size
Cyberbullying Frequency of ICT use Internet 8 .17 12.14**
perpetration Social media 8 .24
Electronic media 2 .10
Cyberbullying Frequency of ICT use Internet 10 .18 23.12***
victimization Social media 8 .27
Electronic media 5 .06
DV: Dependent variable; ICT: information and communications technology; k: number of
studies; effect size: Pearson’s r. Effect size calculations were based on the random effects
model.
**p < .01; ***p < .001.