Cyberbullying 1

RUNNING HEAD: CYBERBULLYING

Cyberbullying among high school students: Cluster analysis of sex and age differences and the

level of parental monitoring.

Paper accepted for publication in the International Journal of Cyber Behavior, Psychology and

Learning (IJCBPL). (in press)

Baylor University

One Bear Place 97301

Waco TX 76798-7301

Ikuko Aoyama Ikuko_aoyama@baylor.edu

Tel. 254-652-5356 /Fax. 254-710-3265

Lucy Barnard-Brak Lucy_Barnard-Brak@baylor.edu

Tel. 254-710-4234 /Fax. 254-710-3265

Tony Talbert Tony_Talbert@baylor.edu

Tel. 254-710-7417 /Fax 254-710-3160

Cyberbullying 2

Abstract

Bullying, a once typical occurrence in schools, has gone digital. As a result, cyberbullying has

become ever more present among youth. The current study aimed to classify high school

students into four groups based on their cyberbullying experiences and to examine the

characteristics of these groups based on the sex and age of the participants and the level of

parental monitoring. Participants were 133 high school students located in central Texas. A

cluster analysis revealed four distinct groups of students who were: ―highly involved both as

bully and victim,‖ ―more victim than bully,‖ ―more bully than victim,‖ or ―least involved.‖

Significantly more girls and more students in lower grades were classified into the ―more victim

than bully group‖ while older students were more likely to be classified into the ―more bully than

victim‖ group. No significant differences were found between cluster membership and the

degree of parental monitoring.

Cyberbullying 3

Introduction

Cyberbullying among youth has been becoming a serious societal and educational

concern internationally. The media, educators, and parents have been paying great attention to

the phenomena for the past few years because researchers from various countries have revealed

the relatively high prevalence of cyberbullying among youth. For example, approximately 30%

of youth (N=384) surveyed in 2004 reported their victimization, and 11% have cyberbullied

others (Hinduja & Patchin, 2009). The more recent study shows that 72% of the youth

(N=1,454) were victimized at least once in the past year, and 13% of them reported frequent

victimization (Juvonen & Gross, 2008).

Theoretical Background of Cyberbullying

Researchers have linked bullying behaviors with theories of human behaviors and

communication. For example, the well- known theory is the social cognitive theory, which

argues that adolescents model their parents or friends’ aggressive behaviors (Duncan, 2004;

Mouttapa, Valente, Gallaher, Rohrbach, & Unger, 2004). ―The effect [of the model] will be

stronger if the observer has a positive evaluation of the model, for example, perceive, him/herself

as tough, fearless, and strong‖ (Olweus, 1993, p. 43). In other words, observing an aggressive

model makes aggressive behaviors less inhibited if observers see a model getting rewarded for

the aggressive actions. In these cases, the reward means the bullies’ victory over the victims.

Thus, all forms of bullying may be learned actions (Hinduja & Patchin, 2008) because bullying

is a type of peer aggression.

One theoretical model that can possibly explain cyberbullying is desinhibited behavioral

effects on the Internet (Hinduja & Patchin, 2009; Kowalski et al., 2008). Joinson (1998) argues

Cyberbullying 4

that people in cyberspace behave in a way they do not in real life because of the effects of

disinhibition: ―Disinhibition means that normal behavioral restraint can become lost or

disregarded‖ (Mason, 2008, p. 328). For example, researchers have demonstrated that people

tend to behave more bluntly when communicating by e-mail or in other electronic venues.

Moreover, misunderstandings, greater hostility, aggressive responses, and nonconforming

behaviors are more likely in computer-mediated communication than in face-to-face

communication (McKenna & Bargh, 2000). In face-to-face interaction, people read the

emotional reactions of others and modulate their own behavior in response to the consequences

(Kowalski et al., 2008). In other words, human behaviors are inhibited by social situations and

public evaluations (Joinson, 1998). In cyberspace context, on the other hand, people have less

social, contextual, and affective signs than in face-to-face communication; thus, they are less

sensitive and remorseful for the types of behaviors that they exhibit (Mason, 2008). In

cyberbullying, perpetrators have no direct social disapproval and punishment for engaging in

bullying others and do not see that victims suffer (Willard, 2007). As a result, their behaviors are

often disinhibited and become ruder, harsher, and more difficult to control (Hinduja & Patchin,

2009).

Disinhibition effects are caused by deindividuation (Joinson, 1998). Deindividuation can

occur when accountability cues are reduced; in other words, anonymity can reduce concerns

about others’ reactions (Joinson, 1998). Deinvididuation also occurs when an individual’s self-

awareness is blocked or reduced by external factors because ―it decreases the influence of

internal (i.e., self) standards of or guides to behavior, and increases the power of external,

situational cues‖ (McKenna & Bargh, 2000, p. 61- 62).

Students’ Status in a Peer Group

Cyberbullying 5

In traditional bullying studies, differences among children have been conceptualized

through categorizing them into four groups: bullies, victim, bully/victims, and not involved

(Espelage & Holt, 2007). However, little research examining the group differences has been

conducted in cyberbullying research. The discrimination among groups is important because

these subgroups exhibit different patterns of aggression and behavioral and internal problems

(Espelage & Holt, 2007). Understanding the group differences is also necessary to deliver an

effective intervention.

Many cyberbullying studies are currently focusing either victims or bullies. However,

bully/victim students who are involved as both bullies and victims are often overlooked.

Researchers have suggested that a child’s status as a bully or victim could be easily

interchanged; for instance, 35.7% of bullies reported experienced being victimized within the

year, and 15.5% of them were currently being victimized as well (Morita et al., 1999). In

addition, traditional bullying studies have shown that bully/victim students have the highest risk

of behavioral and emotional problems because bully/victims experience double negative effects

as both bullies and victims (Marini, Dane, Bosacki, & Ylc-Cura, 2006). In fact, their evaluation

by teachers and peers are low. For example, bully/victims are seen as ―more clumsy and

immature than their peers [and] not only do peers find it difficult to associate with these children,

but teachers and other school personnel frequently report that these children are among the most

difficult to work with in school settings‖ (Kowalski, Limber, & Agatson, 2008, p. 32). Similarly,

bully/victims report a higher rate of depression, somatization, and psychiatric referrals than all of

their peer groups (Ybarra & Mitchell, 2004). Considering all of the fact, the researchers of the

present study believe it is important to identify the distinct subgroups of youth who are

involved/not involved with cyberbullying.

Cyberbullying 6

Sex Differences in Cyberbullying

In traditional bullying, studies have shown that boys were more likely than girls to be

involved in bullying overall; however, more girls experience indirect and psychological types of

bullying such as rumor spreading and social exclusion (Kowalski et al, 2008; Ma, 2002; Olweus,

1993; Raskauskas & Stoltz, 2007). Therefore, researchers have pointed out that cyberbullying is

more prevalent among girls (Anderson & Sturm, 2007; Willard, 2007) because this

cyberbullying is text-based, and girls tend to be more verbal than boys (Hinduja & Patchin,

2009). However, research findings are inconsistent across studies. Some studies found that boys

were more likely to engage in cyberbullying than girls (Dehue , Bolman, & Völlink, 2008;

Katzer, Fetchenhauer, & Belschak, 2009; Shariff, 2008), and girls were more likely to be

victimized online (Dehue et al., 2008; Smith et al., 2008). On the other hand, Li (2006) argues

that more boys reported being cyberbullied than girls. Other researchers, however, find no

significant sex differences (Arıcak, 2009; Beran & Li, 2005).

Age Differences

Research findings on age differences of youth cyberbullying experiences also vary. While

studies in Britain and Canada found no age effects (Beran & Li, 2005; Smith et al., 2008), other

studies identified differences. For example, researchers have argued that cyberbullying peaks

later in middle school or in high school (Hinduja & Patchin, 2009; Kowalski & Limber, 2007).

A survey conducted by Pew/Internet American Life Project (N=935) also reveals that older girls

aged 15 to17 are more likely to report being bullied online than any other age and gender group

(Lenhart, 2007). Likewise, Japanese high school students reported that cyberbullying was more

prevalent among middle school students (Aoyama & Talbert, 2009). In contrast, primary pupils

Cyberbullying 7

in the Netherlands reported their cyberbullying experiences more often than secondary students

did (Dehue et al., 2008).

Gap between Adults and Youth

In spite of the high prevalence of cyberbullying among adolescents, studies suggest that

adults underestimate the incidents. For example, ―The percentage of parents reporting that their

child was engaged in bullying on the Internet or via text messages was considerably lower

(4.8%) than the percentage of children reporting to be engaged in bullying on the Internet or via

text messages (17.3%)‖ (Dehue et al., 2008, p. 219). This finding is consistent with a study by

Bradshaw, Sawyer, and O-Brennan (2007) which indicates that adults estimate the incidents of

traditional bullying: over 49% of children (N=15,185) reported being bullied at least once during

the past month; whereas, 71.4% of staff (N=1,547) estimated that 15% or less of the students at

their school were frequently bullied. In addition, only less than 1% of staff members reported

bullying rates similar to those indicated by students. These findings suggest that adults may not

fully aware of bullying/cyberbullying incidents happening to their children.

Parents’ Monitoring Roles

Traditional bullying studies indicate the association between parental monitoring and

bullying behavior. For example, parents with permissive parenting are less likely to acknowledge

their children’s activities (Marini et al., 2006), and parents of bully-victims often display the

indifferent-uninvolved parenting style, neglect, and inconsistently monitor their children

(Duncan, 2004). In addition, Wienke Totura et al (2009) found that the level of adult monitoring

negatively correlated with bullying behaviors.

Cyberbullying 8

The parental monitoring plays key roles because cyberbullying often occurs at home;

however, parental monitoring strategies do not seem to work well. Mason (2008) said about 30%

of adolescents use the Internet for 3 hours or more daily, and during these hours, more than 50%

of them reported poor parental monitoring. Rosen (2007) also pointed out many parents ―were

unsure what their children were doing online, but didn’t know how to approachthe subject with

their teens‖ (p. 80). Similarly, McQuade et al. (2009) found that 93 percent of parents stated they

established Internet rules for their child’s; however, 37 percent of children reported being given

no rules from their parents on the Internet activity. Likewise, Rosen (2007) found that even

though the majority of parents set limits on their children’s Internet use, they are not actually

monitoring those limits. These findings indicate the difficulty of effective parental monitoring. In

fact, Mesch (2009) reanalyzed a large secondary data of nationally representative youth sample

(N=945) and found that parental mediation and monitoring are not very effective.

Purpose of the study

Cyberbullying research is still in its infancy; thus, research findings are inconsistent

across researchers. It is possible that these mixed findings are due to the lack of knowledge

regarding the level of students’ involvement in cyberbullying. Therefore, the purpose of the

present study was to identify the subgroups of youth who are involved with cyberbullying and to

examine any sex and age differences among these groups. The research questions consist of the

following:

1) Can we classify the students based on their cyberbullying experiences?

2) Are there any sex differences among the groups?

3) Are there any age differences among the groups?

4) If there are age differences among the group, does the association indicate a trend?

Cyberbullying 9

5) Are there any significant relationships between the degree of parental monitoring and

cluster membership?

Method

Participants

Participants were selected from a public high school located in central Texas: 133 high

school students (Male = 52.7%, n = 68, Female = 47.3%, n = 61). Students who were taking

computer classes were invited to participate in the study. 89.5% (n = 119) of the participants

were Caucasian, and the remaining 21% (n = 14) were African American, Hispanic, and others.

Of these students, 57 (43.2%) were ninth graders, 20 (15.2%) were tenth graders, 41 (31.1%)

were eleventh graders, and 14 (10.6%) were twelfth graders, and their mean age was 15.7 years

old (SD = 1.25). The school is located in rural area with 85.7% of the students in this school

being Caucasian (SchoolDataDirect, 2009). Thus, the sample of this present study may be

considered representative as a sample of this high school.

Instrument

The self-report survey uploaded on a web-based online survey management tool was

used. The survey was modified from the one created by Willard (2007) and Smith, Mahdavi,

Carvalho, and Tippett (n.d.). The survey consists of 55 questions, including demographic

information, and open-ended questions. Sixteen questions were used for the analysis of the study.

The duration of the survey was approximately 20 minutes to complete. The participants accessed

the online survey during computer classes at the high school.

Measures

Cyberbullying offending behaviors

Five questions assessed the frequency of cyberbullying offending behaviors (e.g., ―In the

last six months, have you sent mean or nasty messages to someone?‖, and ―In the last six months,

Cyberbullying 10

have you put down someone else online by sending or posting cruel, gossip, rumors, or other

harmful materials?‖). These questions addressed various types of cyberbulllying such as text

message, email, and Internet community. Response choices were ―Yes, 1 to 4 times‖ (coded as

1), ―Yes, more than 5 times‖ (coded as 2), and ―No‖ (coded as 0). Higher scores indicate more

frequent offending behaviors. Cronbach’s alpha was 0.69, and its value is close to the

Cyberbulling Offending Scale (Cronbach’s alpha = 0.76) developed by Hinduja & Patchin (2009).

Cyberbullying victimization

A self-report cyberbullying victimization, including name-calling, social exclusion,

rumor spreading, was measured by the six questions (e.g., ―In the last six months, have you

received online messages that made you fear for your safety?‖ and ―In the last six months, have

you been put down online by someone who has sent or posted cruel, gossip, rumors, or other

harmful materials?‖). Response choices and coding system were the same as above. Higher

scores indicate more frequent victimization. Five questions measured offending behaviors, and

six questions measured victimization; thus, the scores were standardized to provide a standard

metric. Cronbach’s alpha was 0.72, and its value is close to the Cyberbulling Victimization Scale

(Cronbach’s alpha = 0.73) developed by Hinduja & Patchin (2009).

Parental monitoring

Two questions assessed the level of parental monitoring (e.g., ―How often do you discuss

what you are doing online with your parents?‖). Response choices were ―Frequently‖ (coded as

2), ―Occasionally‖ (1), and ―Never‖ (0).

Data Analysis

First, characteristics of the data distribution were evaluated (e.g. skewness). According to

the z-score, both offending and victimization measure data indicated a moderate degree of

Cyberbullying 11

positive skeweness; thus, the data were transformed by square rooting the composite score. This

data transformation procedure has been recommended by Tabachnick and Fidall (2001) when

moderate positive skewness was observed. Second, a k-mean cluster analysis was performed to

classify students into four groups. The number of cluster was determined based on a review of

extant literature. Previous research utilized a k-mean cluster analysis as an appropriate analysis

(Espelage & Holt, 2007). Subsequently, a multivariate analysis of variance (MANOVA) was

conducted to ensure that the cluster analysis had classified the participants accurately. In

conducting our MANOVA, a Box’s test was also conducted to test the assumption of the equality

of covariance matrices. Third, a 2 (sex) x 4 (groups) chi square (χ2) analysis was conducted to

examine the presence of any sex differences. As a measure of effect size, a Phi coefficient (Φ)

was also calculated. Phi values of 0.1, 0.3, and 0.5 may be interpreted as small, medium, and

large association between groups respectively (Green & Salkind, 2004). Then, standardized

residuals in each cell greater or lesser than 1.96 were considered as being statistically significant

at the 0.05 level or less. Fourth, a one-way analysis of variance (ANOVA) was performed to

examine the relationship between age and cluster group membership. A Levene’s F test was

conducted to examine if the assumption of homogeneity of variances was met. Finally,

nonresponse items were handled by a pair-wise deletion method because missing data consisted

of only about 10% for both offending and victimization measures. SPSS 16.0 was used for all

data analyses.

Results

Correlations

A statistically significant, positive correlation between offending and victim scale score

was found (r = 0.62, p < 0.01). This result indicates that students who cyberbullied others are

Cyberbullying 12

likely to be also victimized. In addition, a statistically significant, negative correlation between

offending scale score and sex was observed (r =- 0.24, p < 0.01). This result indicates that girls

are less likely than boys to cyberbully others. As for the level of parental monitoring, age showed

statistically significant, negative correlation (r =- 0.24, p < 0.01) and sex showed a statistically

significant, positive correlation (r =0.21, p < 0.05). These results indicate that the level of

parental monitoring is higher for younger children and girl. The level of parental monitoring was

also negatively correlated with offending scale score (r =- 0.19, p < 0.05). This result indicates

that students are less likely to cyberbully others as the level of parental monitoring increases.

There were no statistically significant correlations among sex, age, and victimization scale score.

Cluster Analysis

It was hypothesized from literature on traditional bullying that four clusters would

emerge. Cluster one was termed the ―least involved‖ group, and scored the lowest on both

offending and victim scale scores. The group included 68 students (51.1 % of the sample).

Cluster two was termed the ―highly involved both as a bully and a victim‖ group, and scored the

highest scores on both offending and victim scale scores. The group included 17 students

(12.8 % of the sample). Cluster three was termed the ―more bully than victim‖ group and scored

the second highest score on offending scale score and the second lowest on victim scale score.

The group included 14 students (10.5 % of the sample). Cluster four was termed the ―more

victim than bully‖ group and scored the second highest on victim scale score and the lowest on

offending scale score. The group included 13 students (9.8 % of the sample). Of the 133 students,

21 students did not complete the survey; thus, they are treated as missing data. Table 1 contains

the descriptive statistics according to cluster membership while Figure 1 contains a graphic

display of offending and victimization scale scores according to cluster membership.

Cyberbullying 13

[Insert Table 1]

[Insert Figure 1]

Further, results from the MANOVA indicated the distinctions among four subgroups

were significant on both bullying (F(3,108) = 351.72, p < 0.01, η2 = 0.90) and victimization

measures (F(3,108) = 240.65, p < 0.01, η2 =0.87). The Box test was significant (F(3,4) = 6.82, p

< 0.01), thus the assumption of the equality of covariances was violated. As a result, the

MANOVA statistic reported was the Wilks’ Lambda, which was significant (Λ = .02, F(6,214)=

0.20, p < 0.01, η2 = 0.85).

Sex differences

A 2 (sex) x 4(cluster) chi-square analysis revealed statistically significant differences for

sex, χ2 (3, N =133) =11.63, p < 0.05, Φ =0.36). The value of Phi indicated a medium strength of

association between sex and student group membership. In examining the standardized residuals

for each cell in the chi-square analysis, results indicated that significantly more girls (n =11, Std

residual = 2.1) than boys (n = 2, Std residual = -1.9) were classified into the ―more victim than

bullys‖ group. However, no other significant sex differences emerged.

Age differences

A Levene’s F test indicated that the assumption of homogeneity of variance was not met,

F(3, 124) = 3.21, p < 0.05. A one-way ANOVA revealed statistically significant differences for

age (F(3, 124) = 3.96, p < 0.05, η2= 0.87), suggesting that cluster membership was associated

with students’ age. Younger students (M = 14.92, SD = 0.33) were more likely to be in the ―more

victim than bully‖ group, but as they become a little older, the trend seemed to split into two

directions: ―Highly involved‖ (M = 15.76, SD = 0.29) or ―Least involved‖ (M = 15.61, SD =

0.13). Then, older students (M = 16.39, SD = 0.28) were more likely to be in the ―more bully

Cyberbullying 14

than victim‖ group. Both linear and quadratic trend appear to be significant; however, quadratic

trend seems to fit the data slightly better, F(3, 124) = 3.96, p < 0.05. Figure 2 contains a graphic

display of the relationship between age and cluster membership.

[Insert Figure 2]

The level of parental monitoring

A Levene’s F test indicated that the assumption of homogeneity of variance was not met, F(3,

124) = 3.04, p < 0.05. A one-way ANOVA revealed statistically nonsignificant differences for

the level of parental monitoring (F(3, 124) = 2.08, p> 0.05), suggesting that cluster membership

was not associated with the level of parental monitoring.

Discussion

The purpose of the present study was to identify the subgroups of youth who are

involved with cyberbullying and to examine sex and age differences and the level of parental

monitoring among these groups. A cluster analysis identified four groups: ―least involved‖,

―highly involved both as a bully and victim‖, ―more bully than victim‖, and ―more victim than

bully‖ group. Although the majority of students were in the ―least involved‖ group (51.1 %),

about 10% of the students were in the ―highly involved both as a bully and victim‖ group‖, and

the rest of the students (21%) had also experienced cyberbullying at least once. In traditional

bullying studies, differences among children have been conceptualized through categorizing

them into four groups: bullies, victim, bully/victims, and not involved (Espelage & Holt, 2007).

However, the analyses indicate that it is rare for high school students to be pure cyberbullies

and/or cybervictims.

As for sex differences, girls were more likely to be in the ―more victim than bully‖ group

than boys. This result is consistent with findings arguing that more girls experience indirect types

Cyberbullying 15

of bullying than boys (Kowalski et al, 2008; Ma, 2002; Olweus, 1993), and girls were more

likely to be victimized online (Dehue et al., 2008; Smith et al., 2008). The result also found

significant age differences among groups. Younger students were more likely to be in the ―more

victim than bully‖ group, and older students were more likely to be in the ―more bully than

victim‖ group. In traditional bullying, researchers have argued that younger children are

victimized more often than older children because younger victims were bullied both by older

and same-age pupils. On the other hand, older victims were bullied mainly by same-age (Smith,

Madsen, & Moody, 1999).

Moreover, it is also possible that older students are more technology savvy than younger

students and know how to protect themselves from being cybervictims: blocking unwanted

contacts or limiting friends’ network on social networking sites. Therefore, older students were

less likely to be victimized compared to younger students. In addition, older students have a

larger social circle than younger students at high school. Even though cyberbullying can happen

anonymously, this type of harassment often occurs within the circle of friends. In fact, Smith et

al. (n.d.) report that most of the cyberbullying is done by students in the same class, or in a same

year different class in their study. Similarly, Kowalski & Limber (2007) found that victims were

cyberbullied by a student at school. In other words, students are victimized by someone they

know, not by strangers in a cyberspace. Thus, older students who know more people at school

possible have higher risk of being highly involved as a bully or victim.

In addition, younger students have not acquired assertive skills yet: ―There is evidence

that some victims of bullying come from enmeshed, overprotective family backgrounds in which

skills of assertiveness are not practiced‖ (Smith, Madsen, & Moody, 1999, p. 282). Research

suggests the usefulness of assertive training for victimized students (Smith, Madsen, & Moody,

Cyberbullying 16

1999); thus, teaching younger students how to be assertive and protect themselves online would

be important.

Finally, the level of parental monitoring was not associated with cluster membership. Our

finding is consistent with Mesch (2009)’s study. This result indicates parental monitoring

strategies are not effective and protective as McQuade et al. (2009) pointed out.

Implications & Limitations of the Study

This research extends previous cyberbullying studies by classifying students and

examining sex and age differences and the level of parental monitoring based on the subgroups.

Past research indicated mixed findings on sex and age differences possibly because the

methodology ignored students’ cyberbullying subgroups (e.g., Aricak, 2009; Raskauskas &

Stoltz, 2007; Li, 2006). As discussed earlier, the classification is important for schools to

implement effective prevention and intervention strategies.

In addition, this study also found sex and age differences. Unlike traditional bullying,

physical strength and age do not seem to be a significant predictor in cyberbullying contexts;

however, girls and younger students are still more likely to be victimized than boys and/or older

students. Even though it seems easier for cybervictim to fight back, some victims may not know

how to protect themselves. Therefore, it can be concluded that cyberbullying victimization

pattern is similar to traditional bullying.

Finally, some limitations of the study also need to be addressed. First, this is a cross-

sectional correlational study; thus, causality inferences cannot be made. Second, the sample size

may be considered small, however our analyses revealed acceptable levels of statistical power (1

– β = .95 to .99) and the majority of the students are Caucasian who live in the rural area of

middle class families. Therefore, future studies can include individuals with different ethnic and

Cyberbullying 17

socioeconomic backgrounds. Finally, age differences are compared only among high school

students in the current study. If middle school or elementary school students were included,

analyses can show more distinct age differences and trend. Future studies can replicate the study

with larger and more diverse samples. In addition, future researchers also can examine other

variables such as educational achievement and psychological traits to distinguish among groups.

Despite these limitations, this research adds to a growing literature on cyberbullying among

youth.

Acknowledgement

The authors thank Dr. Willard and Dr. Smith for letting us use part of their

questionnaires for our study.

Cyberbullying 18

References

Anderson, T., & Sturm, B. (2007). Cyberbullying: From playground to computer. Young Adult

Library Services, 5, 24-27.

Aoyama, I., & Talbert, L. T. (2009, April). A cross-cultural study on cyber-bullying among high

school students in the Unites States and Japan. Paper presented at the meeting of the

American Educational Research Association (AERA), San Diego, CA.

Arıcak, O. T. (2009). Psychiatric symptomatology as a predictor of cyberbullying among

university students. Eurasian Journal of Educational Research, 34, 167-184.

Barnard-Brak, L., & To, Y. (2009). Examining parental nonresponse to stimulant treatment

questions according to ethnicity. Journal of child and adolescent psychopharmacology,

19, 301-304.

Beran, T., & Li, Q. (2005). Cyber-harassment: A study of a new method for an old behavior.

Journal of Educational Computing Research, 32, 265-277.

Bradshaw, C. P., Sawyer, A.L., & O-Brennan, L. (2007). Bullying and peer victimization at

school: Perceptual differences between students and school staff. School Psychology

Review, 36, 361-382.

Dehue, F., Bolman, C., & Völlink, T. (2008). Cyberbullying: Youngsters' experiences and

parental perception. CyberPsychology & Behavior, 11, 217-223.

Duncan, D. R. (2004). The impact of family relationships on school bullies and victims. In D. L.

Esspelage & S. M. Swearer (Eds.), Bullying in American schools. (pp. 227-244). London:

Lawrence Erlbaum Associates, Publishers.

Espelage, L. D., & Holt, K., M. (2007). Dating violence & sexual harassment across the bully-

Cyberbullying 19

victim continuum among middle and high school students. Journal of Youth and

Adolescence, 36, 799-811.

Green, S. B., & Salkind, J. W. (2004). Using SPSS for Windows and Macintosh: Analyzing and

Understanding Data (4th Edition). London: Pearson.

Hinduja, S., & Patchin, J. W. (2007). Offline consequences of online victimization: School

violence and delinquency. Journal of School Violence, 6, 89-112.

Hinduja, S., & Patchin, J. W. (2008). Cyberbullying: An exploratory analysis of factors related to

offending and victimization. Deviant Behavior, 29, 129-156.

Hinduja, S., & Patchin, J. W. (2009). Bullying beyond the schoolyard: Preventing and

responding to cyberbullying. Thousand Oaks, CA: Crown Press.

Joinson, A. (1998). Causes and implications of behavior on the Internet. In Gackenbach, J. (Ed);

Psychology and the Internet: Intrapersonal, interpersonal, and transpersonal

implications (pp. 43-60). San Diego, CA US: Academic Press.

Juvonen, J., & Gross, E. F. (2008). Extending the school grounds?—Bullying experiences in

cyberspace. Journal of School Health, 78, 496-505.

Katzer, C., Fetchenhauer, D., & Belschak, F. (2009). Cyberbullying: Who are the victims?: A

comparison of victimization in internet chatrooms and victimization in school. Journal of

Media Psychology: Theories, Methods, and Applications, 21, 25-36.

Kowalski, M. R., & Limber, P. S. (2007). Electronic bullying among middle school students.

Journal of Adolescent Health, 41, 22-30.

Kowalski, M. R., Limber, P. S., & Agatson, W. P. (2008). Cyberbullying: Bullying in the digital

Cyberbullying 20

age. Malden, MA: Blackwell Publishing.

Li, Q. (2006). Cyberbullying in schools: A research of gender differences. School Psychology

International, 27, 157-170.

Ma, X. (2002). Bullying in middle school: Individual and school characteristics of victims and

offenders. School Effectiveness and School Improvement, 13, 63 – 89.

Marini, Z., Dane, A., Bosacki, S., & Ylc-Cura, Y. (2006). Direct and indirect bully-victims:

differential psychosocial risk factors associated with adolescents involved in bullying and

victimization. Aggressive Behavior, 32, 551-569.

Mason, K. (2008). Cyberbullying: A preliminary assessment for school personnel. Psychology in

the Schools, 45, 323-348.

McKenna, K. Y. A., & Bargh, J. A. (2000). Plan 9 from cyberspace: The implications of the

internet for personality and social psychology. Personality & Social Psychology Review,

4, 57-75.

McQuade, C. S., Colt, P. J., Meyer, B. N. (2009). Cyber bullying. Protecting kid and adults from

online bullies. Westport, Connecticut: Praeger.

Mesch, S. G. (2009). Parental Mediation, Online Activities, and Cyberbullying.

CyberPsychology & Behavior, 12, 387-393.

Morita, Y., Soeda, H., Soeda, K., & Taki, M. (1999). Japan. In P, Smith., Y, Morita., J, Junger-

Tas., D, Olweus., R, Catalano., & P, Slee. (Eds.), The nature of school bullying: A cross-

national perspective (pp. 309-323). Florence, KY US: Taylor & Frances/Routledge.

Cyberbullying 21

Mouttapa, M., Valente, T., Gallaher, P., Rohrbach, L. A., & Unger, J. B. (2004). Social network

predictors of bullying and victimization. Adolescence, 39, 315-335.

Olweus, D. (1993). Bullying at school. What we know and what we can do. Malden, MA:

Blackwell Publishing.

Raskauskas, J., & Stoltz, A. D. (2007). Involvement in traditional and electronic bullying among

adolescents. Developmental Psychology, 43, 564-575.

Rosen, L. D. (2007) Me, MySpace, and I: Parenting the Net Generation. New York: Palgrave

Macmillan.

School Data Direct Retrieved on Nov 1, 2009, from http://www.schooldatadirect.org/

Shariff, S. (2008). Cyber-Bullying: Issues and solutions for the school, the classroom and the

home. New York: Routledge.

Smith, P. K., Madsen, K. C., & Moody, J. C. (1999). What cause the age decline in reports of

being bullied at school? Towards a developmental analysis of risks of being bullied.

Educational Research, 41 267-285.

Smith, P., Mahdavi, J., Carvalho, M., and Tippett, N. (n.d.). An investigation into cyberbullying,

its forms, awareness and impact, and the relationship between age and gender in

cyberbullying. Retrieved July 11, 2008, from http://www.anti-

bullyingalliance.org.uk/downloads/pdf/cyberbullyingreportfinal230106_000.pdf

Smith, P. K., Mahdavi, J., Carvalho, M., Fisher, S., Russell, S., & Tippett, N. (2008).

Cyberbullying: Its nature and impact in secondary school pupils. Journal of Child

Psychology & Psychiatry, 49, 376-385.

Tabachnick, G. B., & Fidall, S. L. (2001). Using multivariate statistics. (4th

ed.). Boston: Allyn

& Bacon.

Cyberbullying 22

Wienke Totura, M. C. et al. (2009). Bullying and victimization among boys and girls in middle

school: The Influence of perceived family and school contexts. The Journal of Early

Adolescence, 29, 571-609.

Willard, N. (2007). Cyberbullying and cyber-threats: responding to the challenge of online

social aggression, threats, and distress. Champaign, IL: Research Press.

Ybarra, M. L. (2004). Linkages between depressive symptomatology and internet harassment

among young regular internet users. CyberPsychology & Behavior, 7, 247-257.

Ybarra, M. L., & Mitchell, K. J. (2004). Online aggressor/targets, aggressors, and targets: A

comparison of associated youth characteristics. Journal of Child Psychology & Psychiatry,

45, 1308-1316.

Cyberbullying 23

Figure 1. Graphic display of offending and victimization scale scores

More

victim

than

bully

Least

Involved

Highly

Involved

More

bully

than

victim

Offending Score -0.55 -0.55 2.05 1.04

Victimization Score 1.07 -0.68 1.63 0.32

-1

-0.5

0

0.5

1

1.5

2

2.5

Offending Score

Victimization Score

Cyberbullying 24

Table 1. Descriptive statistics according to cluster membership

Offending Score Victimization Score

M SD M SD N

More victim than bully -0.55 0 1.07 0.56 13

Least Involved -0.55 0 -0.68 0 68

Highly Involved 2.05 0.83 1.63 0.34 14

More bully than victim 1.04 0.29 0.32 0.68 17

Cyberbullying 25

Figure 2. The relationship between age and cluster membership

14.92

15.61

15.76

16.39

14

14.5

15

15.5

16

16.5

17

More victim

than bully

Least

Involved

Highly

Involved

More bully

than victim

Mean age