Computers in Human Behavior 35 (2014) 12–21

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: www.elsevier.com/locate/comphumbeh

Cyber victimization in middle school and relations to social emotional

outcomes
Christina F. Brown, Michelle Kilpatrick Demaray ⇑, Stephanie M. Secord
Psychology Department, Northern Illinois University, DeKalb, IL, USA

a r t i c l e i n f o a b s t r a c t

Article history: Cyber or electronic bullying is a growing problem among school-aged children and research on cyberbul-
Available online 16 March 2014 lying is still relatively young. The primary purposes of the current study were: (a) to investigate gender
and grade level differences among cyber victims (b) to investigate the overlap between traditional vic-
Keywords: timization and cyber victimization and (c) to investigate the associations among cyber victimization
Online victimization and social emotional outcomes. Data were collected through self-report questionnaires on cyber victim-
Cyber victimization ization, traditional victimization, and social-emotional outcomes in a school-based sample of 106 middle
Cyberbully
school students. Results demonstrated that levels of cyber victimization did not differ by grade or by gen-
Gender differences
Adolescents
der, cyber victimization and traditional victimization are distinct but related constructs, and relations
between cyber victimization and social emotional outcomes varied by gender, with girls suffering more
than boys. This study also confirmed that traditional bullying continues to be significantly related to a
number of negative outcomes for all students. These findings, as well as implications and direction of
future research, are discussed.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction recruited, how or if cyber victimization is defined, and how these

Cyber or electronic bullying is a growing problem among
school-aged children. As technology becomes even more accessible
and relied upon, victims of cyberbullying will likely become even
more common. Current estimates are that as many as 20–35% of
children and adolescents report experiencing cyberbullying
(Diamanduros, Downs, & Jenkins, 2008; Kowalski & Limber,
2007). Because research on cyber victimization is still relatively
young, there is not a clear understanding of basic information such
as gender and developmental differences in experiences of cyber
victimization. From a socio-ecological framework, it is important
to consider the role of both traditional (face-to-face) victimization
and cyber victimization in understanding the relation to negative
outcomes. The primary purposes of the current study were: (a)
to investigate gender and grade level differences among cyber vic-
tims (b) to investigate the overlap between traditional victimization
and cyber victimization and (c) to investigate the associations
among cyber victimization and social emotional outcomes.
To date, studies investigating cyber victimization have had sig-
nificant methodological limitations including how participants are
⇑ Corresponding author. Address: Psychology Department, Psychology-Computer
Science Building, Rm. 400. Northern Illinois University, DeKalb, IL 60115, USA. Tel.:
+1 (815) 753 7077.
E-mail address: mkdemaray@niu.edu (M.K. Demaray).
behaviors are assessed. The development of psychometrically
sound measures of cyber victimization has been limited. Many
researchers have developed their own measure or questionnaire,
but provide limited information on the reliability and validity of
the instrument (Beran & Li, 2005; Dehue, Bolman, & Trijntje,
2008; Hinduja & Patchin, 2008; Kowalski & Limber, 2007). Several
early studies have relied on outdated national surveys; therefore, it
is difficult to generalize findings from studies. Samples also vary
across studies, as some have utilized online convenience samples
of Internet users, cross sectional databases, and others have admin-
istered measures in school settings. The methodological limitations
of cyber victimization research may, in part, explain the inconsis-
tent findings across studies. Thus, the current study also developed
a measure of cyber victimization and documented evidence of reli-
ability and validity for use of this measure to answer the study’s re-
search questions. Furthermore, because amount of time online has
been found to be related to cyber victimization (Hinduja & Patchin,
2008), the current study controlled for time spent online.
1.1. Cyber victimization
Currently there is not a standard definition of cyber victimiza-
tion, but all the definitions generally contain elements of inten-
tional and repeated harm inflicted through the use of technology.

http://dx.doi.org/10.1016/j.chb.2014.02.014
0747-5632/Ó 2014 Elsevier Ltd. All rights reserved.
 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21
For example, Hinduja and Patchin (2008) define cyberbullying as
‘‘willful and repeated harm inflicted through the medium of elec-
tronic text’’ (p. 131). Mason (2008) defined cyberbullying as ‘‘an
individual or a group willfully using information and communica-
tion involving electronic technologies to facilitate deliberate and
repeated harassment or threat to another individual or group by
sending or posting cruel text and/or graphics using technological
means’’ (p. 323). Individuals who are on the receiving end of cyber-
bullying behaviors are considered cyber victims.
1.1.1. Gender differences
To date, gender differences in cyber victimization are not con-
sistent. Kowalski and Limber (2007) found that middle school girls
were more often cyber victims (15% of girls versus 7% of boys) and
cyber bully/victims (10% of girls versus 4% of boys) than middle
school boys. Dempsey, Sulkowski, Nichols, and Storch (2009) found
similar gender differences for cyber victimization in a large middle
school sample. On the other hand, Li (2006) found that 25% of
males and 25.6% of females reported that they were cyberbullied
in a junior high school sample. Several other studies have not
found significant gender differences for cyberbullying or cyber vic-
timization (Beran & Li, 2005; Hinduja & Patchin, 2008; Williams &
Guerra, 2007).
Inconsistent gender related findings may be related to a number
of factors, including researchers examining different age groups,
students in different countries, and utilizing different reference
time periods in their assessments. Some researchers asked partic-
ipants to comment on online experiences in the past year, while
others asked about the past 30 days, the past few months, or even
lifetime experiences. It is difficult to compare results from studies
which examine differing reference periods. Furthermore, since
technology changes so quickly, it is important to utilize data that
is no more than 10 years old. Finally, samples vary across studies.
Hindjua and Patchin (2008) relied on an online convenience sam-
ple of Internet users, while other researchers have used cross sec-
tional databases and still others have administered measures in
school settings. It is clear that additional research is needed to
understand if gender differences are a consistent and important as-
pect of online bullying.
1.1.2. Developmental differences
Results from research on developmental differences have also
been inconsistent. Some studies have found no association be-
tween age and cyber victimization (Beran & Li, 2005; Varjas,
Henrich, & Meyers, 2009), while other studies suggest that cyber
victimization may be more prevalent among middle school stu-
dents. Slonje and Smith (2008) surveyed 360 adolescents ages
12–20 in Sweden and found that rates of cyber victimization were
much higher for students ages 12–15. Williams and Guerra (2007)
looked at Internet bullying in elementary, middle, and high school
and found that Internet bullying and victimization was most fre-
quent in middle school and declined in high school. Kowalski
and Limber (2007) found that among middle school students, 6th
graders were the least likely to be involved in cyberbullying and
victimization. Thus far, it appears that cyber victimization may
be more common in middle school.
1.1.3. Overlap between traditional and cyber victimization
Researchers have applied a social-ecological framework
(Bronfenbrenner, 1979) to traditional bullying, a social phenomena
that can be understood within the larger social context in which it
occurs (Espelage & Swearer, 2004). Bullying does not occur in iso-
lation and bullying behaviors are supported or not supported based
on the environments in which they occur. According to the social-
ecological perspective, a complex interplay among variables at all
levels of the model combine and interact to influence or inhibit
13
the individual from participating in bullying behaviors. Unlike tra-
ditional bullying, which is limited to locations with a bully and a
victim present, cyberbullying and cyber victimization may occur
in any environment in which an individual has access to technol-
ogy. Norms, rules and expectations differ between the two realms.
Online, social cues are less immediate, visual responses are not
part of the interaction, and children likely have easier access to
cyber victims. Anonymity and impersonation may or may not be
at play, and kids may feel more powerful and more protected inter-
acting with others from the safety of their bedroom or behind an
electronic device. Because kids communicate through technology
and then see each other in school, on the playground, or in the
community, there are likely overlaps between the two systems.
In addition, outcomes of online behaviors may be realized in the
physical world and dealt with there. On the other hand, since the
rules and norms in the two systems are different some children
may be involved in cyberbullying who would not typically be
involved in traditional bullying (Espelage & Swearer, 2004).
Little is known about how many school-aged children experi-
ence bullying both online and at school. For some youth, online
victimization may be part of a larger spectrum of victimization
experiences (Mitchell, Ybarra, & Finkelhor, 2007). Mitchell et al.
(2007) found that almost three quarters of children ages 10–17
whom reported being victimized online at least once in the past
year also reported being victimized at least once in the past year
in the traditional sense. Erdur-Baker’s (2010) study of adolescents
ages 14–18 in Turkey found that 32% of the students were victims
of both cyber and traditional bullying and 26% of the students bul-
lied others both electronically and traditionally. Experience with
traditional bullying has been found to be a strong predictor for
cyberbullying and cyber victimization (Li, 2007). Hinduja and
Patchin (2008) found that those who admitted to bullying others
and those who were bullied by others were more than 2.5 times
more likely to cyberbully others or be victimized online.
1.1.4. Social emotional outcomes associated with cyber victimization
Social emotional outcomes related to cyber victimization have
been found to be comparable to those related to victimization from
traditional bullying (Smith et al., 2008). Initial research has shown
that being a victim of cyberbullying can negatively impact physi-
cal, social, and cognitive functioning, development, and well-being.
It can also cause psychological, emotional and academic problems
(Beran & Li, 2005; Hinduja & Patchin, 2008; Mitchell et al., 2007;
Ybarra, Espelage, & Mitchell, 2007). A number of studies have
found an association between cyber victims, cyber bullies and
depression and suicidal ideation (Hinduja & Patchin, 2010) and
externalizing behaviors (Ybarra, Diener-West, & Leaf, 2007; Ybarra,
Espelage, et al., 2007). Ybarra and Mitchell (2004) found that males
who reported depressive-like symptoms were more than eight
times as likely to have experienced cyber victimization. Further-
more, victims report higher levels of anger and frustration, espe-
cially towards the aggressor (Patchin & Hinduja, 2006). Research
also shows that both cyberbullies and victims are at higher risk
for school problems (e.g., suspension, truancy, cheating) and other
deviant behaviors (e.g., drug and alcohol abuse) (Patchin &
Hinduja, 2012). However, many of these studies did not use
validated measures of psychosocial functioning in their analysis.
Some research has examined the simultaneous effects of tradi-
tional and cyberbullying and victimization to determine if unique
associations exist with various outcomes. Dempsey et al. (2009)
surveyed 1684 middle school students to assess cyber victimiza-
tion in the past 30 days, anxiety related to social situations in
the last 30 days, and depressive symptoms in the past seven days
using validated measures of depression and social anxiety. Cyber
victimization was found to be weakly related to social anxiety
but not to depression after controlling for traditional victimization.
14 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21
Fredstrom, Adams, and Gilman (2011) found cyber victimization to
be related to lower levels of self-esteem and higher levels of social
stress, anxiety, depressive symptoms, and locus of control after
controlling for traditional victimization. Recently, Bonanno and
Hymel (2013) found cyber victimization to be related to depression
and suicidal ideation above and beyond the contribution of tradi-
tional victimization in a sample of 399 students in 8th to 10th grade.
Cyberbullying may be particularly harmful to victims for sev-
eral reasons. First, it is very difficult to escape from cyberbullying,
as it can occur across settings. Second, the aggressors may be more
volatile due to their ability to remain anonymous. Without having
direct contact with the victim, aggressors may be less likely to feel
empathy or remorse. Third, there is a greater breadth of audience.
The use of technology and the Internet allows the aggressor to
reach large audiences in a short time frame. Finally, there is less
adult supervision on the Internet, reducing the chances of inter-
vention (Dempsey et al., 2009; Smith & Slonje, 2010).
1.2. Current investigation
Research on cyberbullying is still relatively young; therefore,
there is not yet a clear understanding of basic information such
as gender and developmental differences in experiences of cyber
victimization. To date, studies investigating cyberbullying have
yielded inconsistent results regarding gender and developmental
differences regarding the social emotional outcomes of victims of
cyberbullying. Furthermore, studies have had significant methodo-
logical limitations including how participants are recruited, how or
if cyberbullying and victimization is defined, and how these behav-
iors are assessed. Thus, the current study also developed a measure
of cyber victimization and documented evidence of reliability and
validity for use of this measure to answer the study’s research
questions.
Although research on cyber victimization is rapidly growing,
there remains significant gaps and limitations in the current liter-
ature. One major limitation that will be addressed in the current
study includes methodological issues associated with the measure-
ment of cyber victimization. With a psychometrically sound mea-
sure of cyber victimization, more accurate data in the current study
will add to the literature by investigating (a) gender and develop-
mental differences in social emotional outcomes of victims of
cyberbullying and (b) the overlap between cyber and traditional
victimization.
1.2.1. Research questions and predictions
The following research questions guided the current investiga-
tion: (1) Are there gender or grade level differences in prevalence
rates of cyber victimization? While the literature to date is not nec-
essarily conclusive, it does appear that more females, especially in
middle school, are cyber victimized; therefore, it was predicted
that females would have higher cyber victimization scores than
males (Dehue et al., 2008; Dempsey et al., 2009; Kowalski &
Limber, 2007). Furthermore, based on results found in the litera-
ture, it was predicted that 6th graders would report the lowest
overall levels of cyber victimization as compared to 7th and 8th
graders (Kowalski & Limber, 2007; Slonje & Smith, 2008; Smith
et al., 2008; Williams & Guerra, 2007). (2) What is the overlap be-
tween victims of traditional bullying and cyber victims? It was
hypothesized that the majority of students classified as cyber vic-
tims would also be involved in some aspect of traditional victim-
ization. Traditional bullying has been found to be a strong
predictor for cyberbullying and cyber victimization (Hinduja &
Patchin, 2008; Li, 2007; Mitchell et al., 2007; Smith et al., 2008;
Ybarra, Diener-West, et al., 2007; Ybarra, Espelage, et al., 2007).
(3) What social emotional outcomes, above and beyond traditional
victimization, while controlling for time online, are associated with
cyber victimization? It was hypothesized that cyber victimization
would be significantly and positively related to the following com-
posite scales of the BASC-2: Emotional Symptoms Index, Internal-
izing Problems, and School Problems and significantly and
negatively related to Personal Adjustment. It was also predicted
that cyber victimization would be significantly and positively re-
lated to the following BASC-2 subscales: Anxiety, Social Stress,
and Depression, and significantly and negatively related to Inter-
personal Relations and Self-Esteem (Hinduja & Patchin, 2008;
Mitchell et al., 2007; Shariff, 2005; Ybarra & Mitchell, 2004).
2. Method
2.1. Participants
Participants for this study (N = 106) were recruited from a large
middle school in northern Illinois. There were 1099 students en-
rolled in the school at the time so approximately 10% of the stu-
dent population participated in the study. Specifically, there were
42 (40%) 6th graders, 49 (46%) 7th graders, and 15 (14%) 8th graders
in this study. Fifty-nine of the participants were females (56%) and
47 (44%) were males. The sample was predominantly White (73%),
followed by Latino (9%), Asian (5%), and African American (2%), and
Other/Missing (13%). The socioeconomic status (SES) make-up of
the school was 7.2% low income.
2.2. Materials
All students were asked to complete a measure of cyber victim-
ization developed for the current study, the Cyber Victimization
Survey (CVS). Participants also completed the Revised Olweus
Bully/Victim Questionnaire (OBVQ; Olweus, 1996) and the Behav-
ior Assessment System for Children (BASC-2 SRP; Reynolds &
Kamphaus, 2004). Participants were also asked to respond to two
items about the time they spend online on computers and the time
they spend on their cell phone. The response options were: 1 = less
than an hour, 2 = 1–2 h, 3 = 2–4 h, 4 = 4–6 h, 5 = more than 6 h. These
items were summed to get a Total Time Online score. This variable
allowed the researcher to control for time spent online.
2.2.1. Cyber Victimization Survey
The Cyber Victimization Survey (CVS) is a 15-item measure that
was developed for use in the current study to assess victimization
from online bullying. Participants were asked how often certain
online behaviors/incidents happened to them in the last 2–3
months. All items were answered on a 5-point scale, based on
the response options on the Revised Olweus Bully/Victim Ques-
tionnaire, (1 = it hasn’t happened at all in the past couple of months,
2 = only 1 or 2 times in the past couple of months, 3 = 2 or 3 times a
month, 4 = about once a week, and 5 = several times a week).
Initial items for the Cyber Victimization Survey were developed
in a number of ways. A review of current measures was conducted
and several items from Kowalski and Limber’s (2007) and Hinduja
and Patchin’s (2009) cyberbullying measures were used. New
items were also developed based on a review of the literature. Fol-
lowing item development, several school psychology graduate stu-
dents and psychology professors provided feedback, which was
used to improve, revise, and discard specific items. The measure
was pilot tested in a sample (N = 137) of middle and high school
students. Based on pilot study results, several items were dropped,
the wording of several items was improved, and four items were
added to create the version utilized in the current study. Thus,
the Cyber Victimization Survey utilized in the current study ini-
tially had 20 items.
 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21
The psychometric properties of the Cyber Victimization Survey
were confirmed in the current study sample. Inter-item correla-
tions were analyzed. Correlations ranged from .080 to .862. Items
that correlated poorly with other items were deleted from the
measure. An item was considered to correlate poorly with another
item if the correlation was less than .30 (Tabachnick & Fidell,
1996). Any items that correlated poorly with 11 or more other
items were dropped from the measure. Based on this analysis, 4
items were deleted from the scale. All remaining items were
summed and item-total correlations were conducted. Correlations
ranged from .519 to .799. All but one item total correlation was sig-
nificant at the p < .01 level. The remaining 16 items were factor
analyzed with Principal Axis Factoring. An oblique (Promax) rota-
tion was used because it was hypothesized that any factors would
be correlated with one another. Tabachnick and Fidell (1996) char-
acterize a sample size of 250 as ‘‘fair’’. Because this data set was
relatively small, Bartlett’s test of sphericity and Kaiser–Meyer-Olk-
in (KMO) measure of sampling adequacy were examined. As a rule,
Tabachnick and Fidell indicate that a KMO value above .6 is re-
quired for factorability of the items. The use of Bartlett’s test is rec-
ommended when there are fewer than five cases per variable. For
this data set, the (KMO) measure of .863 indicated a high sampling
adequacy for the factor analysis. Bartlett’s test of sphericity, which
tests whether the correlation matrix is an identity matrix, is signif-
icant (p < .001), indicating that the factor model is appropriate.
Initially, the factor analysis was run without forcing any factors.
Results indicated a four factor solution which explained 75% of the
variance. An examination of the scree plot indicated there was
likely one strong factor. An examination of the pattern matrix also
pointed to a single factor solution. The four factor solution did not
make theoretical sense so a second factor analysis was run forcing
a single factor. Results indicated that a single factor explained 50%
of the variance. Factor loadings ranged from .473 to .804. Item 1
had the lowest factor loading (.473) and was a global question, ask-
ing participants if they had ever been cyberbullied. All remaining
items were more specific, therefore, item 1 was dropped and the
factor analysis was rerun. This model accounted for 52% of the var-
iance. Loadings on the single factor ranged from .622 to .806. The
final scale consisted of 15 items with an internal consistency alpha
equal to .924.
To provide evidence of validity, correlations were run between
the Cyber Victimization Survey and Hinduja and Patchin’s (2007)
Cyberbullying and Online Aggression Survey Instrument and
Kowalski and Limber’s (2007) Cyberbullying Measure. The correla-
tion between the Cyber Victimization Survey and Hinduja and Pat-
chin’s Cyberbullying and Online Aggression Survey Total Score was
.589 and the correlation between the CVS and Kowalski and Lim-
ber’s Cyberbullying Total Score was .523. Both were significant at
the p < .01 level.
2.2.2. The Revised Olweus Bully/Victim Questionnaire
The Revised Olweus Bully/Victim Questionnaire (OBVQ) is a 40-
item self-report questionnaire that measures bullying and victim-
ization. The OBVQ is one of the most commonly used measures of
bullying. Internal consistency, test–retest reliability, and construct
validity have been found to be satisfactory (Kyriakides, Kaloyirou,
& Lindsay, 2006; Olweus, 1997). The measure begins with a de-
tailed definition of bullying and students are asked to respond to
items based on a 2–3 month reference period. Response options
are on a 5-point Likert type scale. The measure is comprised of
two global questions. The first asks ‘‘How often have you been bul-
lied at school in the past couple of months?’’ and the second asks
‘‘How often have you taken part in bullying another student(s) at
school in the past couple of months?’’ These items are each fol-
lowed by more specific items related to victimization and bullying
15
(Solberg & Olweus, 2003). Only the total victimization score, the
sum of items 6 through 13, was used in the current study.
2.2.3. Behavior assessment system for children, second edition, self-
report of personality
The BASC-2 SRP is a personality inventory and provides 5 com-
posite scores: School Problems, Internalizing Problems, Inatten-
tion/Hyperactivity, Personal Adjustment and the Emotional
Systems Index, an overall composite score. Two forms, the child
form (ages 8 through 11) and the adolescent form (ages 12 through
21) were used in the current study. Respondents rate questions in
one of two ways. Some items require a True or False response while
others are answered based on a 4-point Likert type scale, ranging
from Never to Almost Always. The child form consists of 139 items
and the adolescent form contains 176 items. The following com-
posite scores: Emotional Symptoms Index, Personal Adjustment,
School Problems, and Internalizing Problems, and subscale scores:
Anxiety, Depression, Interpersonal Relations, Social Stress, and
Self-Esteem were used in the current investigation. There is strong
evidence for psychometric properties of the BASC-2 SRP.
2.3. Procedure
Students from a large Midwest middle school were recruited for
participation in the study. Informed consent letters for voluntary
participation, including a description of the study were sent home
with all students. Only students who obtained parental consent
were allowed to participate in the study. Data were collected by
the primary researcher, research assistants, and teachers during
the students’ study hall. Participating students were given a brief
description of the study which included assurances of anonymity
and confidentiality unless students indicated they were going to
harm themselves or others, and that participation was voluntary.
After the study description was given, students were asked to sign
an assent form, which indicated their willingness to participate in
the study.
2.4. Statistical analyses
In order to answer the first research question (Are there gender
or grade level differences in prevalence rates of cyber victimization?), a
Gender (male, female) by Grade Level (6th, 7th, 8th) ANOVA was
conducted to determine if there were gender and grade level differ-
ences in cyber victimization. In this analysis the independent vari-
ables were Gender and Grade Level and the dependent variable
was cyber victimization.
To answer the second research question (What is the overlap be-
tween victims of traditional bullying and cyber victims?), a correla-
tion between the CVS score (i.e., cyber victimization) and the
OBVQ victimization score (i.e., traditional victimization) was uti-
lized. Additionally, descriptive analyses comparing the top 20% of
cyber victims, as determined by the Cyber Victimization Survey,
and the top 20% of traditional victims, as determined by the Re-
vised Olweus Bully/Victim Questionnaire, were used.
In order to answer the third research question (What social emo-
tional outcomes, above and beyond traditional victimization, while
controlling for time online, are associated with cyber victimization?),
a number of regression analyses were conducted separately for
boys and girls using cyber victimization (i.e., CVS score) as the
independent variable. In addition, the OBVQ Victimization (i.e., tra-
ditional victimization) score and Time Online score were entered
as independent variables to hold traditional victimization and
amount of time spent online constant in all analyses. The following
BASC-2 composite scores, Emotional Symptoms Index, Personal
Adjustment, School Problems, and Internalizing Problems, were
entered as the dependent variables.
16 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21

3. Results 3.2. Main analyses

3.1. Preliminary analyses 3.2.1. Research question one

In order to answer the first research question (Are there gender
Prior to conducting the main analyses that answer the research or grade level differences in prevalence rates of cyber victimization?), a
questions of interest, several preliminary analyses were conducted Gender (male, female) by Grade Level (6th, 7th, 8th) ANOVA was
to examine gender and grade level differences on each of the vari- conducted to determine if there were gender and grade level differ-
ables. These analyses were conducted to better understand any ences in cyber victimization. Results should be interpreted with
gender and grade level differences in all of the study variables. A caution due to a small n in some cells. The dependent variable
series of 2 (Gender) by 3 (Grade Level) ANOVAs were conducted was the Cyber Victimization Survey (CVS) total score. Neither the
on OBVQ Victimization score, Time Online, and the BASC-2 Com- gender main effect, F(1, 105) = .195, p = .660, the grade main effect,
posites scores (i.e., Emotional Symptoms Index, Personal Adjust- F(2, 105) = 1.728, p = .183, nor the interaction, F(2, 105) = 1.082,
ment, School Problems, and Internalizing Problems). Results p = .343, were significant. Means and standard deviations on the
revealed significant differences in traditional victimization by CVS score by gender and grade level are located in Table 2.
Grade Level, F(2, 105) = 3.888, p < .05, with 8th graders (M = 15.49)
scoring significantly higher than 6th graders (M = 13.02) on tradi- 3.2.2. Research question two
tional victimization. A significant difference in School Problems A correlation between the CVS score (i.e., cyber victimization)
by Gender was also found, F(1, 105) = 8.070, p < .01, with boys and the OBVQ victimization score (i.e., traditional victimization)
(M = 46.40) reporting greater levels of school problems than girls was used to answer the second research question (What is the over-
(M = 40.68). No other significant differences were found. See lap between victims of traditional bullying and cyber victims?). The
Table 1 for the means and standard deviations of main study correlation between the two scales (.174) was not significant with
variables by total sample and gender. p = .074. Additionally, the top 20% of cyber victims, as determined
by the Cyber Victimization Survey (a score of 22 or above), and the
top 20% of traditional victims, as determined by the Revised Olw-
eus Bully/Victim Questionnaire (a score of 13 or above), were com-
pared. Twelve students met the criteria on the CVS (i.e., cyber
victims) and 14 met criteria on the OBVQ (i.e., traditional victims).
Table 1
Four students scored in the top 20% on both measures. See Table 3
Means and standard deviations of primary variables by total sample and by gender.
for demographic information for each group of students, cyber vic-
Variable N Mean (SD) tim only, traditional victim only, and cyber victim and traditional
Total sample Time online 106 3.189 (1.91) victim.
CVS total score 106 16.30 (6.44)
OBVQ total score 106 9.65 (2.23)
BASC-2 composite scores 3.2.3. Research question three
Emotional symptoms index 106 42.18 (8.71) In order to answer the third research question (What social emo-
Personal adjustment 106 56.85 (8.85) tional outcomes, above and beyond traditional victimization, while
School problems 106 43.22 (7.96) controlling for time online, are associated with cyber victimization?),
Internalizing problems 106 43.13 (8.78)
BASC-2 subscale scores
a number of regression analyses were conducted separately for
Anxiety 106 44.44 (9.62) boys and girls using cyber victimization (i.e., CVS score) as the
Depression 106 43.91 (7.24) independent variable. In addition, the OBVQ Victimization score
Interpersonal relations 106 55.06 (8.19) and Time Online score were entered as independent variables to
Social stress 106 43.33 (9.61)
hold traditional victimization and amount of time spent online
Self-esteem 106 55.68 (6.75)
constant in all analyses. The following BASC-2 composite scores,
Female Time online 59 3.42 (1.75)
Emotional Symptoms Index, Personal Adjustment, School Prob-
CVS total score 59 15.53 (3.11)
OBVQ total score 59 9.51 (2.11) lems, and Internalizing Problems, were entered as the dependent
BASC-2 composite scores variables.
Emotional symptoms index 59 41.25 (8.55) Results of the BASC-2 composite scores for girls revealed that all
Personal adjustment 59 58.19 (7.79) regressions, with the exception of School Problems, were signifi-
School problems 59 40.68 (6.25)
Internalizing problems 59 41.93 (9.28)
cant. Specifically, the overall regression using the Emotional Symp-
BASC-2 subscales toms Index Score was significant, F(3, 58) = 8.185, p < .001. Time
Anxiety 59 44.27 (10.68)
Depression 59 43.14 (6.74)
Interpersonal relations 59 57.02 (6.17) Table 2
Social stress 59 41.25 (9.01) Means and Standard deviations by gender, grade level, and total sample on the CVS.
Self-esteem 59 55.40 (7.47)
Gender Grade N CVS Score Mean (SD)
Male Time online 47 2.89 (2.08)
CVS total score 47 17.26 (8.99) Female 6th grade 28 15.96 (2.32)
OBVQ total score 47 9.83 (2.37) 7th grade 22 16.92 (3.72)
BASC-2 composite scores 8th grade 9 17.56 (3.81)
Emotional symptoms index 47 43.34 (8.85) Total 59 16.56 (3.14)
Personal adjustment 47 55.17 (9.85) Male 6th grade 14 15.93 (1.98)
School problems 47 46.40 (8.75) 7th grade 27 20.37 (12.32)
Internalizing problems 47 44.64 (7.94) 8th grade 6 16.17 (1.60)
BASC-2 subscales Total 47 18.51 (9.59)
Anxiety 47 44.66 (8.20)
Total 6th grade 42 15.95 (2.19)
Depression 47 44.87 (7.78)
7th grade 49 18.82 (9.56)
Interpersonal relations 45 52.53 (9.73)
8th grade 15 17.00 (3.12)
Social stress 47 45.94 (9.79)
Self-esteem 45 56.04 (5.74) CVS total score 106 17.43 (6.83)
 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21 17

Table 3 accounted for a significant amount of variance in the Personal

Demographics on students who scored high on cyber and traditional victimization. Adjustment Problems for girls, R2 = .187, p < .01; however, no un-
Demographic Top CVS Top OBVQ Top scorers on CVS and ique predictors emerged. Results of analyses using the Internaliz-
variables scorers scorers OBVQ ing Problems Composite score were similarly significant,
Gender F(3, 58) = 10.117, p < .001. Time Online, CVS, and OBVQ scores ac-
Male 7 8 2 counted for a significant amount of variance in the Internalizing
Female 5 6 2 Problems Composite for girls R2 = .356, p < .001. Only the OBVQ
Grade score was a unique, significant predictor (b = .434, p < .01). See
6th 2 3 1 Table 4 for specific regression analyses for girls.
7th 9 7 2
8th 1 4 1
For boys, results of the regression analysis for the BASC-2 com-
posite scores revealed that all regressions were significant. The
Ethnicity
African American 1 1 1
overall regression using the Emotional Symptoms Index Score
Asian American 1 1 0 was significant, F(3, 46) = 6.665, p < . 01. Time Online, the CVS (cy-
White 8 9 3 ber victimization), and OBVQ (traditional victimization) scores ac-
Hispanic 0 1 0 counted for a significant amount of variance in the Emotional
American
Symptoms Index score, R2 = .317, p < .01. The OBVQ score was the
Other 1 0 0
Missing 1 2 0 only significant unique predictor (b = .542, p < .001). Similarly,
the overall regression using the Personal Adjustment Problems
Total 12 14 4
score was significant, F(3, 46) = 5.410, p < . 01. Time Online, the
CVS, and OBVQ scores accounted for a significant amount of vari-
ance, R2 = .274, p < .01. The OBVQ score was the only significant un-
Online, the CVS (cyber victimization), and OBVQ (traditional vic- ique predictor (b = .489, p < .01). Regressions using School
timization) scores accounted for a significant amount of variance Problems, F(3, 46) = 3.778, p < . 05, and Internalizing Problems,
in the Emotional Symptoms Index Score, R2 = .309, p < .001. Further F(3, 46) = 8.496, p < .001, were also significant overall. Time Online,
analysis determined that both the CVS score (b = .282, p < .05) and the CVS, and OBVQ scores accounted for a significant amount of the
the OBVQ scores (b = .352, p < .05) were significant individual pre- variance in the School Problems composite, R2 = .209, p < .05, and
dictors. Results of the regression analyses utilizing the Personal the Internalizing Problems Composite, R2 = .372, p < .001. Time On-
Adjustment Problems Composite score were also significant, line was a significant unique predictor for these scores, b = .439,
F(3, 58) = 4.208, p < .01. Time Online, the CVS and OBVQ scores p < .01, b = .549, p < .001, respectively. The OBVQ score was also a

Table 4
Regression analysis for the CVS total score, the OBVQ total score, and BASC-2 composite and subscale scores for girls.

Dependent variable Independent variable B SE B B R2

BASC-2 composite
Emotional symptoms index Time online .046 .551 .009 .309***
CVS* .768 .361 .282
OBVQ* 1.426 .539 .352
Personal adjustment Time online .437 .544 .098 .187**
CVS .540 .356 .218
OBVQ .917 .532 .249
School problems Time online .818 .455 .229 .119
CVS .326 .298 .164
OBVQ .291 .445 .098
Internalizing problems Time online .473 .577 .089 .356***
CVS .720 .378 .244
OBVQ** 1.907 .564 .434
BASC-2 subscales
Anxiety Time online 1.057 .659 .174 .366***
CVS .853 .431 .251
OBVQ** 2.166 .644 .429
Depression Time online .341 .440 .089 .289***
CVS* .638 .288 .297
OBVQ* 1.024 .430 .321
Interpersonal relations Time online .074 .422 .021 .223**
CVS .275 .276 .140
OBVQ** 1.111 .413 .380
Social stress Time online .251 .574 .049 .323***
CVS* .988 .376 .345
OBVQ* 1.318 .562 .309
Self-esteem Time online .203 .534 .046 .209**
CVS .602 .349 .244
OBVQ .986 .522 .269

OBVQ = Revised Olweus Bully/Victim questionnaire traditional victimization total score.

CVS = cyber victimization total score.
*
p < .05.
**
p < .01.
***
p < .001.
18 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21

significant unique predictor for Internalizing Problems (b = .549,
p < .001). See Table 5 for specific regression analyses for boys.
Results of the regression analyses using the BASC-2 subscales
for girls revealed that all regressions were significant and Time On-
line, the CVS (cyber victimization), and OBVQ (traditional victim-
ization) scores accounted for a significant amount of variance in
the BASC-2 scores in all assessed areas. Specifically, results re-
vealed that overall regression analyses were significant for Anxiety,
F(3, 58) = 10.566, p < .001, Depression, F(3, 58) = 7.464, p < .001,
Interpersonal Relations, F(3, 58) = 5.274, p<.01, Self-Esteem,
F(3, 58) = 4.854, p < .01, and Social Stress, F(3, 58) = 8.764, p < .001.
Together, Time Online, the CVS, and the OBVQ scores accounted
for a significant amount of variance in Anxiety, R2 = .366, p < .001,
Depression, R2 = .289, p < .001, Interpersonal Relations, R2 = .223,
p < .01, Self-Esteem, R2 = .209, p < .01, and Social Stress, R2 = .323,
p < .001. Furthermore, results revealed the OBVQ score to be a sig-
nificant individual predictor of Anxiety, b = .429, p < .01, Depres-
sion, b = .321, p < .05, Interpersonal Relations, b = .380, p < .01,
and Social Stress, b = .309, p < .05. The CVS score was found to be
a significant individual predictor of Depression, b = .297, p < .05,
and Social Stress, b = .345, p < .05. There were no unique predictors
for Self-Esteem. See Table 4 for specific regression analyses for
girls.
Results of the regression analyses using the BASC-2 subscales
for boys revealed that all regressions were significant and Time On-
line, the CVS (cyber victimization), and OBVQ (traditional victim-
ization) scores accounted for a significant amount of variance in
the BASC-2 scores in all assessed areas. Specifically, results re-
vealed that overall regression analyses were significant for Anxiety,
F(3, 46) = 4.674, p < .01, Depression, F(3, 46) = 5.096, p < .01, Inter-
personal Relations, F(3, 46) = 8.372, p < .001, Self-Esteem,
F(3, 46) = 3.374, p < .05, and Social Stress, F(3, 46) = 7.474, p < .001.
Together, Time Online, the CVS, and the OBVQ scores accounted
for a significant amount of variance in Anxiety, R2 = .246, p < .01,
Depression, R2 = .262, p < .01, Interpersonal Relations, R2 = .369,
p < .001, Self-Esteem, R 2 = .191, p < .05, and Social Stress,
R2 = .343, p < .001. Furthermore, results revealed the OBVQ (tradi-
tional victimization) score to be a significant individual predictor
in all regression analyses, Anxiety, b = .496, p < .01, Depression,
b = .505, p < .001, Interpersonal Relations, b = .605, p < .001, Self-
Esteem, b = .414, p < .01, and Social Stress, b = .562, p < .001. Nei-
ther Time Online nor the CVS (cyber victimization) total score was
found to be a unique predictor in any subscale analyses for boys.
See Table 5 for specific regression analyses for boys.
4. Discussion
4.1. Gender and grade level differences among cyber victims
It was predicted that females would report higher levels of cy-
ber victimization compared to males and that cyber victimization
would be more common in 7th and 8th grades as opposed to 6th
grade. Neither of these predictions was supported. No significant
gender or grade level differences were found in the levels of cyber
victimization, indicating that levels of cyber victimization do not
differ by grade or gender. Caution, however, should be used inter-
preting the grade level results due to the small n in some cells.

Table 5
Regression analysis for the CVS total score, the OBVQ total score, and BASC-2 composite and subscale scores for boys.

Dependent variable Independent variable B SE B B R2

BASC-2 composites
Emotional symptoms index Time online .818 .568 .192 .309**
CVS .131 .123 .142
OBVQ*** 2.020 .471 .542
Personal adjustment Time online .509 .652 .107 .274**
CVS .253 .141 .237
OBVQ** 2.031 .542 .489
School problems Time online** 1.851 .604 .439 .209*
CVS .112 .131 .123
OBVQ .777 .502 .211
Internalizing problems Time online* 1.147 .489 .300 .372***
CVS .030 .106 .037
OBVQ*** 1.840 .406 .549
BASC-2 subscales
Anxiety Time online .213 .553 .054 .246**
CVS .019 .120 .022
OBVQ** 1.717 .459 .496
Depression Time online .468 .519 .125 .262**
CVS .080 .113 .098
OBVQ*** 1.655 .431 .505
Interpersonal relations Time online .168 .595 .036 .369***
CVS .046 .129 .046
OBVQ*** 2.460 .494 .605
Social stress Time online .864 .617 .183 .343***
CVS .004 .134 .004
OBVQ*** 2.320 .512 .562
Self-esteem Time online .244 .404 .088 .191**
CVS 1.08 .088 .179
OBVQ** 1.008 .335 .414

OBVQ = Revised Olweus Bully/Victim Questionnaire traditional victimization total score.

CVS = cyber victimization total score.
*
p < .05.
**
p < .01.
***
p < .001.
 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21
Research on gender differences has been mixed. Li (2006) and
Slonje and Smith’s (2008) work found that differences between
males and females who reported being cyber victims were not sta-
tistically significant. Raskauskas (2010) examined bullying via text
only and found no significant victimization differences between
girls and boys. Several other studies have also failed to find signif-
icant gender differences for cyber victimization (Hinduja & Patchin,
2008; Williams & Guerra, 2007). On the other hand, Kowalski and
Limber (2007) found that middle school girls were more often cy-
ber victims and cyber bully/victims than middle school boys.
Dempsey et al. (2009) found that among a large middle school
sample, significantly more females reported being a cyber victim
in the past 30 days than males, but this relationship was weak.
Contrastingly, Erdur-Baker’s (2010) study found that boys were cy-
ber victims more often than girls. It remains unclear if gender dif-
ferences are an important aspect of online bullying. Further
research is needed to better understand how gender may or may
not influence online behavior.
Regarding grade level differences, similar to this study, a Cana-
dian study of 7th, 8th, and 9th graders did not find prevalence nor
frequency differences by grade for cyber victimization (Beran &
Li, 2005). Kowalski and Limber (2007) found that among middle
school students, 6th graders were the least likely to be involved
in cyberbullying. Williams and Guerra (2007) looked at Internet
bullying in elementary, middle and high school and found that
Internet bullying was most frequent in middle school and declined
in high school. It appears that cyber victimization may peak in
middle school, but further research comparing a wider range of
grades (e.g., middle versus high school) is needed. As technology
becomes more accessible and younger children have access to cell
phones, tablets, and computers it will be interesting to see if and
how late-aged elementary school children are impacted by online
aggression.
The lack of consensus on a definition of cyberbullying likely
influences results as well. Many researchers failed to look at com-
prehensive methods or modes of cyberbullying and some limited
responses to specific technological methods (e.g., texting only) or
modes (e.g., cell phones only). Different time reference points, var-
ied sample characteristics, international data, and inconsistent sur-
vey methodology make generalizing prevalence rates difficult.
Some studies asked respondents about online experiences in the
last 7 or 30 days (Dempsey et al., 2009; Hinduja & Patchin,
2007), or the past 5–6 months (Raskauskas, 2010; Sourander
et al., 2010), while others asked about experiences over their life-
time (Li, 2010).
4.2. The overlap between traditional and cyber victims
Regarding the overlap between traditional and cyber victimiza-
tion, a moderate to strong positive correlation was predicted. This
was not supported, indicating that there was not much overlap be-
tween these two types of victimization and they are distinct con-
structs. Additional support for this comes from the comparisons
among students who scored highest on the OBVQ (i.e., traditional
victimization), students who scored highest on the CVS (i.e., cyber
victimization), and students who scored high on both measures.
When we looked at students who scored in the top 20% on cyber
and traditional victimization, only four students, not quite 4%,
met this criteria. This study looked only at victimization and did
not assess online bullying. If both traditional and online bullying
had been considered as well, greater overlap may have been
discovered.
Traditional bullying has been found to be a strong predictor for
cyberbullying and cyber victimization (Hinduja & Patchin, 2008; Li,
2007). There appears to be a certain amount of overlap between
traditional bullying and cyber victimization, but not all students
19
involved in traditional bullying are involved in cyber bullying.
Ybarra, Espelage, et al. (2007) and Ybarra, Diener-West, et al.
(2007) found that the majority of cyber victims were not tradi-
tional victims; however, the cyber victims who were victimized
most often were also traditional victims. A number of studies (Li,
2007; Mitchell et al., 2007; Raskauskas & Stoltz, 2007; Smith
et al., 2008) found that the majority of cyber victims were involved
in traditional bullying behavior as bullies, victims, or bully-victims.
Erdur-Baker (2010) found that almost one-third of respondents
were both traditional and cyber victims. Future research that looks
at traditional and online bullying as well as victimization and fur-
ther explores the overlap between these distinct, yet related con-
structs is needed. Although this study did not find as great a
crossover between traditional and cyber victims, it is clear that stu-
dents who are victims of both traditional and online bullying are at
greater risk than students who are victimized by only one type of
bullying. Further research should be conducted to better under-
stand this population.
4.3. Social emotional outcomes associated with cyber victimization
Time online, traditional victimization, and online victimization
were significantly related to all outcomes for girls, with the excep-
tion of School Problems, and all outcomes for boys. For girls, tradi-
tional victimization was a significant unique predictor for all
significant analyses with the exception of Self Esteem; cyber vic-
timization was a significant predictor of the Emotional Symptoms
Index, Depression, and Social Stress. Time Online was not a signif-
icant predictor for any outcomes.
Similar to the girls, traditional victimization emerged as a un-
ique predictor for most outcomes for boys. Specifically, traditional
victimization predicted the Emotional Symptoms Index, Personal
Adjustment, Internalizing Problems, Anxiety, Depression, Interper-
sonal Relations, Social Stress, and Self Esteem. Unlike results for
girls, cyber victimization was not a unique predictor for any of
the boys’ outcomes while Time Online was a unique, significant
predictor for the Internalizing Problems Composite Score and the
School Problems Composite Score for boys. The Internalizing Prob-
lems Composite score consists of items that measure atypicality,
locus of control, social stress, depression, and sense of inadequacy
while the School Problems Composite consists of items that mea-
sure attention and learning problems. A score indicates possible
academic difficulties, problems with motivation, attention, and
learning (Reynolds & Kamphaus, 2004). Male students who spend
a lot of time online may do so because they have poor social rela-
tionships with peers. These students may have poor social skills
and therefore may choose to spend time alone with technology
or they may opt to seek out virtual relationships.
This is one of the first studies to examine social emotional out-
comes by gender and an interesting pattern emerged. When the
sample was analyzed by gender, cyber victimization did not pre-
dict any social emotional outcomes for males; however, cyber vic-
timization was a significant, unique predictor of Depression, Social
Stress, and the Emotional Symptoms Index score for girls. These re-
sults suggest that girls may be at greater risk of poor emotional and
social outcomes due to what happens online. Some research has
found that girls may experience higher rates of internalizing
distress as a result of bullying (Klomek, Marrocco, Kleinman,
Schonfeld, & Gould, 2007; Turner, Exum, Brame, & Holt, 2013).
Girls may be more susceptible to cyber victimization because
cyberbullying tends to be more relationally aggressive in nature
(Crick & Grotpeter, 1995). Fredstrom et al. (2011) found that, after
controlling for school-based victimization, cyber victimization re-
mained a significant predictor of lower rates of self-esteem, higher
rates of social stress, anxiousness, depressive symptoms, and locus
of control. Results of the current study found that girls experienced
20 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21
similar outcomes. It is possible that girls are more sensitive in their
interpretation of and response to online bullying behaviors.
Agatston, Kowalski, and Limber (2007) conducted focus groups
with middle and high school students. In this study, a majority of
female students reported that cyberbullying was a problem at their
school. Male students were less likely to report that it was a prob-
lem, suggesting that males and females may interpret and respond
to online behaviors differently. A better understanding of the con-
text and how girls as opposed to boys understand and define
cyberbullying behaviors may allow researchers and practitioners
to better understand these behaviors.
Emotional distress, depression, school problems, substance
abuse, delinquency, and significant psychosocial challenges have
been associated with cyberbullying and victimization (Beran & Li,
2005; Hinduja & Patchin, 2008; Mitchell et al., 2007; Shariff,
2005; Ybarra, Diener-West, et al., 2007; Ybarra, Espelage, et al.,
2007; Ybarra & Mitchell, 2004). However, only a few studies
(Dempsey et al., 2009; Fredstrom et al., 2011) have relied on reli-
able and valid psychometric instruments to measure outcomes. A
study published by Fredstrom et al. (2011), used the BASC-2 to
measure outcomes. Fredstrom’s work was based on a much larger
sample (n = 802) of students who were a bit older (mean age 15.84
years) than the current study. The researchers found cyber victim-
ization to be a significant predictor of low self-esteem, higher
social stress, anxiousness, depressive symptoms, and locus of con-
trol above and beyond the contributions of traditional
victimization. It will be important for future research to continue
to look at the similar, yet distinct social emotional outcomes of tra-
ditional and cyber victimization using psychometrically sound
instruments.
4.4. Limitations of current study
Several limitations of the current study should be noted. The
most significant limitation is sample size. Grade level analyses,
for example, should be interpreted with caution given that there
were only 15 students in the 8th grade. In addition, the sample
was predominantly White non-Hispanic and from a rural area in
the Midwest. Additionally, all data collected were self-report. Some
may consider self-report data to be a limitation. However, cyber-
bullying research to date has relied on self-report data and the ano-
nymity and sometimes covert nature inherent in some cyber
behaviors may mean that self-report data is necessary to examine
and understand this construct. Self-selection bias may also have
impacted the results. Results should be interpreted with these lim-
itations in mind.
The manner of categorizing students as cyber or traditional vic-
tims might be considered a limitation, as well. In this study, stu-
dents were considered victims if they scored in the top 20% of
the sample. The prevalence rate of cyber victimization in the sam-
ple is fairly low so it is possible that a few respondents who were
classified as victims may have been victimized minimally. Other
researchers have considered a respondent a victim based on spe-
cific responses to certain items (Fredstrom et al., 2011; Wang, Ian-
notti, & Nansel, 2009) or scoring above a certain score (Raskauskas,
2010).
Despite these limitations, a strength of the study was the use of
instruments with strong psychometric properties to measure on-
line and traditional victimization, and social emotional outcomes.
Since the development of psychometrically sound measures of
cyberbullying has been limited, the Cyber Victimization Survey
(CVS) was developed to investigate cyber victimization and answer
the study’s research questions of interest. The CVS shows promis-
ing evidence of reliability and validity and future research
should continue to investigate the psychometric properties of this
measure.
4.5. Implication of findings
Although a number of predictions were only partially supported
in this study, a number of significant and important results
emerged. It was found that the relations between cyber victimiza-
tion and social emotional outcomes differed by gender such that
for girls, cyber victimization was related to internalizing problems,
depression, and social stress while relations were not found for
boys. Students who are victimized both traditionally and electron-
ically are a small, but at risk population. School professionals and
practitioners should be aware that the risk for these students is
greater than the risk to children suffering only one form of victim-
ization. Due to the increasing use of technology among adoles-
cents, it may be beneficial for school-based bullying prevention
programs to focus on both cyberbullying and traditional bullying
behaviors.
Future research should delve into the conflicting gender results
in terms of prevalence, levels, and social emotional outcomes asso-
ciated with online victimization. Better understanding of why
some students experience more negative outcomes due to the
additive effects of traditional and online victimization would help
schools and practitioners better identify at-risk students, intervene
early, and implement prevention programs. Better understanding
the overlap between traditional and online victims will help pro-
fessionals identify and assist this small, but at risk population.
There is much more to learn about how technology, aggression,
and communication intersect to make students vulnerable to
negative outcomes from both a victimization and a bullying
perspective.
4.6. Conclusions
The results of this study indicate that levels of cyber victimiza-
tion do not differ by grade or by gender, cyber victimization and
traditional victimization are distinct but related constructs, and
relations between cyber victimization and social emotional out-
comes vary by gender, with girls suffering more than boys. This
study also confirmed that traditional bullying continues to be sig-
nificantly related to a number of negative social emotional out-
comes for all students. This is one of few studies to examine
cyber victimization outcomes by gender. While there were no dif-
ferences in the levels of cyber victimization for boys and girls, cy-
ber victimization alone significantly predicted a number of
negative outcomes for girls and not for boys. This may mean the
additive effect of traditional and online bullying is more problem-
atic for girls. It is also possible that traditional bullying remains the
primary means of significant, negative peer interactions among
middle school males such that any additive effect of online bully-
ing does not more negatively affect most boys. Further research
should continue to explore developmental and gender differences
of online bullying to better identify at risk students and to develop
effective prevention and intervention programs.
References
Agatston, P. W., Kowalski, R., & Limber, S. (2007). Students’ perspectives on cyber
bullying. Journal of Adolescent Health, 41(6), S59–S60. http://dx.doi.org/10.1016/
j.jadohealth.2007.08.017.
Beran, T., & Li, Q. (2005). Cyber-harassment: A study of a new method for an old
behavior. Educational Computing Research, 32(3), 265–277. http://dx.doi.org/
10.2190/8YQM-B04H-PG4D-BLLH.
Bonanno, R. A., & Hymel, S. (2013). Cyber bullying and internalizing difficulties:
Above and beyond the impact of traditional forms of bullying. Journal of Youth
and Adolescence, 42, 685–697. http://dx.doi.org/10.1007/s10964-013-9937-1.
Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA:
Harvard University Press.
Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-
psychological Adjustment. Child Development, 66, 710–722. http://dx.doi.org/
10.2307/1131945.
 C.F. Brown et al. / Computers in Human Behavior 35 (2014) 12–21
Dehue, F., Bolman, C., & Trijntje, V. (2008). Cyberbullying: Youngsters’ experiences
and parental perception. CyberPsychology & Behavior, 11, 217–223. http://
dx.doi.org/10.1089/cpb.2007.0008.
Dempsey, A. G., Sulkowski, M. L., Nichols, R., & Storch, E. A. (2009). Differences
between peer victimization in cyber and physical settings and associated
psychosocial adjustment in early adolescence. Psychology in the Schools, 46,
962–972. http://dx.doi.org/10.1002/pits.
Diamanduros, T., Downs, T., & Jenkins, S. J. (2008). The role of school psychologists
in the assessment, prevention, and intervention of cyberbullying. Psychology in
the Schools, 45, 693–704. http://dx.doi.org/10.1002/pits.20335.
Erdur-Baker, O. (2010). Cyberbullying and its correlation to traditional bullying,
gender and frequent and risky usage of internet mediated communication
tools. New Media & Society, 12, 109–125. http://dx.doi.org/10.1177/
1461444809341260.
Espelage, D. L., & Swearer, S. M. (Eds.). (2004). Bullying in North American schools.
New Jersey: Lawrence Erlbaum.
Fredstrom, B. K., Adams, R. E., & Gilman, R. (2011). Electronic and school-based.
victimization: Unique contexts for adjustment difficulties during adolescence.
Journal Youth Adolescence, 40, 405–415. http://dx.doi.org/10.1007/s10964-010-
9569-7.
Hinduja, S., & Patchin, J. W. (2007). Offline consequences of online victimization:
School violence and delinquency. Journal of School Violence, 6(3), 89–112. http://
dx.doi.org/10.1300/J202v06n03_06.
Hinduja, S., & Patchin, J. W. (2008). Cyberbullying: An exploratory analysis of factors
related to offending and victimization. Deviant Behavior, 29, 129–156. http://
dx.doi.org/10.1080/01639620701457816.
Hinduja, S., & Patchin, J. W. (2009). Bullying beyond the schoolyard preventing and
responding to cyberbullying. Thousand Oaks, CA: Corwin Press.
Hinduja, S., & Patchin, J. W. (2010). Bullying, cyberbullying, and suicide. Archives
of Suicide Research, 14, 206–221. http://dx.doi.org/10.1080/13811118.
2010.494133.
Klomek, A. B., Marrocco, F., Kleinman, M., Schonfeld, I. S., & Gould, M. S. (2007).
Bullying, depression, and suicidality in adolescents. Journal of the American
Academy of Child & Adolescent Psychiatry, 46(1), 40–49. http://dx.doi.org/
10.1097/01.chi.0000242237.84925.18.
Kowalski, R. M., & Limber, S. P. (2007). Electronic bullying among middle school
students. Journal of Adolescent Health, 41, S22–S30. http://dx.doi.org/10.1016/
j.jadohealth.2007.08.017.
Kyriakides, L., Kaloyirou, C., & Lindsay, J. (2006). An analysis of the revised Olweus
bully/victim questionnaire using the Rasch measurement model. British
Journal of Educational Psychology, 76, 781–801. http://dx.doi.org/10.1348/
000709905X53499.
Li, Q. (2006). Cyberbullying in schools: A research of gender differences.
School Psychology International, 27(2), 157–170. http://dx.doi.org/10.1177/
0143034306064547.
Li, Q. (2007). New bottle but old wine: A research of cyberbullying in schools.
Computers in Human Behavior, 23, 1777–1791. http://dx.doi.org/10.1016/
j.chb.2005.10.005.
Li, Q. (2010). Cyberbullying in high schools: A study of students’ behaviors and
beliefs about this new phenomenon. Journal of Aggression, Maltreatment and
Trauma, 19, 372–392. http://dx.doi.org/10.1080/10926771003788979.
Mason, K. L. (2008). Cyberbullying: A preliminary assessment for school personnel.
Psychology in the Schools, 45, 323–348. http://dx.doi.org/10.1002/pits.20301.
Mitchell, K. J., Ybarra, M., & Finkelhor, D. (2007). The relative importance of online
victimization in understanding depression, delinquency, and substance use.
Child Maltreatment, 5, 312–324. http://dx.doi.org/10.1177/1077559507305996.
Olweus, D. (1996). The revised Olweus Bully/Victim Questionnaire, Mimeo. Bergen,
Norway: Research Center for Health Promotion (HIMIL), University of Bergen.
Olweus, D. (1997). Bully/victim problems in school: Knowledge base and an
effective intervention program. The Irish Journal of Psychology, 18, 170–190.
http://dx.doi.org/10.1001/jama.285.16.2094.
21
Patchin, J. W., & Hinduja, S. (2006). Bullies move beyond the schoolyard: A
preliminary look at cyberbullying. Youth Violence and Juvenile Justice, 4(2),
148–169. http://dx.doi.org/10.1177/1541204006286288.
Patchin, J. W., & Hinduja, S. (2012). Cyberbullying: An update and synthesis of
research. In J. W. Patchin & S. Hinduja (Eds.), Cyberbullying prevention and
response (pp. 13–35). New York, NY: Taylor & Francis.
Raskauskas, J. (2010). Text-bullying: Associations with traditional bullying and
depression among New Zealand adolescents. Journal of School Violence, 9(1),
74–97. http://dx.doi.org/10.1080/15388220903185605.
Raskauskas, J., & Stoltz, A. D. (2007). Involvement in traditional and electronic
bullying among adolescents. Developmental Psychology, 43, 564–575. http://
dx.doi.org/10.1037/0012-1649.43.3.564.
Reynolds, C. R., & Kamphaus, R. W. (2004). Behavior assessment system for children
(2nd ed). Circle Pines, MN: AGS Publishing.
Shariff, S. (2005). Cyber-dilemmas in the new millennium: School obligations to
provide student safety in a virtual school environment. Journal of Education, 40,
467–487. <http://www.proquest.com>.
Slonje, R., & Smith, P. (2008). Cyberbullying: Another main type of bullying?
Scandinavian Journal of Psychology, 49, 147–154. doi: http://dx.doi.org/10.1111/
j.14679450.2007.00611.x.
Smith, P. K., Mahdavi, J., Carvalho, M., Fisher, S., Russell, S., & Tippett, N. (2008).
Cyberbullying: Its nature and impact in secondary schools pupils. Journal of
Child Psychology and Psychiatry, 49, 376–385. http://dx.doi.org/10.1111/j.1469-
7610.2007.01846.
Smith, P. K., & Slonje, R. (2010). Cyberbullying: The nature and extent of a new kind
of bullying, in and out of school. In S. R. Jimerson, S. M. Swearer, & D. L. Espelage
(Eds.), Handbook of bullying in schools: An international perspective
(pp. 249–262). New York, NY: Taylor & Francis.
Solberg, M., & Olweus, D. (2003). Prevalence estimation of school bullying with the
Olweus Bully/Victim Questionnaire. Aggressive Behavior, 29, 239–268. http://
dx.doi.org/10.1002/ab.10047.
Sourander, A., Klomek, A. B., Ikonen, M., Londroos, J., Luntamo, T., Koskelainen, M.,
et al. (2010). Psychosocial risk factors associated with Cyberbullying among
adolescents. Archives of General Psychiatry, 67, 720–728. http://dx.doi.org/
10.1001/archgenpsychiatry.2010.79.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (4th ed.).
Needham Heights, MA: Allyn and Bacon.
Turner, M. G., Exum, M. L., Brame, R., & Holt, T. J. (2013). Bullying victimization and
adolescent mental health: General and typological effects across sex. Journal of
Criminal Justice, 41(1), 53–59. http://dx.doi.org/10.1016/j.jcrimjus.2012.12.005.
Varjas, K., Henrich, C. C., & Meyers, J. (2009). Urban middle school students’
perceptions of bullying, cyberbullying, and school safety. Journal of School
Violence, 8, 159–176. http://dx.doi.org/10.1080/15388220802074165.
Wang, J., Iannotti, R. J., & Nansel, T. R. (2009). School bullying among adolescents in
the United States: Physical, verbal, relational, and cyber. Journal of Adolescent
Health, 45, 368–375. http://dx.doi.org/10.1016/j.jadohealth.2010.07.012.
Williams, K. R., & Guerra, N. G. (2007). Prevalence and predictors of internet
bullying. Journal of Adolescent Health, 41, S14–S21. http://dx.doi.org/10.1016/
j.jadohealth.2007.08.018.
Ybarra, M. L., Diener-West, M., & Leaf, P. J. (2007). Examining the overlap in internet
harassment and school bullying: Implications for school intervention. Journal
of Adolescent Health, 41, S42–S50. http://dx.doi.org/10.1016/j.jadohealth.
2007.09.004.
Ybarra, M. L., Espelage, D. L., & Mitchell, K. J. (2007). The co-occurrence of Internet
harassment and unwanted sexual solicitation victimization and perpetration:
Associations with psychosocial indicators. Journal of Adolescent Health, 41(6),
S31–S41. http://dx.doi.org/10.1016/j.jadohealth.2007.09.010.
Ybarra, M. L., & Mitchell, K. J. (2004). Youth engaging in online harassment:
Associations with caregiver-child relationships, internet use, and personal
characteristics. Journal of Adolescence, 27, 319–336. http://dx.doi.org/10.1016/
j.adolescence.2004.03.007.