GAME ON? ONLINE FRIENDSHIP, CYBERBULLYING, AND PSYCHOSOCIAL ADJUSTMENT IN HONG KONG CHINESE CHILDREN

Journal of Social and Clinical Psychology, Vol. 32, No. 2, 2013, pp. 159-185

  Leung and mcbride-chang
  GAME ON?

Game On? Online Friendship,
Cyberbullying, and Psychosocial
Adjustment in Hong Kong
Chinese Children
Angel Nga-man Leung
Hong Kong Community College, The Hong Kong Polytechnic University

Catherine McBride-Chang
The Chinese University of Hong Kong

Across 626 Hong Kong Chinese fifth and sixth graders, children’s experiences of
victimization and bullying in online and real life contexts were compared. Chil-
dren reported their best friendships at school and online when playing massively
multiplayer online games (MMOGs). With demographic variables, computer
gaming habits, school victimization and real life friendship measures statistically
controlled, online victimization uniquely and negatively explained variance in
friendship satisfaction, while online friendship positively and significantly ex-
plained additional variance in children’s social competence, friendship satisfac-
tion, self esteem, and life satisfaction. This research demonstrates theoretical and
practical importance of investigating social experiences (both negative, i.e., be-
ing cyber-bullied, and positive, i.e., building up online friendship) for early ado-
lescents’ psychosocial adjustment.

The present study focused on friendship and bullying experienc-
es of early adolescents, both at school and in an online computer
gaming context. We aimed to explore the associations of friendship
and bullying to more general aspects of children’s psychosocial ad-
justment. As students are exposed to the internet at younger and

  Address correspondence to Catherine McBride-Chang, Room 539, Sino Building, the
Chinese University of Hong Kong, Shatin, Hong Kong; E-mail: cmcbride@psy.cuhk.
edu.hk

© 2013 Guilford Publications, Inc.

159

160	LEUNG AND MCBRIDE-CHANG

younger ages, the central focus of the present study was the way
in which online social experiences might shape different aspects of
children’s sense of well-being.
  Some researchers tend to believe there is a negative displacement
relationship between computer game playing and social develop-
ment (e.g., Kraut et al., 1998) and several past studies (e.g., Blais,
Craig, Pepler, & Connolly, 2008; Ho & Lee, 2001; Holder, Boivin,
Vitaro, & Bukowski,s 2009; Nie, 2001) have demonstrated that com-
puter game playing is negatively associated with various aspects of
psychosocial adjustment. In contrast, the stimulation theory argues
that internet activities can stimulate relationship quality (Valken-
burg & Peter, 2007). Correspondingly, studies have found positive
relationships between computer game playing and psychological
well-being, with players having a better self concept (e.g., Durkin
& Barber, 2002), and playing computer games may be a useful so-
cial experience (Durkin & Aisbett, 1999) as children can further de-
velop social relationships within games, an important facet of psy-
chological well-being (e.g., Griffiths, 2010; Raney, Smith, & Baker,
2006; Valkenburg & Peter, 2007). More specifically, the role of online
games has been studied extensively in more recent studies. Bar-
nett and Coulson (2010) suggested that in online games, especially
in massively multiplayer games (MMOG), players often interact
or work towards a goal, and friendship can be formed. Social ties
(in terms of both online clan social capital and offline social and
emotional support) can be predicted positively by the level of how
gamers perceive the physical proximity, social proximity, and famil-
iarity of other gamers (Trepte, Reinecke, & Juechems, 2012). Zhong
(2011) suggested that social interactions and positive social expe-
rience in MMOGs positively predict online relationships because
of the strong ties being formed in games. In the present study, we
tested these ideas across particular game playing experiences (i.e.,
MMOGs) in relation to friendship and bullying.
  The potential for bullying in cyber space particularly worries
parents and educators (Franek, 2004). Cyberbullying has various
definitions. In summarizing across research on cyberbullying, Ma-
son (2008) defined it as “an individual or a group willfully using
information and communication involving electronic technologies
to facilitate deliberate and repeated harassment or threat to another
individual or group by sending or posting cruel text and/or graph-
ics using technological means” (p. 323). Theoretically, then, children
and adolescents might experience both friendship and victimiza-

GAME ON?                                                            161

tion in virtual space, especially when they spend a large amount
of time playing the virtually realistic MMOGs. However, how the
different domains of online friendship, online bullying, and online
victimization may be related to general psychological well-being
with development has been under-explored to date. Therefore, one
goal of the present research was to compare the experiences of three
aspects of social development, i.e., friendship, victimization, and
bullying, across two social contexts of daily life, i.e., school and the
internet in adolescents. We looked particularly at how online inter-
actions with friends and/or bullies are associated with children’s
overall ratings of self-esteem and life satisfaction.
  With the popularity of the internet, children or adolescents can
form friendships with people they have met at schools or on the
playground, or with people they have met online. This could oc-
cur in part because children spend increasing time online. Gentile,
Lynch, Linder, and Walsh (2004) found that young American boys
and girls reported spending on average 13 and 5 hours, respectively,
on gaming per week. In Hong Kong, self-reports indicate that teens
and young adults spend an average of between 2 and 4 hours on
the internet per day, with approximately 10 hours per week spent
on online games (Breakthrough Youth Research Archives, 2003;
Commission on Youth, 2008). Importantly for the present study, one
cross-cultural survey of Shanghai, Hong Kong, and American ado-
lescents also found that 64% from Shanghai, 46% from Hong Kong,
and 22% of American teens report playing Massively Multiplayer
Online Games (MMOGs; Digital Communities, 2007).
  In the present study, we focused on social interactions in MMOGs
because of their ever-widening appeal to children and adolescents.
MMOGs involve many players playing simultaneously via the in-
ternet (Barnett & Coulson, 2010). This appeal stems largely from the
designs of these online games, particularly those of MMOGs, with
their extensive, expanding, and dramatic graphics (Chan & Vorder-
er, 2006), creating alternative visually realistic, 3-dimensional, and
persistent worlds. The interactive nature of MMOGs has captured
the attention of researchers, who have argued that MMOGs facili-
tate the formation of social networks and social interactions (Cole &
Griffiths, 2007). Most MMOGs emphasize role playing, while some
may involve fighting and first person shooting; other popular types
of MMOGs among teens involve a focus on racing, sports games,
music, or dancing.

162	LEUNG AND MCBRIDE-CHANG

   MMOGs provide youth with the opportunities to exercise a va-
riety of cognitive and social skills that are crucial for development
(e.g., Barnett & Coulson, 2010; Raney et al., 2006; Sailsch, Oppl, &
Kristen, 2006). For example, players exchange knowledge and skills
related to computer games, including different game strategies and
tactics of the games. Such conversations can help to improve their
social and communication skills (Goldstein, 2003; Vandeventer,
1998). Playing games is considered social capital by which to build
up social networks, and this may be especially true among boys
(Raney et al., 2006). MMOGs also offer chances for teens to build
up identity and friendship by interacting with others within the
games themselves. Forming relationships and building up a social
network are major developmental tasks for adolescents (Gowers,
2005). Gamers compete or cooperate with others in the games, and
they form teams to play together, which fulfills their needs for af-
filiation with others. Players have a new opportunity to form social
networks with people that they have not met in reality (Gennaro
& Dutton, 2007). Indeed, in one survey, 18% of adolescents aged
14–17 reported that they had made friends online, and 10% of them
had then gone on to meet their online friends in person (Gennaro &
Dutton, 2007).
   Players usually form groups in MMOGs to play and they ex-
change help with one another in order to advance in levels. These
groups tend to back each other against rivals, promoting feelings of
trust and loyalty. They may also exchange ideas on various topics
via the instant messaging function. Instrumental support is likely
to be strongly emphasized because players can achieve goals much
more easily by staying with others; therefore, they may be willing
to help each other in order to level up in a game. MMOGs gam-
ers enjoy the social aspects of the gamers and they tend to social-
ize with other players (e.g., Barnett & Coulson, 2010). Overall, the
characteristics of MMOGs seem to facilitate instrumental, compan-
ionship, informational, and even esteem support for and to ado-
lescents. All of these are important elements of adolescents’ friend-
ship. The anonymous nature of online friendship may also facilitate
self-disclosure about their true selves, which, for some, affords a
kind of intimate relationship that cannot be experienced elsewhere
(McKenna, Green, & Gleason, 2002; Valkenburg & Peter, 2007).
   Given the specific nature of MMOGs, the present article aimed at
exploring online friendship built in MMOGs. However, apart from
the potential positive linkage between online friendship and inter-

GAME ON?                                                            163

active online games, there are also risks for children and teens in
the social ambiguity of online meetings. These might include the
ganging together of bad kids, meeting ill-intentioned adults pre-
tending to be teens, and other risks of being cyberbullied. The pres-
ent research, thus, also focused on the issues of cyberbullying and
cyber-victimization, in addition to friendship. While estimates vary
across studies and cultures, approximately 9% to 25% of children
and adolescents have experienced cyberbullying, while 15% to 28%
have cyberbullied others across cultures (e.g., Li, 2007; Patchin &
Hinduja, 2006).
  Importantly, there are several differences between traditional
(school) bullying and cyberbullying. First, in traditional school bul-
lying, bullies usually know their victims, but this may not apply to
the cyber-world. Li (2007) suggested that 40% of cybervictims did
not know their cyberbullies.
  In addition, cyberbullies are not typically confronted with the re-
actions or consequences of bullying others. Thus, the inhibition ef-
fect is much less for online bullying, making adolescents feel more
free to bully others (Mason, 2008). However, at the same time, it is
also less likely for the victims to get hurt physically in a cyberbully-
ing context. Moreover if they are bullied in some particular forums,
such as online games, victims of cyberbullying can choose to quit
such forums to escape from the bullying, while such an escape is
almost impossible for bullying in real life.
  Furthermore, Dehue, Bolman, and Vollink (2008) noted that par-
ents are often unaware of new technology, so they may neglect the
issue of online bullying. Ybarra and Mitchell (2004a) suggested that
more than half of adolescents claim that the monitoring from par-
ents about online activities is poor. For instance, Li (2007) found
that only 30% of teenagers reported such bullying to adults. Slonjie
and Smith (2008) suggested that most students think that for pic-
ture/video clip cyberbullying, adults are likely to notice that, but
not for other types of cyberbullying, such as text messages, emails,
or phone calls bullying. In fact, parents, teachers and educators may
all fail to deal with this issue in part because they lack the knowl-
edge structure or capacity to understand and help (Berman & Li,
2005). This lack of surveillance from parents may reinforce the be-
havior of cyberbullying. Indeed, Ybarra and Mitchell (2004b) found
that adolescents with poor relationships with their parents are twice
as likely to become online bullies as are those with solid relation-
ships with parents.

164	LEUNG AND MCBRIDE-CHANG

  Both school and online victimizations have similar psychologi-
cal consequences for victims. Victims of traditional bullying might
have some health problems such as suicidal ideation, eating disor-
ders, and chronic illness, and they tend to be depressed and have
poorer self esteem (Hawker & Boulton, 2000; Mason, 2008). Dehue
et al. (2008) also found that cybervictims tend to lose trust in others
and feel angry. Low self esteem, depression, suicidal ideation, and
poorer academic performance are also related to the experience of
being cyberbullied (Finkelhor, Mitchell, & Wolak, 2005; Patchin &
Hinduja, 2010).
  Despite the many differences between traditional versus online
bullying, both types of bullying involve intentional aggression to-
wards others; both aim at causing harm to others. Katzer, Fetch-
enhauer, and Belschak (2009) suggested that, among 1,700 second-
ary school students in Germany, school victims are more likely to
become online victims as well; the predictors for both online and
school victimization share similarities. Smith, Shu, and Madsen
(2001) suggested that, despite a decline in reports of victimization
throughout childhood to adolescence, some children remain vic-
tims for many years. Some children struggle to attain the necessary
social skills to deal with bullies; this is one reason that they remain
victims. This is in line with characteristic self-blaming (Graham &
Juvonen, 1998) on the part of some victims.
  Victimization and friendship are sometimes studied together,
because both represent different forms of social relationships that
children and adolescents may experience and that may together
contribute to the psychological outcomes of children’s well-being
(e.g., Ladd, Kochenderfer, & Coleman, 1997). Friendship has a pro-
tective function over victimization, probably because friendship
offers a context in which children can learn and practice different
social skills, as well as build up self concepts and self esteem as
they gain emotional support (Hodges, Boivin, Vitaro, & Bukowski,
1999). Having one or more friends is a protective factor for the ex-
perience of being bullied (e.g., Bukowski, Sippola, & Boivin, 1995;
Hodges et al., 1999). With the consensus that real life friendship can
buffer victimization experiences, can online friendship also provide
similar buffering functions?
  There are a few research goals of the present study. Firstly, as re-
search on both online friendship and cyberbullying remains scarce,

GAME ON?                                                           165

especially in Chinese populations, the first goal of the study was to
measure the prevalence rate of cyberbullying and cyber-victimiza-
tion in Hong Kong and to investigate the extent to which the pat-
tern of victimization and bullying might change from the real life
to the online context. As not all Hong Kong Chinese children can
regularly access the internet, it was assumed at the outset that the
prevalence rate for online victimization and online bullying should
be lower than for school victimization and bullying. However, we
expected to see similar patterns of bullying and victimization pat-
terns across online and real life contexts.
   Second, we hypothesized that both school and cyber-victimiza-
tion experiences would be negatively associated with psychological
well-being. Third, given previous research suggesting that friend-
ship may be protective against bullying in real life settings, we
tested the same idea in online contexts. This was a relatively new
idea, since past studies on this are scarce. We hypothesized that the
quality of online friendship in MMOGs could also positively and
uniquely explain the overall psychological well-being of children
and early adolescents, after statistically controlling the role of real
life friendship, traditional (school) victimization, and online victim-
ization. Only MMOGs gamers were included in the analyses for the
third research goal, because we believed that online friendship in
MMOGs could be formed easily and might be comparable to real
life friendship.
   For the measurement of psychological well-being, self esteem,
social competence, friendship satisfaction, and life satisfaction
were used. Self esteem is an important agentic need of adolescents
(Buhrmester, 1996), and it has been used as a measurement of well-
being of adolescents in previous work (e.g., Berndt & Miller, 1993).
Past studies have also shown that victims tend to be more anxious
and have lower self-esteem than are nonvictims (e.g., Hawker &
Boulton, 2000; Mason, 2008). Social competence is another impor-
tant skill that teenagers might achieve from friendship. Being com-
petent socially implies that a teen is better at communicating with
others and better able to adapt to more challenging social roles in
the future. Friendship satisfaction was operationalized here as the
general satisfaction that children and early adolescents can obtain
from friendship; it is the amount of satisfaction that children can
gain from their social interactions with friends. Finally, victims of
traditional bullying often feel sad and angry. Thus, we included

166	LEUNG AND MCBRIDE-CHANG

overall life satisfaction as another indicator of psychological well-
being.

Method
Participants

Six hundred and twenty-six (318 boys; 308 girls) students in grades
5 and 6 from four primary schools in two districts of Hong Kong
participated in this study. Their ages ranged from 9 to 15, with an
average age of 10.81 (SD = .83) years.

Procedure

Invitation letters were sent to principals of fifteen schools in Hong
Kong. Principals of 4 schools agreed to participate. Upon obtaining
parental consent, students filled out an online questionnaire which
included several batteries of scales. The quality of the data collected
by computer-based questionnaires may be equal to or even better
than paper-and-pencil versions, especially in terms of complete-
ness of data, given that web-based questionnaires can be designed
to include a default checking function to remind participants to fill
out any missed questions (Kongsved, Basnov, Holm-Christensen,
& Hjollund, 2007). Students accessed the online questionnaire
through a specific link that only invited students could access. Clear
instructions with pictorial indications were given on the first page
of the questionnaire, with audio clips which recorded the words
of the instructions and every item throughout the whole question-
naire. Students could, thus, click to hear the words if and when they
found any difficult terms to read; this facilitated their completion of
the questionnaire on their own. All children could read Chinese so
all the measures were administered in Chinese.

Measures

  Demographics. Information on flat size and parents’ education
levels was gathered to tap basic demographics of the participants.
Most Hong Kong people live in flats, and flat size can be one of the
indicators to reflect the SES of the children’s family.

GAME ON?                                                           167

  Computer Game Playing Habits. The estimated frequency and time
spent on playing different kinds of computer games were estimat-
ed, i.e., interactive online games (e.g., MMOGs), solo-PC games,
handheld video games (e.g., NDS) and family video console (e.g.,
Playstation3). Nevertheless, items on habits related to types of com-
puter games other than MMOGs were mainly descriptive in nature
for the sake of understanding the overall pattern of computer game
playing. MMOGs was the major focus. A logic link was set in the
online questionnaire so that only students who indicated that they
played MMOGs would answer these follow-up questions. Follow-
up questions included whether their best friend in the MMOGs
played was the same best friend they named for the real life context
and whether their favorite MMOGs were violent or not. This sec-
tion also included a set of questions about the friendship quality of
their best online friend in online games.
  Friendship Qualities. The friendship quality scale of Bukowski et
al. (1994) is well-established and has been used widely with early
adolescent populations. Four subscales, namely companionship,
help, security, and closeness, of the Friendship Qualities scale by Bu-
kowski et al. (1994) were used. Following Demir and Urberg (2004),
we used a composite score of the four subscales for both online and
real life contexts. Participants were asked to think about one of their
best friends in real life (e.g., in school, playground) and then rate
the items related to friendship qualities of that friend. Students who
indicated that they had formed online friendships in MMOGs were
also asked to think about one of their best friends in the MMOG
context, and then answer questions about that particular friend in
considering friendship qualities of the online MMOG friend. The
instructions were clear about that this person could be the same or
different across the two contexts.
  To balance the effect of the order in answering questions related
to the online friendship or the real life friendship first, the online
questionnaire system randomized the order of asking the two sets
of questions. Participants were asked to rate on a 5-point likert scale
how much they agreed with each item, with 1 as strongly disagree
to 5 as strongly agree. A sample item of the scale is “My friend
thinks of fun things for us to do together.” The reliabilities of the
friendship quality scale for both contexts were .94.
  Parental Monitoring of Computer Use. Parental monitoring of chil-
dren’s use of the computer is an important factor to consider in

168	LEUNG AND MCBRIDE-CHANG

this research because of previous work demonstrating associations
between parental relationships and online bullying (e.g., Ybarra &
Mitchell, 2004b). Thus, this variable was included as a control mea-
sure in regression analyses. A parental monitoring scale was devel-
oped by adopting the parental knowledge scale from Kerr and Stat-
tin (2000) to form a 6-item scale measuring parents’ monitoring of
computer use. A sample item is, “I tell my parents what I am doing
on the internet.” Participants were asked to rate on a 6-point likert
scale, from 1 as strongly disagree to 6 as strongly agree for each
of these items. The 6 items were then subjected to EFA, and 1 fac-
tor was extracted, with an eigenvalue = 3.53, which accounted for
58.85% of the total variance. The reliability of the scale was .86.
  Social Competence. The Social Competence subscale of the Per-
ceived Competence Scale for Children (Harter, 1982), which was ad-
ministered as a 5-point Likert scale, was used to assess participants’
self perceived social competence. A sample item is “I am easily liked
by others.” The reliability of the scale was .88.
  Friendship Satisfaction. The friendship satisfaction subscale of the
Multidimensional Students’ Life Satisfaction Scale by Huebner,
Laughlin, Ash, and Gilman (1998) was used. This is a 9-item mea-
sure of the general satisfaction that children gain from friendship.
It measures the sense of satisfaction, rather than the quality of a
particular friend, on a 5-point Likert scale. A sample item is “My
friends are good to me.” The obtained internal consistency reliabil-
ity of the scale was .83.
  Self Esteem. The Rosenberg Self-Esteem Scale (Rosenberg, 1965),
a 10-item measure of global personal self-esteem, presented on a
five-point scale was used. A sample item is “I have plenty of good
qualities.” The reliability of the scale in the present study was .76.
  Satisfaction with Life. This scale assesses children’s overall per-
ceived subjective well-being on a seven-point scale. It has good
cross-cultural validity (Diener, Emmons, Larsen, & Griffin, 1985). A
sample item is “My life is close to ideal.” The obtained reliability of
the scale was .83.
   School Victimization. The 5-item peer victimization scale by
Schwartz, Chang, and Farver (2001) was used to tap the experience
of being victimized in traditional settings. This questionnaire has
been used with Chinese fifth and sixth graders before. One example
item is “Others tease or make fun of me?” Participants rated the

GAME ON?                                                           169

frequency of these things happening to them on a 1–5 Likert scale,
from 1 as never, to 5 as everyday. The reliability of the scale in this
sample was .89.
  School Bullying. The 5-item peer victimization scale by Schwartz et
al. (2001) can also be used to measure bullying behavior by chang-
ing the wordings slightly. For instance, “Others tease or make fun
of me” was changed to “I tease or make fun of others?” to measure
bullying behavior. The original authors also changed the wordings
in this way to tap school bullying in their study. The students in the
present study answered the same 5-point Likert scale as in school
victimization. The reliability in this sample was .91.
   Cyber-Victimization. Five items from Schwartz et al. (2001) were
initially used, with the item “Others hit or push me” changed to
“Others hit or push my avatars in online games deliberately.” Three
additional items of the verbal aggression and three items from the
indirect aggression scale from the Direct and Indirect Aggression
Scales (DIAS), developed by Bjorkqvist, Lagerspetz, and Osterman
(1992) were chosen for inclusion in the scale as well. These items
asked about behaviors that were not covered by the Schwartz et
al. (2001) peer victimization scale. Some minor modifications of the
items were made to fit into the online context. For example, the item
“Others insult me” was changed to “Others insult me by swear-
ing at me in online games/on the internet” because being exposed
to foul language may be relatively common in an online context.
Two more items from the relational aggression scale of Crick and
Grotpeter (1995) were also added to broaden the scope of the newly
developed scale in terms of relational aggression. Finally, five ad-
ditional items which cover other possible aspects of cyberbullying
were added. Examples of these are “Others steal my account or my
belongings (e.g., money, weapons) in online games/on the inter-
net.” A total of 18 items were initially included for the Cyber-vic-
itmization scale. The sample was then randomly split into halves,
an exploratory principal components factor analysis (EFA) was run,
and then a confirmatory factor analysis (CFA) was run on another
sample to validate it. After trimming down the items to fit with the
rule-of-thumb suggested by researchers (e.g., Kline, 2005), eight
items remained. A model has acceptable fit if the value of CFI and
NNFI are in the upper .80s (Bentler, 1992), with a larger value indi-
cating a better fit. RMESA and SRMR should not be larger than .1
for acceptable models (Browne & Cudeck, 1993). When this one fac-

170	LEUNG AND MCBRIDE-CHANG

tor model with eight items was applied to the full sample with 626
students, the chi-square of the model was: χ2 (20, N = 626) = 128.80,
p < .01; and the goodness-of-fit indices were as follows: CFI =.97;
NNFI = .96; SRMR = .027; RMR = .019; RMSEA = .09. It is clear that
all the goodness-of- fit indices worked even better in the full sample
(for details of running EFA, CFA, and factor loadings, please see
Leung, 2010). The final eight-item scale is listed in Appendix A, and
the items together accounted for 66.9% of the total variance when
subjected to EFA with the 626 participants. The obtained reliability
of the scale was .93.
   Cyberbullying. Past studies of victimization and bullying have
tended to adopt identical items, although the wordings are neces-
sarily changed in describing the actions by the participants as bul-
lies or victims (e.g., Schwartz et al., 2001). Therefore, to facilitate
easy comparison between the two scales and for the ease of ad-
ministration in using the two scales in future studies, for the scale
of cyberbullying, the same eight victimization items, with differ-
ent wordings were retained for the CFA of the cyberbullying scale.
The model for this 8-item structure of the cyberbullying scale was:
χ2 (20, N = 626) = 141.14, p < .01; goodness-of-fit indices were as
follows: CFI = .98; NNFI = .97; SRMR = .02; RMR = .01; RMSEA =
.10. This eight-item scale is also listed in Appendix A. The items ac-
counted for 76.7% of the total variance when subjected to EFA with
626 participants. The obtained reliability of the scale was .96.

Results

Most students (91.8%) lived in a flat with at least 2 rooms. A 2-room
flat roughly ranges between 40–50 square meters, which roughly
corresponds to the average flat size in Hong Kong. Approximate-
ly 13.8% of the children’s fathers and 10.4% of their mothers had
earned a bachelor’s degrees or above; the average percentage of
those with a bachelor’s degrees or above in Hong Kong overall
is approximately 17.3% (Census and Statistics Department, 2009).
Thus, these demographics from the present study roughly reflect a
lower-middle class group as compared to the demographics of all
of Hong Kong.
  The (mean) average time (and standard deviation) that the stu-
dents reported spending across the four types of games was 1.12
hours (1.44) for the MMOGs, .73 (1.09) for solitary computer games,

GAME ON?                                                                                       171

   TABLE 1. Means, SD, Correlations Among School Victimization, Bullying and Cyber-
                     Victimization and Cyberbullying (N = 626)
                                               School        School       Cyber-
Scale (alpha)                  Mean (SD)    victimization   bullying   victimization Cyber-bullying
School victimization(.89)      1.98 (.82)        –
School bullying(.91)           1.63 (.70)      .53**           –
Cyber-victimization(.93)       1.38 (.66)      .34**         .36**          –
Cyberbullying(.96)             1.27 (.60)      .24**         .43**        .69**            –
Note. **p < .01, two-tailed.

.73 (1.21) for handheld video games such as Nintendo DS or Play
Station, .52 (1.09) hours for home video consoles. Fifty-six percent
of the sample had reportedly played MMOGs in the previous three
months. Boys played significantly more than girls in MMOGs, soli-
tary computer games, and handheld video games, but did not dif-
fer significantly in time spent on home video consoles, MMOGS, t
(596.32) = -5.79 p < .00, Cohen’s d = .47; solitary computer games,
t (600.70) = -3.21, p < .00, Cohen’s d = .26; handheld video games
(e.g., NDS, PSP), t (549.05) = -4.08, p < .00, Cohen’s d = .35; home
video consoles, t (605.18) = 1.87, p > .05. The effect sizes ranged from
small to medium, with the biggest gender difference in time spent
on MMOGs. There were no significant grade differences in terms of
time spent on the three types of games, but sixth graders spent more
time than fifth graders on MMOGs, t (585.70) = 3.18, p < .00, Cohen’s
d = .26, with grade 6 students spending more time on MMOGs than
grade 5 students.
   Means, SDs, and correlations among school victimization, bully-
ing, and cyber-victimization and cyberbullying and the respective
prevalence rates are listed in Tables 1 and 2. Among the four types
of school and online bullying and victimization, the most preva-
lent was school victimization, followed by school bullying, cyber-
victimization and cyberbullying. Victimization was more prevalent
than bullying in both contexts. If we consider those with a mean
score higher than 3 as frequent bullies/victims, proportionally,
among children who reported having been victimized in schools,
only 8% of them were frequently victimized in the school context,
but there was a larger proportion (11.2%) of those who reported be-
ing heavily victimized online among those who had experienced
cyberbullying; the same was true for bullying. Only 2.76% of those
who rated themselves as school bullies claimed to have done this

172	LEUNG AND MCBRIDE-CHANG

  TABLE 2. Prevalence Rate of School and Online Bullying and Victimization (N = 626)
                             Percentage of       Percentage of students Percentage of students
                       students saying they have saying they have fre-    saying they have very
                            experienced it       quently experienced it frequently experienced it
Scale (alpha)             (score of scale >1)      (score of scale >2)     (score of scale >3)
School victimization             80.5                    36.1                      6.5
School bullying                  65.2                    21.9                      1.8
Cyber-victimization              47.3                    16.3                      5.3
Cyberbullying                    31.2                    12.1                      4.0

very frequently, but 12.82% of those who rated themselves as cyber-
bullies said that they engaged in this behavior very frequently.
   The second research question concerning whether both school
and online victimization is negatively correlated with children’s
and early adolescents’ psychological well-being was addressed by
means of correlation analyses. For both gamers and nongamers, a
significant and negative association was found between traditional
school victimization and psychological well-being (social compe-
tence: r = -.32, p < .01, friendship satisfaction: r = -.38, p < .01; self-
esteem: r = -.33, p < .01, life satisfaction: r = -.20, p < .01). Significant
negative, but less strong associations were found between cyber-
victimization and psychological well-being (social competence: r =
-.10, p < .01, friendship satisfaction: r = -.19, p < .01; self-esteem: r =
-.18, p < .01, and life satisfaction: r = -.12, p < .01).
   To answer the third research question, the combined effects of
friendship and victimization in real life and online contexts (partic-
ularly referring to MMOGs gamers as they would experience both
friendship and bulling in games easily) were examined. regression
analyses were carried out. Gender differences emerged for time
spent on online games, real life friendship and cyber-victimization.
To further investigate the role of gender in moderating how time
spent on online games, and victimization and friendship in both
contexts explained variance in psychological outcomes, five inter-
action terms were formed by multiplying gender with school vic-
timization, school bullying, online victimization, online bullying,
real life friendship, online friendship, and average time spent on
computer games. However, only the interaction term of real life
friendship was significantly associated with some of the psycho-
logical well-being indicators. Therefore, only this interaction term
is shown in Table 3.

GAME ON?                                                           173

   After the effects of school, grade, age, parental education, number
of rooms at home, and parental monitoring of computer use were
statistically controlled, school victimization was negatively related
to the four psychological constructs. Indeed, school victimization
uniquely explained all of the psychological constructs in these anal-
yses. Correspondingly, real life friendship was positively correlated
with social competence, friendship satisfaction, and life satisfac-
tion; it contributed unique variance in explaining all three of these
measure, but not self esteem. After controlling all the real life so-
cial experiences, time spent on MMOGs, types of MMOGs (violent
vs. nonviolent), and whether the best friend identified within the
MMOGs was the same or different from the best friend in reality,
cyber-victimization still explained 1% additional variance in friend-
ship satisfaction, and it was negatively correlated with friendship
satisfaction. Importantly, online friendship contributed 2–3% ad-
ditional significant unique variance across all four psychosocial
adjustment constructs, and it was positively related to the four of
them as well. Gender only moderated the relationship between real
life friendship and social competence and friendship satisfaction,
but boys benefit from a good quality real life friendship more than
girls do in this context. However, no moderating role of gender was
found for online friendship, suggesting that both boys and girls
benefit to the same degree from online friendships.

Discussion

Although the MMOG context is apparently an important environ-
ment in which friendship development and bullying might take
place among children and adolescents, there are very few studies
on this environment. Despite the fact that, in the existing literature,
victimization and friendship tend to be studied together to illus-
trate their relative importance in explaining the psychological well-
being of children, such a strategy has rarely, if ever, been applied to
the online context. The present study was a first attempt to fill these
gaps in the existing literature. In this work on Hong Kong Chinese
children and early adolescents, we examined the prevalence rate
of online and real life bullying and victimization, looked at overall
computer game playing habits, and tested the relative predictive
power of cyber-victimization and online friendship for explaining
psychological well-being on four dimensions.

TABLE 3. Hierarchical Regression Analysis for Predicting Social Competence, Friendship Satisfaction, Self-Esteem, and Life Satisfaction (N = 326)
                                       Social competence                  Friendship satisfaction                Self-esteem                    Life satisfaction
Step                             R2       ΔR2     Beta (t-value)    R2         ΔR2     Beta (t-value)     R2    ΔR2    Beta (t-value)     R2      ΔR2     Beta (t-value)
1      School 2                  .05      .05      –.10(–1.48)      .03        .03     –.01(–.13)        .06    .06*   –.15(–2.09)*      .08    .08**    –.12(–1.7)
       School 3                                     .03(.47)                           –.01(–.08)                      –.06(–.92)                        –.04(–.56)
       School 4                                    –.16(–2.31)*                        –.08(–1.20)                     –.12(–1.75)                       –.14(–2.06)*
       Grade                                       –.08(–1.16)                          .05(.71)                       –.01(–.08)                        –.03(–.45)
       Age                                          .10(1.42)                           .01(.13)                        .06(.90)                          .10(1.43)
       Gender                                       .05(.70)                           –.06(–.95)                      –.00(–.02)                         .02(.31)
       Parental Education                           .01(.10)                            .03(.59)                        .14(2.34)*                        .08(126)
       Number of rooms at                           .02(.35)                           –.05(–.86)                      –.04(–.74)                         .00(.05)
        home
       Parental monitoring                          .08(1.36)                           .07(1.33)                       .06(.96)                          .19(3.18)**
2      School Victimization      .16      .11**    –.29(–5.16)**    .17        .14** –.29(5.31)**       .14    .08**   –.23(–3.87)**     .10    .02**    –.11(–.18)
3      Real life friendship      .19      .03**     .19(2.67)**     .22        .05**    .21(2.99)**     .14    .01      .02(.20)         .13    .03**     .13(1.70)
4      Time spent on MMOGS       .20      .01       .03(.64)        .23        .01      .06(1.09)       .15    .00     –.08(–1.38)       .14    .02      –.15(–2.43)*
       Type of MMOGS                               –.09(–1.44)                         –.05(–.84)                      –.02(–.30)                        –.05(–.78)
       Best friend                                 –.00(–.05)                          –.06(–1.17)                      .01(.11)                          .05(.81)
5      Cyber-victimization       .20      .00     –0.01(–.22)       .24        .01*    -.13(–2.26)*     .16    .01     –.10(–1.70)       .14    .00      –.05(–.87)
6      Online friendship         .22      .02**    0.17(2.83)**     .27        .03**    .19(3.22)**     .18    .02**    .18(2.90)**      .16    .02*      .15(2.36)*
7      Real life friendship X    .24      .01*    –0.14(-2.18)*     .28        .01*    -.12(-1.96)*     .18    .00      -.04(-.63)       .16    .00       -.03(-.45)
         Gender
Notes. Type of MMOGS: violent games coded as 0, nonviolent as 1; Best friend refers to if the best friend in real life & MMOGS is the same person, the same person is
coded as 0, not the same person is coded as 1; Gender: boys are coded as 0, girls as 1.
*p < = .05; ** p < .01
                                                                                                                                                                           174	LEUNG AND MCBRIDE-CHANG

GAME ON?                                                           175

   In the present study, more than half of those surveyed reported
being MMOG players. This finding underscores the growing im-
portance of this environment for understanding social interactions.
Moreover, more than 93% of these participants reported having
formed online friendship in MMOGs. This suggests that the social
functioning of children within MMOGs is important and is an en-
tirely new context in which friendship formation can occur from
traditional models. Despite the fact that online victimization may be
less serious in scope than is school victimization for most children,
this does not imply that cyber-victimization is not an important is-
sue. Indeed, cyber-victimization still explained additional variance
in overall friendship satisfaction, a crucial measure of adjustment in
early adolescents, once real life and school social experiences were
statistically controlled in the present study. Thus, cyber-victimiza-
tion is apparently a unique measure for understanding overall psy-
chosocial adjustment in typical children who play games online.
For children who have lower friendship quality, as well as both
school victimization and online victimization, this buffering effect
may be lower, hampering their psychological well-being. Dehue
et al. (2008) suggested that for those who already exhibit depres-
sive symptoms, being additionally cyberbullied might bring them
additional negative consequences compared with youngsters who
are not depressed. One’s real identity is shielded by a virtual and
anonymous identity on the internet (Mason, 2008), coupled with
the fact that online bullies are not typically confronted by online
victims, the consequences of bullying of others are further reduced.
Willard (2007) suggested that there is a concept of harmful bystand-
ers in cyberbullying. Harmful bystanders are “those who encourage
and support the bully or watch the bullying from the sidelines, but
do nothing to intervene or help the target” (Willard, 2007, p. 6). This
is a notably painful aspect of school bullying as well (e.g., Cowie,
2002).
   Despite not having been an aspect of the present research, the role
of harmful bystanders could be one of the sources that contribute
to frequent and intensive online bullying. One practical implication
for educators and parents concerning the means by which they may
consider trying to reduce cyberbullying would be to raise students’
awareness that being a harmful bystander likely exerts harm to the
victims. Students should be taught to raise awareness of their roles
in this issue, and as suggested by Willard (2007), students could be
encouraged to be helpful online bystanders, by helping to stop the

176	LEUNG AND MCBRIDE-CHANG

online bullying from happening. Ways to accomplish this include
encouraging them to report any suspected cases of cyberbullying
to the administrators of online forums or online games and perhaps
contacting school teachers if they know the victims/bullies.
   There is also a positive aspect of social development in going on-
line. Griffiths (2010) noted that despite some concerns about com-
puter games, a number of recent studies suggested that children
and adolescents can meet friends when they go online. Building up
online friendship in MMOGs can be positive for teens. Indeed, our
findings suggested that online friendship quality was positively re-
lated to all four psychological constructs measured. Moreover, the
standardized beta in the final model (as shown in Table 3) was com-
parable to the strength of the beta of real life friendship in the pres-
ent study. Having more friends, be it from online or offline sources,
is beneficial to the development of children and early adolescents.
While there is already consensus that having at least one good
friend could buffer the negative impacts of victimization (Bukowski
et al., 1995; Hodges et al., 1999), the present research suggested that
if children could also enjoy online friendship, the benefits may be
additive for psychological well-being.
   As suggested by Ferguson, Coulson, and Barnett (2011), cyberbul-
lying, as for other pathological gaming patterns, as well as the new
emerging online games themselves may bring the moral panic to
scholars. This generally negative impression towards children and
adolescents playing online games may partly result from the likely
exaggeration of the actual prevalence rate of pathological gaming
because different past studies might have used different measures
or scales to estimate it. In a recent meta-analysis, it was found that
with precise measures, the overall prevalence rate of pathological
gaming is 3.1% (Ferguson et al., 2011). This moral panic may lead us
to believe that children are at risk once they are in touch with com-
puters. However, this study suggested that although children can
be exposed to cyberbullying, when they play MMOGs, friendships
can also be formed. Combining both the issues of online victimiza-
tion and online friendship and the psychological consequences of
children and early adolescents, it would be an over-simplification to
say that, because of the potential threat of being cyber-bullied when
children go online, playing online games is bad for children. We
should keep an open mind when evaluating the potential risks and
benefits that children and adolescents may have in the cyber world.

GAME ON?                                                         177

Ho and Lee (2001) suggested that time spent online should not be
the sole consideration when investigating the positive or negative
associations of computer playing and social well-being. They found
that time spent on computer usage did not compete for the time
spent on other types of recreation activities, and did not make ado-
lescent players more socially withdrawn than the nonplayers. Re-
sults of the present study have suggested that, for players who play
MMOGs, the most popular game type with the most interactive
elements, they may develop online friendship, which is positively
related to children’s social development. Cyber-victimization is also
a risk.
  The social consequences of online game-playing should be con-
sidered in context. Just as when television was gaining in popular-
ity among middle class households more than 45 years ago in the
United States, there were worries about its negative consequences
on children’s development, there is, correspondingly, an ongoing
debate on the relation of computer game playing to the social de-
velopment of children. Related to these concerns, evidence remains
inconsistent regarding whether violent video games may increase
aggression (e.g., Coyne, Padilla-Walker, Stockdale, & Day, 2011;
Markey & Markey, 2010), decrease aggression (Bennerstedt, Ivars-
son, & Linderoth, 2012; Colwell & Kato, 2003) or have no effect at
all on player aggression (Ferguson, San Miguel, Garza, & Jerabeck,
2012; von Salish, Vogelgesang, Kristen, & Oppl, 2011). However,
most scholars agree that addiction issues may be evidenced in a
small number of players (Griffiths, 2000). However, the positive so-
cial aspects of computer games, especially online games, cannot be
overlooked (e.g., Chan & Vorderer, 2006). The present study sug-
gested that, based on a Hong Kong Chinese population, it would be
within the norm for a child to spend 1 to 2.5 hours per day playing
MMOGs. The issue of friendship and bullying within MMOGs is,
therefore, an ongoing one to consider in understanding social de-
velopment in modern youth.
  The present study has suggested a more comprehensive approach
in looking into MMOGs in particular. MMOGs are the most popular
type of computer games among teenagers; their interactive nature
provides increased social interactions for children in the games,
making them different from solitary computer games. Many past
studies (e.g., Blais et al., 2008; Ho & Lee, 2001; Holder, Coleman,
& Sehn, 2009) have found negative correlations between computer

178	LEUNG AND MCBRIDE-CHANG

game use or time spent on computer games in relation to well-be-
ing or a less social life style. However, these researchers did not
look into what children were actually doing in the computer games.
Blais et al. (2008) found that, computer usage for communication
purposes (e.g., chat rooms or MSN) was positively related to best
friendship in real life, but computer usage for entertainment and
gaming purposes was negatively related to real life best friendship.
Without differentiating what kinds of games children are playing
over the internet, it is difficult to make this distinction.
   To answer the question of whether computer game playing fits
into the reduction/displacement theory (Kraut et al., 1998), in
which internet activities reduce social activities, or, in contrast, fits
the stimulation theory, which states that internet activities can serve
to stimulate relationship quality (Valkenburg & Peter, 2007), in fu-
ture studies, time spent and also enjoyment of different categories
of computer games should also be included. Comparisons could be
made between the psychological consequences of players and non-
players of different types of games to illustrate the respective social
functions of different types of games. Also, future studies should
include a longitudinal comparison of online friendship in MMOGs
versus friendship in real life to provide a clear pattern of whether
online friendship could actually replace or reduce the quality of real
life friendship or not.
   There were several limitations of the present study. First of all, our
measures on cyberbullying and cyber-victimization were largely
limited to the realms of verbal and indirect aggression. The scales
were influenced by considerations of what typically happens in
school, since most past research measuring bullying and victimiza-
tion took place within the school environment. However, the more
serious cyberbullying (such as receiving messages with explicit
sexual content, or being continuously threatened) may be rare and
is less likely to happen among children and teens as compared to
adults, though they may occasionally occur. This side of cyberbul-
lying was not well-captured by the present scale. Future studies
should incorporate such bullying practices as well.
   Moreover, although we measured children’s item by item re-
sponses to particular instances of bullying and victimization, we
did not explicitly ask children about their subjective experiences of
each at school and online separately. Future studies should further
include children’s feelings about being cyberbullied and being bul-

GAME ON?                                                           179

lied in school, to compare the difference in the subjective serious-
ness of the two victimization experiences in the two contexts. For
example, Bond, Wolfe, Tollit, Butler, and Patton (2007) suggested
that what researchers and adults believe to be bullying and victim-
ization could be different from how children and adolescents view
it.
   It would be important to include questions to investigate what ac-
tions on the internet may attract attacks from other people, to better
prevent it from happening. Finally, the present study was correla-
tional in nature. Thus, no causal conclusions could be drawn. Bidi-
rectional associations between psychosocial adjustment and online
friendships and bullying in children can only be revealed by future
longitudinal research.
   Despite these limitations, however, this research has made sev-
eral important contributions to understanding of children’s social
development. Practically, this research investigated a popular but
relatively understudied topic—online game playing. The study
may help in addressing some public concerns about whether com-
puter gaming has a positive or negative relationship to emotional
and relational development in children. Such findings may influ-
ence educational practices and potentially shape parenting strate-
gies concerning computer game monitoring. This work might also
raise public awareness about the issue of cyberbullying, a potential
danger in the cyber millennium. Theoretically, this research bridges
the gap between existing research on school bullying and friend-
ship and online measures of similar constructs; together, these areas
facilitate greater understanding of the relationship between online
friendship and the broader well-being of children. Social experi-
ences in MMOGs are important to the social development of chil-
dren growing up in wired societies, and online friendship explains
unique variance in the well-being of students over everyday life
social experiences, implying that in future studies, online social ex-
periences cannot be overlooked. Despite the fact that online friend-
ship is something virtually based, the positive aspects of online
friendships are real for the players. This also suggests that it is im-
portant to have a balanced view of both the positive and negative
associations of online game playing in relation to the psychological
development of children and early adolescents.

180	LEUNG AND MCBRIDE-CHANG

         Appendix A. Items of Cyber-Victimization Scale and Cyberbullying Scale
              Cyber-victimization scale                                  Cyberbullying scale
1. Others gossip or say mean things about me in        1. I gossip or say mean things about others in
   online games/on the internet.                          online games/on the internet.
2. Others say “If you don’t do what I say, I will stop 2. I say “If you don’t do what I say, I will stop liking
   liking you” in online games/on the internet.           you” to others in online games/on the internet.
3. Others get mad at me, then they ignore or stop      3. I get mad at others, then I ignore or stop talking
   talking to me in online games/on the internet.         to others in online games/on the internet.
4. Others steal my account or my belongings            4. I steal others’ account or belongings (e.g.,
   (e.g., money, weapons) in online games/on the          money, weapons) in online games/on the
   internet.                                              internet.
5. Others pretend to be my friends as a kind of        5. I pretend to be others’ friends as a kind of
   revenge in online games/on the internet.               revenge in online games/on the internet.
6. Others tell bad stories about me, saying that I am 6. I tell bad stories about others, saying that others
   a bad kid, in online games/on the internet.           are bad kids, in online games/on the internet.
7. Others try to get others to dislike me because      7. I try to get others to dislike someone else
   they are angry at me in online games/on the            because I am angry at them in online games/on
   internet.                                              the internet.
8. Others say that they would help me but then         8. I say that I would help others but then betray
   betray me in online games/on the internet.             others in online games/on the internet.

References
Barnett, J., & Coulson, M. (2010). Virtually real: A psychological perspective on mas-
       sively multiplayer online games. Review of General Psychology, 14, 167–179.
Bennerstedt, U., Ivarsson, J., & Linderoth, J. (2012). How gamers manage aggres-
       sion: Situating skills in collaborative computer games. Computer Supported
       Collaborative Learning, 7, 43–61.
Bentler, P. (1992). EQS program manual. Los Angeles: BMDP Statistical Software.
Bernan, T., & Li, Q. (2005). Cyber-harassment: A new method for an old behavior.
       Journal of Educational Computing Research, 32, 137–153.
Berndt, T. J., & Miller, K. A. (1993). The assessment and correlates of adolescents’ friend-
       ships. Unpublished manuscript, Prudue University.
Bjorkqvist, K., Lagerspetz, K.M.J., & Osterman, K. (1992). Direct & indirect aggres-
       sion scales. Retrieved May 29, 2008, from http://www.vasa.abo.fi/svf/up/
       kaj.htm.
Blais, J. J., Craig, W. M., Pepler, D., & Connolly, J. (2008). Adolescents online: The
       importance of internet activity choices to salient relationships. Journal of Youth
       and Adolescence, 37, 522–536.
Bond, L., Wolfe, S., Tollit, M., Butler, H., & Patton, G. (2007) A comparison of the
       gatehouse bullying scale and the peer relations questionnaire for students in
       secondary school. Journal of School Health, 77, 75–79.
Breakthrough Youth Research Archives. (2003). Youth Survey on Cyberrisk. Retrieved
       May 29, 2008, from http://www.breakthrough.org.hk/ir/researchlog.htm.

GAME ON?                                                                             181

Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K.
      Bollen & J. Long (Eds.), Testing structural equation models (pp. 136–162). New-
      bury Park, CA: Sage.
Buhrmester, D. (1996). Need fulfilment, interpersonal competence, and the develop-
      mental contexts of early adolescent friendship. In W.H. Bukowski, A. F. New-
      comb, & W.W. Hartup (Eds.), The company they keep: Friendship in childhood and
      adolescence (pp. 158–185). New York: Cambridge University Press.
Bukowski, W. M., Boivin, M., & Hoza, B. (1994). Measuring friendship quality dur-
      ing pre-and early adolescence: The development and psychometric proper-
      ties of the Friendship Qualities Scale. Journal of Social and Personal Relation-
      ships, II, 471–484.
Bukowski, W. M., Sippola, L. K., & Boivin, M. (1995, March). Friendship protects
      “at risk” children from victimization by peers. In J. M. Price (Chair), The role
      of friendship in children’s developmental risk and resilience: A developmental psy-
      chopathology perspective. Symposium conducted at the biennial meeting of the
      Society for Research in Child Development, Indianapolis, IN.
Census and Statistics Department. (2009). Distribution of educational attainment of
      population aged 15 and over. Retrieved January, 11, 2011, from http://www.
      censtatd.gov.hk/FileManager/EN/Content_809/education.pdf.
Chan, E., &Vorderer, P. (2006). Massively multiplayer online games. In P. Vorderer
      & J. Bryant (Eds.), Playing video games (pp. 77–113). Mahwah, NJ: Lawrence
      Erlbaum Associates.
Cole, H., & Griffiths, M. D. (2007). Social interactions in massively multiplayer on-
      line role-playing gamers. CyberPsychology & Behavior, 10, 575–583.
Colwell, J., Grady, C., & Rhaiti, S. (1995). Computer games, self-esteem, and gratifi-
      cation of needs in adolescents. Journal of Community and Applied Social Psychol-
      ogy, 5, 195–206.
Colwell, J., & Kato, M. (2003). Investigation of the relationship between social isola-
      tion, self-esteem, aggression and computer game play in Japanese adoles-
      cents. Asian Journal of Social Psychology, 6, 149–158.
Commission on Youth. (2008). Executive Summary. Retrieved June 26, 2010, from
      http://www.coy.gov.hk/eng/report/it.htm
Cowie H. (2002) Not bystanding but standing by: Strategies for pupils to cope with
      bullying. New Era in Education, 83, 41–43.
Coyne, S. M., Padilla-Walker, L. M., Stockdale, L., & Day, R. D. (2011). Game on…
      girls: Associations between co-playing video games and adolescent behav-
      ioral and family outcomes. Journal of Adolescent Health, 49, 160–165.
Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-
      psychological adjustment. Child Development, 66, 710–722.
Dehue, F., Bolman, C., & Vollink, T. (2008). Cyberbullying: Youngsters; experiences
      and parental perception. CyberPsychology & Behavior, 11, 217–223.
Demir, M., & Urberg, K. A. (2004). Friendship and adjustment among adolescents.
      Journal of Experimental Child Psychology, 88, 68–82.
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life
      scale. Journal of Personality Assessment, 49, 71–75.
Digital Communities. (2007). Key insights on youth in Shanghai, Hong Kong and the
      U.S. revealed in comparison study. Retrieved June 26, 2010, from http://www.
      govtech.com/dc/articles/237784