Content uploaded by Vivian Hsueh Hua Chen
Author content
All content in this area was uploaded by Vivian Hsueh Hua Chen on Mar 31, 2015
Content may be subject to copyright.
This article was downloaded by: [Nanyang Technological University]
On: 13 March 2015, At: 00:25
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Click for updates
Behaviour & Information Technology
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/tbit20
Group identification as a mediator of the effect of
players’ anonymity on cheating in online games
Vivian Hsueh Hua Chena & Yuehua Wub
a Wee Kim Wee School of Communication and Information, Nanyang Technological University,
31 Nanyang Link, 637718, Singapore
b School of Media and Design, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai
200240, Shanghai, China
Accepted author version posted online: 25 Oct 2013.Published online: 06 Dec 2013.
To cite this article: Vivian Hsueh Hua Chen & Yuehua Wu (2013): Group identification as a mediator of the effect of players’
anonymity on cheating in online games, Behaviour & Information Technology, DOI: 10.1080/0144929X.2013.843721
To link to this article: http://dx.doi.org/10.1080/0144929X.2013.843721
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
Behaviour & Information Technology, 2013
http://dx.doi.org/10.1080/0144929X.2013.843721
Group identification as a mediator of the effect of players’ anonymity on cheating in online games
Vivian Hsueh Hua Chena∗and Yuehua Wub
aWee Kim Wee School of Communication and Information, Nanyang Technological University, 31 Nanyang Link, 637718, Singapore;
bSchool of Media and Design, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai 200240, Shanghai, China
(Received 13 February 2013; accepted 6 September 2013)
This study aims to add to the discussion about the applicability of the classical deindividuation theory and social identity
model of deindividuation effects (SIDE) in explaining online behaviours. It explores the effect of anonymity in facilitating
social influence of group identity in online game cheating. A nationally representative survey was conducted face to face.
Results from the survey administered in Singapore confirm predictions derived from the SIDE and challenge the classical
deindividuation theory. Specifically, it was concluded that the frequency of gaming with online strangers (anonymous gaming)
significantly predicted the frequency of cheating in online games. The effect of anonymity on game cheating was found to be
significantly mediated by the group identification with online gaming communities/groups. Gender differences were found.
Male gamers cheated more frequently than female gamers. Female gamers are more likely to cheat as a consequence of group
identification than male gamers. Implications and future research are discussed.
Keywords: game cheating; social identity model of deindividuation effects; virtual community
1. Introduction
As the Internet becomes widely accessible, worldwide
computer-mediated communication (CMC) has emerged as
a routine means of communication and social interaction.
Compared with traditional face-to-face communication,
CMC is characterised by a number of new features, includ-
ing reduced social cues and social anonymity (Kiesler et al.
1984). The psychological and social implications of the (rel-
ative) anonymity afforded by the Internet have been widely
discussed and examined (see Christopherson 2007, for a
review). Specifically, some researchers have sought to use
anonymity to account for online problematic behaviours,
such as ‘flaming’ and ‘grief gaming’ (Thompsen 1996,
Moor 2007,Chen et al. 2009). To add to the relevant
literature in the field, this study purports to examine the
effect of anonymity on one relative under-explored prob-
lematic behaviour – game cheating (Kimppa and Bissett
2005,Kücklich 2008). Game cheating has been preva-
lent ever since the invention of video games (Consalvo
2005). Cheating in games can ruin the ‘fairness’ of
gaming and affect the interests of other game players
(Kimppa and Bissett 2005,Yan and Randell 2009)as
well as game companies. Rampant game cheating often
leads to bad game reputation and loss of game players
or subscribers in the case of multiplayer online games
(Zetterström 2005,Duh and Chen 2009). Subsequently,
it affects game companies’ profits. Considering the time
and effort invested in both game development and game
play, cheating deprives game players and game companies
of their labour and can be regarded as a moral offence
(Kimppa and Bissett 2005). The prevalence of cheating in
multiplayer computer games (Consalvo 2007,Webb and
Soh 2007) therefore poses new moral and legal questions for
both the gaming and broader Internet community (Kimppa
and Bissett 2005,Zetterström 2005,Parker 2007). Hence,
game cheating is an important topic for research.
The geographic separation and anonymity afforded by
online games, among other unique features of networked
computer games, is believed to be one of the reasons for an
increase in the possibility and frequency of game cheating
behaviours (Parker 2007). However, this argument lacks
solid empirical evidence. Based upon the social identity
model of deindividuation effects (SIDE) (Reicher et al.
1995), the present study looks at how anonymous game
playing influences game cheating and examines the under-
lying mechanism of this effect. Specifically, it takes into
account the effect of group identification in the relationship
between anonymous gaming and game cheating.
2. Literature review, theoretical framework, and
hypotheses
2.1. Game cheating
Cheating can be generally understood as breaking rules
(Schwieren and Weichselbaumer 2010). However, rule-
breaking in games cannot always be easily defined. There
∗Corresponding author. Email: snowvc@gmail.com
© 2013 Taylor & Francis
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
2V.H.H. Chen and Y. Wu
are no clear-cut rules applicable to all games, game players,
and game communities. Different people view and define
game cheating in distinct ways (Consalvo 2005). While
some players find certain game practices acceptable, others
may label those same practices as cheating. What consti-
tutes cheating in games is often debatable depending on
different actors, technologies (Botvich et al. 2010), and sit-
uations involved. In this study, game cheating is defined
as strategies that a player uses to gain an unfair advantage
over his/her peer players or to achieve a target which is not
supposed to be achieved according to the game rules or at
the discretion of the game operator (Yan and Randell 2009).
Despite increasing research attention to game cheating,
the empirical investigation of cheating is still limited. The
extant literature on game cheating primarily focuses on
methods/classification of cheating (e.g. Kimppa and Bissett
2005,Webb and Soh 2007,Yan and Randell 2009), motiva-
tions to cheat (Consalvo 2007), and ways to combat cheating
(Zetterström 2005,Hu and Zambetta 2008,Botvich et al.
2010). The current literature, however, does not fully spec-
ify the social psychological mechanisms of the cheating
behaviour in video games. To fill this gap, the purpose of
the present study is to explore how the anonymity afforded
by the online environment and relevant online social activ-
ities influence cheating behaviours in networked computer
games. Detailed hypotheses are presented in the following
sections.
2.2. Anonymity, deindividuation, and the social identity
model of deindividuation effects
Anonymity is one of the major characteristics of online
environments (Johnson 1997). Being anonymous online
gives individuals’ opportunities to behave in socially unde-
sirable or even harmful ways without being sanctioned
(Kiesler et al. 1984,Sproull and Kiesler 1986), as in
the cases of online flaming (insulting, swearing, or using
offensive language on the Internet) and game cheating.
Deindividuation theory is one of the major theories
describing the effects of anonymity. Originating from
Le Bon’s (1895, trans. 1947) concept of ‘submergence’,
the theory suggests that immersion and anonymity in groups
might evoke a deindividuated state, which causes a decrease
in self-observation, self-evaluation, and concern for social
comparison and evaluation, leading to weakened inter-
nalised controls and anti-normative behaviours (Zimbardo
1969). The theory of deindividuation has been extended to
the domain of CMC. It was argued that certain features of
CMC, such as anonymity and immersion in the medium,
produce the classic deindividuating conditions of reduced
self-awareness and disinhibition (Reicher et al. 1995).
Hence, researchers have utilised this theory to account for
deviant online behaviours such as grief gaming (Chen et al.
2009).
With regard to online problematic behaviour and the
‘deindividuation’ effects of anonymity within the group,
an alternative explanation is offered by the SIDE (Reicher
et al. 1995). Building on self-categorisation theory (Turner
1987), SIDE argues that the ‘deindividuated’ state evoked
by anonymity does not lead to the loss of individual-
ity or self-awareness; rather, it increases the salience of
social/group identity and hence leads to greater confor-
mity to group norms (Reicher et al. 1995,Postmes et al.
2001). Being anonymous means that there is a lack of indi-
viduating cues for people to be identified as individuals,
which, according to self-categorisation theory, induces a
psychological state of ‘depersonalisation’ (Turner 1987).
Once depersonalised, the self-concept shifts from personal
identity to social identity (Turner 1987,Reicher et al. 1995).
Individuals tend to behave according to how they believe
other people from the same (or higher) social category or
group will behave (Chan 2010).
A meta-analysis of 60 previous studies found that SIDE
received a far stronger empirical support and was a better
predictor of CMC behaviour than classical deindividuation
theory (Postmes and Spears 1998). For example, online
flaming was found to be influenced by social norms within
a group (Postmes et al. 2000).Postmes et al’s (2001)
experimental study also revealed that anonymity increased
the influence of group norms in discussions online. In
brief, rather than displaying uninhibited and anti-normative
behaviours, SIDE predicts that people in anonymous CMC
tend to conform to perceived group norms. It is within this
theoretical framework that the present study attempts to
explore the effect of anonymity on game cheating. The
primary intention for this study is not to fully test the
SIDE itself. Rather, we use this model to help interpret and
understand certain aspects of game cheating behaviour in
massively multiplayer online games.
2.3. Anonymous gaming, group identification, and
game cheating
Previous studies based on SIDE demonstrate that individ-
uals tend to act according to perceived group norms and
expectations when social group identity is salient and fellow
group members are deindividuated in computer-mediated
groups (Postmes and Spears 1998,Postmes et al. 2001,
Chan 2010). Group norms often times are not explicitly
stated as online groups are not formed based on a ‘prede-
fined social structure’ (Postmes et al. 2001). Instead, norms
are frequently inferred from the common behaviours or
predominant attributes of typical group members (Reicher
1987). Past literature has shown that game cheating can be
considered as a norm among video game players. Consalvo
(2007) concluded from her in-depth interviews and a survey
of game players and developers that cheating was a daily
practice for gamers and most players ‘engage in the practice
on a regular basis’ (p. 93), despite the negative connota-
tions associated with the term ‘cheating’. Yee (2006) listed
‘competing unfairly’ as one of the motivations for players
to engage in multiplayer online games. In a similar vein,
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
Behaviour & Information Technology 3
De Simone et al. (2012) argued that a proclivity to cheat is
part of human nature. When given an option, players will
typically cheat in a video game. A common perception of
the online (gaming) environment as a space for experimen-
tation on breaking rules (Consalvo 2005) as well as the
perception that bending rules in video games has almost no
detrimental effect (Kimppa and Bissett 2005), also promotes
cheating behaviours among gamers.
Game cheating, despite being anti-normative, is already
embedded in the normative culture of play in video games.
Many gaming magazines, websites and forums tell players
the exact ways of how to cheat in video games, for example,
offering cheat codes or telling their audiences how to exploit
certain features (see Kücklich 2004). Hence, game cheating
can be considered a predominant attribute of typical gam-
ing community members. In this sense, cheating in games
can be understood as a normative behaviour among gamers
who belong to the same gaming community. It is impor-
tant to clarify the use of the term ‘normative’ in describing
game cheating. As mentioned in the previous section, game
cheating is considered anti-normative and problematic in a
general sense. The fact that game cheating is perceived to
be a normative behaviour within a gaming community does
not justify or change the fact that game cheating is still con-
sidered as an anti-normative behaviour in general. In this
sense, game cheating is an anti-normative behaviour that
is being recognised as normal and common within given
gaming communities.
As noted previously, SIDE predicts that anonymity rein-
forces the influence of group norms in online communities.
As a result anonymous gaming could increase the likelihood
of game cheating behaviour, perceived as a group norm,
influencing gamers’ in-game behaviour. In other words,
when people play games under anonymous conditions, they
are more likely to follow what they believe other players are
doing, and performing cheats can be one of them. Therefore,
we propose the following hypothesis.
H1: The more often a person plays games anonymously, the more
often he/she cheats in games.
Salience of social identity (often measured by group
identification) is another major concept in SIDE and SIDE-
based research (Reicher 1984,Postmes et al. 2001,Lee
2006). According to self-categorisation theory (Turner
1987), social influence is cognitively mediated by one’s
self-categorisation as a group member. Empirical studies
also demonstrated that heightened in-group identity pro-
duces greater adherence to group norms (e.g. Mackie 1986).
Ren et al. (2012) conducted a field experiment on a movie-
related online community, where they found that increased
group identification led to more frequent site visits and
fostered stronger identity attachment as compared with
other conditions. Building upon self-categorisation theory,
SIDE argues, with empirical support, that anonymity in the
group can enhance identification with the group and thereby
increase conformity to group norms. In other words, group
identification mediates the effect of anonymity on individual
behaviours (Postmes et al. 2001,Lee 2006). In the context
of playing games within a group, cheating can be consid-
ered as a kind of normative behaviour or a group feature.
Hence, one can argue that anonymous gaming can foster or
enhance a player’s social in-group identification in an online
gaming community/group. The salience of gaming group
identity is a mediator of the effect of anonymous gaming
on game cheating. Specifically, we will test the following
hypotheses.
H2: The more often a person plays games anonymously, the more
salient the group identification he/she demonstrates in a game
community.
H3: Gaming group identification is a mediator of the relationship
between anonymous game playing and game cheating.
3. Method
Studies based on SIDE and deindividuation theory have
been primarily conducted in laboratory conditions involv-
ing ‘zero-history groups discussing hypothetical scenarios
with no real-life consequences’, where ‘experimental con-
trol came at the cost of limited ecological validity’ (Lee
2006, p. 443). Considering the possible shortcomings of
experimental research, SIDE researchers emphasise the
need to apply the theory and assess its use in more real-
istic contexts, which is to study anonymous conditions that
occur in naturalistic settings on the Internet instead of con-
ditions created in a lab (Postmes et al. 1998). Although
the self-reporting survey approach is also criticised for
its limitations in validity due to potential response bias,
some studies have found a relatively high rate of self-
reported deviant behaviours such as drug use (Zanis et al.
1994). Given the right questionnaire design and context that
assures anonymity and/or confidentiality, survey respon-
dents can give honest reports about criminal or various
deviant behaviours (see Farrington 1973, Harrison 1994).
Thus, we could reasonably argue that a survey design allows
a closer access to real-world attitudes and behaviours than
experimental design (Northrup 1996). This improvement
in ecological validity is an advantage over studies based
on experiments. Hence, this study employed a survey to
examine the influence of anonymity and salience of group
identification in gaming on game cheating behaviours.
3.1. Sampling
The data for this study were collected as part of a larger
survey study of online game behaviours of adolescents in
Singapore. Data collection took place over a period of three
weeks in May and June of 2009. Based on the data pro-
vided by the Singapore Department of Statistics most recent
census in 2008, stratified sampling was employed to recruit
respondents according to the population distribution in Sin-
gapore. The number of samples collected from each Mass
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
4V.H.H. Chen and Y. Wu
Rapid Transit station and neighborhood center was repre-
sentative of the population ratio of the nearby living area
to the entire population. A total of 1400 paper–pencil sur-
veys were randomly distributed at all neighborhood centers
and the nearest Mass Rapid Transit railway stations in Sin-
gapore. A set of screening questions was used to identify
qualified respondents. Respondents had to be 18 years of
age or below, a Singaporean citizen or permanent resident,
had to live within the target residential area, and play video
games. A total of 961 valid responses (69% response rate)
were collected for data analysis. Participants received no
incentives.
3.2. Measures
Aside from demographic information, the survey included
measures for the three most played games, frequency of
gaming, anonymity, gaming group identification, and game
cheating. Most of the variables involved in this study were
measured with items using a seven-point Likert scale. All
the seven-point scale items used the same endpoints (1 =
never and 7 =very frequently).
To measure the dependent variable of game cheating,
respondents were asked to self-report on the frequency
they used cheats or hacked a game, and ‘used mods or
other player-generated code that changed something in the
game’. Mean scores of these two items were obtained to rep-
resent respondents’ game cheating frequency (Cronbach’s
α=.63).
Zimbardo (1969) defined anonymity as the inability of
others to identify or single out an individual such that the
individual cannot be evaluated, criticised, judged, or pun-
ished. A common feature of online communication is the
relative anonymity of contact with others, especially dur-
ing initial interactions. The lack of visual and social cues
in the online environment provides opportunities of pre-
tense and deceit even with real-life friends, which arguably
maintains a certain level of anonymity. However, people
may also knowingly play games with real-life friends over
the Internet. Taking both conditions into consideration, the
extent of anonymity during game play is measured by the
frequency of playing games with strangers online. The more
frequently a gamer plays games with complete strangers, the
more frequently they are anonymous in game. Specifically,
respondents were asked to rate the frequency with which
they played with people they only met inside the online
games. Even if gamers later develop close relationships
with strangers whom they first met in games online, the
lack of identifiable cues may still persist. Following Tanis
and Postmes (2008), ‘anonymity’ in this study has a lim-
ited meaning, pertaining to gamers who play online games
anonymously with others not in their interpersonal real-life
social networks, but who belong to the same overarching
gaming community.
Gaming group identification is measured by respon-
dents’ involvement in gaming communities or groups.
Group identification here can be equated with group
salience, common in SIDE literature.1Specifically, a four-
item scale was used to gauge the salience of group identifi-
cation. Respondents were asked to report the frequency of
(1) playing games as a part of a guild or other online group
within the game, (2) reading or visiting websites, reviews
or discussion boards related to the games they play, (3)
writing or contributing to websites, reviews or discussion
boards related to the games they play, and (4) organising or
managing game groups or guilds. (Cronbach’s α=.79).
4. Data analysis and results
4.1. Descriptive statistics
During data screening, unreasonable input, such as people
reporting gaming for 24 hours a day and 7 days a week, and
people reporting gaming for more than 7 days a week, 24
hours a day and missing data were excluded. A total number
of 941 cases were kept for final analyses.
Descriptive statistics showed that the participants in this
study aged from 13 to 18, with the mean age of 16 and the
16–18 age range being the largest group (65%). The sam-
ple had more males (57%) than females (43%). In terms
of ethnic background, respondents were primarily Chinese
(81%), with Malays (11%) and Indians (5%) constituting
the next two major ethnic groups. This generally reflects the
population makeup in Singapore. The highest level of edu-
cation attained was secondary school (55%), junior college
(23%), polytechnic (17%), institute of technical education
(4%), and university (2%). The average gaming time was
14.15 hours (SD =14.2) per week. All participants in this
study played games.
In terms of cheating in games, 30% of the participants
reported never cheating in games, the majority (70%) of
gamers cheated at least occasionally. The mean score on
game cheating was 2.80 (SD =1.80). Seventeen per cent
of the participants fell into the category of frequent cheaters
(score of 5 or above). Consistent with the existing literature
(Consalvo 2007,Webb and Soh 2007), these results indicate
that game cheating is a common behaviour among young
gamers in this study. Moreover, the researchers conducted
six focus group sessions to study game cheating with late
adolescents nine months after data collection for this survey
study had been done. Participants in the focus group were
experienced gamers who were familiar with game cheating.
Their age ranged from the age of 17 to 29 (M=23) and their
gaming experiences were between 1 and 20 years (M=
10.5). Out of 29 participants, 18 indicated that their friends
and they themselves cheated very frequently, 2 said they
did it sometimes, 5 said they did it infrequently, and 3 said
they had never cheated. The focus group result also showed
that all participants viewed game cheating as something
‘everyone is doing’ and ‘If you don’t do it, you will lose out’.
This further validated our assumption that game cheating is
a group norm in the gaming community.
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
Behaviour & Information Technology 5
Path c’ (c, H1)
Path a (H2) Path b
Group
identification
Game
cheating
Anonymity
Figure 1. Diagram of paths in the mediation model of game cheating (H3).
The mean anonymity score was 3.43 (SD =2.10) out
of 7. Twenty-nine percent of the participants reported that
they never played games with people whom they had met
online for the first time (complete strangers), whereas 33%
played with complete strangers quite frequently (scoring 5
or above on the 1–7 frequency scale). The mean gaming
group identification score was 3.20 (SD =1.50) out of 7.
An additional factor of interest for this study was gender.
Most studies found significant gender differences (Jensen
et al. 2002, Bianco and Schmelz 2007) in the incidence
of cheating and unethical behaviours in general. In this
study, gender was first controlled when regression analy-
ses were conducted to test the hypotheses. Subsequently,
moderated mediation analysis was conducted to establish
whether gender acts as a significant moderator effect.
4.2. Regression analyses and results
Regression analyses were performed to test the hypotheses.
H1 and H2 were tested first as they are the prerequisites
of H3. In a series of regression analyses, a path analysis –
displayed in Figure 1– was conducted to establish the medi-
ating role of group identification in the relation between
the predictor variable (anonymous gaming) and the out-
come variable (game cheating) (Baron and Kenny 1986,
Frazier et al. 2004). Correlations between the variables in
the regression analysis are reported in Table 1.
The first regression analysis had to establish whether
there was a significant relationship between anonymous
gaming and game cheating (H1) (see Path c in Figure 1).
Several researchers (Zhao et al. 2010,Rucker et al. 2011
and references therein) question the necessity of this step
as significant mediation effects can be found even when the
regression coefficient of Path c is non-significant.2Nev-
ertheless, the authors find it useful to report all of the
significant regression coefficients in the analysis. The sec-
ond regression analysis was needed to establish whether
there was a significant relationship between anonymous
gaming and gaming group identification (the mediator) (H2)
(see Path a in Figure 1). In the third multiple regression
analysis, the outcome variable was regressed on both the
predictor (Path c) and the mediator. To validate the medi-
ation model, the mediator of group identification must be
significantly related to the outcome variable of game cheat-
ing (see Path b in Figure 1) after controlling for the effect of
anonymous gaming on game cheating. According to Frazier
et al. (2004), a complete mediation would be established if
the effect of anonymous gaming on game cheating (c) did
not differ from zero after controlling for the effect of group
identification. If group identification was a partial medi-
ator, the relation between anonymous gaming and game
cheating (c) would be significantly smaller (but still be
greater than zero) when group identification was included
than when the outcome was only regressed on anonymous
gaming (c). Rucker et al. (2011), however, caution that full
mediation can be established only when all possible media-
tors and suppressor variables are identified and included in
the model. Therefore, no claims of full mediation are made
here. A bootstrap interval (Preacher and Hayes 2004) was
calculated to test the significance of the mediated effect.
Table 2shows the results of testing the three hypotheses
delineated above. Gender as a control variable was found to
be significant in all the three regression analyses, suggesting
a significant gender effect on gamers’ cheating behaviour.
Specifically, male gamers cheated significantly more
often than female gamers and had a higher score for
Table 1. Correlation table for variables in regression analysis.
Anonymous Gaming group Game
Variables Gender gaming identification cheating
Anonymous gaming −.24∗∗
Gaming group identification −.26∗∗ .42∗∗
Game cheating −.20∗∗ .23∗∗ .40∗∗
∗∗p<.01.
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
6V.H.H. Chen and Y. Wu
Table 2. Testing mediation model of game cheating using multiple regression.
Testing steps in mediation model bSE bβFdf regression df residual Adjusted R2
Step 1 (Path c) 36.11∗∗ 2 924 .07
Outcome: game cheating
Predictor: anonymous gaming .16 .03 .19∗∗
Control: gender −.55 .12 −.15∗∗
Step 2 (Path a) 129.79∗∗ 2 924 .22
Outcome: group identification
Predictor: anonymous gaming .29 .02 .40∗∗
Control: gender −.48 .09 −.16∗∗
Step 3 (Path b and c) 69.63∗∗ 3 923 .18
Outcome: game cheating
Mediator: group identification .46 .04 .38∗∗
Predictor: anonymous gaming .03 .03 .04
Control: gender −.34 .11 −.09∗∗
∗∗p<.01.
group identification. Meanwhile, statistical results fully
supported the three hypotheses and the mediation model.
First, anonymous gaming was found to be a significantly
positive predictor of online game cheating behaviour (con-
trolling for gender) (β=.19, t=5.79, p<.01), indicating
that the more often a person played games with people
he/she met online for the first time (online strangers),
the more often he/she cheated in games. Hence, H1 was
supported. In Step 2, gaming group identification (the
hypothesised mediator) was regressed on anonymous gam-
ing, which again revealed a significant regression coefficient
(β=.40, t=13.51, p<.01), suggesting that the fre-
quency of anonymous gaming was positively associated
with gamers’ group identification (controlling for gender),
which provides support for H2.
In Step 3, game cheating frequency was regressed on
both anonymous gaming and gaming group identification.
Gaming group identification was found to be a significant
predictor of game cheating (β=.38, t=11.26, p<.01)
after controlling for anonymous gaming and gender (Path
b), whereas the effect of anonymous gaming was not sig-
nificant (β=.04, t=1.07, p=.29, ns) (controlling for
group identification and gender), which means Path cdid
not differ from zero. Bootstrap 95% CI for the indirect
effect of anonymous gaming on the frequency of cheat-
ing was [0.10, 0.17] indicating a significant mediation
effect of gaming group identification. This result shows
that gaming group identification mediated the relationship
between anonymous gaming and online game cheating was
supported. Therefore, H3 is supported. The percentage of
variance explained by the path model (100 ×R2) with only
anonymous gaming as a predictor of game cheating (7%)
significantly increased when gaming group identification
was added as a predictor (18%).
As the effect of gender was quite pronounced in the
model, we decided to run a moderated mediation regres-
sion analysis using the PROCESS for SPSS macros (Hayes
2013). The result of this analysis is summarised in Table 3.
It can be concluded that gender does not significantly
moderate relationships between the variables in the model
Table 3. Moderated mediation model of game cheating.
Testing steps in mediation model bSE bt Fdf regression df residual Adjusted R2
Path a 86.87 3 923 .22
Outcome: group identification
Predictors: anonymous gaming gender
anonymous gaming ×gender
.27 .03 9.78∗∗
−.62 .17 −3.71∗∗
.04 .04 1.01
Paths b and c42.63 5 921 .19
Outcome: game cheating
Mediators: group identification group
identification ×gender
.39 .05 7.61∗∗
.16 .08 1.95
Predictors: anonymous gaming gender
anonymous gaming ×gender
.04 .04 1.15
−.72 .27 −2.64∗∗
−.03 .06 −0.59
∗∗p<.01.
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
Behaviour & Information Technology 7
in question. Only one interaction, group identification ×
gender, approaches marginal significance (p=.052). This
means that females are more likely to cheat as a consequence
of group identification. Otherwise, gender affects the means
of the variables in the model (females are less likely to
identify with a group, game with strangers, and cheat), but
not the paths between them.
5. Discussion and conclusion
The current study explored the effect of anonymity and
group identification on game cheating behaviours. The
result fully supported the three hypotheses. Specifically, it
was found that gaming with strangers online (anonymous
gaming) significantly increased cheating practices in games
(H1). Group identification was found to be stronger among
gamers who played games more frequently with people they
met online for the first time (H2). The effect of anonymity
on game cheating, however, was found to be mediated by
group identification within gaming communities (H3).
The results hence are consistent with the assumption of
the current study that SIDE accounts better for incidence of
cheating in online games than the classical deindividuation
theory. It is also in congruence with previous findings by
SIDE researchers (Postmes et al. 2001). That is, people in an
anonymous gaming situation tend to conform to perceived
group norms and follow normative behaviours (cheating in
this study) due to the mediation of online gaming group
identification. The reduction of individual visibility as a
result of anonymous gaming accentuates the process of
depersonalisation, and further amplifies the perception that
the individual belongs to a particular social group. Individ-
uals will tend to judge themselves based on stereotypical
group features and behave accordingly (Postmes et al.
1998).
Furthermore, the findings also support another assump-
tion of the current study that game cheating is a normative
behaviour/culture within the online gaming community.
Although previous studies on gaming argued strongly for
this assumption, this study provides the necessary empiri-
cal evidence. The findings also provide further explanations
as to why and how gamers cheat in online games, con-
tributing to the current game cheating literature. In the
context of online games, maintaining anonymity is com-
mon for players. When players immerse themselves in the
gaming community/culture anonymously and identify with
the gaming community, they tend to cheat more frequently
in games. Therefore, over time, game cheating becomes
a common practice within the online gaming commu-
nity, affecting even newly joined game players/community
members. This is not solely due to anonymity but also
the influence of the online gaming community culture
as discussed earlier. This explains how and why players
may cheat in video games, and how anonymity can affect
cheating behaviours from both SIDE and deindividuation
perspectives.
The gender differences found in this study fit with previ-
ous research findings. Females tend to cheat less than males,
in games and in real life. Jensen et al. (2002) found that
male students were more accepting towards lying to their
parents and also lied more to their parents than did female
students. In Calabrese and Cochran’s (1990) study of aca-
demic .cheating in American high schools, it was discovered
that girls were less likely to cheat than boys. However, they
had a stronger tendency to cheat, as compared with boys,
if they were helping others to succeed whereas males were
more likely to cheat for personal success. One plausible
explanation for this finding derived from previous research
(Evans et al. 1993) is that females, as compared with males,
tend to be more concerned about their interpersonal relation-
ships. Hence, females worry more about the social sanction
for a problematic behaviour than do males. If the behaviour
in question has social merit (e.g. better relationship qual-
ity), females will be more likely than males to engage in it.
This explanation also strengthens the argument put forth by
SIDE scholars. With the emphasis on social relationships,
females are more likely to be involved in their communities
and establish a shared social identity than males. Conse-
quently, females are more likely than males to follow a
stereotypical group behaviour.
Postmes and Spears’s (2002) study, also based on SIDE,
provides a slightly different but related perspective. They
argued that anonymity offered by the online environment
was associated with greater gender differences than in
real life because online group members cannot individu-
ate each other. When group members are unable to identify
each other, self-stereotyping occurs. That is, individuals
behave according to not only group stereotypes but also
gender stereotypes. Hence, females may exhibit commonly
recognised female behavioural traits, such as emphasis on
relationships with group members and adherence to group
norms.
Besides being in agreement with current research on
gaming, this study is innovative in several ways. First,
the findings provide new insights into and the underlying
mechanisms of game cheating. It established a relationship
between a form of socially undesirable online behaviour
(game cheating) and participants’ identification with online
gaming groups. This is the first study that establishes such a
relationship. It not only extends the application of SIDE to a
new domain online gaming, but also contributes to gaming
literature by providing an explanation for game cheating
behaviours online from a SIDE perspective. Second, most
of SIDE research has been focused on experimental stud-
ies in lab settings. Only a handful of studies on SIDE, as
mentioned in previous sections, were conducted in field set-
tings and even those were experimental. The current study
utilises self-report survey in the field setting (online gam-
ing groups) and provides evidence that SIDE claims can
still be verified. Third, for researchers who are interested in
studying deviant or anti-social behaviours, the current study
demonstrates that online video gaming is a good testing
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
8V.H.H. Chen and Y. Wu
ground for their theories. The online gaming community is
easily accessible and highly dynamic. In such a research
context, participants can be observed continuously without
intrusion. Gamers are typically willing to report and explain
their (deviant) behaviours in video games they play, espe-
cially if the researcher(s) are part of the gaming community.
To conclude, this study shows that SIDE is applicable for
explaining game cheating behaviour online. It shows that
deviant behaviours online such as game cheating are largely
influenced by the online social groups people feel they
belong to. An online group, despite its fluid, unstable and
imaginary nature, is powerful in constructing and changing
its members’ attitudes and views on behaviours. Hence, a
behaviour that is perceived as problematic and deviant can
be reconstructed with a different interpretation.
6. Limitations and future research
One caveat is that despite the increased ecological valid-
ity offered by this study’s survey of real online gaming
experiences, the method of recalled self-report has its
limitations. The prevalence of game cheating has been
reported in the literature (Consalvo 2005,2007) and can
be considered as a normative behaviour within the gaming
community. However, no existing study, including this one,
can offer definitive conclusions as to whether game cheat-
ing is regarded as a positive or negative behaviours within
gaming communities. In a self-report survey, it is possi-
ble that game cheating can be regarded as an inappropriate
behaviour leading to underreporting or denial among par-
ticipants. Conversely, over-reporting is also likely, if game
cheating is held in high regard in the gaming circles of the
participants. It is also possible that in different kinds of game
and associated gaming communities, certain kinds of cheat-
ing behaviours can be regarded as positive and others as
negative. Future research should take this into consideration
in the questionnaire design.
Since the theoretical basis for the current study is SIDE,
psychological and individual factors that might influence
cheating behaviours, such as personality traits and moral
values, were not included. Future studies may investigate
individual characteristics or moral-value-related variables.
In addition, it is possible that the salience of group iden-
tification mediates the effect of anonymity because it is
correlated with other variables that are the ‘true’ mediators,
which has also been noted by SIDE researchers (Postmes
et al. 2001). Furthermore, in the context of this study, game-
related variables are highly relevant. Some research has
indicated that context and skill can influence a person’s
cheating behaviour (Schwieren and Weichselbaumer 2010).
Hence, game-related factors such as game types, gaming
history, gaming skills, and gaming context should be consid-
ered in future studies. By including factors identified above,
future studies may be able to explain more variance in game
cheating than the current study (18%).
As part of a larger study of online gaming behaviour
of adolescents in Singapore, this study only used an ado-
lescent sample. Although video games are often regarded
as an activity for adolescents or children, computer gam-
ing has become a popular recreational activity for people
of all ages worldwide. Griffiths et al. (2004) found that
there are significant differences in gaming habits and pref-
erences between adolescent and adult gamers. Therefore,
adult game cheating behaviours might differ from those of
adolescents. Hence, to gain an unbiased and complete pic-
ture of game cheating behaviours, adult gamers should also
be studied in the future.
Lastly, the correlational nature of the mediation analy-
ses makes the inferences less strong than when the process
is manipulated directly in experimental research. Conse-
quently, an online gaming experiment can be an effective
method for follow-up studies.
Funding
This study was funded by National Research Foundation in
Singapore [grant number NRF2008IDM-IDM001-014 & 016].
Notes
1. The overlap and distinction of the two terms is beyond the scope of
this paper. Please refer to Turner (1999) and Leach et al. (2008) for
detailed discussion on this issue.
2. We are grateful to an anonymous reviewer for drawing our attention
to the alternatives to Baron and Kenny’s classical model of mediation
analysis.
References
Baron, R.M. and Kenny, D.A., 1986. The moderator-mediator
variable distinction in social psychological research: con-
ceptual, strategic and statistical considerations. Journal of
Personality and Social Psychology, 51 (6), 1173–1182.
Botvich, D., et al., 2010. Integrating players, reputation and rank-
ing to manage cheating in MMOGs. Paper presented at the
Fifth International Conference on the Foundations of Digital
Games, Monterey.
Calabrese, R.L. and Cochran, J.T., 1990. The relationship of alien-
ation to cheating among a sample of American adolescents.
Journal of Research & Development in Education, 23 (2),
65–72.
Chan, M., 2010. The impact of email on collective action: a field
application of the SIDE model. New Media & Society, 12 (8),
1313–1330.
Chen, V.H.-H., Duh, H.B.L., and Ng, C.W., 2009. Players who
play to make others cry: the influence of anonymity and
immersion. Proceedings of the international conference on
advances in computer enterntainment technology, Athens,
Greece, 341–344.
Christopherson, K.M., 2007. The positive and negative impli-
cations of anonymity in Internet social interactions: “On
the Internet, Nobody Knows You’re a Dog.” Computers in
Human Behavior, 23 (6), 3038–3056.
Consalvo, M., 2005. Gaining advantage: how videogame players
define and negotiate cheating. Paper presented at the DIGRA
2005: Changing Views, Worlds in Play. Vancouver, Canada.
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
Behaviour & Information Technology 9
Consalvo, M., 2007. Cheating: gaining advantage in videogames.
Cambridge, MA: The MIT Press.
De Simone, J.J., et al., 2012. Is cheating a human function? The
roles of presence, state hostility, and enjoyment in an unfair
video game. Computers in Human Behavior, 28 (6), 2351–
2358.
Duh, H.B.-L. and Chen, V.H., 2009. Cheating behaviors in online
gaming. Lecture Notes in Computer Science, Vol. 5621.
Berlin: Springer, 567–573.
Evans, E.D., Craig, D., and Mietzel, G., 1993. Adolescents’ cogni-
tions and attributions for academic cheating: a cross-national
study. The Journal of Psychology, 127 (6), 585–602.
Farrington, D.P.,1973. Self-reports of deviant behavior: predictive
and stable? The Journal of Criminal Law and Criminology,
64 (1), 99–110.
Frazier, P.A., Tix, A.P., and Barron, K.E., 2004. Testing modera-
tor and mediator effects in counseling psychology research.
Journal of Counseling Psychology, 51 (1), 115–134.
Griffiths, M.D., Davies, M.N.O., and Chappell, D., 2004. Online
computer gaming: a comparison of adolescent and adult
gamers. Journal of Adolescence, 27 (1), 87–96.
Harrison, Lana D., 1994. The validity of self-reported data on drug
use. Journal of Drug Issues, 2 (1), 1–20.
Hayes, A.F., 2013. Introduction to mediation, moderation, and
conditional process analysis. New York, NY: The Guilford
Press.
Hu, J. and Zambetta, F., 2008. Security issues in massive online
games. Security and Communication Networks, 1 (1), 83–92.
Jensen, L.A., et al., 2002. It’s wrong, but everybody does it: aca-
demic dishonesty among high school and college students.
Contemporary Educational Psychology, 27 (2), 209–228.
Johnson, D.G., 1997. Ethics online. Communications of the ACM,
40 (1), 60–65.
Kücklich, J., 2004. Other playthings – cheating in com-
puter games. Paper presented at the Other players con-
ference. Available from: http://itu.dk/op/proceedings.htm
[Accessed16 November 2011].
Kücklich, J., 2008. Forbidden pleasures: cheating in computer
games. In: M. Swalwell and J.Wilson, eds. The pleasures
of computer gaming: essays on cultural history, theory and
aesthetics. Jefferson, NC: McFarland, 52–71.
Kiesler, S., Siegel, J., and McGuire, T.W., 1984. Social psychologi-
cal aspects of computer-mediated communication. American
Psychologist, 39 (10), 1123–1134.
Kimppa, K.K. and Bissett, A.K., 2005. The ethical significance of
cheating in online computer games. International Review of
Information Ethics, 4 (12), 31–38.
Le Bon, G., 1895, trans. 1947. The crowd: a study of the popular
mind. London: Ernest Benn.
Leach, C., et al., 2008. Self-definition and self-investment: a
multi component model of in-group identification. Jour-
nal of Personality and Social Psychology, 95 (1), 144–165.
doi:10.1037/0022-3514.95.1.144.
Lee, E.-J., 2006. When and how does depersonalization increase
conformity to group norms in computer-mediated communi-
cation? Communication Research, 33 (6), 423–447.
Mackie, D.M., 1986. Social identification effect on group polariza-
tion. Journal of Personality and Social Psychology, 50 (4),
720–728.
Moor, P., 2007. Conforming to the flaming norm in the online
commenting situation. Unpublished thesis. University of
Amsterdam.
Northrup, D.A., 1996. The problem of the self-report in sur-
vey research. Institute for Social Research Newsletter,11
(3). Available from: http://www.math.yorku.ca/ISR/self.htm
[Accessed 10 September 2010].
Parker, J., 2007. Cheating by video game participants. Loading...
1 (1). Available from: http://journals.sfu.ca/loading/index.
php/loading/article/viewFile/24/23 [Accessed 22 January
2012].
Postmes, T. and Spears, R., 1998. Deindividuation and anti-
normative behavior: a meta analysis. Psychological Bulletin,
123 (3), 238–259.
Postmes, T. and Spears, R., 2002. Behavior online: does anony-
mous computer communication reduce gender inequal-
ity? Personality and Social Psychology Bulletin, 28 (8),
1073–1083.
Postmes, T., Spears, R., and Lea, M., 1998. Breaching or
building social boundaries? SIDE-effects of computer medi-
ated communication. Communication Research, 25 (6),
689–715.
Postmes, T., Spears, R., and Lea, M., 2000. The formation of
group norms in computer-mediated communication. Human
Communication Research, 26 (3), 341–371.
Postmes, T., et al., 2001. Social influence in computer-mediated
communication: the effects of anonymity on group behav-
ior. Personality and Social Psychology Bulletin, 27 (10),
1243–1254.
Preacher, K.J. and Hayes, A.F., 2004. SPSS and SAS procedures
for estimating indirect effects in simple mediation models.
Behavior Research Methods, Instruments, & Computers,36
(4), 717–731.
Reicher, S.D., 1984. Social influence in the crowd: Attitudinal and
behavioural effects of de-individuation in conditions of high
and low group salience. British Journal of Social Psychology,
23 (4), 341–350.
Reicher, S.D., 1987. Crowd behaviour as social action. In:
J.C. Turner, et al., eds. Rediscovering the social group:
a self-categorization theory. Oxford: Basil Blackwell,
171–202.
Reicher, S.D., Spears, R., and Postmes, T., 1995. A social identity
model of deindividuation phenomena. In: W. Stroebe and M.
Hewstone, eds. European Review of Social Psychology, Vol.
6. Chichester: Wiley, 161–198.
Ren, Y., et al., 2012. Building member attachment in online
communities: applying theories of group identity and inter-
personal bonds. MIS Quarterly, 36 (3), 841–864.
Rucker, D.D., et al., 2011. Mediation analysis in social psychol-
ogy: current practices and new recommendations. Social and
Personality Psychology Compass, 5 (6), 359–371.
Schwieren, C. and Weichselbaumer, D., 2010. Does competition
enhance performance or cheating? A laboratory experiment.
Journal of Economic Psychology, 31 (3), 241–253.
Sproull, L. and Kiesler, S., 1986. Reducing social context cues:
electronic mail in organizational communication. Manage-
ment Science, 32 (11), 1492–1512.
Tanis, M. and Postmes, T., 2008. Cues to identity in online dyads:
effects of interpersonal versus intragroup perceptions on per-
formance. Group Dynamics: Theory, Research, and Practice,
12 (2), 96–111.
Thompsen, P.A., 1996. What’s fueling the flames in cyberspace?
A social influence model. In: L. Strate, R. Jacobson, and S.B.
Gibson, eds. Communication and cyberspace: social inter-
action in an electronic environment. Cresskill, NJ: Hampton
Press, 293–315.
Turner, J.C., 1987. The analysis of social influence. In: J.C.
Turner, et al., eds. Rediscovering the social group: a self-
categorization theory. Oxford: Basil Blackwell, 68–88.
Turner, J.C., 1999. Some current issues in research on social
identity and self-categorization thoeries. In: N. Ellemers,
R. Spears, and B. Doosje, eds. Social identity: context,
commitment, content. Oxford: Blackwell, 6–34.
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
10 V.H.H. Chen and Y. Wu
Webb, S.D. and Soh, S., 2007. Cheating in networked com-
puter games: a review. Paper presented at the DIMEA ’07:
proceedings of the 2nd international conference on digital
interactive media in entertainment and arts, Perth, Western
Australia.
Yee, N. 2006. Motivations for play in online games. CyberPsy-
chology and Behavior, 9, 772–775.
Yan, J. and Randell, B., 2009. An investigation of cheating in
online games. IEEE Security and Privacy, 7 (3), 37–44.
Zanis, D.A., McLellan, A.T., and Randall, M. 1994. Can you trust
patient selfreports of drug use during treatment. Drug Alcohol
Depend, 35 (2), 127–132.
Zetterström, J., 2005. A legal analysis of cheating in
online multiplayer games. Unpublished thesis. Göteborg
University.
Zhao, X., Lynch Jr., J.G., and Chen, Q., 2010. Reconsid-
ering Baron and Kenny: myths and truths about medi-
ation analysis. Journal of Consumer Research, 37 (2),
197–206.
Zimbardo, P.G., 1969. The human choice: Individuation, rea-
son, and order versus deindividuation, impulse and chaos.
In: W.J. Arnold and D. Levine, eds. Nebraska Symposium
on Motivation. Lincoln, NB: University of Nebraska Press,
237–307.
Downloaded by [Nanyang Technological University] at 00:25 13 March 2015
