Computers in Human Behavior 29 (2013) 881887

Contents lists available at SciVerse ScienceDirect

Computers in Human Behavior

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

A process model of cyberbullying in adolescence

Lambros Lazuras a, Vassilis Barkoukis b,, Despoina Ourda b, Haralambos Tsorbatzoudis b
a
South-East European Research Centre (SEERC), Thessaloniki, Greece
b
Department of Physical Education and Sport Science, Aristotle University of Thessaloniki, Greece

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

Article history: Cyberbullying is an emerging form of aggression that utilizes information and communication technolo-
Available online 17 January 2013 gies (ICTs). While cyberbullying incidents attract considerable attention, research on the causes and psy-
chosocial predictors of cyberbullying is still limited. The present study used an integrated theoretical
Keywords: model incorporating empathy, moral disengagement, and social cognitions related to cyberbullying.
Bullying Structured questionnaires were administered to 355 randomly selected adolescents (M = 14.7,
Aggression SD = 1.20). Linear regression analysis showed that social norms, prototype similarity and situational
ICT use
self-efcacy directly predicted cyberbullying expectations. Multiple mediation modelling indicated that
normative inuences mediated the effects of moral disengagement and affective empathy on cyberbul-
lying expectations. These ndings provide valuable information regarding the effect of both distal and
proximal risk factors for cyberbullying in adolescence, highlight the relationship between normative pro-
cesses and moral self-regulation, and set the basis for related educational and preventive interventions.
2012 Elsevier Ltd. All rights reserved.

1. Introduction Because cyberbullying has only recently attracted research

1.1. Cyberbullying
Cyberbullying1 is an emerging form of aggression that takes
place in cyberspace and is utilized by contemporary information
and communication technologies (ICTs). Unlike traditional face-to-
face bullying, cyberbullying provides total anonymity to the aggres-
sor, and can reach a wide audience (e.g., a humiliating video against
another person posted on social networking or le sharing websites
can become visible to millions of web users; Beran & Li, 2007; Pat-
chin & Hinduja, 2006). Most importantly, cyberbullying can have a
signicant psychological impact on the victim, by leading to with-
drawal and social exclusion, lower self-esteem and academic
achievement, or even depression and suicide ideation and attempts
(Hinduja & Patchin, 2010; Juvonen & Gross, 2008; Klomek, Souran-
der, & Gould, 2010; Li, 2007). The rates of cyberbullying range be-
tween 12% and 25% in Europe, USA, and Canada, while the overall
rates of other forms of online aggression may be even higher (Pat-
chin & Hinduja, 2006; Slonje & Smith, 2008; Ybarra & Mitchell,
2004).
Corresponding author. Address: Department of Physical Education and Sport
Science, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece.
E-mail address: bark@phed.auth.gr (V. Barkoukis).
1
The term cyberbullying used in this manuscript is based on Olweus (1993)
denition of bullying, and is therefore dened as intentional and repeated form of
aggression from individuals or groups of people against a victim or a group of victims
utilizing contemporary ICTs (see also Dunkels, Franberg, & Hallgren, 2011; Slonje &
Smith, 2008).
attention (most empirical studies on the subject being published
after 2008), there are still important questions to be answered
and accordingly inform evidence-based preventive strategies (Li,
2007; Patchin & Hinduja, 2006). One such question is whether
cyberbullying can be explained solely by individual characteristics
and traits, or by the interplay between traits and social cognitions
that facilitate behaviour initiation. So far, empirical research has
feed out some relevant traits for cyberbullying, such as empathy,
and has also identied the role of cognitive processes like moral
disengagement, and personal beliefs, including attitudes, norma-
tive beliefs, and demographic characteristics, such as age and gen-
der (Ang & Goh, 2010; Pornari & Wood, 2010; Walrave & Heirman,
2011). Nevertheless, researchers have yet to examine the interplay
among these risk factors, and, accordingly provide an integrated
behavioural model for cyberbullying in young people.
1.2. Empathy
Empathy is a cardinal aspect of human behaviour that facilitates
and eases social interaction by allowing people to identify and
communicate each others emotions (Cohen & Strayer, 1996; Davis,
1994; Preston & de Waal, 2002). Researchers have argued that
empathy should be treated as a relatively stable attribute in a per-
sons life time that may affect different types of social behaviours
(Loudin, Loukas, & Robinson, 2003; Strayer, 1987). Studies have
shown that empathy comprises two rather distinct processes: a
cognitive process reecting ones ability to identify and cognitively
process another persons emotional states, and an affective process

0747-5632/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.chb.2012.12.015
882 L. Lazuras et al. / Computers in Human Behavior 29 (2013) 881887
that facilitates emotional understanding and communication
through an emotional and less cognitively-bound channel, also
termed vicarious emotional sharing (Davis, 1983; Shamay-Tso-
ory, Aharon-Peretz, & Perry, 2009).
In relation to bullying, several studies have shown that lower
levels of empathy are associated with higher frequency of bullying
behaviours in children and adolescents (Bartholow, Sestir, & Davis,
2005; Endresen & Olweus, 2002; Joliffe & Farrington, 2006; Lovett
& Shefeld, 2007; Olweus, 1993). In a similar fashion, recent stud-
ies conrmed that empathy plays an important role in cyberbully-
ing behaviour. Specically, Ang and Goh (2010) showed that both
male and female adolescents with lower empathy levels, reported
higher cyberbullying scores, and Schultze-Krumbholz and Scheit-
hauer (2009) found that both cyberbullying perpetrators and vic-
tims reported lower empathy levels, as compared to individuals
not involved in cyberbullying. In a similar vein, Steffgen, Knig,
Pfetsch, and Melzer (2011) found that cyberbullies had signi-
cantly lower scores on empathy than non-cyberbullies. It is note-
worthy that the aforementioned studies did not employ the same
measures of empathy. Thus, the ndings actually show that the
relationship between cyberbullying and empathy is independent
of the methods used to assess this effect.
1.3. Moral disengagement
In their course of life, individuals engage in behaviours that are
in discord with their moral or personal values. In order to cope with
and resolve this dissonance they cognitively re-process the moral
values attached to the behaviours and accordingly initiate a moral
disengagement mechanism (Bandura, 1986, 1991). This strategy al-
lows the cognitive moralization of actions that would otherwise be
considered immoral or against personal moral norms (Bandura,
Barbaranelli, Caprara, & Pastorelli, 1996; McAlister, Bandura, &
Owen, 2006). Thus, moral disengagement can soothe the mental
discomfort associated with disputes, arguments, and even more ex-
treme forms of aggressive behaviours that may occur in the course
of social interaction. Indeed, several studies have shown that there
is a positive correlation between higher levels of moral disengage-
ment and higher levels of aggressive behaviours (Bandura, 2002;
Bandura et al., 1996). In relation to bullying behaviour, the ndings
are mixed with some studies reporting that moral justication pre-
dicted only traditional bullying but not cyberbullying (Bauman &
Pero, 2011; Perren & Gutzwiller-Helfennger, 2012), whereas oth-
ers have reported a signicant correlation between moral disen-
gagement and both traditional bullying and cyberbullying
(Pornari & Wood, 2010). Given that cyberbullying studies have only
recently emerged, further research is needed in order to establish
the role of disengagement in the process of cyberbullying.
1.4. Attitudes, norms, regret and intentionality
Cyberbullying is dened as a goal-directed behaviour that is in-
tended to hurt others (Patchin & Hinduja, 2006; Pyzalski, 2011).
Cyberbullying may include a wide range of actions, including post-
ing offensive and insulting messages on the web, harassment and
mistreatment with online means (e.g., texting, instant messaging)
altering or hacking personal accounts and information in social
networking sites, and even posting embarrassing videos online,
or creating libellous blogs against someone (Juvonen & Gross,
2008; Li, 2007). At the very least, such actions require some sort
of strategic delegation of time and effort. Thus, intentionality plays
a key role in the occurrence of cyberbullying, and distinguishes
cyberbullying from other more general forms of aggression in ado-
lescence (Pyzalski, 2011; Slonje & Smith, 2008).
Goal intentions, and their psychosocial predictors, are impor-
tant in understanding premeditated behaviours. Research on the
Theory of Planned Behaviour (TPB; Ajzen, 1991, 2002) has shown
that attitudes, social norms, and self-efcacy beliefs explain a great
deal of intention-formation across behavioural domains in adoles-
cence (Conner & Armitage, 1998; Hamilton & White, 2008; McMil-
lan & Conner, 2003), including aggressive acts like peer sexual
harassment and abuse (Li, Frieze, & Tang, 2010). Nevertheless,
the correspondence between intentions and actual behaviour is
far from being perfect (e.g., Webb & Sheeran, 2006). Therefore,
researchers have suggested that the traditional TPB approaches
are enriched with theory-driven variables that can explain specic
behaviours in specic situations and social contexts (Armitage &
Conner, 2001; Conner & Armitage, 1998).
To this end, several studies have shown that the tripartite of
attitudes-social norms-self-efcacy can better predict intentions
and behaviour if additional variables are assessed, such as antici-
pated regret, which reects the feeling of remorse from following
(or abstaining from) a specic course of action (Abraham & Sheer-
an, 2004; Conner, Sandberg, McMillan, & Higgins, 2006). Antici-
pated regret predicts intentions, and strengthens the link
between intentions and behaviour (Abraham & Sheeran, 2004;
Conner & Armitage, 1998; Perugini & Bagozzi, 2001).
Also, considering the role of descriptive norms (i.e., judgments
of frequency and prevalence of target behaviours) over subjective
norms (i.e., perceived social approval of target behaviours), might
further enhance the predictive validity of normative inuences on
intentions (Rivis & Sheeran, 2003a). Normative inuences can be
understood in terms of stored social representations or proto-
types, whereby more favorable evaluations of these prototypes
predict stronger intentions to engage in prototype-relevant
behaviours (Rivis & Sheeran, 2003b). Prototype evaluation has
been studies in the context of TPB (e.g., Norman, Armitage, &
Quigley, 2007; Rivis, Sheeran, & Armitage, 2006), and is also a
main component of the Prototype/Willingness Model which is
used to predict adolescent risk taking (Gerrard, Gibbons, Stock,
Vande Lune, & Cleveland, 2005; Gibbons, Gerrard, Blanton, &
Russell, 1998).
Finally, researchers have argued that the intention concept it-
self needs to be changed in order to better understand adolescent
risk-taking. In particular, instead of asking questions referring to
concrete plans (e.g., I intent to do X), it is advisable to assess inten-
tionality through questions of behavioural expectations (e.g., I ex-
pect to do X), because people often do not expect what they
intend to do, and vice versa (Davis & Warshaw, 1992, p. 392). Un-
like personal planning, therefore, when asked about the perceived
likelihood of performing a target action, adolescents may consider
potential barriers to action, external inuences, and personal skills
and competences; thus, making behavioural expectations more va-
lid predictors of future behaviour, than behavioural intentions (Da-
vis & Warshaw, 1992; Rhodes & Matheson, 2005; Warshaw &
Davis, 1986).
1.5. A process-model approach to cyberbullying
Psychologists often conduct research to establish whether and
to what extent one variable affects another. However, the dis-
covery that two variables are related to each other is only one
small part of the aim of psychology. Deeper understanding is
gained when we comprehend the process that produces the
effect. (Preacher & Hayes, 2008, p. 717)
Preacher and Hayes (2008) assertion is highly relevant to the
study of cyberbullying for the following reasons. Firstly, related re-
search has already identied some psychosocial correlates of
cyberbullying, but we need to put these associations in context
in order to better understand the causal processes and mechanisms
 L. Lazuras et al. / Computers in Human Behavior 29 (2013) 881887
underlying the behaviour in question. Secondly, cyberbullying is a
recent phenomenon, and related theories are still in its infancy.
Therefore, by using theoretical developments in other elds of re-
search (e.g., health behaviour, decision-making, and risk-taking)
we can inform subsequent research and theory development in
cyberbullying. Finally, by understanding the causal mechanisms
of cyberbullying we can accordingly inform evidence-based pre-
ventive strategies. In the present study we suggest that such an
integrative process-model approach will be of benet to cyberbul-
lying research, as it will shed light to the causal process that shape
intentions to engage in cyberbullying. This is highly relevant to
past research on the subject, because such a process-model ap-
proach will integrate previous ndings and provide a theory-dri-
ven model of cyberbullying behaviour.
To this end, the present study employs an integrated theoretical
model to assess the predictors of adolescents cyberbullying inten-
tions. This model pertains to the direct and indirect effects of indi-
vidual traits and self-regulatory processes, and social cognitions
relevant to cyberbullying (e.g., attitudes, norms, self-efcacy,
anticipated emotions). Empathy is seen as a personality trait that
can inuence the decision to engage in cyberbullying directly, or
indirectly, through the effects of more proximal predictors, such
as personal attitudes or social norms. Past research on the theory
of planned behaviour and related theoretical approaches have
shown that individual traits can indeed inuence intentions both
directly and indirectly (e.g., Conner & Armitage, 1998; Fishbein,
2009; Rhodes & Courneya, 2003).
Furthermore, according to Bandura et al. (1996) moral disen-
gagement induces aggressive behaviour indirectly, through aggres-
sion proneness. In this line it could be argued that moral
disengagement could induce cyberbullying behaviour by encour-
aging cyberbullying intentions (seen as eagerness/proneness for
cyberbullying). In this regard, it would be interesting and theoret-
ically relevant to assess whether moral disengagement is associ-
ated directly with cyberbullying intentions, after controlling for
the effects of other predictors of intentionality, such as relevant so-
cial cognitions (e.g., attitudes, norms, and efcacy). The need to
further examine the relationships between moral and social cogni-
tive mechanisms in relation to adolescent bullying behaviour was
also addressed by previous research (e.g., Gini, 2006). Finally, it
would be of interest to assess the unique direct effects of social
cognitions towards cyberbullying. Several studies have shown that
attitudes or social norms inuence cyberbullying intentions and
behaviour, but there are limited studies assessing the effects of
these variables by taking into account the effects of other predic-
tors of intentions (e.g., anticipated regret, prototype similarity).
In line with these assumptions, it was hypothesized that:
H1. Lower empathy scores and higher levels of moral disengage-
ment would predict cyberbullying intentions.
H2. Attitudes, descriptive social norms, prototype favourability
and similarity, anticipated regret, and self-efcacy would directly
predict cyberbullying intentions.
H3. Empathy and moral disengagement would predict cyberbully-
ing intentions indirectly, through the effects of social cognitions.
2. Methods
2.1. Participants
The present study is part of a larger-scale European Project and
presents ndings from a baseline measurement of cyberbullying
883
intentions, empathy, moral disengagement, and social cognitions
in a sample of Greek adolescent students from two randomly se-
lected secondary schools in Athens and Thessaloniki, the two larg-
est cities in Greece (totalling approx. 70% of the Greek population).
Overall, 355 out of 500 (71% response rate) questionnaires were re-
turned. The age of participants ranged between 13 and 17 years
(M = 14.7, SD = 1.20), and 55.5% were females. Permission to con-
duct the study and ethics approval was obtained from the Ministry
of Education, and informed consent for the participation in the
study was granted by the respective school authorities in Greece.
2.2. Measures
A structured questionnaire was developed for the purposes of
the study, including measures of demographics (age and gender),
awareness and reporting of cyberbullying incidents, empathy,
moral disengagement, attitudes, social norms and prototypes,
self-efcacy beliefs, anticipated regret, and behavioural expecta-
tions. It is noteworthy, that the denition of cyberbullying given
by Slonje and Smith (2008) was included in the rst page of the
questionnaire and participants were requested to complete the
survey with this denition in mind.
Regarding age, respondents were asked to write down the year
of their birth, which was then deducted from the year of survey
completion (i.e., 2010) to generate a chronological age score. Gen-
der was assessed with a single question (what is your gender) fol-
lowed by two response options (1 = male, 2 = female).
Awareness of cyberbullying was assessed with a single item
asking students to report if they had heard or witnessed cyberbul-
lying against others, and if they had been the victims of cyberbul-
lying themselves, and responses were given on a yes/no format.
The participants who admitted cyberbullying awareness, were fur-
ther prompted to state if they reported (Cyberbullying referral) the
incident (yes/no), and to whom they reported (open ended).
2.2.1. Empathy
Empathy was assessed with Joliffe and Farrington (2006) Basic
Empathy Scale (BES). This scale includes a total of 20 items, with
nine items reecting the cognitive empathy facet (e.g., I can under-
stand my friends happiness when she/he does well at something, I
nd it hard to know when my friends are frightened), and 11 items
reecting affective empathy (e.g., My friends emotions do not affect
me much, I get caught up in other peoples feelings easily). After re-
verse scoring eight items, all cognitive and all affective items were
summed to respectively produce the cognitive empathy and affec-
tive empathy scores. Responses were given on standard 5-point
Likert type scale (1 = strongly disagree, 5 = strongly agree), and high-
er scores indicated higher levels of empathy on each facet. The
internal consistency reliability scores were at acceptable levels
for both cognitive (Cronbachs a = 0.65) and affective empathy
(Cronbachs a = 0.74) facets.
2.2.2. Moral disengagement
Moral disengagement was assessed with the 24-item respective
measurement by Bandura et al. (1996), which reected six mecha-
nisms of moral disengagement: moral justication (e.g., It is alright
to ght to protect your friends), advantageous comparison (e.g.,
Stealing some money is not too serious compared to those who steal
a lot of money), displacement of responsibility (e.g., If kids are liv-
ing under bad conditions they cannot be blamed for behaving aggres-
sively), diffusion of responsibility (e.g., A kid in a gang should not be
blamed for the trouble the gang causes), distorting consequences
(e.g., It is okay to tell small lies because they do not really do any
harm), and attribution of blame (e.g., If kids ght and misbehave
in school it is their teachers fault). Each mechanism was assessed
with the mean of four items, scored on a continuous 3-point scale
884 L. Lazuras et al. / Computers in Human Behavior 29 (2013) 881887
(from 1 = agree to 3 = disagree), and higher scores reected lower
levels of moral disengagement. Based on the recommendations
by Bandura et al. (1996), we summed the responses to generate
a total score of moral disengagement. Accordingly, the internal
consistency reliability for the 24-item version was acceptable
(Cronbachs a = 0.71).
2.2.3. Attitudes
Attitudes to engage in cyberbullying were assessed with the
stem proposition To me, cyberbullying is. . . followed by four items
(harmless/harmful, immoral/moral, good/bad, wise thing to do/stupid
thing to do) scored on a 7-point continuous scale. A mean score was
computed and higher scores reected more negative attitudes to-
wards cyberbullying (Cronbachs a = 0.82).
2.2.4. Social norms
Norms were assessed with four distinct items reecting infor-
mational inuence (also termed descriptive norms) on cyberbul-
lying. Specically, the rst item (classmate norms) asked
participants to estimate how many of their classmates engage in
cyberbullying behaviour (responses ranged from 1 = nobody to
5 = almost all of them); the second item assessed cyberbullying
behaviour among the ve closest friends (close friend norms, re-
sponses ranged from 0 = nobody, to 5 = all ve of them); the third
item assessed perceived prevalence of cyberbullying among
same-age peers (perceived prevalence norms) in Greece (i.e., Out
of 100%, how many people your age in Greece do you think engage
or have engaged in cyberbullying? responses given in an open-
ended format); and the fourth item asked participants whether
they had witnessed or heard of other same-aged peers engaging
in cyberbullying (peer norms, responses ranged from 1 = never
to 5 = very often).
2.2.5. Prototypes
Following the recommendations by Gibbons et al. (1998), a def-
inition of prototypes was given, and students were asked to evalu-
ate the prototype of a typical same-age adolescent who engages in
cyberbullying. Prototype favorability was assessed with 12 items
reecting both positive (e.g., smart, popular, cool, and indepen-
dent) and negative attributes (e.g., confused, careless, immature,
and dull), rated along a continuous 7-point scale (1 = not at all,
7 = very much). A mean score was calculated for each prototype fa-
cet, and internal consistency scores were satisfactory for both po-
sitive (Cronbachs a = 0.66) and negative prototype attributes
(Cronbachs a = 0.61). Also, prototype similarity (i.e., perceived
similarity with the cyberbullying prototype) was assessed with a
single question In general, how similar do you think you are to same
age peers who engage in cyberbullying behaviour? scored on a 7-
point continuous scale (from 1 = not at all, to 7 = very much).
2.2.6. Situational self-efcacy
The situational self-efcacy measure was adopted by Lazuras,
Eiser, and Rodanos (2009). The stem proposition I will be tempted
to engage in cyberbullying when. . . was followed by four different
response options respectively assessing four distinct situations
(i.e., I am with other friends who do so, others laugh at me for not
doing it, I feel that most of my friends engage in cyberbullying, and
my friends ask me to do so). Responses were recorded on a 5-point
scale from 1 = denitely not to 5 = denitely yes. A mean score was
generated and higher scores reected less efcacy to resist (or
stronger temptation to conform to) social inuences (Cronbachs
a = 0.80).
2.2.7. Anticipated regret
The stem proposition If I hurt or cause problems to someone by
engaging in cyberbullying, I would... was followed by four items
(not regret it at all/denitely regret it; not feel disappointed with
myself/feel disappointed with myself; not feel bad at all/feel very
bad; not feel worried/feel very worried) scored along a 7-point
continuous scale, with higher scores reecting more anticipated
regret. A mean score was computed and the internal consistency
reliability of the measure was high (Cronbachs a = 0.89).
2.2.8. Expectations to engage in cyberbullying
Behavioural expectations for cyberbullying were assessed with
the mean of three items reecting the likelihood/expectation to en-
gage in cyberbullying in the next 2 months. The responses were re-
corded on a 7-point continuous scale, from 1 = denitely not to
7 = denitely yes, and higher scores reected a greater expectation
to engage in cyberbullying (Cronbcachs a = 0.84).
2.3. Procedure
Following approval from the relevant school authorities, a
group of researchers visited the selected school units, explained
the purposes of the research to students, informed students about
anonymity, condentiality and their right to withdraw from the
study at any point without any foreseeable consequences, and
handed out the surveys during regular teaching hours. In most
cases the teacher was absent, and researchers assisted students
in case they had questions about survey completion. Students re-
turned the completed questionnaires to the researchers, and sur-
vey completion lasted approximately 15 min.
3. Results
Table 1 presents the intercorrelations, means and standard
deviations scores, and, where appropriate, the internal consistency
reliability scores for the measures used in the study.
3.1. Experiencing and reporting cyberbullying
While 32.4% students reported they had either witnessed or
personally experienced cyberbullying (victimized), only 12.5% re-
ported the incident to someone. Those who reported the cyberbul-
lying incident preferred to do so to their friends (44.7%), parents
and siblings (44.7%), and to a lesser extent to ofcial authorities,
such as the police (10.5%). There were no gender differences in
either experiencing/witnessing or reporting cyberbullying.
3.2. Direct effects on expectation to engage in cyberbullying
A hierarchical linear regression was used to assess the direct ef-
fects of age, gender, moral disengagement, empathy, and social
cognitions (attitudes, norms, prototype favourability and similar-
ity, anticipated regret and self-efcacy) on cyberbullying expecta-
tions. The analysis was completed in two steps and an overall
signicant model emerged predicting (AdjR2) 54.9% of the variance
(F (16, 284) = 22.6, p < 0.001) a large multivariate effect size
according to Cohen (1992). Tolerance levels were high (>0.508)
and above the criterion of 0.300 for multicollinearity. Thus, the
predictor variables in the present study were independent from
each other, and multicollinearity risk was low.
In the rst step of the analysis we assessed the effects of age,
gender, empathy facets, and moral disengagement. Moral disen-
gagement (b = 0.360, p < 0.001) and affective empathy (b =
0.243, p < 0.001) signicantly predicted cyberbullying expecta-
tions. Adding social cognitions in the second step of the analysis
reduced the effects of moral disengagement, and turned the effect
of affective empathy non-signicant, thus suggesting a mediation
effect. The signicant predictors of expectations for cyberbullying
 L. Lazuras et al. / Computers in Human Behavior 29 (2013) 881887 885

Table 1
Intercorrelations, means, standard deviation, and internal consistency reliability scores of the measures used in the study.

1 2 3 4 5 6 7 8 9 10 11 12 13 14
* * * * * * * *
1. BES cognitive 0.39 0.28 0.04 0.20 0.05 0.07 0.16 0.23 0.18 0.27 0.21 0.19 0.09
2. BES affective 0.25* 0.08 0.21* 0.07 0.07 0.16* 0.30* 0.25* 0.41* 0.41* 0.29* 0.28*
3. MD 0.16* 0.24* 0.06 0.08 0.21* 0.27* 0.24* 0.24* 0.31* 0.31* 0.39*
4. Classmate norms 41* 0.33* 0.29* 0.16* 0.16* 0.28* 0.24* 0.26* 0.32* 0.40*
5. Close friend norm 0.29* 0.26* 0.29* 0.28* 0.44* 0.34* 0.39* 0.40* 0.41*
6. Perceived norm 0.41* 0.13* 0.09 0.22* 0.10 0.09 0.19* 0.26*
7. Peer norms 0.07 0.10 0.17* 0.13* 0.19* 0.29* 0.31*
8. PF (pos.) 0.35* 0.26* 0.25* 0.23* 0.34* 0.26*
9. PF (neg.) 0.31* 0.36* 0.39* 0.21* 0.27*
10. PS 0.32* 0.49* 0.43* 0.64*
11. Attitudes 0.56* 0.42* 0.37*
12. AR 0.45* 0.48*
13. Self-efcacy 0.51*
14. CB expectation
M 34.37 39.38 50.65 1.97 1.42 33.62 2.50 2.95 4.74 1.80 5.90 5.75 1.62 2.13
SD 5.04 6.68 6.84 0.92 1.08 22.03 0.95 1.17 1.17 1.36 1.25 1.47 0.81 1.50
Cronbachs a 0.65 0.74 0.71 0.66 0.61 0.82 0.89 0.80 0.84
*
p < 0.05; MD = moral disengagement; BES = Basic Empathy Scale; PF = prototype favorability; PS = prototype similarity; AR = anticipated regret; CB = cyberbullying;
Cronbachs alpha scores are relevant to the multi-item measures used in the study.

Table 2
Psychosocial predictors of expectations to engage in cyberbullying (N = 355).

Step Predictors B 95% CI (B) b AdjR2

1 Age 0.016 0.146 to 0.114 0.013 0.203
Gender 0.156 0.510 to 0.197 0.052
Moral disengagement 0.078 0.102 to 0.054 0.360***
BES affective 0.057 0.085 to 0.028 0.243***
BES cognitive 0.033 0.003 to 0.069 0.106
2 Age 0.227 0.826 to 0.373 0.009 0.549
Gender 0.04 0.189 to 0.110 0.017
Moral disengagement 0.591 0.399 to 0.783 0.173***
BES affective 0.035 0.036 to 0.106 0.087
BES cognitive 0.506 0.349 to 0.663 0.052
Classmate norms 0.17 0.007 to .333 0.101*
Close friend norms 0.051 0.198 to 0.097 0.035
Perceived prevalence 0.002 0.005 to 0.008 0.025
Peer norms 0.158 0.005 to 0.311 0.097*
Prototype favorability (pos.) 0.085 0.035 to 0.204 0.064
Prototype favorability (neg.) 0.076 0.051 to 0.202 0.056
Prototype similarity 0.521 0.404 to 0.639 0.446***
Attitudes 0.074 0.203 to 0.056 0.059
Anticipated regret 0.061 0.172 to 0.051 0.06
Self-efcacy 0.301 0.1000.502 0.154**

BES = Basic Empathy Scale.

*
p < 0.05.
**
p < 0.005.
***
p < 0.001.

in the nal step of the analysis included moral disengagement
(b = 0.173, p < 0.001), prototype similarity (b = 0.446, p < 0.001),
perceived prevalence of cyberbullying among classmates or class-
mate norms (b = 0.101, p = 0.04), frequency of witnessing or being
aware about cyberbullying incidents committed by same age peers
or peer norms (b = 0.097, p = 0.04), and situational self-efcacy
(b = 0.154, p = 0.003). The ndings from the regression analysis
are summarized in Table 2.
3.3. Analysis of indirect effects
Multiple mediation analysis was used to assess the indirect ef-
fects of affective empathy and moral disengagement on cyberbul-
lying expectations. According to Preacher and Hayes (2008), 95%
condence intervals and bootstrapping with 1000 resamples were
used. The ndings showed that the effect of affective empathy on
expectations was mediated by prototype similarity (z = 4.36,
p < 0.001) and situational self-efcacy (z = 3.89, p < 0.001), but
not peer and classmate norms. Accordingly, the effect of moral dis-
engagement was mediated by prototype similarity (z = 4.30,
p < 0.001), situational self-efcacy (z = 3.55, p < 0.001), and class-
mate norms (z = 2.22, p = 0.02).
4. Discussion
The present study set out to assess an integrated theory-driven
process-model of cyberbullying incorporating individual traits
(empathy), self-regulatory cognitive processes (moral disengage-
ment), and social cognitions related to cyberbullying. The ndings
partially supported the hypothesis about the direct effects of social
cognitions on intentions. Specically, only social norms (peer and
classmate norms), prototype similarity and situational self-efcacy
predicted cyberbullying intentions. Variables with well-estab-
lished effects on intention-formation, such as attitudes, anticipated
regret, and prototype favourability did not exert a signicant effect
at all. This is in contrast with the basic tenets of theories like the
TPB (Ajzen, 1991) and the P/W model (Gibbons et al., 1998), and
shows that, in the context of cyberbullying, intentionality tends
886 L. Lazuras et al. / Computers in Human Behavior 29 (2013) 881887
to be driven more by descriptive norms, perceived similarity to the
prototype, and efcacy to resist coercive and normative pressure
than by self-centred beliefs, such as personal attitudes and antici-
pated regret. Thus, interventions to tackle cyberbullying behaviour
could benet by targeting normative pressures, as well as coping
skills (e.g., refusal efcacy) and methods to resist social inuences.
Additionally, according to our expectations moral disengage-
ment directly predicted cyberbullying intentions, even after con-
trolling for social cognitions and empathy. This is in accordance
with Bandura et al. (1996) who argued that moral disengagement
leads to aggression proneness, and with past research demonstrat-
ing positive relationships between moral justication mechanisms
and cyberbullying (Pornari & Wood, 2010). In terms of a causal
mechanism the ndings from the mediation analysis indicated that
moral disengagement may underlie a normative process not cap-
tured in previous cyberbullying studies. Specically, potential
aggressors perceive cyberbullying as more prevalent in specic ref-
erence groups (e.g., classmates), view themselves as more similar
to cyberbullying perpetrators, and believe they can hardly resist
endorsing cyberbullying acts under normative and situational
pressures. This argument further extends our understanding of
moral disengagement, by suggesting that this mechanism is not
only relevant to misplacing responsibility, and misattributing
blame or consequences, but is likely to underlie a more general
normative belief-formation process that justies the enactment
of harmful behaviours. In short, individuals who morally disengage
seem also more prone to situational inuences for cyberbullying
behaviour.
Past research (e.g., Ang & Goh, 2010; Steffgen et al., 2011) has
shown that empathy plays an important role in cyberbullying, as
adolescents engaging in cyberbullying behaviour tend to score
lower in empathy measures. Nevertheless, in line with the mediat-
ing hypothesis of the TPB (e.g., Conner & Armitage, 1998; Rhodes &
Courneya, 2003), the present study showed that the effect of empa-
thy in cyberbullying intentions turned non-signicant when other
predictors, such as normative beliefs and self-efcacy, were taken
into account. This should not be taken as evidence against the
importance of empathy, but rather as a need to reconsider the role
of empathy in cyberbullying research, and study its interactions
with more proximal predictors of intentions and behaviour. Specif-
ically, it appears that individuals with lower affective empathy
scores are more likely to develop maladaptive normative and
self-efcacy beliefs, which in turn, predict proclivity for cyberbul-
lying behaviour.
To conclude, this is one of the rst empirical studies to assess an
integrated theoretical model of cyberbullying behaviour account-
ing for both distal and proximal risk factors. Our ndings set the
basis for future educational interventions and preventive strategies
against cyberbullying by highlighting the importance of moral
training, targeting maladaptive normative beliefs, and building
self-efcacy skills.
Acknowledgment
This study was nancially supported by a research grant (JLS/
2008/DAP3/AG/1211) under the DAPHNE III Programme of the
European Commission, DG Justice, Freedom, and Security. We are
grateful to the students who participated in this study.
References
Abraham, C., & Sheeran, P. (2004). Deciding to exercise: The role of anticipated
regret. British Journal Health Psychology, 9, 269278.
Ajzen, I. (1991). The theory of planned behaviour. Organizational Behaviour and
Human Decision Processes, 50, 179211.
Ajzen, I. (2002). Residual effects of past on later behaviour: Habituation and
reasoned action perspectives. Personality and Social Psychology Review, 6,
107122.
Ang, R. P., & Goh, D. H. (2010). Cyberbullying among adolescents: The role of
affective and cognitive empathy, and gender. Child Psychiatry and Human
Development, 41, 387397.
Armitage, C. J., & Conner, M. (2001). Efcacy of the theory of planned behaviour: A
meta-analytic review. British Journal of Social Psychology, 40, 471499.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.
Bandura, A. (2002). Selective moral disengagement in the exercise of moral agency.
Journal of Moral Education, 31, 101119.
Bandura, A., Barbaranelli, C., Caprara, G. V., & Pastorelli, C. (1996). Mechanisms of
moral disengagement in the exercise of moral agency. Journal of Personality and
Social Psychology, 71, 364374.
Bandura, A. (1991). Social cognitive theory of moral thought and action. In W. M.
Kurtines & J. L. Gewirtz (Eds.), Handbook of moral behavior and development. NJ:
Erlbaum, Hillsdale.
Bartholow, B. D., Sestir, M. A., & Davis, E. B. (2005). Correlates and consequences of
exposure to video game violence. Hostile personality, empathy, and aggressive
behavior. Personality and Social Psychology Bulletin, 31, 15731586.
Bauman, S., & Pero, H. (2011). Bullying and cyberbullying among deaf students and
their hearing peers: An exploratory study. Journal of Deaf Studies and Deaf
Education, 16, 236253.
Beran, T., & Li, Q. (2007). The relationship between cyberbullying and school
bullying. Journal of Student Wellbeing, 1, 1533.
Cohen, D., & Strayer, J. (1996). Empathy in conduct disordered and comparison
youth. Developmental Psychology, 32, 988998.
Conner, M., & Armitage, C. A. (1998). Extending the theory of planned behaviour: A
review and avenues for further research. Journal of Applied Social Psychology, 28,
14291464.
Conner, M., Sandberg, T., McMillan, B., & Higgins, A. (2006). Role of anticipated
regret, intentions, and intentions stability in adolescent smoking initiation.
British Journal of Health Psychology, 11, 85101.
Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a
multidimensional approach. Journal of Personality and Social Psychology, 44,
113126.
Davis, M. H. (1994). Empathy: A social psychological approach. Madison, WI: Brown &
Benchmark Publishers.
Davis, F. D., & Warshaw, P. R. (1992). What do intention scales measure? The Journal
of General Psychology, 119, 391407.
Dunkels, E., Franberg, G. M., & Hallgren, C. (Eds.). (2011). Youth culture and net
culture: Online social practices. New York: Information Science Reference.
Endresen, I. M., & Olweus, D. (2002). Self-reported empathy in Norwegian
adolescents: Sex-differences, age trends, and relationship to bullying. In D.
Stipek & A. Bohart (Eds.), Constructive and destructive behavior. Implications for
family, school, & society. Washington: American Psychological Association.
Fishbein, M. (2009). An integrative model for behavioral prediction and its
application to health promotion. In R. J. DiClemente, R. A. Crosby, & M. C.
Kegler (Eds.), Emerging theories in health promotion practice and research
(pp. 215234). New York: John Wiley & Sons.
Gerrard, M., Gibbons, F. X., Stock, M. L., Vande Lune, L. S., & Cleveland, M. J. (2005).
Images of smokers and willingness to smoke among African American pre-
adolescents: An application of the prototype/willingness model to adolescent
health risk behaviour to smoking initiation. Journal of Pediatric Psychology, 30,
305318.
Gibbons, F. X., Gerrard, M., Blanton, H., & Russell, D. W. (1998). Reasoned action and
social reaction: Willingness and intention as independent predictors of health
risk. Journal of Personality and Social Psychology, 74, 11641180.
Gini, G. (2006). Social cognition and moral cognition in bullying: Whats wrong?
Aggressive Behavior, 32, 528539.
Hamilton, K., & White, K. M. (2008). Extending the theory of planned behavior: The
role of self and social inuences in predicting adolescent regular moderate-to-
vigorous physical activity. Journal of Sport and Exercise Psychology, 30, 5674.
Hinduja, S., & Patchin, J. W. (2010). Bullying, cyberbullying, and suicide. Archives of
Suicide Research, 14, 206221.
Joliffe, D., & Farrington, D. (2006). Development and validation of the basic empathy
scale. Journal of Adolescence, 29, 589611.
Juvonen, J., & Gross, E. (2008). Extending the school grounds? Bullying experiences
in cyberspace. Journal of School Health, 78, 496505.
Klomek, B. A., Sourander, A., & Gould, M. (2010). The association of suicide and
bullying in childhood to young adulthood: Review of cross-sectional and
longitudinal research ndings. Canadian Journal of Psychology, 55, 282288.
Lazuras, L., Eiser, J. R., & Rodanos, A. (2009). Predicting Greek adolescents
intentions to smoke: A focus on normative processes. Health Psychology, 28,
770778.
Li, Q. (2007). New bottle but old wine: A research of cyberbullying in schools.
Computers in Human Behavior, 23, 17771791.
Li, M., Frieze, I. H., & Tang, C. (2010). Understanding peer sexual abuse: Using the
theory of planned behavior. Sexual Abuse: A Journal of Research and Treatment,
22, 157171.
Loudin, J. L., Loukas, A., & Robinson, S. (2003). Relational aggression in college
students: Examining the roles of social anxiety and empathy. Aggressive
Behaviour, 29, 430439.
Lovett, B. J., & Shefeld, R. A. (2007). Affective empathy decits in aggressive
children and adolescents: A critical review. Clinical Psychology Review, 27, 113.
 L. Lazuras et al. / Computers in Human Behavior 29 (2013) 881887
McAlister, A. L., Bandura, A., & Owen, S. V. (2006). Mechanisms of moral
disengagement in support of military force. The impact of Sept.9.11. Journal of
Social and Clinical Psychology, 25, 141165.
McMillan, B., & Conner, M. (2003). Using the theory of planned behavior to
understand alcohol and tobacco use in students. Psychology, Health and
Medicine, 8, 317328.
Norman, P., Armitage, C. J., & Quigley, C. (2007). The theory of planned behavior and
binge drinking: Assessing the impact of binge drinker prototypes. Addictive
Behaviors, 32, 17531768.
Olweus, D. (1993). Bullying at school: What we know and what we can do. Oxford:
Blackwell.
Patchin, J. W., & Hinduja, S. (2006). Bullies move beyond the schoolyard. Youth
Violence and Juvenile Justice, 4, 148169.
Perren, S., & Gutzwiller-Helfennger, E. (2012). Cyberbullying and traditional
bullying in adolescence. Differential roles of moral disengagement, moral
emotions, and moral values. European Journal of Developmental Psychology, 9,
195209.
Perugini, M., & Bagozzi, R. P. (2001). The role of desires and anticipated emotions in
goal-directed behaviors: Broadening and deepening the theory of planned
behavior. British Journal of Social Psychology, 40, 7998.
Pornari, C. D., & Wood, J. (2010). Peer and cyber aggression in secondary school
students: The role of moral disengagement, hostile attribution bias, and
outcome expectancies. Aggressive Behaviour, 36, 8194.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for
assessing and comparing indirect effects in multiple mediator models. Behavior
Research Methods, 40, 879891.
Preston, S. D., & de Waal, F. B. M. (2002). Empathy: Its ultimate and proximate bases.
Behavioral and Brain Sciences, 25, 171.
Pyzalski, J. (2011). Electronic aggression among adolescents: An old house with a
new facade (or even a number of houses)? In E. Dunkels, G. M. Franberg, & C.
Hallgren (Eds.), Youth culture and net culture: Online social practices
(pp. 278295). New York: Information Science Reference.
Rhodes, R. E., & Courneya, K. S. (2003). Relationships between personality, an
extended theory of planned behaviour model, and exercise behaviour. British
Journal of Health Psychology, 8, 1936.
Rhodes, R. E., & Matheson, D. H. (2005). Discrepancies in exercise intentions and
expectation: Theoretical and applied issues. Psychology and Health, 20,
6378.
887
Rivis, A., & Sheeran, P. (2003a). Social inuences and the theory of planned
behaviour: Evidence for a direct relationship between prototypes and young
peoples exercise behaviour. Psychology and Health, 18, 567583.
Rivis, A., & Sheeran, P. (2003b). Descriptive norms as an additional predictor in
the theory of planned behaviour: A meta-analysis. Current Psychology, 22,
218233.
Rivis, A., Sheeran, P., & Armitage, C. J. (2006). Augmenting the theory of planned
behaviour with the prototype/willingness model: Predictive validity of actor
versus abstainer prototypes for adolescents health protective and health-risk
intentions. British Journal of Health Psychology, 11, 483500.
Schultze-Krumbholz, A., & Scheithauer, H. (2009). Social-behavioural correlates of
cyberbullying in a German student sample. Journal of Psychology, 217,
224226.
Shamay-Tsoory, S. G., Aharon-Peretz, J., & Perry, D. (2009). Two systems for
empathy: A double dissociation between emotional and cognitive empathy in
inferior frontal gyrus versus ventromedial prefrontal lesions. Brain, 132,
617627.
Slonje, R., & Smith, P. (2008). Cyberbullying: Another main type of bulling?
Scandinavian Journal of Psychology, 49, 147154.
Steffgen, G., Knig, A., Pfetsch, J., & Melzer, A. (2011). Are cyberbullies less
empathic? Adolescents cyberbullying behavior and empathic responsiveness.
Cyberpsychology, Behavior, and Social Networking, 14, 643648.
Strayer, J. (1987). Empathy and its development. In N. Eisenberg & J. Strayer (Eds.),
Cambridge studies in social and emotional development. New York: Cambridge
University Press.
Walrave, M., & Heirman, W. (2011). Cyberbullying: Predicting victimisation and
perpetration. Children and Society, 25, 5972.
Warshaw, P. R., & Davis, F. D. (1986). The accuracy of behavioural intention versus
behavioural expectation for predicting behavioural goals. The Journal of
Psychology, 119, 599602.
Webb, T. L., & Sheeran, P. (2006). Does changing behavioural intentions engender
behaviour change? A meta-analysis of the experimental evidence. Psychological
Bulletin, 132, 249268.
Ybarra, M., & Mitchell, K. (2004). Online aggressor/targets, aggressors, and targets: A
comparison of associated youth characteristics. Journal of Child Psychology and
Psychiatry, 45, 13081316.