You are on page 1of 7

Computers in Human Behavior 109 (2020) 106340

Contents lists available at ScienceDirect

Computers in Human Behavior


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

The role of self-efficacy in defending cyberbullying victims


Madeleine Clark , Kay Bussey *
Department of Psychology, Macquarie University, Sydney, Australia

A R T I C L E I N F O A B S T R A C T

Keywords: Due to the lasting and pervasive effects of cyberbullying on cybervictims, current literature has examined factors
Cyberbullying which decrease the prevalence of cyberbullying episodes. One factor known to mitigate cyberbullying rates is
Adolescents bystander intervention. However, despite many bystanders witnessing acts of cyberbullying, few bystanders elect
Cyber defending
to intervene. There is currently minimal literature examining which factors influence bystander intervention and
Cyber bystanders
Empathic self-efficacy
therefore which factors can be targeted in interventions. To address this gap in the literature, the present study
Defending self-efficacy examines the role of self-efficacy, specifically defending self-efficacy and empathic self-efficacy and how these
beliefs influence defending behavior while controlling for grade, gender and previous experiences of cyberbul­
lying and cybervictimization. To examine these relationships, 540 Australian students aged 11 to 15 years
completed a questionnaire assessing cyberbullying behavior and self-efficacy beliefs in the past school term.
Results revealed that both defending self-efficacy and empathic self-efficacy were significantly associated with
frequency of defending behavior. Specifically, higher rates of defending self-efficacy and higher levels of
empathic self-efficacy were associated with higher rates of defending behavior during cyberbullying episodes.
Cybervictimization was also significantly associated with higher levels of cyber defending behavior. These results
suggest the importance of cyberbullying interventions targeting defending and empathic skills and beliefs in
bystanders to increase defending behavior and decrease cyberbullying episodes.

1. Introduction intervening and actively defending the victim (Bastiaensens et al., 2015;
Moxey & Bussey, 2019; Salmivalli, 2010). Defending includes
Cyberbullying is defined as an act of recurring and intentional anti-bullying behaviors such as comforting the victim or confronting the
aggression towards someone less capable of defending themselves using perpetrator (Salmivalli, 1999). In addition to effectively ceasing cyber­
electronic means (Menesini et al., 2012). Estimates of cyberbullying bullying episodes, victims who are defended also report higher rates of
prevalence vary greatly depending on the definition used. Some scholars peer acceptance, self-esteem and perceived popularity compared to their
argue that an act of aggression can qualify as cyberbullying regardless of non-defended peers (Sainio, Veenstra, Huitsing & Salmivalli, 2010).
whether it is a recurring behavior (DeSmet et al., 2014; Nocentini et al., Despite playing a significant role as moderators of cyberbullying
2010) and regardless of its impact (Dredge, Gleeson & Garcia, 2014; episodes (Obermaier, Fawzi & Koch, 2016; Olenik-Shemesh, Heiman &
Menesini et al., 2012). As a result, some studies report rates as high as Eden, 2017) it is estimated that between 9 and 50 percent of bystanders
65% while other studies that adopt a more stringent definition report defend cyberbullying victims (Dillon & Bushman, 2015; Freis & Gurung,
cyberbullying rates as low as 5% (Jadambaa et al., 2019). However, the 2013; Huang & Chou, 2010; Lenhart et al., 2011). The small percentage
most commonly reported rate of adolescent cyberbullying victimization of defenders is surprising given the amplified opportunity to defend
is between 20 and 40% (Tokunaga, 2010). Furthermore, cyberbullying during cyberbullying episodes. For example, it is not necessary for the
rates are increasing as adolescents gain greater access to the internet bystander to be present during the cyberbullying episode as bystanders
(Kowalski, Limber, & McCord, 2019). This is concerning given the may witness the incident after it has transpired or have the content
extensive and lasting psychological, behavioral, social and academic directly forwarded to them (Li, Smith & Cross, 2012). The lack of
effects it has on those who are victimised (Cassidy, Faucher, & Jackson, defending behavior is thought to be explained by uncertainty as to what
2013; Kim, Colwell, Kata, Boyle & Georgiades, 2017). One factor known constitutes cyberbullying (Alipan, Skues, Theiler & Wise, 2019) along­
to attenuate cyberbullying and its effects is a bystander or bystanders side the bystander effect (Machackova, Dedkova & Mezulanikova, 2015;

* Corresponding author.
E-mail address: kay.bussey@mq.edu.au (K. Bussey).

https://doi.org/10.1016/j.chb.2020.106340
Received 2 August 2019; Received in revised form 5 February 2020; Accepted 11 March 2020
Available online 13 March 2020
0747-5632/© 2020 Elsevier Ltd. All rights reserved.
M. Clark and K. Bussey Computers in Human Behavior 109 (2020) 106340

Obermaier et al., 2016; You & Lee, 2019). According to Darley and victims of in cyberbullying episodes and self-efficacy to talk with the
Latane’s (1968) bystander model, bystanders are less likely to intervene bully, although in this study defending self-efficacy was not measured as
when other individuals are present due to diffusion of responsibility a domain construct but instead a global measure of self-efficacy.
(believing the responsibility to act is divided among those present), Therefore, the role of defending self-efficacy in defending victims in
evaluation apprehension (worrying that other bystanders will nega­ the cyber context is unclear. It is vital to explore the influence of
tively judge your actions) and pluralistic ignorance (observing others self-efficacy on individuals’ defending if their behavior has the potential
not acting and concluding this is appropriate). This effect is concerning for reducing cyberbullying occurrences (Salmivalli, Voeten & Pos­
as passive bystanders provide no support for the victim and their lack of kiparta, 2011).
action may be misperceived by the perpetrator as support for their
cyberbullying behavior (Bastiaensens et al., 2014). Given that poten­ 1.3. Empathic self-efficacy
tially large audiences can witness incidences of cyberbullying either at
the time of the occurrence or after it has occurred, it is important to A further factor shown to increase a bystander’s willingness to
uncover which factors are associated with those bystanders who defend intervene is empathy (Barlin � ska, Szuster, & Winiewski, 2013). Given
cyber victims. that defending is a challenging situation, it is important that an indi­
vidual not only empathizes with the victim, but that they also have a
1.1. Social cognitive theory strong belief in their capacity to engage in social interactions demanding
such empathy (Di Giunta et al., 2011). Empathic self-efficacy is defined
To investigate the behavior of bystanders during cyberbullying epi­ as an individual’s self-perceived ability to sense the emotions of others
sodes, Bandura’s (1986) social cognitive theory was drawn on as this and respond empathetically, including to their distress and hardship
model highlights the need to consider contextual factors influencing a (Eklund, Loeb, Hansen & Andersson-Wallin, 2012). To date, empathic
person’s behavior. This is particularly relevant for the complex social self-efficacy has not been examined in relation to defending behavior in
context in which bullying occurs whereby both the peer group and the the cyberbullying context. This is surprising given that empathy has
social networking platform play a pivotal contextual role (Allison & been demonstrated as an essential precursor to defending behavior in
Bussey, 2016). Social Cognitive Theory which proposes a triadic model traditional bullying episodes (Barchia & Bussey, 2011; Nickerson, Mele
in which personal, behavioral and environmental elements operate by & Princiottta, 2008). Furthermore, it has been shown that empathic
reciprocally influencing each other has been adapted to the cyberbul­ self-efficacy is a better predictor of prosociality than any other
lying context (Allison & Bussey, 2017). This approach provides a self-efficacy variable examined in the current literature (Caprara, Ales­
comprehensive framework for understanding the complexity of bullying sandri, & Eisenberg, 2012; Eklund et al., 2012). Consequently, it is
dynamics and the social situations in which they occur. One of the most predicted that it is important that bystanders possess heightened
pivotal components of social cognitive theory when applied to defend­ empathic self-efficacy in order to increase their motivation and likeli­
ing behavior is personal agency, which refers to an individuals’ ability to hood of intervening in a cyberbullying episode.
develop and direct their actions towards a task (Bandura, 1986). While the effects of defending self-efficacy and empathic self-efficacy
Although performance skills play a role in personal agency, the most have been examined within the traditional bullying literature, it cannot
central factor is self-efficacy, which, is known to influence an in­ be assumed that the same effects would generalize to the cyber context.
dividual’s actions as well as the skills they acquire (Bandura, 1986, Although there is significant overlap between traditional bullying and
1997). Self-efficacy refers to an individual’s belief in their ability to cyberbullying, the two forms of aggression are still arguably unique in
organise the motivation, cognitive resources, skills and action required ways which may influence bystander behavior. Specifically, in the
to execute an undertaking successfully (Bandura, 1997). For individuals cyberbullying context, individuals including perpetrators and by­
with low self-efficacy beliefs relating to intervening in bullying episodes, standers are able to remain anonymous which may increase or decrease
it is predicted that despite a sense of responsibility or desire to intervene, the propensity of defending from that observed in offline contexts.
they will elect to distance themselves from the situation in order to cope Furthermore, in some cases it is more difficult for authority figures such
rather than defend the victim (Pozzoli & Gini, 2013). as teachers, parents or the police to detect and cease cyberbullying ep­
isodes due to privacy settings (Cassidy et al., 2013). Finally, the role of
1.2. Defending self-efficacy bystanders in a cyber context is regarded as more variable and contex­
tually dependent depending on privacy and anonymity factors (DeSmet
Defending self-efficacy, defined as an individual’s self-perceived et al., 2014). Therefore, it is imperative to examine the influence of these
ability to successfully intervene in a bullying episode and help the self-efficacy beliefs in a cyber specific context.
victim (Thornberg, Wa €nstro
€m, Hong & Espelage, 2017), is therefore
proposed as a crucial factor involved in predicting a bystander’s 1.4. Research questions and hypotheses
response. Research examining the role of defending self-efficacy in
influencing bystanders’ course of action is unclear, however. While the The aim of the present study is to examine the role of defending self-
majority of literature reports that defending self-efficacy is positively efficacy and empathic self-efficacy as predictors of defending behavior
associated with defending and negatively associated with passive during cyberbullying episodes. First, it was hypothesised that defending
behavior during traditional bullying episodes (Po €yho
€nen, Juvonen & self-efficacy will be positively associated with frequency of defending,
Salmivalli, 2010; Po €yho€nen & Salmivalli, 2008; Thornberg & Jungert, such that high levels of defending self-efficacy will be associated with
2012) other studies have found no such association (Barchia & Bussey, higher rates of reported defending. Second, it was hypothesised that
2011). Furthermore, some studies have reported that this relationship is empathic self-efficacy will be uniquely and positively associated with
moderated by perceived popularity, so that at low perceptions of frequency of defending behavior, over and above defending self-
popularity defending self-efficacy did not influence defending behavior efficacy, such that high levels of empathic self-efficacy will be associ­
(Po€yho€nen, Juvonen & Salmivalli, 2010). ated with higher rates of defending behavior. Third, it was hypothesised
In addition to these inconsistencies, the role of defending self- that the positive relationship between defending self-efficacy and
efficacy in relation to intervention in cyberbullying contexts has rarely defending behavior will be stronger fo participants with high empathic
been examined. One study by DeSmet et al. (2014) also reported that self-efficacy. That is, while defending self-efficacy is postulated as
low defending self-efficacy was a predictor of bystander behavior when necessary for defending behavior within a cyber context, it is proposed
the bystander perceived low levels of personal social support. Further­ that this belief is more likely to be acted on at higher levels of empathic
more, DeSmet et al. (2012) reported an association between defending self-efficacy.

2
M. Clark and K. Bussey Computers in Human Behavior 109 (2020) 106340

No specific predictions were made for gender and grade differences Cyberbullying and Cyber victimization. Ga�mez-Guadix, Villa-George
in defending behavior due to the inconsistent findings in the literature. and Calvete’s (2014) Cyberbullying Questionnaire Revised Scale was
Some studies have reported that girls engage in more defending used to measure cyberbullying and cyber victimization. Participants
behavior than boys (Bastiaensens et al., 2014; Olenik-Shemesh et al., rated on a six-point scale how frequently they had engaged in each of 14
2017; Mazzone, Camodeca & Salmivalli, 2016), whereas other studies items relating to cyberbullying and 9 items relating to cyber victimiza­
have found no gender differences in defending behavior (Barlin
� ska et al., tion. Total scores were calculated by averaging all items in the subscale,
2013, 2015, 2018). For age differences, some studies have reported that with higher scores indicating more frequent experiences of cyberbully­
younger adolescents are more likely to engage in defending behavior ing and cyber victimization. The cyberbullying subscale achieved
than older adolescents (Allison & Bussey, 2017; DeSmet et al., 2016; excellent reliability (α ¼ 0.97) and the cyber victimization scale ach­
Pozzoli & Gini, 2010) whereas Bussey et al., (2015) found that older ieved good reliability (α ¼ 0.85).
students engaged in more defending behavior than younger students. As
age and gender may be associated with defending behavior, but were not 2.3. Procedure
the focus of this study, these variables will be controlled in analyses
predicting defending behavior. Ethics approval for this study was granted by the Human Research
Ethics Committee at the authors’ university. Participants aged between
2. Methods 11 and 15 years (grades 7 and 9) were recruited from 14 co-educational
independent schools in NSW. Consent was obtained from the school
2.1. Participants principals, parents and students. Approximately 46% of eligible students
participated in the study. The questionnaire was administered in class
A total of 540 students participated in this study, including 344 grade groups during term two of the school year with an average group of 19
7 students (194 females, Mage ¼ 12.65 years; SD ¼ 0.42 years) and 196 students under the supervision of teachers and/or researchers present.
grade 9 students (110 females, Mage ¼ 14.63 years; SD ¼ 0.45 years). As Three schools (17.4% of participants) completed a paper version of the
this study formed part of a larger, longitudinal project only students in study due to lack of access to electronic devices while the remaining 11
grade 7 and 9 were invited to participate as students in year 8 and 10 schools (82.6% of participants) completed it online through school
were asked to participate in the following year. This sample was selected facilities.
as cyberbullying has been shown to peak during middle school (Suchy & Each participant was provided with a unique identification number,
Tomasino, 2014). The uneven sample distribution between grades login code and URL code to access the survey. After logging in, partic­
resulted from disparities in return rates between grades due to greater ipants were presented with an online (or paper version for those
academic commitments required of the older students. Participants were completing the paper format of the questionnaire) consent form where
recruited from 14 co-educational independent schools in New South students were required to agree to complete the questionnaire. Students
Wales, Australia and the majority of participants were from first completed questions regarding their demographics. Students were
upper-middle class families (Australian Curriculum Assessment and then presented with a brief definition of cyberbullying adapted from
Reporting Authority, 2016). Most participants identified as Anglo/Celtic Solberg and Olweus (2003). The remaining measures were presented in
(67.2%), with a substantial minority from European (15.9%) and one of two random orders for all students including11 other measures
East/South East Asian (6.7%) backgrounds. not used in this study, as this study formed part of a larger, longitudinal
research project examining adolescent cyberbullying. At the completion
2.2. Measures of the questionnaire participants were presented with a debrief state­
ment which informed them to indicate if they wished to make an
Defending self-efficacy. Barchia and Bussey’s (2011) Defender appointment with their school counsellor to discuss any aspect of the
Self-Efficacy Scale used in a traditional bullying context was adapted to questionnaire or subsequent distress resulting from participation. Of the
measure self-efficacy to engage in common defending strategies total 540 participants, 14 opted to make an appointment with their
employed in cyberbullying episodes. Participants rated on a 7-point school counsellor. This was then arranged between the researcher, the
Likert scale how well they could enact nine specific defending behav­ school counsellor and the participant.
iors (e.g. “talk to a parent about the situation”). This scale was con­
textualised to incorporate common defending measures employed in 2.4. Exploratory factor analysis
cyberbullying episodes (e.g. report the post). A total score was calcu­
lated by summing all items, with higher scores indicating higher levels An exploratory factor analysis was conducted on the defending self-
of defending self-efficacy. This scale achieved acceptable reliability (α ¼ efficacy and empathic self-efficacy measures to determine whether the
0.82). two scales measure conceptually different constructs (see Table 1) as
Empathic self-efficacy. Barchia and Bussey’s (2011) Empathy Index these scales were modified from previous research for the cyber context.
Scale was adapted to measure self-efficacy to empathize. Participants All nine items from each measure were included in the factor analysis
rated on a 4-point Likert scale how well they could enact nine specific with a principal axis extraction and Oblimin rotation selected as corre­
empathic behaviors (e.g. “recognize when your friend needs help”). This lation between the factors was expected. Results yielded a two-factor
scale was adapted to measure self-efficacy using Bandura’s (2005) guide model which loaded onto defending self-efficacy and empathic self-
for constructing self-efficacy scales by asking participants how well they efficacy. One of the items relating to defending self-efficacy (“ignore
could engage in these behaviors, rather than using a dichotomous the situation”) demonstrated poor loading onto the defending self-
true/false response scale. A total score was calculated by summing all efficacy factor with a loading below 0.20. This suggested it was not
items, with higher scores indicating higher levels of empathic conceptually meaningful and therefore this item was removed from the
self-efficacy. This scale achieved good reliability (α ¼ 0.82). analysis. Both the self-efficacy scales demonstrated good internal con­
Defending Behavior. Allison and Bussey’s (2017) Cyberbullying sistency, with Cronbach’s alphas of .82 for each scale.
Bystanding and Intervention Scale was used to measure defending fre­
quency. Participants rated on a 6-point Likert scale how frequently they 2.5. Data management
tried to help victims of cyberbullying in the last term that were cyber­
bullied in 14 different ways. A total score was calculated by averaging all Missing Data. Small amounts of data were missing at the item level
items, with higher scores indicating more frequent defending. This scale ranging from 0.2% to 8.3%. Missing data were imputed at the item level
achieved excellent reliability (α ¼ 0.91). by using the expectation-maximisation procedure in SPSS where

3
M. Clark and K. Bussey Computers in Human Behavior 109 (2020) 106340

Table 1 SPSS to investigate potential clustering effects of response methodology


Factor structure and factor loadings for defending self-efficacy and empathic (paper format versus online format). Results indicated no random effects
self-efficacy items. for response format methodology.
Factor
Scale Items Loadings 3.3. Grade and gender effects
1 2

1. Defending Self-Efficacy
A MANOVA was run to examine grade and gender effects for each of
Talk to a parent about the situation .71 the measures in order to control the Type 1 error rate. Descriptive sta­
Tell a teacher about the situation .72 tistics are displayed in Table 2. Results revealed that grade seven stu­
Talk to a friend about the situation .56 dents reported higher levels of defending self-efficacy than did grade
Ignore the situation .16
nine students. Additionally, males reported more frequent cyberbullying
Tell the bully to stop .71
Provide comfort to the kid who is being cyberbullied .62 and lower empathic self-efficacy than did females. No grade by gender
Tell the bully that what they are doing is not okay .75 interactions were significant for any of the measures. Due to the grade
Defend the kid who is being cyberbullied by being mean to the bully .53 and gender main effects, grade and gender were controlled for in sub­
Report the post .62 sequent analyses.
2. Empathic Self-Efficacy
Know what your friends would like when they are upset .66
Feel the same way that a person in trouble feels when they are .57 3.4. Correlations among measures
experiencing it
Understand the effect your actions have on others feelings .63 Partial Pearson’s correlational analyses were run to examine the
Quickly find out things that new people you meet like and dislike .53
Recognize when someone wants comfort and emotional support .73
strength of relationship between the variables while controlling for the
Recognize when a person likes you .51 effects of grade and gender. Results are displayed in Table 3. Defending
Recognize whether a person is annoyed with you .58 self-efficacy, empathic self-efficacy and defending were all significantly
Recognize when a person is scared .67 and positively correlated with each other. Cyber victimization was
Recognize when a friend needs your help .73
significantly and positively correlated with defending. Cyberbullying
was significantly and positively correlated with cyber victimization and
imputation was based on individual items being used as predictors. The defending and significantly and negatively correlated with empathic
expectation-maximisation procedure was selected as it has been self-efficacy.
demonstrated as the preferred method when imputing data that is not
missing completely at random, compared to methods such as list-wise 3.5. Hierarchical regression analyses
deletion, pair-wise deletion or means substitution (Alison, 2002;
Enders, 2001; Schafer & Graham, 2002). Hierarchical regression analyses were run to examine the relation­
ships between self-efficacy beliefs (defending self-efficacy and empathic
3. Results self-efficacy) and their interaction with defending behavior. The results
are presented in Table 4. A five-step model was created in which grade
3.1. Data analytic plan and gender were entered at step one as control variables, as the MAN­
OVA results had indicated differences between these groups for multiple
The results of this study are presented in three sections. First, a variables. Cyberbullying and cyber victimization were entered at step
multivariate analysis of variance (MANOVA) was completed to examine two due to their observed correlations with variables of interest.
grade and gender effects for each measure. Second, correlational ana­ Defending self-efficacy was entered at step three, empathic self-efficacy
lyses were computed. Third, hierarchical regression analyses were un­ was entered at step four and the interaction term for defending self-
dertaken to examine the relationships between self-efficacy beliefs efficacy with empathic self-efficacy was entered at step five. Prior to
(defending self-efficacy, empathic self-efficacy and their interaction conducting the regression analyses the independent and control vari­
term) and frequency of defending behavior during cyberbullying ables were all mean-centred. Preliminary analyses demonstrated that no
episodes.
Table 3
Correlations analyses of independent and dependent variables controlling for
3.2. Clustering effects grade and gender.
1 2 3 4 5
Prior to examining grade and gender effects preliminary analyses
using the linear mixed model procedure in SPSS were conducted to 1. Cyberbullying – .50** -.02 -.15** .28**
2. Cyber victimization -.06 -.05 .54**
investigate potential clustering effects of responses within schools. This

3. Defending Self-Efficacy – .38** .14**
was done as the data presented are nested (students within grades within 4. Empathic Self-Efficacy – – .14**
schools). Results indicated no random effects for schools and therefore 5. Defending Frequency –
school effects were not controlled for in any subsequent analyses.
*p < .05 **p < .01.
Furthermore, a linear mixed model procedure was also conducted in

Table 2
Estimated marginal means and standard errors of all measures by gender and by grade.
Males Females F Grade 7 Grade 9 F Total number of items α
M (SD) M (SD) M (SD) M (SD)

1. Cyberbullying 1.25 (.03) 1.07 (.03) 14.95*** 1.17 (.03) 1.15 (.04) .11 14 .97
2. Cyber victimization 1.31 (.04) 1.28 (.03) .50 1.29 (.03) 1.30 (.04) .04 9 .85
3.Defending Self-efficacy 35.69 (.66) 36.63 (.58) 1.13 37.16 (.53) 35.16 (.71) 5.13* 9 .82
4. Empathic Self-efficacy 26.38 (.30) 27.69 (.26) 10.92** 27.15 (.24) 26.92 (.32) .32 9 .82
5. Defending Behavior 1.47 (.04) 1.44 (.04) .29 1.43 (.03) 1.47 (.04) .55 14 .91

Note. M ¼ mean; SD ¼ standard deviation; α ¼ Cronbach’s alpha. *p < .05 **p < .005 ***p < .0005.

4
M. Clark and K. Bussey Computers in Human Behavior 109 (2020) 106340

Table 4
Hierarchical regression analysis of the associations between defending self-efficacy, empathic self-efficacy and defending.
Frequency of Defending Behavior
Variable
Step 1 β Step 2 β Step 3 β Step 4 β Step 5 SE 95% CI
β

Grade .02 .02 .03 .02 .01 .02 [-.02, .07]


Gender -.02 .01 .01 -.01 -.01 .04 [-.09, .08]
Cyberbullying .07 .06 .09 .08 .08 [.46, .75]
Cyber victimization .58*** .60*** .59*** .59*** .11 [-.11, .34]
Defending self-efficacy .01** .01* .01* .00 [.00, .01]
Empathic self-efficacy .02** .02** .01 [.01, .03]
Defending self-efficacy x Empathic self-efficacy .00 .00 [-.00, .00]
Total R2 .00 .28 .32 .33 .33

Note. Based on centred and bootstrapped continuous variables. Gender was coded as 0 ¼ Male and 1 ¼ Female; Grade was coded as 0 ¼ Grade 7 and 1 ¼ Grade 9.
*p < .05 **p < .005 ***p < .0005.

assumptions of regression were violated except, the P–P plots which cease the cyberbullying cycle.
indicated non-normal distribution of the residuals. The frequency of By contrast, the results of this study did not support the third hy­
defending behavior scale was positively skewed while both self-efficacy pothesis, where the positive relationship between defending self-
scales were negatively skewed. To take account of this, bootstrapping efficacy and defending behavior was not observed to be significantly
analyses were completed to provide confidence intervals and bias- stronger in participants with higher empathic self-efficacy. Instead, the
corrected p-values. Bootstrapping was conducted because it has been results show that defending self-efficacy is related to defending behavior
shown to produce robust results for data where the assumption of at all levels of empathic efficacy. Taken together, the results of this study
normality is not met and is preferred over other methods such as suggest that while the effect of each of these self-efficacy domains is not
transformations and relying on the robustness of F-tests and t tests (Field contingent on the other, defending self-efficacy and empathic self-
& Wilcox, 2017). efficacy both have independent effects on defending behavior.
Results from the hierarchical regression analysis indicate that the These results revealed several practical implications. Given the
overall model was highly significant (F (7, 532) ¼ 37.59, p < .0005). In impact of both defending self-efficacy and empathic self-efficacy on
the final model, cyber victimization, defending self-efficacy and defending behavior, it would be important for intervention programs to
empathic self-efficacy were all significant predictors of defending fre­ target these beliefs. Bandura (1986) proposes four sources of
quency. Furthermore, defending self-efficacy was uniquely, significantly self-efficacy, which, should be targeted in order to increase defending
and positively associated with defending over and above empathic self- and empathic self-efficacy. These are: enactive attainment (personal
efficacy and at all levels of empathic self-efficacy. The interaction be­ mastery experiences), vicarious experiences (observing similar in­
tween defending self-efficacy and empathic self-efficacy was not dividuals perform certain behaviors where their success or failure will
significant. impact on one’s self - efficacy to perform the same behavior.), verbal
persuasion (attempting to persuade someone they have the skills and
4. Discussion capabilities to achieve their desired goal) and physiological state (when
people interpret their physiological state such as somatic arousal as in­
The present study builds on the established literature by extending dicators of vulnerability to dysfunction and failure) (Zimmerman,
the examination of the relationship between defending self-efficacy, 2000). Therefore, intervention programs should include verbal persua­
empathic self-efficacy and frequency of defending behavior to cyber­ sion discussing the benefits of intervening and role play exercises tar­
bullying episodes. The results of this study supported both the first and geting mastery and vicarious experiences in defending as well as
second hypothesises. Specifically, defending self-efficacy and empathic reduction in physiological stressors via means of exposure exercises.
self-efficacy were each positively associated with frequency of defending
behavior in the cyberbullying context. Furthermore, empathic self- 4.1. Strengths
efficacy was uniquely associated with defending behavior over and
above the effects of defending self-efficacy, suggesting both defending This study was the first to examine the relationships between
self-efficacy and empathic self-efficacy uniquely contributed to defending self-efficacy, empathic self-efficacy and frequency of
defending behavior. These results are consistent with findings from the defending behavior within a cyberbullying context. Moreover, it did so
traditional bullying literature which report that defending self-efficacy while controlling for the effects of grade, gender and previous experi­
(Po€yho€nen, Juvonen & Salmivalli, 2010; Po €yho €nen & Salmivalli, ence of cyberbullying and cyber victimization. These relationships were
2008; Thornberg & Jungert, 2012) and empathic self-efficacy (Barchia examined using pre-validated measures relevant to self-efficacy beliefs
& Bussey, 2011; Nickerson, Mele & Princiottta, 2008) are positively and the cyberbullying context. This study also consisted of a large
associated with defending behavior. sample size including both males and females and therefore provided
These results are also consistent with Bandura’s (1986) social adequate statistical power to conduct formal tests of regression.
cognitive theory, which, proposes that higher levels of self-efficacy
result in higher levels of motivation to act which in turn results in
4.2. Limitations
higher rates of behavioral engagement. Previous experiences with cyber
victimization also emerged as a significant predictor of defending
The present study is not without limitations, however. Firstly, there
behavior. These results align with research which reports a positive
are limitations within the study design. Namely, the study relied on self-
association between previous experiences with cyber victimization and
report data, meaning the validity of information may have been
defending (Van Cleemput et al., 2014) and may be linked to greater
compromised by a number of factors including careless responding,
empathy activation, which has been shown to be a strong predictor of
acquiescence, social desirability effects and deliberately exaggerated
defending behavior in previous research (Van Cleemput et al., 2014).
responses (Chan, 2009). Future research should seek to include reports
These results highlight the importance of cyber victims having high
from multiple sources including teachers, parents and peers to improve
defending self-efficacy in order to ensure they effectively intervene and
the validity of the supplied data. Furthermore, the study design was

5
M. Clark and K. Bussey Computers in Human Behavior 109 (2020) 106340

cross-sectional thereby limiting directional inferences regarding the Cassidy, W., Faucher, C., & Jackson, M. (2013). Cyberbullying among youth: A
comprehensive review of current international research and its implications and
relationship between the self-efficacy factors and defending behavior.
application to policy and practice. School Psychology International, 34(6), 575–612.
Future research should seek to explore these relationships employing https://doi.org/10.1177/0143034313479697.
longitudinal methodology. Additionally, the sample of students included Chan, D. (2009). So why ask me? Are self-report data really that bad? In C. Lance, &
in this study were all recruited from independent schools within R. Vandenberg (Eds.), Statistical and methodological myths and urban legends: Doctrine,
verity and fable in the organizational and social sciences (1st ed., pp. 309–333). New
Australia, therefore the geographical and socioeconomic attributes of York: Taylor & Francis Group.
the participants may limit the generalisability of the findings. Future Darley, J. M., & Latan�e, B. (1968). Bystander intervention in emergencies: Diffusion of re-
research should seek to use a cross-national sample. sponsibility. Journal of Personality and Social Psychology, 8, 377–383. https://doi.
org/10.1037/h0025589.
DeSmet, A., Bastiansens, S., VanCleemput, K., Poels, K., Vandebosch, H., & De
Bourdeaudhu, I. (2012). Mobilizing bystanders of cyberbullying: An exploratory
4.3. Conclusion study into behavioral determinants of defending the victim. Studies in Health
Technology and Informatics, 181, 58–63. https://doi.org/10.3233/978-1-61499-121-
2-58.
Despite these limitations, the study is firmly embedded in the theo­ DeSmet, A., Veldeman, C., Poels, K., Bastiaensens, S., Van Cleemput, K., Vandebosch, H.,
retical framework of social cognitive theory (Bandura, 1986, p. 1997) et al. (2014). Determinants of self-reported bystander behavior in cyberbullying
and the results of this study represent a significant contribution to the incidents amongst adolescents. Cyberpsychology, Behavior, and Social Networking, 17
(4), 207–215. https://doi.org/10.1089/cyber.2013.0027.
literature on defenders of cyberbullying victims. In sum, the current Di Giunta, L., Eisenberg, N., Kupfer, A., Steca, P., Tramontano, C., & Caprara, G. (2011).
findings suggest that both defending self-efficacy and empathic Assessing perceived empathic and social self-efficacy across countries. European
self-efficacy are vital beliefs required for maximising frequency of Journal of Psychological Assessment, 26(2), 77–86. https://doi.org/10.1027/1015-
5759/a000012.
bystander intervention during cyberbullying episodes. This study
Dillon, K., & Bushman, B. (2015). Unresponsive or un-noticed?: Cyberbystander
therefore provides a strong foundational basis for further empirical intervention in an experimental cyberbullying context. Computers in Human Behavior,
research as well as designing future intervention programs aimed at 45, 144–150. https://doi.org/10.1016/j.chb.2014.12.009.
Dredge, R., Gleeson, J., & de la Piedad Garcia, X. (2014). Cyberbullying in social
promoting defending of cyberbully victims.
networking sites: An adolescent victim’s perspective. Computers in Human Behavior,
36, 13–20. https://doi.org/10.1016/j.chb.2014.03.026.
Funding Eklund, J., Loeb, C., Hansen, E. M., & Andersson-Wallin, A. C. (2012). Who cares about
others?: Empathic self-efficacy as an antecedent to prosocial behavior. Current
Research in Social Psychology, 2, 31–41.
This research did not receive any specific grant from funding Freis, S., & Gurung, R. (2013). A Facebook analysis of helping behavior in online
agencies in the public, commercial or not-for-profit sectors. bullying. Psychology of Popular Media Culture, 2(1), 11–19. https://doi.org/10.1037/
a0030239.
G�amez-Guadix, M., Villa-George, F., & Calvete, E. (2014). Psychometric properties of the
CRediT authorship contribution statement cyberbullying questionnaire (CBQ) among Mexican adolescents. Violence & Victims,
29(2), 232–247. https://doi.org/10.1891/0886-6708.vv-d-12-00163r1.
Huang, Y., & Chou, C. (2010). An analysis of multiple factors of cyberbullying among
Madeleine Clark: Conceptualization, Methodology, Formal anal­ junior high school students in Taiwan. Computers in Human Behavior, 26(6),
ysis, Writing - original draft. Kay Bussey: Conceptualization, Method­ 1581–1590. https://doi.org/10.1016/j.chb.2010.06.005.
ology, Formal analysis, Writing - original draft. Jadambaa, A., Thomas, H., Scott, J., Graves, N., Brain, D., & Pacella, R. (2019).
Prevalence of traditional bullying and cyberbullying among children and adolescents
in Australia: A systematic review and meta-analysis. Australian and New Zealand
References Journal of Psychiatry, 53(9), 878–888. https://doi.org/10.1177/
0004867419846393.
Kim, S., Colwell, S., Kata, A., Boyle, M., & Georgiades, K. (2017). Cyberbullying
Alipan, A., Skues, J., Theiler, S., & Wise, L. (2019). Defining cyberbullying: A
victimization and adolescent mental health: Evidence of differential effects by sex
multifaceted definition based on the perspectives of emerging adults. International
and mental health problem Type. Journal of Youth and Adolescence, 47(3), 661–672.
Journal of Bullying Prevention. https://doi.org/10.1007/s42380-019-00018-6.
https://doi.org/10.1007/s10964-017-0678-4.
Allison, K., & Bussey, K. (2017). Individual and collective moral influences on
Kowalski, R., Limber, S., & McCord, A. (2019). A developmental approach to
intervention in cyberbullying. Computers in Human Behavior, 74, 7–15. https://doi.
cyberbullying: Prevalence and protective factors. Aggression and Violent Behavior, 45,
org/10.1016/j.chb.2017.04.019.
20–32. https://doi.org/10.1016/j.avb.2018.02.009.
Australian Curriculum Assessment and Reporting Authority. (2016). My school website.
Lenhart, A., Madden, M., Smith, A., Purcell, K., Zickuhr, K., & Rainie, L. (2011). Teens,
Retrieved from www.myschool.edu. (Accessed 20 February 2019).
kindness and cruelty on social network sites. Washington, DC: Pew Research Centre.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Li, Q., Smith, P. K., & Cross, D. (2012). Research into cyberbullying: Context. In Q. Li,
Englewood Cliffs, NJ: Prentice-Hall.
D. Cross, & P. K. Smith (Eds.), Cyberbullying in the global playground: Research from
Bandura, A. (1997). Self-efficacy. New York, NY: Freeman.
international perspectives (pp. 3–12). West Sussex: Wiley-Blackwell.
Barchia, K., & Bussey, K. (2011). Predictors of student defenders of peer aggression
Machackova, H., Dedkova, L., & Mezulanikova, K. (2015). Brief report: The bystander
victims: Empathy and social cognitive factors. International Journal of Behavioral
effect in cyberbullying incidents. Journal of Adolescence, 43, 96–99. https://doi.org/
Development, 35(4), 289–297. https://doi.org/10.1177/0165025410396746.
10.1016/j.adolescence.2015.05.010.
Barli�nska, J., Szuster, A., & Winiewski, M. (2013). Cyberbullying among adolescent
Menesini, E., Nocentini, A., Palladino, B., Fris� en, A., Berne, S., Ortega-Ruiz, R., et al.
bystanders: Role of the communication medium, form of violence, and empathy.
(2012). Cyberbullying definition among adolescents: A comparison across six
Journal of Community & Applied Social Psychology, 23(1), 37–51. https://doi.org/
European countries. Cyberpsychology, Behavior, and Social Networking, 15(9),
10.1002/casp.2137.
455–463. https://doi.org/10.1089/cyber.2012.0040.
Barli�nska, J., Szuster, A., & Winiewski, M. (2015). The role of short- and long-term
Moxey, N., & Bussey, K. (2019). Styles of bystander intervention in cyberbullying
cognitive empathy activation in preventing cyberbystander reinforcing
incidents. International Journal of Bullying Prevention. https://doi.org/10.1007/
cyberbullying behavior. Cyberpsychology, Behavior, and Social Networking, 18(4),
s42380-019-00039-1.
241–244. https://doi.org/10.1089/cyber.2014.0412.
Nickerson, A., Mele, D., & Princiotta, D. (2008). Attachment and empathy as predictors
Barli�nska, J., Szuster, A., & Winiewski, M. (2018). Cyberbullying among adolescent
of roles as defenders or outsiders in bullying interactions. Journal of School
bystanders: Role of affective versus cognitive empathy in increasing prosocial
Psychology, 46(6), 687–703. https://doi.org/10.1016/j.jsp.2008.06.002.
cyberbystander behavior. Frontiers in Psychology, 9. https://doi.org/10.3389/
Nocentini, A., Calmaestra, J., Schultze-Krumbholz, A., Scheithauer, H., Ortega, R., &
fpsyg.2018.00799.
Menesini, E. (2010). Cyberbullying: Labels, behaviors and definition in three
Bastiaensens, S., Vandebosch, H., Poels, K., Van Cleemput, K., DeSmet, A., & De
European countries. Australian Journal of Guidance and Counselling, 20(2), 129–142.
Bourdeaudhuij, I. (2014). Cyberbullying on social network sites. An experimental
https://doi.org/10.1375/ajgc.20.2.129.
study into bystanders’ behavioral intentions to help the victim or reinforce the bully.
Obermaier, M., Fawzi, N., & Koch, T. (2016). Bystanding or standing by? How the
Computers in Human Behavior, 31, 259–271. https://doi.org/10.1016/j.
number of bystanders affects the intention to intervene in cyberbullying. New Media
chb.2013.10.036.
& Society, 18(8), 1491–1507. https://doi.org/10.1177/1461444814563519.
Bastiaensens, S., Vandebosch, H., Poels, K., Van Cleemput, K., DeSmet, A., & De
Olenik-Shemesh, D., Heiman, T., & Eden, S. (2017). Bystanders’ behavior in
Bourdeaudhuij, I. (2015). ‘Can I afford to help?’ How affordances of communication
cyberbullying episodes: Active and passive patterns in the context of personal-socio-
modalities guide bystanders’ helping intentions towards harassment on social
emotional factors. Journal of Interpersonal Violence, 32(1), 23–48. https://doi.org/
network sites. Behavior & Information Technology, 34(4), 425–435. https://doi.org/
10.1177/0886260515585531.
10.1080/0144929x.2014.983979.
P€
oyh€ onen, V., Juvonen, J., & Salmivalli, C. (2010). What does it take to stand up for the
Caprara, G. V., Alessandri, G., & Eisenberg, N. (2012). Prosociality: The Contribution of
victim of bullying?: The interplay between personal and social factors. Merrill-Palmer
traits, values, and self-efficacy beliefs. Journal of Personality and Social Psychology,
Quarterly, 56(2), 143–163, 1353/mpq.0.0046.
102, 1289–1303. https://doi.org/10.1037/a0025626.

6
M. Clark and K. Bussey Computers in Human Behavior 109 (2020) 106340

P€
oyh€onen, V., & Salmivalli, C. (2008). New directions in research and practice Suchy, C. M., & Tomasino, K. K. (2014). Types of defending behaviors among late
addressing bullying: Focus on defending behavior. In D. J. Pepler, & W. M. Craig elementary and middle school students. In Paper presented at the national conference
(Eds.), Understanding and addressing bullying: An international perspective (pp. 26–42). of undergraduate research. Lexington: University of Kentucky.
Bloomington: Author House. Thornberg, R., & Jungert, T. (2012). Bystander behavior in bullying situations: Basic
Pozzoli, T., & Gini, G. (2013). Why do bystanders of bullying help or not? A moral sensitivity, moral disengagement and defender self-efficacy. Journal of
multidimensional model. The Journal of Early Adolescence, 33(3), 315–340. https:// Adolescence, 36(3), 475–483. https://doi.org/10.1016/j.adolescence.2013.02.003.
doi.org/10.1177/0272431612440172. Thornberg, R., W€ anstr€
om, L., Hong, J., & Espelage, D. (2017). Classroom relationship
Sainio, M., Veenstra, R., Huitsing, G., & Salmivalli, C. (2010). Victims and their qualities and social-cognitive correlates of defending and passive bystanding in
defenders: A dyadic approach. International Journal of Behavioral Development, 35(2), school bullying in Sweden: A multilevel analysis. Journal of School Psychology, 63,
144–151. https://doi.org/10.1177/0165025410378068. 49–62. https://doi.org/10.1016/j.jsp.2017.03.002.
Salmivalli, C. (1999). Participant role approach to school bullying: Implications for Tokunaga, R. (2010). Following you home from school: A critical review and synthesis of
interventions. Journal of Adolescence, 22(4), 453–459. https://doi.org/10.1006/ research on cyberbullying victimization. Computers in Human Behavior, 26(3),
jado.1999.0239. 277–287. https://doi.org/10.1016/j.chb.2009.11.014.
Salmivalli, C. (2010). Bullying and the peer group: A review. Aggression and Violent Van Cleemput, K., Vandebosch, H., & Pabian, S. (2014). Personal characteristics and
Behavior, 15(2), 112–120. https://doi.org/10.1016/j.avb.2009.08.007. contextual factors that determine “helping,” “joining in,” and “doing nothing” when
Salmivalli, C., Voeten, M., & Poskiparta, E. (2011). Bystanders matter: Associations be- witnessing cyberbullying. Aggressive Behavior, 40(5), 383–396. https://doi.org/
tween reinforcing, defending, and the frequency of bullying behavior in classrooms. 10.1002/ab.21534.
Journal of Clinical Child and Adolescent Psychology, 40, 668–676. https://doi.org/ You, L., & Lee, Y. (2019). The bystander effect in cyberbullying on social network sites:
10.1080/15374416.2011.597090. Anonymity, group size, and intervention intentions. Telematics and Informatics, 45,
Solberg, M. E., & Olweus, D. (2003). Prevalence estimation of school bullying with the 101284. https://doi.org/10.1016/j.tele.2019.101284.
Olweus bully/victim questionnaire. Aggressive Behavior, 29, 239–268. https://doi. Zimmerman, B. (2000). Self-Efficacy: An essential motive to learn. Contemporary
org/10.1002/ab.10047. Educational Psychology, 25(1), 82–91. https://doi.org/10.1006/ceps.1999.1016.

You might also like