Advertisement

The Influence of School Climate and Empathy on Cyberbystanders’ Intention to Assist or Defend in Cyberbullying

  • Anja Schultze-KrumbholzEmail author
  • Pavle Zagorscak
  • Markus Hess
  • Herbert Scheithauer
Original Article

Abstract

Bystanders play a crucial role in aggressive behavior in group contexts. Cyberbystanders can react in proactive ways such as assisting a cyberbully or defending the victim. Since school relationships spill over to the online world, the school context is likely to influence students’ online behavior. In the present study, we examine the influence of school climate on active cyberbystanding behavior beyond the influence of individual social competences. Participants were 726 students from a more comprehensive study, who were classified as non-victims and non-bullies (i.e., pure bystanders). They were from 36 classes in five schools in Germany (Grades 7–10; Mage = 13.37 years, SDage = 1.01 years, 53.3% female). Two hypothetical scenarios were used to operationalize active bystanding: one for assisting in cyberbullying and one for defending the cybervictim. Separate multilevel analyses were conducted to predict assisting in cyberbullying and defending the victim on the individual level, respectively. Individual-level independent variables were affective and cognitive empathy. Class-level independent variables were assessed with six subscales of the Inventory of School Climate (ISC-S; Brand et al. 2003). Age, gender, class-level offline and online bullying, and offline and online victimization were controlled for. Results showed cognitive and affective empathy, lack of positive peer interactions, and offline bullying to predict assisting. Defending was predicted by cognitive and affective empathy, lack of safety problems, and teacher support. School plays a role in students’ online bystander behavior and interventions should aim to foster a supportive school climate.

Keywords

Cyberbullying Bystander Empathy School climate Adolescence 

Background

Cyberbullying is a subtype of aggressive behavior often defined as “any behavior performed through electronic or digital media by individuals or groups that repeatedly communicates hostile or aggressive messages intended to inflict harm or discomfort on others” (Tokunaga 2010, p. 278). It also includes an imbalance of power between perpetrator and victim with the perpetrator being superior to the victim due to one of multiple possible characteristics, but especially through publishing materials that are not easy for the victim to remove. While cyberbullying displays a large conceptual overlap with offline bullying, the use of omnipresent internet-ready devices for cyberbullying allows for aggressive behaviors that are performed anyplace, anytime, with large audiences, potentially anonymously, and with decreased inhibition due to a lack of direct feedback (Slonje and Smith 2008). As a result, individuals perceive cyberbullying scenarios as worse or more threatening than traditional bullying scenarios (Sticca and Perren 2013).

Cyberbullying is a common phenomenon in adolescence with reported median prevalence rates across studies of 15% for cyberbullying perpetration and 23% for cybervictimization (Hamm et al. 2015). Further, a meta-analysis highlights that both perpetration and victimization are associated with a series of internalizing and externalizing problems, emphasizing the phenomenon’s impact on adolescents’ mental and physical health and the need for effective preventive strategies (Fisher et al. 2016).

Findings from the offline bullying context suggest that preventive interventions need to take the group dynamics of peer aggression into account to be efficacious (Pepler et al. 2010; Salmivalli et al. 1996). Adolescents that repeatedly initiate aggressive acts (i.e., “perpetrators”) against others (i.e., “victims”) are oftentimes supported by other individuals that fulfill the role of an active audience (e.g., by liking and sharing the perpetrator’s aggressive content) or that take part in the behavior initiated by the perpetrator. Further, there are individuals who are labeled as “bystanders” or “outsiders”, who are neither targeted by the aggression nor do they participate in it. Recent empirical studies have confirmed the existence of perpetrators/victims, assistants, and different forms of bystanders in the cyberbullying context (e.g., Schultze-Krumbholz et al. 2018a). Prevention researchers have especially focused on the role of bystanders due to their influence on social norms of the entire group and the indirect reinforcement of aggressive acts through a perceived passive consent to what is happening (Paluck and Shepherd 2012). Encouraging bystanders to leave their roles as passive onlookers (i.e., passive bystanders) in order to become defenders of the victims (i.e., active bystanders) is seen as a promising strategy to reduce the incidence and the consequences of cyberbullying (High and Young 2018; Kazerooni et al. 2018).

However, knowledge about environmental, personal, or circumstantial factors that influence which participant role an individual takes in the cyberbullying context is limited. Studies on these factors are required to foster environments in schools that contribute to an increase in bystander behavior supporting the victims and discouraging the perpetrators of cyberbullying. Although research focusing on cyberbystanding behavior is still limited, according to result from traditional bullying and bullying bystander research, one can assume that an interplay of personal and environmental factors is relevant for bullying and bystanding behaviors and that—on the person level—especially skills are relevant that are necessary for the understanding of and interaction in social situations, such as perspective taking and empathy (e.g., Ettekal et al. 2015; Gini et al. 2007), “self-oriented personal competencies” (Zych et al. 2019a, p. 4), or affect- and empathy-related variables (e.g., Trach and Hymel 2019).

From a theoretical perspective, the concurrent inclusion of individual and contextual factors into the analysis of behavior in cyberbullying situations is based on assumptions of the socio-ecological model proposed by Bronfenbrenner (e.g., Bronfenbrenner and Crouter 1983) where individual competencies and dispositions are embedded in different systems such as the school (class) environment. On a more applied level, the social and emotional learning (SEL) framework offers a valuable approach to justify the simultaneous considerations of individual and contextual aspects in explaining social behavior in general. Similar to the socio-ecological approach, SEL highlights the importance of the interplay between individual characteristics like social awareness and school-level aspects like a safe and positive school climate for effectively promoting social and emotional learning and respective behavior (Weissberg et al. 2015).

School Climate, Cyberbullying, and Cyberbystander Behavior

Recent trends in the (cyber)bullying research program are to move away from the perspective that behavior in bullying situations is solely associated with individual factors and towards a more social-ecological perspective, in which factors of the social context play a stronger role in explaining students’ roles and actions in (cyber)bullying (Smith et al. 2019). One of these social contexts for adolescents is the school where they spend a significant amount of their time and which is characterized by the school climate. Moreover, school climate is a factor that can actively be changed by a school personnel (Ma 2010). The term school climate comprises all aspects that are related to the school experience such as teaching and learning quality, social relationships, structural features, and beliefs, attitudes, values, and norms, but it is not used consistently in research on bullying and cyberbullying. School climate is often viewed as a multidimensional construct and has been categorized into four dimensions: academic, community (referring to relationships), safety, and institutional environment (Wang and Degol 2016). However, it is important to separate school-level from classroom-level aspects of social climate at school. Whereas the former focuses on aspects such as the feeling of safety at school, the latter highlights the role of the protective quality of social interactions within the classroom (Thornberg et al. 2018).

In recent years, the amount of studies that relate perceptions of school climate to cyberbullying involvement has increased. Previous studies have shown that students who are involved in cyberbullying rated their school environment less positively (Bayar and Ucanok 2012) and that both cybervictims and perpetrators feel less safe at their schools (Sourander et al. 2010). More specifically, cybervictims reported a lower sense of belonging to their schools and lower commitment to school than non-victims. They also have difficulties in building a positive school relationship (Patchin and Hinduja 2010). This negative perception might lead to social withdrawal and drop-out from school (Wachs 2012). In a recent meta-analysis, Guo (2016) found an effect size of r = .13 for the predictive value of negative school climate on cyberbullying perpetration and a similar effect size of r = .15 for cyberbullying victimization. However, in comparison with individual predictors like behavioral problems or prior offline bullying experiences, the effect of the contextual factor “negative school climate” was rather low. To gain more insight into underlying processes that link school climate to cybervictimization, Veiga Simão et al. (2017) investigated cybervictims’ perceptions of school climate with 218 early and middle adolescents. The main assumption was that in a positive school environment, cybervictims are more willing to tell teachers about their experiences (Smith et al. 2004). The results corroborated earlier findings that cybervictimization is related to a perception of negative school climate and confirmed their hypothesis that victims who perceive their school climate as more positive are more willing to seek help from their teachers. However, as most of the studies in the field were cross-sectional, the causal relationship between perception of school climate and involvement in cyberbullying remains unclear. One prospective study by Cappadocia et al. (2013) revealed no predictive value of school climate on cybervictimization 1 year later. In a more recent longitudinal study with two measurement points conducted with early adolescents, Holfeld and Leadbeater (2017) applied cross-lagged panel modeling to investigate the causal relationship between school climate and cybervictimization. Controlling for offline bullying, they found that positive perceptions of overall school climate as well as different facets of positive school climate (e.g., teacher-pupil interaction) at the beginning of a school year predict less cybervictimization at the end of the school year. In summary, even though individual factors might show higher relations to cybervictimization and perpetration, school climate seems to offer incremental and unique contributions to their explanation.

To date, research on the link between school climate and bystander behavior in bullying and cyberbullying episodes is limited. For example, Thornberg et al. (2018) investigated the role of an authoritative classroom climate (characterized for example as warm, demanding, cohesive, and supportive) on offline victimization and bystander behavior. Results of their study with early adolescents (mean age 11.5 years) showed that authoritative classrooms are related to more positive reactions of bystanders witnessing a bullying episode.

Although school staff and administrators frequently believe that cyberbullying intervention and prevention lies outside of their responsibility (Englander 2012), research demonstrated that students continue their school relationships outside of school and that they communicate through digital media when not in school (Subrahmanyam et al. 2006). Cyberbullying has shown ties to school relationships in the past (Hinduja and Patchin 2012). Moreover, the significant overlap between offline and cyberbullying and offline and cybervictimization (Waasdorp and Bradshaw 2015; del Rey et al. 2012) as well as the findings that a substantial part of cyberbullies were actually classmates of the victims (e.g., Juvonen and Gross 2008) imply that school factors might actually play a significant role in the dynamics of cyberbullying.

There are no studies that we are aware of that have explicitly examined the relationship between cyberbystanding and school climate so far. This might seem logical when assuming that there is little to no connection between the offline school world and the online context, which is supposedly private.

Domínguez-Hernández et al. (2018) collated constructs in their systematic review, which could be viewed as indicators of or to be related to school climate. For example, peer relationships, especially friendships, influence a cyberbystander’s response depending on who the bystander is friends with: the victim or the perpetrator. Group norms, attitudes, and expectations of peers and teachers also seem to be relevant determinants: groups endorsing pro-aggression attitudes, for example, are more likely to reinforce joining in or assisting in cyberbullying.

Cognitive and Affective Empathy as Predictors of Cyberbystander Behavior

Empathy in general is discussed as one of the key components of social-emotional competence promoting appropriate social behavior (Eisenberg et al. 2010). Thus, cognitive and affective aspects of empathy represent individual factors, which have repeatedly been associated with cyberbullying involvement in research (Zych et al. 2019a). As the present study tries to investigate the impact of social context above and beyond individual social-emotional competence contributing to the level of involvement in cyberbullying incidents, cognitive an affective empathy was included in the analysis as a main individual characteristic guiding social behavior. In general, the construct of empathy can be divided into a cognitive component (i.e., understanding the emotions of others) and an affective component (i.e., the emotional response to others’ emotions) (cf. Zurek and Scheithauer 2017).

Robust empirical evidence exists for empathy as a general correlate of cyberbullying perpetration (e.g., Schultze-Krumbholz and Scheithauer 2013). According to a recent meta-analysis, which synthesizes the results of 25 studies, perpetrators of cyberbullying show reduced average scores on affective and cognitive empathy when compared with non-affected individuals, whereas victims of cyberbullying do not seem to differ from non-affected individuals in that regard (Zych et al. 2019b). Further, first efficacy trials suggest that preventive interventions can simultaneously increase affective empathy and reduce cyberbullying stressing the importance of the construct in the prevention context (e.g., Schultze-Krumbholz et al. 2016).

Compared with the number of studies on the associations between empathy and cyberbullying perpetration or cybervictimization, there are fewer findings on the influence of empathy on bystander behavior and the results are partly inconsistent. For example, Macháčková and Pfetsch (2016) found that affective empathy, but not cognitive empathy, predicted support for victims of cyberbullying by bystanders, whereas another study found only cognitive empathy to be predictive of favorable bystander responses (Barlińska et al. 2018). In contrast, a latent class analysis by Schultze-Krumbholz et al. (2018a) found significant associations between both affective and cognitive empathy and membership in a prosocial bystander class.

Concordantly, two recent reviews on factors influencing bystander behavior summarized that empathy is an important correlate of bystander response to cyberbullying (Allison and Bussey 2016; Domínguez-Hernández et al. 2018). However, Allison and Bussey (2016, p. 189) additionally cautioned that “more conclusive evidence is needed to determine whether empathy inductions can reliably increase bystander intervention through increasing empathy”.

An additional issue especially relevant for informing preventive strategies refers to the concurrent contribution of individual versus contextual factors in predicting bystander behavior. As discussed above, recent efforts to reduce cyberbullying have highlighted the role of social environmental factors compared with the impact of individual factors like empathy (Smith et al. 2019). The behavior of individuals in a group cannot be attributed solely to individual characteristics, but might be subject to group influences, which model the adequate behavior within this group even if it is of the antisocial nature. Therefore, the level of (cyber)bullying and (cyber)victimization in the group should also be accounted for because it can function as a descriptive norm about typical behavior in this group. At the same time, bystander behavior cannot be examined without taking into account individual empathy, which has been shown to be a core predictor of bystander behavior. However, recent studies, which included both individual factors like empathy and aspects of school climate to predict either aggressive or prosocial behavior, have found direct links between empathy and behavior but not between aspects of school climate and behavior (Barr and Higgins-D’Alessandro 2007; Batanova and Loukas 2016). These studies again highlight the role of empathy in predicting aggressive and prosocial behavior. The present study might further advance the knowledge about the concurrent role of the individual factor empathy and contextual aspects of school climate in predicting bystander behavior in cyberbullying incidents.

The Present Study

In the present study, we examine the influence of school climate on active bystanding behavior, i.e., assisting the bully or defending the victim, over and above individual social competences using multilevel modeling. Thereby, we focus on the school climate dimensions “community” and “safety”. We investigate the question whether positive school climate predicts positive bystander behavior when witnessing cyberbullying episodes that also take place outside the school context. In cyberbullying, it is difficult to differentiate the reinforcers from the assistants, as proposed by Salmivalli et al. (1996) for offline bullying. Because reinforcing behavior like giving “likes” and commenting on a cyberbully’s comments is also active bystander behavior, we focus on one form and example of negative active bystanding. Dependent variables were assisting (i.e., negative active bystanding) and defending behavior (i.e., positive active bystanding) at the individual level. As associations between bystanding and empathy were found in previous analyses of the present data (Schultze-Krumbholz et al. 2018a) and other studies, and to be able to analyze the concurrent contribution of individual and contextual factors to different bystander behavior, cognitive and affective empathy on the individual level were included as independent variables, while age and gender were used as control variables. At the class level, the independent variables were different subscales of school climate (teacher support, consistency and clarity of rules and expectations, negative peer interactions, positive peer interactions, support for cultural pluralism, and safety problems) with offline/online bullying and victimization as control variables. Based on previous research mentioned above, we expected that high cognitive and affective empathy predict positive bystander behavior like defending and that low cognitive and affective empathy predict negative bystander behavior like assisting. Based on the reviewed research, we expected positive school climate indicators to predict positive bystander behavior and negative school climate indicators to predict negative bystander behavior. We did not have directed hypotheses regarding the impact of the contextual vs. the individual factors.

Method

Participants

The present data were collected as part of a more comprehensive 3-wave evaluation study of a school-based cyberbullying prevention program (Schultze-Krumbholz et al. 2018b). Only the data from the first wave (baseline assessment) were utilized for the present analyses in order to prevent distortions caused by intervention effects. Initially, 897 seventh- to tenth-grade students from 36 classes in five schools in a major German city participated in the present study. Only those classified as non-involved (cut-off score for cybervictimization and cyberbullying: “once or twice a month” or more often) were included in the present analyses leaving a total sample of 726 adolescents. They were on average 13.37 years old (SDage = 1.01, range 11 to 17 years), and 53.3% were female, 44.6% were male, and 2.1% did not indicate their gender. Overall, 683 (missing rate = 5.9%) adolescents provided complete data on all indicators included in the assisting behavior analysis and 679 (missing rate = 6.5%) provide complete data for the defending behavior analyses.

Procedure

In line with the local regulations for empirical studies in schools, active and informed parental consent was collected for students aged 14 and younger. Moreover, active and informed consent was obtained from all individual participants included in the study. Standardized self-report paper-and-pencil questionnaires were distributed and completed during regular school lessons. A member of the research team was present during data assessment in each class, was available for questions, and collected the questionnaires. All participants were guaranteed anonymity and voluntariness. The procedure was approved by the responsible school administration of the federate state, where data collection took place.

Measures

The overall study assessed a variety of indicators believed to be related to cybervictimization and cyberbullying. In the present study, cognitive and affective empathy on the individual level were used to predict individual assisting and defending, and indicators of school climate on the class level were used to examine effects on individual assisting and defending. Gender and age were included as control variables on the individual level, and offline bullying and victimization were included as control variables on the class level.

Student-Level Measures

Assisting in cyberbullying was assessed using a single item with a seven-point answer scale (1 = not at all likely to 7 = very likely). The item was developed to assess behavioral willingness based on Gibbons et al. (1998) and read “Suppose your friends are thinking about sending a threatening message or posting an insulting video or photo of a person that none of you likes on the internet. How likely is it that you like the idea and assist them?”

Analogous to assisting, defending against cyberbullying was measured using a single item (1 = not at all likely to 7 = very likely) developed to assess behavioral willingness based on Gibbons et al. (1998). The item was “Suppose you see that someone is being threatened, insulted or made fun of in an online group or chatroom. All the group or chat members dislike this person. How likely is it that you will intervene and try to defend this person?” We chose this approach rather than a multiple-item scale to present ecologically valid situations with more context information to students where they had to make a response decision.

Cognitive empathy was measured with the 8-item perspective-taking subscale of the Interpersonal Reactivity Index (Davis 1980; German translation: Lamsfuss et al. 1992) using a 5-point answer scale (1 never true to 5 almost always true), Cronbach’s α = .85. An example is “I sometimes try to understand my friends better by imagining how things look from their perspective”.

Affective empathy was operationalized as adolescents’ tendency to show affective empathy in a given situation. This was assessed by presenting a stimulus situation from the Sympathy Reactivity Questionnaire (Volland et al. 2008), which we adapted for the cyber context: “Imagine you are on Facebook and come across a hate group against a person where others post mean and insulting comments. You find out that the person knows about this group. What would you think?” After reading this stimulus situation, adolescents answered questions about how they would react emotionally on a 6-spoint scale (1 not at all true to 6 completely true), e.g., “Because I found out about this group I am sad myself.” Only four of the initial seven items were included into scale computation (being touched by the situation, caring about whether the person is doing better soon, being sad oneself about the situation, and worrying about the person) to include only the affective components in the analyses; Cronbach’s α = .80. Negatively worded items were reverse coded.

Class-Level Measures

School climate was assessed using a selected number of subscales of the Inventory of School Climate-Student (ISC-S; Brand et al. 2003); specifically teacher support, consistency and clarity of rules and expectations, negative peer interactions, positive peer interactions, support for cultural pluralism, and safety problems. All scales were aggregated on the class level using mean scores. The subscale “teacher support” consisted of six items (e.g., “Teachers go out of their way to help students.”) and used a 5-point answer scale (1 never to 5 always), Cronbach’s α = .84. Consistency and clarity of rules and expectations comprised five items (e.g., “When teachers make a rule, they mean it.”) with a 5-point answer scale (1 never to 5 always), Cronbach’s α = 76. Negative peer interactions measured answers to five items like “Students in this school are mean to each other.” on a 5-point answer scale (1 never to 5 always), Cronbach’s α = 78. Positive peer interactions referred to five statements like “Students enjoy working together on projects in classes.” (5-point answer scale; 1 never to 5 always), Cronbach’s α = 76. Support for cultural pluralism consisted of four items, which were to be answered on a 4-point answer scale (1 never to 4 often), e.g., “You get to do something which helps you learn about students of different origins and cultures at your school.”, Cronbach’s α = .68. The original instrument used the term “race” instead of “origin”, but this term is not commonly used nor socially acceptable in Germany. Safety problems were assessed with six items (4-point answer scale, 1 never to 4 often), e.g., “How often have you been afraid that someone will hurt or bother you at school?”, Cronbach’s α = .85.

Offline bullying and victimization were measured with the bullying and victimization scales of the European Cyberbullying Intervention Project Questionnaire (ECIPQ; Brighi et al. 2012; Del Rey et al. 2015). Each scale comprised seven items rated on 5-point answer scales (0 never to 4 more than once a week), e.g., “I said mean things to someone or insulted them” and “Someone said mean things to me or insulted me”; Cronbach’s α = .80 for bullying and α = .77 for victimization. To assess offline bullying and victimization on a class level, the mean score was computed for each class.

Cyberbullying and cybervictimization were assessed using the respective scales of the ECIPQ (Brighi et al. 2012; Del Rey et al. 2015). Each scale consisted of 11 self-report behavior-based items to be answered on a 5-point answer scale (0 never to 4 more than once a week), e.g., “I said nasty things to someone or called them names using texts or online messages” and “Someone said nasty things to me or called me names using texts or online messages”; Cronbach’s α = .88 for cyberbullying and α = .70 for cybervictimization. This measure was used to classify students as involved vs. non-involved and to select the subsample for the present analyses based on cut-off scores (sum scores of 2 or less). To assess cyberbullying and cybervictimization on a class level, the mean score was computed for each class.

Data Analysis

Separate multilevel analyses were conducted to examine the influence of class-level variables on assisting in cyberbullying and defending the victim on the individual level, respectively. We computed different successive models. The baseline model (1) did not contain any predictors, but was used to estimate the intra-class correlations and the proportion of variance at the different levels. The second model (2a) was a random intercepts model estimating class and individual effects on defending and assisting, respectively. Additionally, this model was computed again, this time including class-level offline bullying and victimization as control variables (2b). Random intercepts and random slopes models (3) were computed in a third step to examine effects of the class-level variables on the associations between the individual-level variables. All independent variables were centered around their grand mean before being entered into the prediction. Due to skewness of some of the variables, the maximum likelihood robust (MLR) estimator was used. Gender and age were included as control variables at the individual level.

To predict cyberbystander behavior, only participants were included in the analytical sample, who indicated “never” or “only once or twice” being involved in cyberbullying as a perpetrator or victim. Class-level scores were computed based on the data from all participants (i.e., including cyberbullies and cybervictims) to adequately depict the atmosphere in the class. Descriptive statistics were computed using the statistics software SPSS 25 (IBM Corp 2017). Multilevel modeling was conducted with the statistics program Mplus 8 (Muthén and Muthén 1998-2017).

Results

Descriptive Results

The descriptive results (see Table 1) showed that on average, non-involved students perceived themselves as rather unlikely to assist in cyberbullying behavior with a mean of 2.57, which is clearly below the theoretical scale mean score of 4.0. In contrast, non-involved students reported to be more likely to defend others in cyberbullying situations with the mean score of 3.95 being only slightly below the theoretical scale mean score of 4.0. The mean scores showed that offline bullying and victimization were not highly prevalent among adolescents, who were not involved in cyberbullying, given that the means ranged from .07 to .48 for offline bullying and from .13 to .62 for offline victimization, while a score between 0 and 4 was theoretically possible.
Table 1

Means and standard deviations of the study variables

Variable

M

SD

Range

Level 1 (Individual) (n = 726)

  Assisting in cyberbullying

2.57

1.53

1–7

  Defending against cyberbullying

3.95

1.66

1–7

  Cognitive empathy

3.21

0.68

1–5

  Affective empathy

3.68

1.14

1–6

Level 2 (Class) (n = 36)

  Teacher support

2.79

0.34

1–5

  Clarity of rules

3.57

0.30

1–5

  Negative peer interactions

2.60

0.22

1–5

  Positive peer interactions

3.43

0.27

1–5

  Support for cultural pluralism

2.98

0.19

1–4

  Safety problems

1.33

0.19

1–4

  Offline bullying

0.34

0.15

0–4

  Offline victimization

0.40

0.13

0–4

  Cyberbullying

.10

.10

0–4

  Cybervictimization

.10

.05

0–4

Range refers to the potential answer range provided by the number of answer categories

Due to skewness of the data on some of the variables, correlations between the study variables were computed using Kendall’s Tau (Table 2). The correlations showed associations of age with assisting, empathy, and almost all of the level-2 variables. School climate was rated more negatively with increasing age. Gender correlated with all level-1 variables except defending, and no level-2 variables except support for cultural pluralism. The outcome variable “assisting” was significantly associated with almost all variables except safety problems and cyberbullying and victimization, while defending correlated with the level-1 predictors except the control variables age and gender and only significantly correlated to a small extent with some of the level-2 variables. Assisting and defending correlated negatively, but weakly with each other. Cognitive empathy was not related to any level-2 variable, while affective empathy was associated with some of the climate subscales. Except in two instances, the climate subscales were all significantly related to each other. Offline bullying correlated with the climate subscales more often than did offline victimization. Cyberbullying only correlated with the control variable age and the level-2 variables teacher support, safety problems, and offline bullying. Finally, cybervictimization correlated with age, positive peer interactions and safety problems, and all of the bullying and victimization indicators.
Table 2

Level 1 and Level 2 correlations between the study variables (n = 726)

Variable

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Level 1

               

  1. Age

              

  2. Gender

.02

             

  3. Assisting

.10**

.10**

            

  4. Defending

− .04

− .02

− .17***

           

  5. Cognitive empathy

.04

− .17***

− .19***

.22***

          

  6. Affective empathy

− .10**

− .28***

− .21***

.21***

.31***

         

Level 2

               

  7. Teacher support

− .40***

.00

− .12***

.09**

.01

.08**

        

  8. Clarity of rules

− .36***

.00

− .12***

.06*

− .02

.09**

.44***

       

  9. Negative peer interactions

.08**

.02

.11***

− .04

− .04

− .07*

− .31***

− .34***

      

  10. Positive peer interactions

− .22***

− .01

− .14***

.07**

.05

.08**

.50***

.41***

− .43***

     

  11. Support for cultural pluralism

− .20***

− .07*

− .06*

.01

.01

.04

.17***

.40***

− .09**

.21***

    

  12. Safety problems

− .10**

− .04

.05

− .02

− .03

− .01

.17***

.00

.28***

.01

.17***

   

  13. Offline bullying

.02

.01

.09**

.05

− .02

− .04

− .04

− .24***

.30***

− .13***

− .09**

.23***

  

  14. Offline victimization

− .23***

.00

.06*

.04

− .03

.00

.10***

.00

.21***

− .04

.04

.34***

.33***

  

  15. Cyberbullying

.21***

− .03

.03

.05

.02

0.20

− .13***

− .01

.02

.03

.03

.07**

.27***

.04

 

  16. Cybervictimization

.20***

− .02

.03

.03

.04

.02

− .04

.00

− .02

.06*

− .03

.14***

.14***

.25***

.47***

*p < .05, **p < .01, ***p < .001; correlation coefficients are Kendall-Tau-b

Model 1—Baseline Model

Analyses of the baseline models showed that the ICC for assisting was .049 (σ2 = 2.239, p < .001; τ00 = 0.115, p = .07), indicating that 4.9% of differences between the participating adolescents identified as non-involved could be attributed to differences between classes while 95.1% were due to individual differences. The ICC for defending was .024 (σ2 = 2.688, p < .001; τ00 = 0.065, p < .05), indicating that 2.4% of the differences in defending behavior between the adolescents were due to classroom influences, while 97.6% of the variance in defending could be ascribed to individual differences.

Following recommendations of Lai and Kwok (2015), we computed the design effect based on Kish (1965). The authors conclude from their simulation study that multilevel modeling should be implemented when design effects are larger than 1.1 or when researchers are interested in level-2 effects. The design effects were 1.9 for assisting and 1.5 for defending.

Models 2a and 2b—Random Intercepts

On the individual level, both lower levels of cognitive and of affective empathy predicted willingness to assist in cyberbullying when friends are thinking about harassing a person they dislike on the internet (see Table 3). None of the class variables exerted an influence on individual assisting behavior. When offline and online bullying and victimization were included as control variables (Model 2b, see Table 3), the negative coefficient for positive peer interactions became significant, i.e. individual assisting in cyberbullying was predicted by being in a class where students perceived fewer positive peer interactions. Further, assisting in the cyber context was predicted by higher levels of offline bullying on the class level. Lack of cognitive and affective empathy on the individual level remained significant predictors for a student’s assisting behavior in cyberbullying.
Table 3

Model 2 (random intercept) individual and class-level variables predicting assisting and defending

 

Model 2a

Model 2b

Assisting

Defending

Assisting

Defending

B

SE

B

SE

B

SE

B

SE

Individual variables (L1)

  Age

0.060

0.071

0.073

0.063

0.085

0.082

− 0.005

0.084

  Gender (0 = female, 1 = male)

0.066

0.121

0.248

0.127

0.062

0.121

0.237

0.126

  Cognitive empathy

− 0.358***

0.102

0.550***

0.095

− 0.362***

0.102

0.546***

0.095

  Affective empathy

− 0.264***

0.066

0.276***

0.058

− 0.263***

0.066

0.270***

0.056

Class variables (L2)

  Teacher support

− 0.232

0.529

0.898**

0.347

− 0.410

0.463

0.785*

0.351

  Clarity of rules

0.280

0.502

0.031

0.365

0.502

0.471

0.161

0.397

  Negative peer interactions

− 0.155

0.371

0.840*

0.344

− 0.598

0.452

0.440

0.397

  Positive peer interactions

− 0.635

0.347

0.046

0.433

− 0.727*

0.323

− 0.154

0.347

  Support for cultural pluralism

− 0.390

0.248

− 0.285

0.467

− 0.334

0.259

− 0.266

0.394

  Safety problems

0.626

0.663

− 0.982*

0.388

0.372

0.544

− 1.378**

0.449

  Offline bullying

    

1.235*

0.574

1.317

0.715

  Offline victimization

    

0.482

0.592

0.181

0.667

  Cyberbullying

    

− 0.437

1.162

0.747

1.381

  Cybervictimization

    

0.049

1.739

0.567

1.982

*p < .05, **p < .01, ***p < .001; reported regression coefficients are unstandardized

In contrast, higher scores of cognitive and affective empathy predicted individual willingness to defend a person everyone dislikes in the cyber context. On the class level, higher scores of perceived teacher support and lower scores of perceived safety problems predicted defending behavior. Moreover, higher scores of perceived negative peer interactions predicted higher scores of self-reported willingness to defend in the cyber context. When offline bullying and victimization and cyberbullying and victimization were included as control variables (Model 2b, see Table 3), the coefficient for teacher support remained significant as well as the lack of safety problems. In contrast, negative peer interactions were no longer a significant predictor. The influence of individual cognitive and affective empathy on defending behavior did not change.

Age and gender were controlled for in the analyses. They did not have a significant effect on assisting or defending in cyberbullying on the individual level.

Model 3—Random Intercepts and Random Slopes

To examine whether the associations between variables were equal across classes, we added random slopes for assisting/defending and cognitive and affective empathy to the models. There were no significant effects on the slopes for assisting in cyberbullying as a non-involved adolescent. For defending, the model showed estimation problems. Tentative results suggested that there might be influences of negative peer interactions and offline bullying on the association between affective empathy and defending.

Discussion

The present analyses aimed to examine the influence of class-level perceptions of school climate on active cyberbystanding behavior of otherwise non-involved adolescents while controlling for prominent known individual-level predictors, namely cognitive and affective empathy, as well as age and gender. The results showed that the class influence is comparatively small: 4.9% of the variance in assisting behavior and 2.4% of the variance in defending behavior were explained by the class-level predictors. Thus, both behaviors seem to be much more strongly influenced by individual factors, defending even more so than assisting in cyberbullying. That is, defending is a behavior much more determined by personal competences, while assisting underlies group influences to a larger extent. It seems that class-level factors have a stronger influence on adolescents’ antisocial behavior than on their prosocial behavior, even outside the classroom.

Concerning negative bystander behavior, i.e., assisting in cyberbullying, there was no gender effect, which is consistent with previous studies that simultaneously examined the influence of cognitive and affective empathy (e.g., Barlińska et al. 2013). The experimental studies by Barlińska et al. (2013) moreover showed that activating cognitive or affective empathy reduced willingness to assist in cyberbullying. Similar results were found in our study: higher levels of cognitive and affective empathy on the individual level were associated with lower levels of assisting cyberbystander behavior. However, this stands in contrast to other studies that found no significant associations between both dimensions of empathy and reinforcement of cyberbullying (e.g., Macháčková and Pfetsch 2016). From our results, we tentatively conclude that cognitive and affective empathy might be seen as inhibitory factors against actively supporting cyberbullying behavior. Regarding influences of the social environment, being in a class where students perceived less positive peer interactions fostered assisting cyberbystanding behavior. This could be interpreted in light of results regarding relational quality between cybervictims and cyberbystanders. For example, Brody and Vangelisti (2016) found that cyberbystanders were less likely to defend or support a victim when they did not feel close to them. Similar results were found also in other studies (e.g., DeSmet et al. 2012). Bayar and Ucanok (2012) in turn reported results that perceptions of a negative school climate fostered cyberbullying in general.

For willingness to defend a victim of cyberbullying, the individual-level predictors age and gender were not significant. As was to be expected from previous research, increased levels of cognitive and affective empathy predicted positive bystander behavior. Being able to understand and feel another person’s emotions seems to foster willingness to defend others against cyberbullying. Empathy induction might therefore be a good starting point for prevention and intervention measures, although not all studies showed both dimensions of empathy to be relevant (Macháčková and Pfetsch 2016; Barlińska et al. 2018). Even after controlling for offline and online bullying and victimization, teacher support positively predicted defending behavior. Teacher support might be related to an overall positive climate in the class and the school. Eliot et al. (2010) were able to show that high perceived school and teacher support increased willingness to seek help in cases of bullying. A significant predictor of willingness to defend was reduced safety problems. It seems that in order to defend others, it is helpful to perceive the school as a safe place where possible retaliation need not be feared.

Class-level offline bullying, i.e., self-identification of perpetrators, not perceived prevalence of bullying in class, was a significant predictor of assisting. Classes, which include more bullies, also foster negative online bystander behavior. This may be due to group pressure dynamics, which were also implied by the measure that was used. Bastiaensens et al. (2016), for example, found that bystanders’ perception of friends’ approval of cyberbullying increased the perception of social pressure to participate in cyberbullying. This could possibly be transferable to approval of bullying in the offline context communicating general approval of and social pressure for aggressive behavior.

Limitations

The results of the present analyses contribute to an area of cyberbullying that has not been addressed extensively yet. However, there are some limitations to the study. One of them is that by excluding regular cyberbullies from the analyses, we may have artificially limited our subsample of (potential) assistants. Assisting might be part of a more general aggressive behavioral pattern and some specific behaviors of assisting might have overlapped with the cyberbullying measure, especially if they were conducted repeatedly. Our approach does not allow for overlapping roles. This highlights a conceptual problem of the cyberbullying construct where it is difficult to differentiate between bystanders such as non-involved outsiders versus passive witnesses, for example, but also between active negative bystanding and cyberbullying perpetration. We focused on “pure witnesses” (students with no or very low-frequency experience or behavior) and did also not include victims and their potential defending behavior, because defending behavior of victims is likely intertwined with their own experience. However, we were interested specifically in those students who do not have frequent own perpetration or victimization experience as they are usually the largest group. We were interested in how the class context influences them over and above personal social competences when they have no or little personal (recent) experiences, which to draw upon. This approach still left us with almost 81% of the total sample.

As we reported, the ICCs were very small. This may be an artifact of the measure we used which assessed perceptions of school climate by referring to “the teachers (in this school)” rather than to an individual classroom teacher. Also, we used self-report measures which are likely prone to social desirability effects, especially when asking about prosocial behavior and self-incrimination in antisocial behavior. We tried to reduce these effects by ensuring complete anonymity of the questionnaires and by providing a quiet setting without discussions between participants about which answers they chose. Finally, intended behavior is not real behavior. We used hypothetical self-report examples; we did not ask about actual behavior in situations that the students had experienced. When comparing studies assessing intended behavior with those examining actual behavior, Lindstrom Johnson et al. (2013) found that many adolescents report intentions to defend bullied peers, but rates of actual defending are clearly lower (20%). Also, by using very specific examples, we assessed assisting and defending with only one item each and limited the potential behavioral range as well as the technology used for the cyberbullying incidents. This is due to time and space constraints of the data assessments. The advantage of using ecologically valid items with very specific item content may become a disadvantage in studies including participants not being involved in using the devices described in the items. We decided to use these specific items because these communication channels allow outsiders to respond and to get involved in whatever way they choose as compared with private communication channels (e.g., text messaging, private messaging). Still, the results described in the present study probably cannot be extrapolated to other samples. Therefore, the present results should be treated with caution. However, we did not focus on the prevalence of assisting and defending, but rather on predictors for these behaviors (and the respective intentions). Also, we did not use subscales from the same instrument for cognitive and affective empathy, but rather adapted a reactivity measure for affective empathy because we were not interested in dispositions, but emotional reactions in a cyberbullying-specific situation. Comparisons of affective and cognitive empathy among each other are therefore problematic. However, our focus was not on examining, which of these plays a greater role, but whether they each play a role at all.

Conclusions and Future Directions

The results, which we presented here, are in line with those that stress the importance of individual over contextual factors (DeSmet et al. 2016); although, we would like to emphasize that in the present study, we only examined the influence of the school/classroom environment as a contextual variable. Contextual factors may not influence bystanding behavior directly to a very large extent, but they may mediate and moderate certain associations so that fostering a positive school climate should still be a part of prevention and intervention strategies. More distal, contextual factors may still play an important role for behavior in cyberbullying situations because bullying often shows hybrid forms (i.e., a combination of online and offline behavior) (Domínguez-Hernández et al. 2018, p. 2). In order for teachers to be able to be supportive, efforts may need to be made to improve their diagnostic competence in recognizing cyberbullying among their students. This needs more research efforts in the future as well as efforts in developing efficacious intervention and prevention strategies. Schools need to be aware that they do not only play a role in offline bullying but that they also have the opportunity and the duty to become active against cyberbullying because they also affect students’ online behavior.

Notes

Funding Information

This research was financially supported by a research grant from the DAPHNE III program to combat violence against children, young persons and women of the European Commission (Action number: JLS/2008/DAP3/AG/1211-30-CE-0311025/00-69; project title “Cyberbullying in Adolescence: Investigation and Intervention in Six European Countries” granted to the University of Bologna, Italy).

Compliance with Ethical Standards

Conflict of Interest

The authors declare that they have no conflict of interest.

Disclaimer

The views expressed in this article are ours and do not represent the granting agency.

References

  1. Allison, K. R., & Bussey, K. (2016). Cyber-bystanding in context: a review of the literature on witnesses' responses to cyberbullying. Children and Youth Services Review, 65, 183–194.  https://doi.org/10.1016/j.childyouth.2016.03.026.CrossRefGoogle Scholar
  2. Barlińska, J., Szuster, A., & Winiewski, M. (2013). Cyberbullying among adolescent bystanders: role of the communication medium, form of violence, and empathy. Journal of Community & Applied Social Psychology, 23(1), 37–51.  https://doi.org/10.1002/casp.2137.CrossRefGoogle Scholar
  3. Barlińska, J., Szuster, A., & Winiewski, M. (2018). Cyberbullying among adolescent bystanders: role of affective versus cognitive empathy in increasing prosocial cyberbystander behavior. [original research]. Frontiers in Psychology, 9(799).  https://doi.org/10.3389/fpsyg.2018.00799.
  4. Barr, J. J., & Higgins-D’Alessandro, A. (2007). Adolescent empathy and prosocial behavior in the multidimensional context of school culture. The Journal of Genetic Psychology, 168(3), 231–250.  https://doi.org/10.3200/GNTP.168.3.231-250.CrossRefGoogle Scholar
  5. Bastiaensens, S., Pabian, S., Vandebosch, H., Poels, K., Van Cleemput, K., DeSmet, A., et al. (2016). From normative influence to social pressure: how relevant others affect whether bystanders join in cyberbullying. Social Development, 25(1), 193–211.  https://doi.org/10.1111/sode.12134.CrossRefGoogle Scholar
  6. Batanova, M., & Loukas, A. (2016). Empathy and effortful control effects on early adolescents’ aggression: when do students’ perceptions of their school climate matter? Applied Developmental Science, 20(2), 79–93.  https://doi.org/10.1080/10888691.2015.1067145.CrossRefGoogle Scholar
  7. Bayar, Y., & Ucanok, Z. (2012). School social climate and generalized peer perception in traditional and cyberbullying status. Educational Sciences: Theory and Practice, 12(4), 2352–2358.Google Scholar
  8. Brand, S., Felner, R., Shim, M., Seitsinger, A., & Dumas, T. (2003). Middle school improvement and reform: development and validation of a school-level assessment of climate, cultural pluralism, and school safety. Journal of Educational Psychology, 95(3), 570–588.  https://doi.org/10.1037/0022-0663.95.3.570.CrossRefGoogle Scholar
  9. Brighi, A., Ortega, R., Pyzalski, J., Scheithauer, H., Smith, P. K., Tsormpatzoudis, H., et al. (2012). European cyberbullying intervention project questionnaire - ECIPQ.Google Scholar
  10. Brody, N., & Vangelisti, A. L. (2016). Bystander intervention in cyberbullying. Communication Monographs, 83(1), 94–119.  https://doi.org/10.1080/03637751.2015.1044256.CrossRefGoogle Scholar
  11. Bronfenbrenner, U., & Crouter, A. (1983). The evolution of environmental models in developmental research. In P.-H. Mussen (Ed.), Handbook of child psychology, volume I: history, theory, and methods (4th ed., pp. 357–414). New York, USA: Wiley & Sons.Google Scholar
  12. Cappadocia, M. C., Craig, W. M., & Pepler, D. (2013). Cyberbullying: prevalence, stability, and risk factors during adolescence. Canadian Journal of School Psychology, 28(2), 171–192.  https://doi.org/10.1177/0829573513491212.CrossRefGoogle Scholar
  13. Davis, M. H. (1980). A multidimensional approach to individual differences in empathy. JSAS Catalog of Selected Documents in Psychology, 10, 85.Google Scholar
  14. del Rey, R., Elipe, P., & Ortega-Ruiz, R. (2012). Bullying and cyberbullying: overlapping and predictive value of the co-occurrence. Psicothema, 24(4), 608–613.Google Scholar
  15. Del Rey, R., Casas, J. A., Ortega-Ruiz, R., Schultze-Krumbholz, A., Scheithauer, H., Smith, P., et al. (2015). Structural validation and cross-cultural robustness of the European Cyberbullying Intervention Project Questionnaire. Computers in Human Behavior, 50, 141–147.CrossRefGoogle Scholar
  16. DeSmet, A., Bastiaensens, S., Van Cleemput, K., Poels, K., Vandebosch, H., & De Bourdeaudhuij, I. (2012). Mobilizing bystanders of cyberbullying: an exploratory study into behavioural determinants of defending the victim. Studies in Health Technology and Informatics, 181, 58–63.Google Scholar
  17. DeSmet, A., Bastiaensens, S., Cleemput, K., Poels, K., Vandebosch, H., Cardon, G., et al. (2016). Deciding whether to look after them, to like it, or leave it: a multidimensional analysis of predictors of positive and negative bystander behavior in cyberbullying among adolescents., 57, 398–415.  https://doi.org/10.1016/j.chb.2015.12.051.
  18. Domínguez-Hernández, F., Bonell, L., & Martínez-González, A. (2018). A systematic literature review of factors that moderate bystanders’ actions in cyberbullying. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 12(4), article 1,  https://doi.org/10.5817/CP2018-4-1.
  19. Eisenberg, N., Eggum, N. D., & Di Giunta, L. (2010). Empathy-related responding: associations with prosocial behavior, aggression, and intergroup relations. Social Issues and Policy Review, 4, 143–180.  https://doi.org/10.1111/j.1751-2409.2010.01020.x.CrossRefGoogle Scholar
  20. Eliot, M., Cornell, D., Gregory, A., & Fan, X. (2010). Supportive school climate and student willingness to seek help for bullying and threats of violence. Journal of School Psychology, 48(6), 533–553.  https://doi.org/10.1016/j.jsp.2010.07.001.CrossRefGoogle Scholar
  21. Englander, E. K. (2012). Spinning our wheels: improving our ability to respond to bullying and cyberbullying. Child and Adolescent Psychiatric Clinics, 21(1), 43–55.  https://doi.org/10.1016/j.chc.2011.08.013.CrossRefGoogle Scholar
  22. Ettekal, I., Kochenderfer-Ladd, B., & Ladd, G. W. (2015). A synthesis of person- and relational-level factors that influence bullying and bystanding behaviors: toward an integrative framework. Aggression and Violent Behavior, 23, 75–86.  https://doi.org/10.1016/j.avb.2015.05.011.CrossRefGoogle Scholar
  23. Fisher, B. W., Gardella, J. H., & Teurbe-Tolon, A. R. (2016). Peer cybervictimization among adolescents and the associated internalizing and externalizing problems: a meta-analysis. Journal of Youth and Adolescence, 45(9), 1727–1743.  https://doi.org/10.1007/s10964-016-0541-z.CrossRefGoogle Scholar
  24. 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(5), 1164–1180.  https://doi.org/10.1037/0022-3514.74.5.1164.CrossRefGoogle Scholar
  25. Gini, G., Albiero, P., Benelli, B., & Altoè, G. (2007). Does empathy predict adolescents’ bullying and defending behavior? [article]. Aggressive Behavior, 33(5), 467–476.  https://doi.org/10.1002/ab.20204.CrossRefGoogle Scholar
  26. Guo, S. (2016). A meta-analysis of the predictors of cyberbullying perpetration and victimization. Psychology in the Schools, 53(4), 432–453.  https://doi.org/10.1002/pits.21914.CrossRefGoogle Scholar
  27. Hamm, M. P., Newton, A. S., Chisholm, A., & et al. (2015). Prevalence and effect of cyberbullying on children and young people: a scoping review of social media studies. JAMA Pediatrics, 169(8), 770–777,  https://doi.org/10.1001/jamapediatrics.2015.0944.
  28. High, A. C., & Young, R. (2018). Supportive communication from bystanders of cyberbullying: indirect effects and interactions between source and message characteristics. Journal of Applied Communication Research, 46(1), 28–51.  https://doi.org/10.1080/00909882.2017.1412085.CrossRefGoogle Scholar
  29. Hinduja, S., & Patchin, J. W. (2012). School climate 2.0: preventing cyberbullying and sexting one classroom at a time. Thousand Oaks, CA: Corwin Press.  https://doi.org/10.1007/s13224-012-0134-z.CrossRefGoogle Scholar
  30. Holfeld, B., & Leadbeater, B. J. (2017). Concurrent and longitudinal associations between early adolescents’ experiences of school climate and cyber victimization. Computers in Human Behavior, 76, 321–328.  https://doi.org/10.1016/j.chb.2017.07.037.CrossRefGoogle Scholar
  31. IBM Corp. (2017). IBM SPSS Statistics for Windows, version 25.0. Armonk, NY: IBM Corp.Google Scholar
  32. Juvonen, J., & Gross, E. F. (2008). Extending the school grounds?- Bullying experiences in cyberspace. Journal of School Health, 78(9), 496–505.  https://doi.org/10.1111/j.1746-1561.2008.00335.x.CrossRefGoogle Scholar
  33. Kazerooni, F., Hardman Taylor, S., Bazarova, N. N., & Whitlock, J. (2018). Cyberbullying bystander intervention: the number of offenders and retweeting predict likelihood of helping a cyberbullying victim. Journal of Computer-Mediated Communication, 23(3), 146–162.  https://doi.org/10.1093/jcmc/zmy005.CrossRefGoogle Scholar
  34. Kish, L. (1965). Survey sampling. Oxford, England: Wiley.  https://doi.org/10.2307/2091346.Google Scholar
  35. Lai, M. H. C., & Kwok, O. (2015). Examining the rule of thumb of not using multilevel modeling: the “design effect smaller than two” rule. The Journal of Experimental Education, 83(3), 423–438.  https://doi.org/10.1080/00220973.2014.907229.CrossRefGoogle Scholar
  36. Lamsfuss, S. M., Silbereisen, R. K., & Boehnke, K. (1992). Empathie und Motive sozialen Handelns. In Berichte aus dem Institut für Allgemeine und Vergleichende Erziehungswissenschaft Nr. 10 (2nd ed.). Berlin: Arbeitsbereich Empirische Erziehungswissenschaft, Freie Universität Berlin.Google Scholar
  37. Lindstrom Johnson, S., Waasdorp, T. E., Debnam, K., & Bradshaw, C. P. (2013). The role of bystander perceptions and school climate in influencing victims’ responses to bullying: to retaliate or seek support? Journal of Criminology, 2013, 10–10.  https://doi.org/10.1155/2013/780460.CrossRefGoogle Scholar
  38. Ma, X. (2010). Bullying in middle school: individual and school characteristics of victims and offenders. School Effectiveness and School Improvement, 13(1), 63–89.  https://doi.org/10.1076/sesi.13.1.63.3438.CrossRefGoogle Scholar
  39. Macháčková, H., & Pfetsch, J. (2016). Bystanders’ responses to offline bullying and cyberbullying: the role of empathy and normative beliefs about aggression. [article]. Scandinavian Journal of Psychology, 57(2), 169–176.  https://doi.org/10.1111/sjop.12277.CrossRefGoogle Scholar
  40. Muthén, L. K., & Muthén, B. O. (1998-2017). Mplus user’s guide (Eight ed.). Los Angeles, CA: Muthén & Muthén.Google Scholar
  41. Paluck, E. L., & Shepherd, H. (2012). The salience of social referents: a field experiment on collective norms and harassment behavior in a school social network. [article]. Journal of Personality & Social Psychology, 103(6), 899–915.  https://doi.org/10.1037/a0030015.CrossRefGoogle Scholar
  42. Patchin, J. W., & Hinduja, S. (2010). Cyberbullying and self-esteem. Journal of School Health, 80(12), 614–621.  https://doi.org/10.1111/j.1746-1561.2010.00548.x.CrossRefGoogle Scholar
  43. Pepler, D., Craig, W., & O’Connell, P. (2010). Peer processes in bullying: informing prevention and intervention strategies. In Handbook of bullying in schools: an international perspective (pp. 469–479). New York, NY, US: Routledge/Taylor & Francis Group.Google Scholar
  44. Salmivalli, C., Lagerspetz, K. M. J., Björkqvist, K., Österman, K., & Kaukiainen, A. (1996). Bullying as a group process: participant roles and their relations to social status within the group. Aggressive Behavior, 22(1), 1–15.CrossRefGoogle Scholar
  45. Schultze-Krumbholz, A., & Scheithauer, H. (2013). Is cyberbullying related to lack of empathy and social-emotional problems? International Journal of Developmental Science, 7(3–4), 161–166.  https://doi.org/10.3233/DEV-130124.Google Scholar
  46. Schultze-Krumbholz, A., Schultze, M., Zagorscak, P., Wölfer, R., & Scheithauer, H. (2016). Feeling cybervictims’ pain—the effect of empathy training on cyberbullying. Aggressive Behavior, 42(2), 147–156.  https://doi.org/10.1002/ab.21613.CrossRefGoogle Scholar
  47. Schultze-Krumbholz, A., Hess, M., Pfetsch, J., & Scheithauer, H. (2018a). Who is involved in cyberbullying? Latent class analysis of cyberbullying roles and their associations with aggression, self-esteem, and empathy. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 12(4), article 2.  https://doi.org/10.5817/CP2018-4-2.CrossRefGoogle Scholar
  48. Schultze-Krumbholz, A., Zagorscak, P., Roosen-Runge, A., & Scheithauer, H. (2018b). Medienhelden - Unterrichtsmanual zur Förderung von Medienkompetenz und Prävention von Cybermobbing (2nd ed.). München: Reinhardt Verlag.Google Scholar
  49. Slonje, R., & Smith, P. K. (2008). Cyberbullying: another main type of bullying? Scandinavian Journal of Psychology, 49, 147–154.  https://doi.org/10.1111/j.1467-9450.2007.00611.x.CrossRefGoogle Scholar
  50. Smith, P. K., Pepler, D., & Rigby, K. (2004). Bullying in schools: how successful can interventions be? Cambridge, UK: Cambridge University Press.CrossRefGoogle Scholar
  51. Smith, P., Hymel, S., Kyobe, M., Salmivalli, C., Søndergaard, D. M., & Veenstra, R. (2019). Where is the bullying research program going, and where should it be going? Paper presented at the world anti-bullying forum, Dublin, Ireland, 4.-6. June.Google Scholar
  52. Sourander, A., Brunstein Klomek, A., Ikonen, M., Lindroos, J., Luntamo, T., Koskelainen, M., Ristkari, T., & Helenius, H. (2010). Psychosocial risk factors associated with cyberbullying among adolescents - a population-based study. Archives of General Psychiatry, 67(7), 720–728.  https://doi.org/10.1001/archgenpsychiatry.2010.79.CrossRefGoogle Scholar
  53. Sticca, F., & Perren, S. (2013). Is cyberbullying worse than traditional bullying? Examining the differential roles of medium, publicity, and anonymity for the perceived severity of bullying. Journal of Youth and Adolescence, 42(5), 739–750.  https://doi.org/10.1007/s10964-012-9867-3.CrossRefGoogle Scholar
  54. Subrahmanyam, K., Smahel, D., & Greenfield, P. (2006). Connecting developmental constructions to the internet: identity presentation and sexual exploration in online teen chat rooms. Developmental Psychology, 42(3), 395–406.  https://doi.org/10.1037/0012-1649.42.3.395.CrossRefGoogle Scholar
  55. Thornberg, R., Wänström, L., & Jungert, T. (2018). Authoritative classroom climate and its relations to bullying victimization and bystander behaviors. School Psychology International, 39(6), 663–680.  https://doi.org/10.1177/0143034318809762.CrossRefGoogle Scholar
  56. Tokunaga, R. S. (2010). Following you home from school: a critical review and synthesis of research on cyberbullying victimization. Computers in Human Behavior, 26(3), 277–287.  https://doi.org/10.1016/j.chb.2009.11.014.CrossRefGoogle Scholar
  57. Trach, J., & Hymel, S. (2019). Bystanders’ affect toward bully and victim as predictors of helping and non-helping behaviour. Scandinavian Journal of Psychology, 0(0),  https://doi.org/10.1111/sjop.12516.
  58. Veiga Simão, A. M., Ferreira, P. C., Freire, I., Caetano, A. P., Martins, M. J., & Vieira, C. (2017). Adolescent cybervictimization – who they turn to and their perceived school climate. Journal of Adolescence, 58, 12–23.  https://doi.org/10.1016/j.adolescence.2017.04.009.CrossRefGoogle Scholar
  59. Volland, C., Ulich, D., Kienbaum, J., & Hölzle, E. (2008). Doing gender by doing emotion? Die geschlechtsspezifische Entwicklung der Mitgefühlsbereitschaft im Jugendalter. Psychologie in Erziehung und Unterricht, 55, 27–38.Google Scholar
  60. Waasdorp, T. E., & Bradshaw, C. P. (2015). The overlap between cyberbullying and traditional bullying. Journal of Adolescent Health, 56(5), 483–488.  https://doi.org/10.1016/j.jadohealth.2014.12.002.CrossRefGoogle Scholar
  61. Wachs, S. (2012). Moral disengagement and emotional and social difficulties in bullying and cyberbullying: differences by participant role. [article]. Emotional & Behavioural Difficulties, 17(3/4), 347–360.  https://doi.org/10.1080/13632752.2012.704318.CrossRefGoogle Scholar
  62. Wang, M.-T., & Degol, J. L. (2016). School climate: a review of the construct, measurement, and impact on student outcomes. [journal article]. Educational Psychology Review, 28(2), 315–352.  https://doi.org/10.1007/s10648-015-9319-1.CrossRefGoogle Scholar
  63. Weissberg, R. P., Durlak, J. A., Domitrovich, C. E., & Gullotta, T. P. (2015). Social and emotional learning: past, present, and future. In R. P. Weissberg, J. A. Durlak, C. E. Domitrovich, & T. P. Gullotta (Eds.), Handbook of social and emotional learning: research and practice (pp. 3–19). New York, NY, US: The Guilford Press.Google Scholar
  64. Zurek, P. P., & Scheithauer, H. (2017). Towards a more precise conceptualization of empathy: an integrative review of literature on definitions, associated functions, and developmental trajectories. International Journal of Developmental Science, 11(3/4), 57–68.  https://doi.org/10.3233/DEV-16224.CrossRefGoogle Scholar
  65. Zych, I., Baldry, A. C., Farrington, D. P., & Llorent, V. J. (2019a). Are children involved in cyberbullying low on empathy? A systematic review and meta-analysis of research on empathy versus different cyberbullying roles. Aggression and Violent Behavior, 45, 83–97.  https://doi.org/10.1016/j.avb.2018.03.004.CrossRefGoogle Scholar
  66. Zych, I., Farrington, D. P., & Ttofi, M. M. (2019b). Protective factors against bullying and cyberbullying: a systematic review of meta-analyses. Aggression and Violent Behavior, 45, 4–19.  https://doi.org/10.1016/j.avb.2018.06.008.CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Department of Educational Psychology, Institute of EducationTechnische Universität BerlinBerlinGermany
  2. 2.Department of Education and PsychologyFreie Universität BerlinBerlinGermany
  3. 3.German University of Health & Sports (DHGS)BerlinGermany
  4. 4.Department of Education and PsychologyFreie Universität BerlinBerlinGermany

Personalised recommendations