The Role of Emotions, Moral Disengagement and Gender in Supporting Victims of Bullying

: Previous research shows that classmates supporting victims’ defence is fundamental to combat bullying. To ﬁnd a suitable response for the bullying problem, we must bear in mind how all the victim’s classmates respond and what variables can determine their helping behaviour. Moral disengagement has been demonstrated to be a factor that explains behaviour when faced with bullying. Emotions have also been shown to be relevant for bullying behaviour. This research aimed to gain knowledge of how adolescents behave when faced with bullying and to analyse how their behaviour relates to moral disengagement and both positive and negative emotions, speciﬁcally supportive behaviour for victims. In the present study 1029 students participated, all of whom came from Secondary Education, Training Cycles and Higher Secondary Education Stages in Spain. The regression analysis conﬁrmed that being male increased the likelihood of performing active and passive behaviours. Conversely, being female involved displaying more proactive behaviours. Feeling positive / pleasant emotions about bullying increased active behaviour. Feeling negative / unpleasant emotions about bullying increased behaviour in the proactive behavior group. Adolescents should be aware that stopping bullying is their personal responsibility, which depends on their behaviour with the victim. Interventions must aim to reduce moral disengagement and positive emotions about bullying by increasing negative emotions about such behaviour. Author Contributions: Conceptualization, M.C.C. and E.L.; methodology, M.C.C.; software, E.L.; validation, E.L. and S.Y.; M.C.C. and E.L.; investigation, M.C.C.; resources, S.Y.; curation, E.L.;


Introduction
Traditional bullying has been defined as an undesired aggressive behaviour that causes harm and unease and is also repeated over time as part of a relationship which clearly implies unequal power. When bullying is inflicted using electronic devices, it is known as cyberbullying [1]. Traditional bullying and cyberbullying are a problem that schools should address given its serious consequences with regard to academic performance [2] and health [3].
Following the ecological model [4], friends form a support and protection group [5], but this group can also reinforce bullying [6]. As bullying is conceived as a group process, apart from the victim/bully dyad, other classmates become key actors in preventing and intervening in bullying [7]. Therefore, the relationship between the bully and the victim is affected by other relationships which the rest of the classmates form part of [8,9]. Research has shown that classmates defending a victim are fundamental to confront bullying [10], and they become a protective factor against victimisation [11]. If this defence is absent, a defenceless situation arises, which may also make the bullying situation worse because bullies take it to mean implicit support [12]. Olweus considered that bullying can be reduced, or even eliminated, if the number of classmates who do nothing significantly lowers and the number of defenders (classmates who support bullying victims) increases [13].
So all students must be aware of the role they play for a programme against bullying to be efficient [14]. Peers can allow bullying to continue if they remain silent or can stop it altogether if they support victims [15]. Several studies point out that strategies which seek social support are efficient, and they manage to block the bully's aggression [16,17], whereas responses involving directly facing the bully do not seem efficient. The same can be stated of those that consist in doing nothing at all or ignoring the situation [18].
Most studies have focused on determining which strategies are the most widely used ones [16,19,20]. In gender terms, some studies found that more females adopt support strategies [21][22][23], while males prefer to aggressively respond [24,25]. Other studies have found no gender differences [26].
To provide a suitable response to the bullying problem, we must bear in mind all the victim's classmates' responses and which variables can determine their helping behaviour. Indeed, very little is generally known about why and how students choose the way they respond [27].
Bandura's Social Cognitive Theory identifies moral disengagement as a cognitive process by means of which people justify their aggressive behaviour or distort the consequences it can imply for others [28]. Several studies have identified that those students who display moral disengagement with bullying are more engaged in these behaviours [29,30]. Research has also related moral disengagement to responding to a bullying situation [31].
Along with moral disengagement, research has been focused on emotions as factors that explain the bullying behaviour [15,32]. The results reveal that the positive emotions felt with bullying [15,33], satisfaction [32] and lack of negative emotions [34] all reinforce bullies' behaviour. We are unaware of previous studies that have analysed the link between the emotional component and the behaviour performed when faced with bullying.

Research Problem
Many previous surveys have analysed the relationship between bullying behaviour in conjunction with moral disengagement as well as both positive and negative emotions. A preliminary exploratory study is necessary to investigate the relationship between the cognitive and emotional component and the different ways in which adolescents react towards bullying. The objective of the current study was to analyse how active, passive and proactive behaviour in the face of bullying, and specifically supportive behaviour towards the victims, with moral disengagement and emotions (positive and negative), relate to one another.
Logistic regression was considered more appropriate than confirmatory techniques because this was an exploratory study. The study aim was to analyse how moral disengagement and emotional component were associated with the different bullying coping strategies. Additionally, the multinomial logistic regression allowed us to know the probability of association between the study variables for each alternative response to bullying. Results of the multinomial logistic regression can facilitate the elaboration of prevention and intervention proposals.
This study offers new insights from previous research. To help victims of bullying it is necessary to increase the support behaviour, but also to reduce the active and passive behaviours. The current study was focused on the different ways of reacting towards bullying, simultaneously considering active, passive and proactive behaviour. Moreover, the different proactive answers were analysed to become aware of the relation between moral disengagement and emotions of helping behaviour towards the victim.

Research Questions
This study aims to learn how adolescents behave when facing bullying and to analyse how this behaviour, specifically defensive behaviour in support of victims, is related to the cognitive (moral disengagement) and emotional (positive and negative emotions) components.
According to former research, we expected to find gender differences in the responses to bullying, namely more helping behaviour being shown by females (Hypothesis H1); significant differences in moral disengagement and emotions according to bullying behaviour (Hypothesis H2); and that the students who help victims would display low moral disengagement and experience negative emotions (unpleasant) (Hypothesis H3).
Therefore, the following null hypotheses were proposed: Null Hypothesis H1. There is no gender difference in response to bullying, namely more helping behaviour being shown by females.
Null Hypothesis H2. There is no difference in moral disengagement and emotions according to bullying behaviour.
Null Hypothesis H3. There is no difference in the students who support victims would display low moral disengagement and experience negative emotions (unpleasant) in the face of bullying.

Participants
Incidental sampling took place. Five high schools with a total amount of 1258 pupils contributed to the survey. However, 3.73% of the parents did not approve the participation of their children in the research (n = 47). When the questionnaire was given out to pupils, 25 of them were absent. Of the total implemented questionnaires, 13.3% (n = 1186) revealed one or more uncompleted answers. The study included 1029 students from Secondary Education, Training Cycles and Higher Secondary Education Stages. They completed all the items in the research questionnaires; 46% were male (n = 474), 54% were female (n = 555) and 96% were aged between 12 and 19 years (M = 15.18, SD = 2.01).

Instruments
To study behaviours in response to bullying, the items from the research by Menesini et al. [15] were used and the specific item "helping the victim" was added. Responses were grouped into three categories following the classification of Cuevas and Marmolejo [35]. Active behaviour was characterised by assaulting and offering positive feedback to the bully or urging or encouraging bullying by means of laughing or making gestures that support aggression (this included the items "he/she also bullies", "join those who bully" and "not participating because it is fun watching it"). Passive behaviour was defined as keeping out of the way or ignoring what goes on (which corresponded to "not engaging in bullying"). In proactive behaviour, people act to defend victims, they seek help from peers, teachers or other adults, and they attempt to stop acts of bullying (constituted by items "telling bullies to stop", "asking an adult to help" and "supporting the victim"). Participants had to mark what they do/would do if they see/saw someone bullying a classmate. It was a multiple-choice questionnaire, which was categorized into three exclusive categories of bullying behaviour.
The moral disengagement survey by Bussey, Fitzpatrick and Raman [36] was applied. It consists in eight specific moral disengagement items for bullying behaviours in a 5-point Likert-type scale: 1 = completely disagree, 2 = disagree, 3 = not sure, 4 = agree, and 5 = completely agree. One item represented each moral disengagement mechanism: moral justification, euphemistic language, advantageous comparison, displacement of responsibility, diffusion of responsibility, distorting consequences, attribution of blame, and dehumanization. The average of the eight items was the total score for moral disengagement. The scale's consistency was adequate, at α = 0.68.
To evaluate the emotional component, we used an adaptation of the scale by Larrañaga, Navarro and Yubero, which included five negative/unpleasant emotions (anger, sadness, shame, worry and unease) and three positive/pleasant emotions (joy/fun, well-being and satisfaction) [32].
Students had to identify the emotional intensity of each proposed emotion according to their bullying experience. It was measured using a Likert-type scale with five items ranging from "1 = not at all" to "5 = very much". The rate of variables was the weighted average of the emotions which constituted the scale. The reliability for the pleasant emotions was α = 0.86 and for the unpleasant emotions was α = 0.82.

Procedure
The invitation to take part in the survey was sent by the Delegations for Education and Culture of the involved provinces. The principal office teams of the five high schools, which accepted their offer of participation, were contacted. The goal was a concrete determination on objectives of the research as well as fixing the date of data collection, trying to disturb as little as possible the high school rhythm. First of all, informed consent was obtained from the parents of those students who were minors. The scales were handed out in classrooms. Students were explained the study objective and informed about their voluntary participation and the anonymity of responses. The mean time taken to answer the battery of questions was 20 min. This study was conducted in accordance with the Declaration of Helsinki. The Clinical Research Ethics Committee of the Virgen de la Luz Hospital in Cuenca (Spain) approved the study protocol (No. PI0519).

Data Analysis
First, the percentages of the responses of behaviours when faced with bullying were analysed by studying gender differences by the Chi-squared test. Second, differences in the study variables (moral disengagement, positive and negative emotions) were studied by analysis of variance statistics in the three bully behaviour groups. Finally, the relation of gender, moral disengagement and (positive and negative) emotions with behaviours when faced with bullying was analysed by logistic regression analyses. All the analyses were carried out with the SPSS programme.

Differences in Study Variables (Moral Disengagement, Positive and Negative Emotions) between Behaviour Groups When Faced with Bullying
As Table 2 shows, the difference between groups is significant in the three study variables: moral disengagement (F = 4.97, p < 0.001), positive emotions (F = 3.18, p < 0.05) and negative emotions (F = 8.17, p < 0.001). The post hoc analyses showed more moral disengagement in the passive behaviour group with a statistically significant difference compared to active behaviour (p < 0.05) and proactive behaviour (p < 0.01). Positive/pleasant emotions were higher in the active behaviour group with a statistically significant difference compared to the passive behaviour (p < 0.01) and proactive behaviour (p < 0.01) groups. This last behaviour group presented more negative/unpleasant emotions than the passive behaviour group (p < 0.05).

Relation of Gender, Moral Disengagement and Emotions with Behaviour When Faced with Bullying
The binary logistic regression analyses were performed to explain the relation of moral disengagement and the positive and negative emotions with behaviour when faced with bullying, specifically with behaviours that support victims.
The obtained results (Tables 3 and 4) show how the study variables were involved in students' responses to bullying. A higher probability of displaying active and passive behaviours was found for males, while females showed more proactive behaviours. Moral disengagement increased passive behaviours and reduced behaviours that involved seeking help from an adult and asking bullies to stop. Feeling positive/pleasant emotions when faced with bullying increased active behaviour, whereas feeling negative/unpleasant emotions about bullying increased victims' support.

Discussion
The purpose of this research was to jointly analyse the link between moral disengagement and emotions in behaviour when faced with bullying. Most studies have been focused on analysing the response of victims or groups of bystanders, but we believe it is relevant to know which variables can determine all classmates' behaviour when faced with bullying. Salmivalli et al. [7] previously considered that the role of all classmates was a determining factor, because those who did not watch bullying while it took place knew about it.
Our research objective was to learn how adolescents would behave when faced with bullying and how this behaviour is related to the cognitive (moral disengagement) and emotional (positive and negative emotions) components. We intended to analyse how moral disengagement and emotions can lead adolescents to perform behaviours so as to support victims.
In general terms, behaviour tendency was confirmed, just as other authors have also reported [21,26]. The weight of the percentage of positive behaviours to bullying (proactive) was heavier than that of negative behaviours (active and passive). Several studies have pointed out that seeking social support is one of the strategies most widely used by victims [16,17], and revealed that most sought support from a friend. Most victims stated having received support from at least one classmate [37], being rejected less and displaying higher self-esteem than those victims who did not count on such help. Hence the effects of bullying are weaker when victims find support. Over 60% of those herein surveyed indicated that they would support victims (proactive behaviours). Being able to rely on classmates' support, and not just support from friends, would prove more effective against bullying and would have fewer negative effects on victims. Some studies have also pointed out that very few victims tell their parents or teachers about what happened [38,39]. We were able to verify that more than half the classmates indicated that they would seek help from an adult, which would involve witnessing bullying and victims being able to rely on support from their teacher and parents.
We expected to find gender differences in the obtained responses to bullying, with more support behaviours shown by females (Hypothesis H1). The results obtained did reject Null Hypothesis H1. The results confirmed that females indicated more support behaviours [21][22][23], as almost 80% of them reported proactive behaviours. As in previous research [24,25], a higher percentage of active behaviours was represented by males (17%), although 12% of the females also stated that they would support aggression. Passive behaviours were reported by most male participants.
Null Hypothesis H2 was rejected. Our results confirmed significant differences in moral disengagement and emotions depending on the behaviour shown in response to bullying (Hypothesis H2). Those students who informed that they would engage in passive behaviours in response to bullying also showed a higher level of moral disengagement. Emotions marked the differences found in active and proactive behaviours. Positive emotions to bullying were more often found in those students who informed about active behaviour, while negative emotions came over more clearly in the proactive behaviour group.
The regression analyses confirmed the relation of gender, moral disengagement and emotions with behaviour when faced with bullying. In line with other studies [21][22][23], gender seems to be an important variable when opting for a type of behaviour in response to bullying. Being male increased the probability of carrying out active and passive behaviours; conversely, being female yielded higher levels for all proactive behaviours.
The results showed evidence in favour of the Null Hypothesis H3. We expected the students who supported victims to indicate low moral disengagement and to experience negative (unpleasant) emotions about bullying (Hypothesis H3). Moral disengagement was significantly found only with active behaviour, although negative emotions entered the victims' support behaviour model.
The obtained data indicated that moral disengagement was related to the response shown to bullying [31]. Although former research has pointed out that moral disengagement increases aggressive behaviour [29,30,32], a high level of moral disengagement was related to passive behaviour when faced with bullying. For moral disengagement, responsibility shifted as people perceived that they had very little personal responsibility and assumed that others had to act. This means that they expected someone else instead of them to intervene in favour of victims. A low level of moral disengagement was also related to the proactive behaviours of asking an adult for help and asking bullies to stop.
Moreover, feeling positive/pleasant emotions about bullying increased active behaviour. These results fall in line with those reported in previous studies, which have pointed out that positive emotions towards bullying influence aggressive behaviour [15,32,33]. However, lack of negative emotions, which have been related to aggression [34], did not prove significant in behaviour when faced with bullying. Feeling negative/unpleasant emotions when faced with bullying increased the specific behaviour to support victims. In an experimental situation, previous research has indicated that for helping behaviours to be displayed, students need to think about the harm that bullying can cause, and they must put themselves in the victim's place to make them aware of their emotional state [40]. Our results pointed out that students' emotions also played a relevant role in determining the behaviour they would display to face bullying.

Limitations
One limitation of this study is the type of instruments employed, because self-report measures can lead to a social desirability bias in responses. Although the sample survey is broad, incidental sampling was used; hence, an adequate representation of the population of analysis was not possible. This fact can limit the generalization of the results obtained. The present study is correlational and based on reports made about what youths would do when faced with bullying and, therefore, allows us to know only their behavioural intention when it moves in a given direction, rather than their actual behaviour. According to the statistical analysis that we carried out; the logistic regression allowed us to know the probability of association between the study variables for each alternative response to bullying. The logistic regression facilitated the elaboration of proposals for the intervention. However, we consider that the use of structural equations to perform a pooled analysis with the possible relationships between hypotheses could be very interesting. Another study limitation lies in this study's cross-sectional design. It would be necessary to replicate the study using a longitudinal methodology to continue to examine in-depth the mediator action of emotions in bullying behaviours. This research has determined that the victim's gender also seems to influence students' behaviour when faced with bullying [41]. It would be interesting to know the differences in classmates' behaviour according to this variable. Supplementing the current information through interviews and/or discussion groups with the teenagers to understand more deeply their responses to bullying would be very useful for the research.

Conclusions and Practical Implications
To conclude, our results revealed that being female and feeling negative emotions about bullying increased behaviour in support of victims. Being male involved higher levels of negative behaviours being shown towards bullying, which would lead to continued bullying. Feeling positive emotions about bullying involved more active behaviour being displayed. Passive behaviour increased as moral disengagement did. However, while moral disengagement lowered, seeking help from an adult and asking bullies to stop increased.
This knowledge could guide us in the intervention against bullying. We must make adolescents understand that stopping bullying is their personal responsibility, and that it depends on their behaviour towards the victim [14,15]. All of them as class group members can determine the validity or invalidity of bullying [42] through their behaviour. Defending victims is an efficient measure to stop bullying [10,11]. Leaving victims alone supports bullying [12]. Intervention must address reducing moral disengagement and positive emotions when bullying occurs by increasing negative emotions about this behaviour.
Hazler [43] considered that lack of intervention might also be due to not knowing how to act. It is important that adolescents know which efficient behaviours can be adopted against bullying. Perren et al. suggested that successful measures against bullying are determined by three factors: reducing the risk, capacity to stop the problem and cushioning negative impacts on victims [34]. Only proactive behaviour is suitable to act against bullying. Whitson [44] concluded that adolescents must always clearly and meaningfully support the vulnerable classmate. Thus, it is important for adolescents to be quite clear about behaviours by means of which they can help victims. Only with knowledge about efficient intervention measures against bullying, and about the variables that determine them, shall we be able to work on this problem.
Interventions could consist in creating protection conditions through a group of classmates displaying proactive behaviours, especially to support victims. Agreeing with other research papers [32,45], the obtained results allowed us to conclude that the way in which the peer group manages its emotions when bullying occurs can explain the bullying situation that takes place among adolescents. Other studies state that those programmes which include emotional aspects are more effective against bullying [46,47]. The classroom study of Fotopoulou, Zafeiropoulos and Alegre concluded that emotional education programmes are a good alternative to improve the characteristics of the whole social group [48].
Those interventions that address improving student support relationships and awakening negative emotions about bullying can be useful for lowering bullying incidence. Nonetheless, bullying is a complex phenomenon, and we must admit that although classmates are a fundamental reference group it is not the only socialising agent involved. Some students are afraid that intervening can make them new victims, which favours not intervening at all [49]. It is the whole education community's responsibility to engage classmates to collaborate in order to eradicate bullying and to not be afraid of intervening to help victims. Teachers and parents play a key role in adolescents' socialisation. To make intervention efficient, we must work in the same direction with adolescents, teachers and families. It is important that a rejection of bullying is present in all contexts. Teachers are a fundamental tool to help victims' classmates intervene and not be afraid to support victims. In turn, parents' advice must also clearly support victims and not accept aggression.
Education centres' organisational and educational proposals and activities stress the need to perform actions that favour co-existence. However, the complexity of the theme means that intervening involves multidisciplinary and multilevel work, and we cannot place all the responsibility on teachers or on adolescents' shoulders. Let us not forget that community prevention programmes are more efficient [50]. Professionals have to act by seeking efficient and ecological interventions based on improvements being made in all those involved and not just in adolescents. Hence responsibility corresponds to everyone: school staff members, parents and those who design social, education and health policies [51]. Social Services form part of the intervention network of any town or city, which means that any gender violence between peers must form part of the intervention strategies adopted and carried out to face this problem. Including social work would address preventing difficulties related to social interaction and changes being made as both individuals and groups [52] and may become a very valuable support for intervening against bullying.