Introduction

In today’s social fabric, aggressive acts against sexual minorities are increasing, especially among adolescents (Espelage et al. 2018a, b), with serious psychological consequences for the victims (Wang et al. 2018). Homophobic bullying is a set of deliberate actions aimed at denigrating or offending one or more people belonging to a sexual minority or attacking (with verbal or physical violence) sexual identity, gender, body, behaviors and desires. In other words, bullying is a form of dehumanization that serves to attach importance to the bully at the expense of others (Rivers 2011). The literature often points out how the bully can have a network of friends (followers) who support him/her (perhaps for fear of being victimized too), or he/she may not have adequate social networks to rely on (e.g., Merrin et al. 2018). Bullies’ attitudes often are based on prejudice, but they can also be the complex result of a construction made up of personal characteristics and inadequate peer relationships. In school settings, the action of bullies can be aimed at confirming their identity, regardless of the consequences of their actions on the victim (Di Maggio et al. 2016; Poteat and Russell 2013). The bully often is not able to maintain peer relationships based on trust: as a consequence, his general aim is to assert himself using the peer group to find supporters and confirm the idea of being the biggest and most powerful (e.g., Orue and Calvete 2018; D’Urso et al. 2017; Salmivalli 2010). In this sense, bullies often show an inability to adapt in the social context, eventually demonstrating a real maladjustment (Menesini et al. 2009). Furthermore, bullies are often unable to recognize authority; they’re characterized by antisocial traits and they don’t tend to seek comfort in relationships with parents and peers (e.g., Petruccelli et al. 2017; Wang et al. 2012).

Extensive studies have suggested that not having a good friendship network based on trust can be a risk factor in the genesis of aggressive, antisocial and delinquent behaviors, especially in adolescence, which is the structuring and definition phase of self and self-esteem (Pace et al. 2018; Waller et al. 2017; Zappulla et al. 2014). The friendship ties based on loyalty, trust and intimacy represent a secure basis, an emotional comfort strongly sought by adolescents, which begins in early childhood to continue throughout life (e.g., Hartup and Stevens 1997; Rotenberg et al. 2014). A recent study conducted by Merrin et al. (2018) confirms this position, finding that there is no substantial connection between actions of bullying and being popular among peers. In other words, bullies are often not able to properly manage the social network for their own return. The literature emphasized how the bully often has poor empathic skills and is not very excitable. A bully is often a person who is not able to regulate his/her emotions, is perceived as not very emotional and therefore may be unable to establish positive relationships with peers (e.g., van Noorden et al. 2017; Morcillo et al. 2015). Previous research has suggested that bullies are not necessarily antisocial, but rather behave in ways that receive peer approval (Rigby and Slee 1991). Recent studies (Zych et al. 2016; Mitsopoulou and Giovazolias 2015) have highlighted how bullies have poor empathic skills and are generally callous or unemotional. Furthermore, the study conducted by Thornberg and Jungert (2017) suggested that bullying behaviors in children are associated with a higher level of callousness and uncaring. In line with the scapegoat theory (Roberts 2015; Rivers 2011), the bully with an aim to shift attention to himself directs it towards a designated victim (e.g., the gay, lesbian or presumed person). In this way, the bully may believe he/she has a social role (even imaginary), or aspires to it, within the school group to which he/she belongs. However, the bully takes out his/her anger and interpersonal deficiencies on the scapegoat to dehumanize the other, and to show his/her poor empathic ability. These poor interpersonal skills often arise from a failure in peer relationships. The literature (e.g., Rotenberg and Boulton 2013; Vitaro et al. 1997; Hodges et al. 1997), underlines how the inability to establish relationships with peers based on trust may be connected with poor interpersonal and expressive abilities. Particularly in adolescence, interpersonal abilities are developed from satisfying friendly networks based on trust (Pace et al. 2016). In contrast, an adolescent who tends not to trust peers will not develop the adequate interpersonal skills for a positive outcome. In this sense, the results of development can be translated through aggressive attitudes among peers, such as bullying (e.g., Laninga-Wijnen et al. 2017). In adolescence, the complementarity between trust in peers and interpersonal traits can establish circle of mutual influence that, if negative, can lead to negative outcomes, and the adolescent may build relationships in which he/she feels powerful only if he/she subjugates a designated victim. In other words, the constructive positive push to the relationship derives from the basic trust that adolescents have with peers, re-proposing the schema of the primary relationship (e.g., Nikiforou et al. 2013). Nevertheless, Pozzoli et al. (2012) suggested that it was necessary to study the different characteristics of bullying through a social‐ecological approach.

In light of the aforementioned argument, the aim of this study is to identify the risk factors related to peer relationships, the individual characteristics of bullies, and how they can explain the genesis of homophobic bullying acts. In particular, in line with the theoretical frameworks, we explore the role of trust between peers and personal characteristics (related to gender‐role attributes) on homophobic bullying.

Participants and Procedure

Participants in this study were 334 adolescents, (141 boys—42.1%- and 193 girls—57.6%) aged from 15 to 20 years (M = 16.50; SD = .87), attending the third and fourth classes of some public high schools situated in Italian cities. Regarding sexual orientation participants were: 94.6% heterosexual, 2.7% predominantly heterosexual but in some circumstances homosexual, .9% predominantly heterosexual but with a strong homosexual component, .9 bisexual, and .9% homosexual. A written informed consent was obtained for all by sending letters to their parents in order to inform them of the study. No parents objected to their child’s involvement in the study. We also obtained assent from all the adolescents involved in the study. Data were collected between October 2017 and January 2018. This research was approved by the ethics committee of “Kore” University of Enna. Therefore, all procedures which involved human participants were performed in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Measures

General Information

We used an anamnestic questionnaire that investigates information related to gender, age, religious orientation (with particular attention to the importance of religion) and political, current relational situation, country of birth, city and school. In addition, the Kinsey Scale (Kinsey et al. 1948) was included in the same questionnaire for the classification of the sexual orientation of the participants (from 1 = exclusively heterosexual to 7 = exclusively homosexual).

Homophobic Bullying

The homophobic bullying scale (Prati 2012) is a questionnaire aimed at measuring homophobic bullying behaviors by students, through three perspectives: witness, bully and victim. Participants were asked to report if they observed or were involved in different homophobic behaviors (isolation/exclusion, spread of lies, homophobic skirmishes, theft or damage of property, physical assault, sexual/electronic harassment) in their schools, in the last 30 days. Response options are on 4-point Likert scale [never (1), only once or twice (2), about once a week (3), several times a week (4)]. In the present study we used the scale of bullying acted towards sexual minorities and/or alleged sexual minorities.

Personal Characteristic

The personal attributes questionnaire (PAQ; Spence and Helmreich 1978) is a self-report questionnaire which measures the degree to which an individual has internalized instrumental and expressive gender‐role attributes. This questionnaire measures these characteristics through two scales “instrumentality” and “expressivity”, commonly considered as masculinity and femininity, respectively. This tool has 24 bipolar items rated on a 5-point scale in which people are asked to indicate the extent to which they can be characterized in terms of various adjectives (e.g., “Not at all aggressive” to “Very aggressive”; “Not at all emotional” to “Very emotional”). The items are scored together to yield three scores: self-assertive or instrumentality (masculinity, eight items), interpersonal traits or expressivity (femininity, eight items) and androgyny (masculinity–femininity, eight items). This last scale has been abandoned. Indeed, it remains the full 24-item scale intact even though they don’t score the androgyny subscale.

Trust Among Peers

We used also the inventory of parent and peer attachment (IPPA; Armsden and Greenberg 1987; Baiocco et al. 2009). This scale contains a three-part self-report questionnaire that assesses adolescent attachment to mother, father, and peers. It is composed 25 items for each significant figure. Each individual’s attachment to a specific person (e.g., mother, father and peers) is assessed via four subscales (trust, communication, disaffection and security). For example, the scale of trust measures the agreement of mutual understanding and respect to significant figure (e.g., peers, mother, father) and relationship with him/her, the scale of communication investigates the extent and quality of spoken communication while the scale of disaffection/alienation detects feelings of anger and interpersonal alienation. Participants must reply to the questionnaire through a 5-point Likert scale (range 1–5), which ranges from 1 = “Never true” to 5 = “Always true”. For the present study we used the subscale related to trust among peers (e.g., “My friends understand me”, “My friends accept me as I am”).

Results

Preliminary Analyses

We conducted a univariate analysis of variance (ANOVA) to explore the role of gender on homophobic bullying, personal characteristics pulled out from PAQ’s scales (interpersonal traits and self-assertive) and trust among peers.

Data showed significant main effects of gender on interpersonal traits or expressivity [F(1331) = 11.00, p < .001], on trust among peers [F(13,331) = 4.72, p < .05], with females scoring lower than males in all cases (Table 1).

Table 1 Descriptive statistics of the study variables
Full size table

Correlational analyses (Table 2) showed how higher level of homophobic bullying was negatively correlated to higher level of interpersonal traits and to trust among peers. However, homophobic bullying was positively correlated to higher level of self-assertiveness. Moreover, the two scales of PAQ (interpersonal traits and self-assertiveness) were positively related to higher level of trust among peers.

Table 2 Intercorrelations between the variables of the mediation model
Full size table

The Mediating Effect of Interpersonal Traits on the Relationship Between Lack of Trust Among Peers and Homophobic Bullying

Following the steps enumerated by Baron and Kenny (1986), on the basis of preliminary results we conducted three set of regression analysis to examine whether Interpersonal Traits mediated the relationship between Trust among Peers and Homophobic Bullying. As the first set is concerned the following steps must be met to establish mediation (see Fig. 1): (1) the independent variable (X; trust among peers) negatively predicts the dependent variable (Y; homophobic bullying); (2) the independent variable (X) predicts the mediator (M; interpersonal traits or expressivity); (3) the mediator (M) and the independent variable (X) predict the dependent variables (Y) with the effect of X on Y that becomes not significant or that decreases when controlling for M. Results from regression analyses are presented in Table 3. The first set of analyses examined whether Interpersonal traits or expressivity mediated the relationship between Trust among peers and Homophobic bullying. The total effect of Trust among Peers on Homophobic bullying (path c of Fig. 1a) was statistically significant (β = .11, t  = − 2.10; p < .05), the effect of Trust among Peers on Interpersonal traits or expressivity (path a of Fig. 1, panel B) was statistically significant (β = .21, t = 3.86; p < .000), and the effect of interpersonal traits or expressivity on homophobic bullying (path b of Fig. 1b) was statistically significant (β = − .16, t  = − 2.98; p < . 05). Finally, the magnitude of the direct effect of trust among peers on homophobic bullying when controlling for the effect of interpersonal traits or expressivity (path c′ of Fig. 1b) has increased (β  = − .08, t  = − 1.5; n.s.) compared with the total effect of trust among peers on homophobic bullying, suggesting a total mediation. Bootstrapping analyses indicated that Trust among Peers has exerted an indirect effect (− .072; ab in Fig. 1b) on homophobic bullying through the intervention of interpersonal traits or expressivity (95% confidence interval [CI] = .08–.20).

Fig. 1
figure 1

a Illustration of the total effect: X affects Y. b Illustration of the mediation design: X affects Y indirectly through M

Full size image
Table 3 Interpersonal traits or expressivity mediating the association between trust among peers and homophobic bullying
Full size table

Discussion and Conclusion

The present study intended to explore, in a group of adolescents, some characteristics of homophobic bullying. In particular, we examined whether the actions of bullying were predicted by a lack or presence of peer trust, which are little-investigated variables in relation to this phenomenon. Furthermore, we verified whether and how personal characteristics are related to the aetiology of these violent and aggressive episodes, and whether they play a role in the relationship between peer confidence and acting bullying.

Preliminary analyses demonstrated that female participants reported higher levels than boys on interpersonal traits and expressivity scales. This result, in line with literature (e.g., Flannery and Smith 2017; Passanisi and Pace 2017; Spence and Helmreich 1978), suggests that girls consider themselves to have more interpersonal characteristics than males. In other words, girls perceive themselves as more sensitive, warmer and open to others. Furthermore, analysis indicates that girls are more capable than boys of establishing trust with their peers. This data also agrees with literature (e.g., Rose and Asher 2017) and is related to previous results. In other words, girls are more available to others, and have greater abilities and willingness to create bonds based on trust, compared to males. However, these perceptions are also connected to gender roles, above all in Western culture. Furthermore, no significant gender differences have emerged on homophobic bullying. This data suggests how this phenomenon can affect both genders.

Subsequently, analysis describes how a lack of trust in peers is connected to homophobic bullying. This result is in line with previous research (e.g., van Noorden et al. 2017), and confirms that those who undertake acts of homophobic bullying tend to be adolescents with little trust in their peers, maybe because they are centered on themselves and on the desire to affirm their social status without considering the possibility of “exploiting” the peer group. Furthermore, bullies are likely to use acts of harassment as a defense. In other words, bullying can be an attempt to seek approval and peer trust (Rigby and Slee 1991). Therefore, the desire to prevail over others can lead to the absence of interest in trust-based friendships; it is also possible that the lack of trust in peers is due to the fact that they cannot find supporters in the social group they belong to. Another possible interpretation is that bullies tend to place little trust in others and, consequently, rely only on themselves. In line with literature (e.g., Nikiforou et al. 2013), trust between peers is related to interpersonal traits. In this sense, adolescents develop more expressive skills if they can maintain relationships with peers based on trust. That’s why trust among peers in adolescence may represent a protective factor related to positive developmental outcomes.

Moreover, the mediation model establishes how a lack of interpersonal traits and expressive characteristics is a very important risk factor in the relationship between poor peer trust and homophobic bullying. According to literature (Thornberg and Jungert 2017; Zych et al. 2016; Mitsopoulou and Giovazolias 2015), among adolescents, personal characteristics should not be neglected; they also play a key role in explaining homophobic bullying behavior. Bullies, although they are not very capable or willing to establish relationships of trust with peers, are more prone to perceive themselves as not very warm or empathic. That’s why this cognitive distortion impacts their disposition to undertake homophobic bullying. In line with the triadic reciprocal determinism model (Bandura 1986), it can be deduced how the adolescent’s characteristics, linked to emotionality and expressiveness, play an active role in the inclination to perform acts of bullying. In this sense, a low level of trust in peers may directly influence homophobic bullying. On the other hand, this is not sufficient to explain this type of deviant behavior. Furthermore, individual variables are more relevant to homophobic bullying than the variables related to social context (trust between peers).

Although this work contributes to extending the literature on risk factors related to homophobic bullying, some limitations should be considered. First, the results obtained for the entire adolescent population cannot be generalized. Future studies could test whether the model presented in this study works in other contexts. Second, it is possible that the use of self-report questionnaires to investigate a delicate phenomenon such as homophobic bullying influenced the participants’ answers regarding social desirability. Future studies could use different methods (e.g., implicit tools) and informants (e.g., teachers and parents). A third limit may be linked to the data collected, because they are not longitudinal. Subsequent studies could investigate the same variables over different time frames to develop an even clearer picture of the predisposing risk factors for homophobic bullying.

In conclusion, the present research represents an important attempt to increase understanding of the relationship between individual factors in the development of homophobic bullying.

This study also adds new information about the impact of interpersonal traits on the outcomes for adolescents, regardless of the fact that bullies have little trust in their peers. The present work shows how a bully’s low confidence in him/her self can be related to the propensity to bully, which is why these results are important to develop specific projects aimed to improve sociality and to promote positive attitudes among peers, specifically towards sexual minorities. In schools, among teachers and adolescents, there is often a lack of awareness that differences enrich society and do not destroy it.